Toma de Decisiones y Seleccio?n de Tiro en Baloncesto - Herramientas para La Evaluacio?n y El Entrenamiento
Toma de Decisiones y Seleccio?n de Tiro en Baloncesto - Herramientas para La Evaluacio?n y El Entrenamiento
Toma de Decisiones y Seleccio?n de Tiro en Baloncesto - Herramientas para La Evaluacio?n y El Entrenamiento
UNIVERSIDAD DE GRANADA
2016
Editorial: Universidad de Granada. Tesis Doctorales
ISBN: 978-84-9125-994-7
URI: http://hdl.handle.net/10481/46449
Prof. Dr. José César Perales López
Universidad de Granada
CERTIFICA:
Universidad de Granada
CERTIFICA:
Directores de la Tesis
Doctorando
Resumen…………………………………………………………………………... 15
Abstract……………………………………………………………….................... 23
Introducción………………………………………………………………………… 29
Objetivos…………………………………………………………………………… 47
Objectives………………………………………………………………………….. 53
Artículos/Articles………………………………………………………………….. 59
Artículo 1………………………………………………………………… 61
Referencias/References…………………………………………………………… 187
Agradecimientos/Acknowledgements……………………………………………… 197
Resumen
RESUMEN
tópicos abordados –desde el clásico fenómeno de hot hand hasta las aproximaciones
complementa con un apartado dedicado a los factores que influyen sobre los procesos
estudios, con los que intentamos solventar algunas de las principales limitaciones de la
En el Artículo 1, Una revisión del fenómeno hot hand como creencia subjetiva y
el fenómeno hot hand (o fenómeno de rachas), quizá el tópico que más investigación y
después de haber encadenado dos o tres aciertos consecutivos). Una buena parte de la
17
investigación se ha centrado en discernir si tal dependencia serial entre ensayos existe,
desde un punto de vista estadístico, o no. En nuestra revisión se deja de lado, sin
observan diferencias en función del nivel de pericia y del marco en el que se plantea el
juicio o la decisión. No obstante, aún queda mucho por investigar sobre los factores
Los estudios sobre las consecuencias de la creencia en rachas, por su parte, muestran
que esa creencia tiene un fuerte impacto en las decisiones. Más específicamente, la
más arriesgada. Sin embargo, queda por aclarar en el futuro el impacto de esa tendencia
situaciones de tiro. Tras cada observación tenían que decidir si el jugador con balón
18
defensiva, Distancia de lanzamiento, Rebote ofensivo, Balance defensivo y Alternativa
claves a las que debían atender (todas excepto distancia), mientras que el grupo
incidental no recibió información previa sobre las claves, y sólo contaban para aprender
grupos. Los resultados mostraron que los participantes incorporaron de forma eficiente
las claves a sus decisiones. Las instrucciones del grupo intencional tuvieron un efecto
beneficioso sobre el aprendizaje de algunas de las claves sobre las que se facilitó
información. Sin embargo, ese grupo utilizó la clave Distancia de forma menos eficiente
que el grupo incidental. Esto muestra que las instrucciones intencionales sobre cuatro de
las cinco claves relevantes bloqueó el aprendizaje sobre la clave no incluida dentro de
las instrucciones.
an exploratory study, damos los primeros pasos hacia el desarrollo de un modelo que
acción del lanzamiento, como por ejemplo anotar o no anotar (y que ocurre a
19
y resultados próximos mostraron que una mayor Oposición y Distancia reducen la
probabilidad de anotar (OR = .81; p < .001 y OR = .89; p = .013); una mejor disposición
al Rebote ofensivo aumenta la probabilidad de coger el rebote (OR = 1.57; p < .001);
recibir un contraataque (OR = 1.27; p < .036). Un segundo bloque de regresiones entre
resultados próximos y distales mostraron que la efectividad del lanzamiento y del rebote
ofensivo predicen positivamente la puntuación total del partido (β = .62; p < .001 y β =
.32; p < .001) y el resultado final (ganar o perder el partido; OR = 1.12; p < .001 y OR =
1.05; p = .021). Finalmente, un análisis del impacto de las claves decisionales sobre los
selección del lanzamiento. Esta evidencia podría ayudar a los entrenadores a la hora de
ofrecer un feedback objetivo sobre el rendimiento del lanzamiento de los jugadores más
afecta al curso del partido. El Estudio 2 demuestra que las decisiones de tiro se vinculan
progresivamente, vía aprendizaje, a las claves que las preceden, en función del feedback
que se recibe de ellas. Y, lo que es más importante, el aprendizaje sobre unas claves u
los que la dirección de la atención por parte del entrenador puede ser al mismo tiempo
20
caracterizar el tiro, ya no sólo desde el punto de vista de sus resultados (e.g., encestar o
no), sino también de las claves decisionales que determinan ese tiro. Este avance
metodológico abre la puerta, tal y como muestra el Estudio 3, a evaluar los tiros reales
de un jugador o un equipo sobre la base de si en esos tiros se han tenido en cuenta las
claves presentes, y si esas claves son las propias de una buena o mala selección (esto es,
sino por la calidad misma de la decisión de tirar, medida objetivamente por la presencia
de las claves estadísticamente más informativas, posibilita: (1) evaluar los resultados
reflexividad) y situacionales (e.g., situación del tiro dentro de una secuencia percibida
21
Abstract
ABSTRACT
This doctoral thesis entitled ―Decision-making and shot selection in basketball: Tools
for assessment and training‖ consists of three main sections: Introduction, articles and
general discussion.
The first section gives a global view of research on judgments and decision-making
in sport, analysing its historical evolution and most important fields of study –from
classical themes like the hot hand phenomenon to more modern perspectives such as
that impact on judgments and decision-making learning is included (the recruitment and
utilization of decisional cues and feedback in generating decisional if-then rules), and
how this knowledge influences different teaching strategies (incidental and intentional).
To finish the section, we justify the studies developed, explaining how we will try to
solve the main limitations of the research carried out to date and provide tools that allow
In Article 1, A review of the hot hand phenomenon as subjective belief and its
probably the most thoroughly studied topic related to judgments and decision-making in
several consecutive shots (for example, an increase in the probability of scoring a field
goal in basketball after having scored two or three times consecutively). A large amount
of the research has focused on discerning whether such sequential dependence between
events exists or not. In this study, however, the discussion related to the existence or
nonexistence of streaks was put aside, focusing on two less studied components of the
25
same: the nature of the hot hand belief and the behavioural implications of that belief.
Studies related to the hot hand belief confirm that people strongly believe in their
existence in most sports. Differences are observed depending on expertise and framing
behavioural point of view, the studies show that the influence of the belief in streaks has
a huge impact on subsequent decisions. The studies reviewed here suggest that the hot
hand belief causes worse and more risky decision-making. However, the impact of this
clarified.
cues, examines the advantages and disadvantages of using intentional guidance to teach
participants how discriminate between good and bad circumstances in which to take a
participants were asked to decide, in each trial, whether the player in possession of the
ball should shoot or not. After each decision, they received feedback on their response
The intentional group was instructed to utilize four of the five cues (all except Distance)
while the incidental group did not receive information on the cues and they learned just
with the feedback provided trial by trial during the experimental task (adequate or
inadequate shot). Crucially, the Distance cue was kept incidental for both groups.
Participants effectively incorporated the cues into their decisions. Instructions of the
intentional group had a beneficial effect on learning on some cues. However, this group
26
utilized Distance less efficiently than the incidental group, i.e., intentional instructions
In the article 3, Towards a decision quality model for shot selection in basketball:
an exploratory study, we take the first steps towards a shot selection quality model in
basketball that incorporates decisional cues that might be predictive, not only of
proximal results (e.g., scoring), but also of distal results (e.g., winning/losing the
systematic observation guidelines. The decisional cues under scrutiny were shooting
opposition, distance and lateral angle, disposition to offensive rebound and disposition
decisional cues and proximal results showed higher opposition and distance to decrease
the probability of scoring (OR = .81; p < .001 and OR = .89; p = .013); a better
disposition towards rebound to increase the chances of catching rebound (OR = 1.57; p
< .001); and better defensive balance disposition to decrease the probability of a fast
break (OR = 1.27; p < .036). A second set of regressions between proximal and distal
results showed shooting and offensive rebound effectiveness to predict total points
scored (β = .62; p < .001 and β = .32; p < .001) and game result (winning/losing the
game; OR = 1.12; p < .001 and OR = 1.05; p = .021). Finally, an analysis of the impact
winning and average team‘s disposition to offensive rebound (OR = 1.18; p = 018).
These results cast light on the actual weights (validities) of the different cues involved
in predicting outcomes of shooting decisions. This evidence could help coaches provide
The studies developed in this thesis have multiple implications for training and
27
existence or not of streaks, the hot hand belief causes offensive and defensive
adaptations, and affects the course of the game. Study 2 shows that shooting decisions
are progressively linked, via learning, to the cues that precede them, depending on the
feedback received from them. And, more importantly, learning about a cue or other
attention by the coach can be both beneficial and detrimental. Along with that, this
study shows the possibility of characterizing shooting, not only from the point of view
of its results (e.g., score or no score), but also by the decisional cues that determine that
shot. This methodological advance allows, as shown in Study 3, to assess the actual
shots of a player based on whether those shots take into account the cues present, and if
those cues imply a good or bad selection (i.e., whether, the presence of these cues
In sum, the possibility of evaluating shot selection, not by its consequences, but for
the decision quality of the shot, measured objectively by the presence of the statistically
most informative cues, enables: (1) to assess progressively the results of different
methods of teaching and training; and (2) to study the influence of individual variables
on the shot selection (e.g. impulsivity or reflectivity) and situational (e.g., shot situation
28
Introducción
Introducción
que tienen una repercusión directa sobre los resultados obtenidos. La planificación de
características propias del deporte (que involucra tanto la cognición como la acción, y
en el que las consecuencias de las acciones están con frecuencia estrictamente acotadas
por criterios preciso de éxito y fracaso) hacen del deporte un potencial laboratorio para
el estudio los procesos generales relacionados con los juicios y la toma de decisiones
(Gilovich, 1984).
los años 40 del siglo pasado en varias direcciones (e.g., social, económica o
estudio elaborado por Gilovich, Vallone y Tversky (1985) sobre el fenómeno hot hand
evidencia de ello, sino que parecía producirse el efecto contrario; a mayor número de
Raab y Bar-Eli, 2013), recientes estudios han utilizado enfoques novedosos que no
31
Introducción
forma ha recobrado fuerza a un debate que dura ya más de treinta años y que no parece
tener una solución cercana. Dado que una buena parte de la investigación asociada al
centrada en los dos componentes del fenómeno hot hand más frecuentemente ignorados:
A pesar del interés generado por el citado artículo (o quizá, precisamente a causa de
toma de decisión hasta la primera década del siglo XXI, infravalorando en parte su
potencial (Araújo, Davids, Chow, Passos y Raab, 2009; Bar-Eli y Raab, 2006). Así,
durante la década los ‗90 destacan los trabajos de Ripoll (1991) sobre la relación entre
(1993), sobre la toma de decisión, sobre todo, en deportes colectivos, desde una
búsqueda del efecto hot hand en distintos deportes, en busca de encontrar dependencia
secuencial entre ensayos (para una revisión, ver Bar-Eli, Avugos y Raab, 2006).
búsqueda estadística del fenómeno obviando la posible repercusión que podría tener
32
Introducción
como por ejemplo Judgement and decision making in sport and exercise: Rediscovery
and new visions (Bar-Eli y Raab, 2006) o Ecological approaches to cognition in sport
and exercise (Araújo y Davids, 2009), así como diversos libros (e.g., Judgement,
decision-making and success in sport; Bar-Eli et al., 2011). Estas publicaciones han
abordado desde múltiples perspectivas y dado lugar a una gran variedad de modelos
(Koehler y Harvey, 2008). Sin embargo, tal como muestran Bar-Eli et al., (2011) tan
sólo una docena ha sido aplicada al ámbito deportivo (ver Tabla 1). El estudio de esta
de las decisiones en el ámbito deportivo (en concreto, la Teoría del campo decisional;
33
Introducción
TABLA 1
RESUMEN DE LAS PRINCIPALES TEORÍAS SOBRE JUICIOS Y TOMA DE DECISIÓN APLICADAS AL DEPORTE
34
Introducción
diversas teorías. Por ejemplo, en nuestro primer artículo de revisión sobre la creencia y
35
Introducción
Kahneman, 1992). Dicho heurístico fue utilizado por Gilovich et al., (1985) para
creencia en rachas, aparece el heurístico take the hot (Burns 2004; Raab, Gula y
heurísticos así como sus repercusiones adaptativas. En nuestro segundo y tercer artículo,
utilizadas cuando un individuo toma una decisión y a las limitaciones o restricciones del
los ya citados Modelos Ecológicos de Sistemas Dinámicos (Araújo et al., 2006; Araújo
adaptación que Hastie y Dawes (2011) hacen del Modelo de Lente de Brunswik (1955).
distintas claves cada una de las cuales predice con una cierta validez el resultado de la
decisión. Que dichas claves tengan un mayor o menor peso a la hora de elaborar un
juicio o tomar una decisión dependerá de hasta que punto el sujeto haya aprendido a
captar su valor predictivo real (validez objetiva). Cuanto mayor sea la coincidencia
entre los pesos de las claves en las decisiones y su validez objetiva, mayor será la
36
Introducción
2007). Así, se han llevado a cabo diversos trabajos relacionados con la enseñanza de la
Moreno y Del Villar, 2014; Memmerth y Roth, 2007). Mientras los primeros se
los dos procedimientos más utilizados en la actualidad son las sesiones de vídeo fuera
de pista y las sesiones in situ que pretenden dar feedback inmediato a los jugadores para
optimizar sus conductas (Bar-Eli et al., 2011). De esta forma, los deportistas pueden
de enseñanza. En función del tipo de instrucciones, estas estrategias han sido divididas
pretende conseguir un aprendizaje dirigido por el entrenador donde éste hace consciente
37
Introducción
(ver, por ejemplo, Shanks, Rowland y Ranger, 2005 o Maddox y Ashby, 2004). Desde
deliberada hace referencia a dos procesos que son extremos de un continuo que varía en
función del grado de automatización o control de las decisiones, sin que exista una toma
claves informativas, o dejar que sea la persona entrenada quien descubra la importancia
38
Introducción
de las mismas por él/ella mismo/a a partir únicamente del feedback de sus decisiones.
Nótese que, con ello, no decimos que el participante no sea consciente de la importancia
de las claves, o que no ponga esfuerzo y recursos en descubrirlas, sólo que dichos
conoce como reglas si-entonces para guiar la conducta de los deportistas (ver, por
ejemplo, Griffin, Mitchell y Oslin, 1997). Estas reglas incluyen una situación si que
implica una conducta a seguir entonces. Utilizaremos como ejemplo una situación que
lanzamiento, el jugador que tiene el balón deberá atender a distintas variables o claves
de un posible lanzamiento (situación si) no hay oponentes cercanos, hay opción de que
los compañeros cojan un rebote ofensivo en caso de fallo y además hay compañeros que
podrían evitar un posible contraataque del equipo contrario en caso de fallo y no coger
En los últimos años, diversos autores han comparado intervenciones con estrategias
de enseñanza intencionales e incidentales para comprobar cuáles son más eficaces. Por
ejemplo, Raab (2003) comprobó que las instrucciones incidentales producen mejores
fueron encontrados por Votsis, Tzetzis, Hatzitaki y Grouios (2009) tras una intervención
39
Introducción
decisionales para incidir en las decisiones de los participantes así como la influencia de
las instrucciones intencionales a la hora de incorporar claves con poca capacidad para
lado u otro en voleibol) y el uso de este tipo de reglas si-entonces se enmarcan dentro de
claves que hacen que el lanzamiento sea apropiado y aquellas que hacen que no lo sea.
De esta forma, los elementos fundamentales en este aprendizaje serán las claves
decisionales a las que hay que atender, el feedback o criterio que indicará al jugador la
calidad del resultado de su decisión y la regla interna (si-entonces) que relaciona las
de estos tres elementos. Por un lado, la saliencia o capacidad para captar la atención de
las claves decisionales así como la atención dirigida intrínsecamente hacia las mismas
hará que el aprendizaje sea más o menos efectivo (Blair y Watson, 2009). Además, el
(Kruschke, 2003). Por otro lado, la demora del feedback influirá cuantitativa y
40
Introducción
primer lugar, del número de claves decisionales; a mayor número de claves, mayor
disposición de claves determinada, habrá una mayor probabilidad de que una decisión
sea correcta, pero no siempre lo será. En tercer lugar, el hecho de que la regla sea lineal
condiciones que por separado indican un signo de feedback positivo, dará lugar a un
feedback). Sin embargo, en las reglas no lineales, la suma de dos condiciones que por
separado indican un signo positivo, puede dar lugar a un signo del feedback negativo
(A→O, B→O, AB→ no O) (Shanks y Darby, 1998). Así, las reglas de mayor
jugador decide y mejora la calidad de las decisiones. Sin embargo, el desarrollo de esta
convicción se ha visto dificultado por la lentitud con la que los avances teóricos
41
Introducción
al considerar qué es una decisión correcta. De hecho, hasta la fecha, uno de los
para decidir qué es una buena o una mala decisión, al menos en la selección e tiro, y la
mayoría de los estudios se han basado en juicios de expertos para evaluar las decisiones
o las claves decisionales que han de tenerse en cuenta a la hora de tomar una decisión
determinada (ver, por ejemplo, Raab, 2003). Si bien la literatura ha comprobado que los
expertos tienen mayor capacidad para captar información del ambiente que los novatos
de calidad objetivo con el que poder comparar sus valoraciones hace que éstas hayan
podido ser desvirtuadas por sesgos o creencias subjetivas erróneas, propias del campo o
investigación sobre este tópico han predominado los estudios basados en tareas de
validez ecológica.
intentos de vincular la selección del lanzamiento a los resultados de juego (e.g., Csapo,
Avugos, Raab y Bar-Eli, 2015a), existe un vacío en la literatura. Este problema será
42
Introducción
sobre las cuales se pueda aprender a tomar mejores decisiones, entendiendo por
―mejores‖ el hecho de maximizar los resultados a corto y largo plazo de tales decisiones
Por tanto, los tres trabajos de esta tesis están unidos (1) por la conciencia de las
obtener criterios objetivos con los que poder definir qué es una decisión correcta y así
En el primer artículo Una revisión del fenómeno hot hand como creencia subjetiva
anterioridad, se analizó el tópico más estudiado en relación con los juicios y toma de
no es casual, y en gran medida nos sirve para ejemplificar cómo los avances propuestos
esas consecuencias conductuales llevan a mejores o a peores decisiones. Sin una forma
43
Introducción
de tiro, y sin una forma de saber si las claves que incorpora son las más informativas
sobre los resultados del tiro, tanto inmediatos como demorados, no será posible avanzar
incidental cues, intentamos comprobar las ventajas y desventajas entre las instrucciones
lanzar a canasta. Para ello, primero, hemos tenido que idear una metodología estadística
que permita evaluar el proceso gradual a través del cual los individuos aprender a tener
en cuenta las condiciones antecedentes del tiro (claves decisionales) para decidir si tirar
En este trabajo utilizamos una tarea experimental de simulación que nos permitía a
nosotros, como experimentadores, decidir a priori qué era un buen o un mal lanzamiento
y dar feedback controlado a los participantes. Durante la tarea éstos observaban una
serie de imágenes de situaciones reales previas a un posible tiro y tras cada observación,
tenían que decidir si el jugador con balón debía lanzar o no. Después de cada decisión,
claves a las que debían atender excepto una (distancia), al grupo incidental no se le
informó sobre ninguna clave, teniendo tan sólo como referencia para aprender el
44
Introducción
atender para tomar la decisión de lanzar o no; ver Perales et al., 2011 y Llorca-Miralles
et al., 2013). Sin embargo, a pesar del avance metodológico y la demostración palpable
una mejor toma de decisiones de tiro, tanto la selección de claves como la valoración de
Por ello, el siguiente paso se encaminó a vincular las claves decisionales con los
resultados reales de juego y así conocer su validez objetiva. Con dicho fin realizamos el
estudio reportado en el artículo 3, Towards a decision quality model for shot selection
in basketball: an exploratory study, donde damos los primeros pasos hacia el desarrollo
baloncesto. En este modelo se incorporaron claves decisionales que predicen no sólo los
resultados próximos a la acción del lanzamiento, como por ejemplo anotar o no anotar,
Igualmente, podría ser utilizado para analizar cómo varía la selección de lanzamiento de
45
Introducción
un jugador durante una racha, proporcionando información útil sobre el fenómeno hot
hand.
Nótese que con este último estudio, de alguna forma, se cierra un ciclo que
comprobar si dos o más aciertos encadenados llevan a los equipos a hacer una mejor o
peor selección del tiro; todo lo más se había intentado ver si los siguientes tiros son más
o menos efectivos en términos de anotación. Sin embargo, si algo queda claro en los
resultados del artículo 3 es que el hecho de que un tiro esté bien o mal seleccionado
depende de factores que van más allá de la efectividad anotadora. En este estudio,
observamos que, por ejemplo, en relación al resultado final del partido, tirar cuando hay
jugadores atacantes al rebote puede ser incluso más importante que tirar con posición
cómoda (sin oposición, o desde un punto relativamente cercano al aro). Sólo con las
del tiro depende de otros factores que van más allá, o interaccionan con la posición del
tiro en una racha: por ejemplo, el perfil de impulsividad del jugador, el estado del
marcador, la presión temporal o la importancia del partido. Esto es, facilita una vía para
de tiro.
46
Introducción
Objetivos
47
OBJETIVOS
En esta tesis doctoral se analizará la decisión de lanzar a canasta a través de tres
estudios. Más allá de los objetivos específicos de cada estudio, el objetivo general será
forma objetiva la calidad de la selección de tiro (esto es, distinguir un tiro mejor
seleccionado de uno peor seleccionado), y, por tanto, no haya tampoco forma de evaluar
identificación del problema, este conjunto de trabajos pretende formular posibles vías de
solución.
Artículo 1: Una revisión del fenómeno hot hand como creencia subjetiva y sus
Objetivo general:
creencia.
Objetivos específicos:
- Resumir las tendencias y los logros principales realizados hasta el momento, así
como las limitaciones de los resultados obtenidos, sobre todo desde un punto de
49
Artículo 2: El coste oculto del entrenamiento: la instrucción intencional sobre la
incidentales.
Objetivo general:
Objetivos específicos:
aprendizaje.
Objetivo general:
Objetivos específicos:
50
Objetivos
51
Objectives
OBJECTIVES
In this thesis, the decision to shoot at the basket will be analysed through three
different studies. Beyond the specific objectives of each study, the main aim of this
work will be to identify some of the disadvantages and practical limitations caused by
the absence of an accepted method to objectively assess the quality of shot selection
(i.e., distinguish between better and worse selected shots), and therefore there is no way
identifying the problem, this set of studies aims to formulate possible solutions.
Article 1: A review of the hot hand phenomenon as subjective belief and its
General objective:
hot hand belief as well as the studies that analyse the behavioural implications
Specific objectives:
- To propose future directions for progress in the study of the hot hand
phenomenon.
55
Article 2, The hidden cost of coaching: intentional shot adequacy discrimination
General objective:
Specific objectives:
intentional.
selection.
exploratory study.
General objective:
- To develop a decision quality model for shot selection in basketball from real
games.
Specific objectives:
rebound and defensive balance) impact on proximal results (scoring, getting the
- To test how shooting proximal results affect distal results (points scored, points
56
Objectives
57
Artículos/Articles
ARTÍCULO 1:
Suárez-Cadenas, E., Cárdenas, D., & Perales, J.C. (en prensa). Una revisión
Impact Factor: 0.487; Applied Psychology‘s Rank: 62/76; Quartile: Q4; JIF Percentile: 19.079
UNA REVISIÓN DEL FENÓMENO HOT HAND COMO CREENCIA
SUBJETIVA Y SUS CONSECUENCIAS CONDUCTUALES EN EL
DEPORTE
ERNESTO SUÁREZ-CADENAS1, DAVID CÁRDENAS1 Y JOSÉ C. PERALES2
1
Departamento de Educación Física y Deportiva, Universidad de Granada
2
Departamento de Psicología Experimental; Centro de Investigación Mente, Cerebro y
Comportamiento, (CIMCYC), Universidad de Granada
Resumen.— El término efecto hot hand (‗fenómeno de estar en racha‘) hace referencia
a un hipotético aumento del rendimiento tras el encadenamiento de varios ensayos
acertados. Se identifican tres tipos de investigaciones en contextos deportivos: (1)
centradas en la (in)existencia de rachas, (2) en los mecanismos psicológicos de la
creencia humana en las rachas y (3) en las posibles consecuencias conductuales
provocadas por la creencia. Esta revisión abarca los dos últimos aspectos: la psicología
de la creencia en rachas y de la conducta provocada por ésta. El proceso de revisión
sistemática se llevó a cabo mediante un protocolo estandarizado basado en las pautas
sugeridas por Fernández-Ríos y Buela-Casal (2009). Los estudios relacionados con la
creencia en rachas confirman que las personas creen firmemente en su existencia en la
mayoría de deportes observándose diferencias en función del nivel de pericia y de
modificaciones de framing. No obstante, aún queda por investigar más profundamente
en los factores situacionales y las características psicológicas que afectan a la
percepción de rachas. Desde un punto de vista conductual, los estudios demuestran que
creer en la existencia rachas tiene un enorme impacto en las decisiones. Los estudios
aquí revisados parecen mostrar que la influencia de la creencia en rachas provoca una
toma de decisiones peor y más arriesgada. Sin embargo, queda por aclarar en el futuro
el impacto de esa tendencia sobre indicadores objetivos de rendimiento en contextos
reales de juego.
1
Dirección de correspondencia: Ernesto Suárez Cadenas, Facultad de Ciencias del
Deporte, Carretera de Alfacar s/n 18071 Granada, España o e-mail
([email protected]).
63
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
El término hot hand (en ocasiones traducido como ‗fenómeno de estar en racha‘)
dependencia entre lanzamientos. O dicho de otra forma, como ocurriría con la serie de
fallar un tiro no variaba en función del resultado del tiro anterior. Esta contradicción
momentum (Bar-Eli et al., 2006; Iso-Ahola y Dotson, 2014). Dado que este tópico hace
referencia a un constructo que va mucho más allá del ámbito deportivo, en esta revisión
nos referiremos exclusivamente al fenómeno hot hand en el deporte, en línea con otros
64
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
En línea con los estudios pioneros, gran parte de la investigación sobre hot hand se
disciplinas deportivas (e.g., Albright, 1993a; Clark, 2005; Newton y Aslam, 2006).
Aunque un estudio meta-analítico mostró que las rachas en el deporte son más una
ilusión que una realidad estadística (Avugos et al., 2013), recientemente, autores como
forma, ha vuelto a tomar fuerza a un debate que dura ya más de treinta años y que no
parece tener una solución cercana, tal como muestran las recientes discusiones entre
comportamiento deportivo
en los mecanismos que llevan a las personas a creer que tras una sucesión de aciertos,
65
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
rachas como una falacia o una ilusión cognitiva tradicionalmente asociada al heurístico
que un evento B, siempre que A sea más representativo que B; Kahneman y Tversky,
Kahneman, 1971).
Scheibehenne, Wilke y Tood, 2011; Todd y Gigerenzer, 2007), esto es, de la posibilidad
aun no siendo estrictamente racional, puede resultar positivo. Estos hallazgos, apoyados
por estudios neurocientíficos (Xue, Lu, Levin y Bechara, 2011), han provocado un
(2012), por ejemplo, han comprobado que la creencia en rachas puede dar lugar a
heurístico take the hot (selecciona al jugador en racha; Burns, 2004) podría ser una regla
de decisión adaptativa.
hot hand en el deporte. Por un lado (1) los centrados en la (in)existencia estadística de
las rachas, y, por otro, (2) los que analizan los mecanismos psicológicos de la creencia
humana en las rachas y (3) los que buscan las posibles consecuencias conductuales –
analizar los otros dos tipos de investigaciones: (i) los mecanismos psicológicos de la
66
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
creencia en rachas y (ii) la conducta provocada por ésta; así como examinar y proponer
MÉTODO
Protocolo
estudios sobre el fenómeno hot hand en el deporte siguiendo los estándares propuestos
sobre la creencia en rachas y estudios sobre la conducta provocada por la anterior. Para
Harrington (2004).
distintos objetivos de los estudios, por tanto, los datos fueron sintetizados siguiendo un
enfoque narrativo (Lucas, Baird, Arai, Law y Roberts, 2007). Este método permite
literatura existente.
Criterios de inclusión
real o simulado, (c) relacionados con secuencias producidas por destrezas deportivas y
67
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Selección de artículos
acuerdo de .85, valorado como muy bueno (Landis y Koch, 1977). Las discrepancias se
textos completos de los artículos para determinar si cumplían con la totalidad de los
se realizó una búsqueda adelante-atrás (Botella y Gambara, 2006) en la que dos autores
68
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
artículos (20 estudios) sobre la creencia en rachas y 12 (19 estudios) sobre la conducta.
Nótese que algunos artículos incluyen varios estudios tanto en creencia como en
conducta.
RESULTADOS
y estimar porcentajes de acierto observados. Por ejemplo, Gilovich et al., (1985), tras
rachas, realizaron un estudio donde espectadores in situ tenían que predecir el resultado
del próximo lanzamiento de un jugador tras haber observado el anterior, revelando que
con mayor probabilidad un acierto tras haber observado un acierto. Resultados similares
Un análisis distinto fue llevado a cabo por Koehler y Conley (2003), examinando la
69
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
NBA. Comprobaron que los comentarios sobre rachas o sobre hot hand obedecían a la
comprobaron que los participantes clasificaban como rachas las secuencias con
aleatorias) y sin embargo, clasificaban las secuencias como aleatorias cuando tenían una
una tarea experimental basada en vídeos de voleibol donde se mostraba un jugador con
rachas y otro sin éstas, observaron que los participantes tendían a estimar un
rendimiento mayor al observado al jugador en racha (percibieron hasta dos puntos más
Köppen y Raab, 2014; Mathews 2012; Tyszka, Zielonka, Dacey y Sawicki, 2008) que
ordenador, ruleta). Los cuatro estudios encontraron que los participantes asocian las
secuencias con más rachas a habilidades humanas y las secuencias con mayor
También encontramos dos artículos que comparan juicios predictivos (evento que
continúa una secuencia) y juicios inferidos al pasado (eventos que precedían a una
significativas entre juicios predictivos del siguiente evento e inferencias del evento que
precedía a una secuencia, mientras que Burns y Corpus (2004) observan una mayor
70
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
necesario para percibir una racha, mostrando que la idea de racha surge a partir del
tercer evento (regla del tres). Castel, Rossi y McGillivary, (2012) muestran que los
adultos mayores creen más en el fenómeno hot hand que los adultos jóvenes, poniendo
en relieve que la edad podría afectar a la creencia. Gula y Köppen (2009) comprobaron
que los expertos en voleibol perciben las rachas de forma más objetiva que los
baloncesto. En esta misma línea, merece mención especial la investigación realizada por
participantes (actor vs. observador) a lo largo de dos estudios. En el primero, los roles
se determinaron mediante instrucciones y se encontró que los actores pasaron más veces
71
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
TABLA 1
RESULTADOS DE LA CREENCIA DE HOT HAND
Muestra: 33 universitarios.
Comparar la Mayor tendencia a clasificar como rachas
Baloncesto, percepción de Instrumento: Autoinforme. secuencias producidas por humanos.
Ayton y
tenis y fútbol Esquemática- rachas producidas Baloncesto-moneda [F=14.26, p<.001]; fútbol-
Fischer Variables: Juicios categóricos. ANOVA
simbólica por mecanismos ruleta [F=3.19, p<.005]; tenis-muerte [F=4.39,
(2004) Escenarios (baloncesto-
inanimados vs. p<.005]. Las secuencias con mayor alternancia
moneda; fútbol-ruleta, tenis-
humanos se atribuyen a procesos de azar inanimados.
muerte).
Muestra: 20 universitarios.
Analizar la
Los participantes percibieron un rendimiento
percepción del Instrumento: Autoinforme.
Gula y superior al real (2 puntos más) en el jugador
rendimiento
Köppen Variables: Juicios de con racha en secuencias largas e imperfectas y
observado en dos
(2009) Voleibol Real-vídeo probabilidad tras observar 10 ANOVA cortas y perfectas [F=12.59, p=.002]. Los
jugadores y
ataques. Duración de la principiantes creían ver más puntos que los
comparar entre
secuencia. Precisión del juicio expertos [F=15.35, p=.001]. Los principiantes
expertos y
tras secuencias de 10 ataques. perciben peor que los expertos.
principiantes.
Peritaje.
Comparar juicios
Muestra: 207 participantes.
pasados y Los participantes creen que hay mayor
predicciones Instrumento: Autoinforme. probabilidad de que las rachas acaben antes en
Matthews Esquemática- futuras y ANOVA la moneda (mecanismo inanimado) que en
Baloncesto Variables: Juicio categórico
(2010) simbólica comprobar la baloncesto (habilidad humana) [F=13.38, p<
futuro. Juicio categórico
influencia de la .001]. No hay diferencias entre juicios pasado
pasado. Escenario (moneda y
naturaleza de la y predicciones futuras.
baloncesto).
secuencia.
Muestra: 66 entrenadores.
El 92,3% de los entrenadores creyeron que un
Instrumento: Autoinforme jugador tenía más probabilidad de anotar un
Esquemática- Analizar las (similar a Gilovich et al., Análisis punto (55%) después de anotar 2 o 3 seguidos
Raab et al., simbólica características de 1985).
Voleibol cualitativo- que después fallar (46%) La creencia es menor
(2012) (II) la creencia en
Variables: Juicios predictivos. descriptivo que en Gilovich et al. (1985). El 84,6%
entrenadores.
Juicios de probabilidad. creyeron que los pasadores deben jugar de
acuerdo a esta creencia.
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Comprobar cómo
Muestra: 433 participantes.
afecta la T de Student para En secuencias con más alternancia se
Mathews Esquemática-
Baloncesto alternancia de la Instrumento: Autoinforme. muestras consideró la racha con menos probabilidad de
(2013) (II) simbólica
secuencia a los independientes continuar.
Variables: Juicios predictivos.
juicios posteriores.
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Comprobar cómo
afectan las Muestra: 202 estudiantes
modificaciones en universitarios. Durante el experimento: Resultados similares
el framing (actor/ al estudio anterior. Los participantes no
Instrumento: Autoinforme. perciben representaciones falsas de las tasas de
observador Análisis
Raab y determinados por Variables: Juicios categóricos acierto/error.
cualitativo-
MacMahon Voleibol Real-vídeo distintas y predictivos sobre secuencias descriptivo Después del experimento: Sólo mostraron
(2015) (II) perspectivas observadas durante una tarea creencia en rachas 31% de los actores y 28%
visuales) en la experimental. Rol (actor/ Chi cuadrado
de los observadores. No se encontró
creencia de rachas observador) determinado correlación entre la creencia en rachas y la
durante y después mediante la modificación de la conducta de rachas.
de una tarea perspectiva visual.
experimental.
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
decisionales
La Tabla 2 indica que totalidad de los estudios analizados (n=19) muestran que la
parte, 4 estudios (desarrollados en los artículos: Burns, 2004; Raab et al., 2012; Csapo y
Raab, 2014 y Csapo, Avugos, Raab y Bar-Eli, 2015a) revelan implicaciones que
En particular, Burns (2004) mostró a través de una simulación por ordenador que la
conducta de hot hand podría dar lugar a obtener más puntos en baloncesto. De forma
similar, Raab et al., (2012) mostraron que tanto entrenadores como jugadores eran
función a éstas de forma adaptativa, pasando más veces el balón al jugador que estaba
en dos estudios recientes, Csapo y Raab (2014) y Csapo et al., (2015a) comprobaron
que tanto entrenadores como jugadores aumentan la presión defensiva de forma efectiva
experimentales).
Cabe resaltar que de los estudios que muestran implicaciones positivas, ninguno
analiza directamente la conducta del jugador en racha. Por el contrario, los estudios que
78
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
artículos: Attali, 2013; Bocskocsky et al., 2014; Csapo, Avugos, Raab y Bar-Eli, 2015b;
baloncesto más complejos y con mayor riesgo de error bajo la influencia de la creencia
en rachas. Dichos estudios mostraron que los jugadores lanzan significativamente más
veces tras acierto que tras fallo (Attali, 2013; Neiman y Loewenstein, 2011), desde
mayor distancia y bajo mayor presión defensiva (Attali, 2013; Bocskocsky et al., 2014;
Csapo et al., 2015b). Además, Attali (2013), comprobó que los entrenadores realizaban
más sustituciones de jugadores que no estaban en racha que aquéllos que si lo estaban,
rachas.
encontró mayor influencia en la conducta por parte de los expertos, diferencias que no
secuenciales, donde los participantes con el rol de actor pasaron más veces al jugador en
Dos estudios se alejan de las tendencias anteriores. Por un lado, Carlson y Shu
(2007) comprobaron que la regla del 3 afecta a la conducta, es decir, que tras tres
79
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
ningún constructo. Sin embargo, mayores valores del control percibido y afrontamiento
para las rachas positivas y valores más bajos de afectividad negativa para las rachas
negativas.
80
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
TABLA 2
RESULTADOS DE LA CONDUCTA DE HOT HAND
Tipo de
Referencia Objetivo Muestra e instrumentos Estadística y variables Resultados y discusión
deporte
Muestra: 1362
Analizar la relación universitarios. Los participantes mostraron alta asociación (χ²=376,
Burns (2004) entre la creencia e p<.001) entre creencia y conducta. La creencia y la
Baloncesto Instrumento: Chi-cuadrado
(II) implicaciones posterior conducta fue mayor en los participantes con
Autoinforme.
conductuales. nivel alto que en el resto.
Reinterpretar
La comparación entre porcentaje de los lanzamientos de
resultados de Gilovich
Muestra: Lanzamientos Filadelfia 76ers (posibilidad de reacciones estratégicas)
et al. (1985).
de los Filadelfia 76ers y Probabilidades y Universidad Cornel (sin posibilidad de reacciones
Comprobar si se
del experimento condicionadas estratégicas) muestran evidencia a favor de la regla del
Carlson y Shu producen cambios
Baloncesto controlado de Gilovich et 3. (β=.11, p<.001). Los porcentajes de lanzamiento se
(2007) conductuales a partir Regresión logística binaria
al., (1985). modifican después de tres aciertos o errores en
del tercer ensayo
ambientes con reacciones estratégicas. Esos patrones no
acertado de forma
ocurren en ambientes donde no son posibles las
consecutiva (regla del
reacciones estratégicas.
3).
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Muestra: 24 deportistas
(expertos y principiantes).
Ambos grupos se vieron influenciados de forma
Instrumento: tarea significativa por el jugador en racha, dando mayor
experimental (rachas número de pases a éste (t(23) = 4.49, p<.001, d=.91)
Probabilidades
Comprobar cómo simuladas).
condicionadas Ningún efecto general fue significativo (mín. p=.92 )
influye el nivel de
Köppen y Raab Variables: Nivel de
Voleibol pericia en la relación T de Student para muestras Los dos grupos mostraron un incremento en la velocidad
(2012) (I) pericia. Número de pases
creencia y conducta de independientes de decisión al pasar al jugador con aciertos
al jugador con racha
rachas. consecutivos, sin encontrarse diferencias entre éstos
positiva o negativa. ANOVA
(d=.60). Tanto los deportistas expertos como los
Tiempo de decisión.
principiantes detectan las rachas y utilizan dicha
Probabilidad condicionada
información para tomar decisiones de pase.
de pase a un jugador u
otro.
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Muestra: 24 deportistas
(individuales y colectivos).
Hubo efecto significativo para los deportistas tanto
Instrumento: tarea individuales como colectivos de perfección de la
experimental (rachas secuencia [F=6.77, p=.017]. Tanto deportistas de
Comprobar cómo simuladas). Probabilidades deportes individuales como colectivos dieron más pases
influye el tipo de condicionadas al jugador con más aciertos consecutivos (t(23)=5.74,
Köppen y Raab Variables: Nivel de
Voleibol práctica deportiva en p<.001). Los dos grupos mostraron un incremento en la
(2012) (II) pericia. Número de pases T de Student
la relación creencia y velocidad de decisión al pasar al jugador con aciertos
al jugador con racha
conducta de rachas. ANOVA consecutivos, sin encontrarse diferencias entre éstos
positiva o negativa.
(d=.81). No hubo diferencias significativas en los pases
Tiempo de decisión.
entre jugadores en racha o no, entre deportistas
Probabilidad condicionada
individuales y colectivos.
de pase a un jugador u
otro.
Comprobar si acertar
un lanzamiento por Se encontraron diferencias significativas en la
Muestra: 173032 pares de
parte de un jugador probabilidad de lanzar un lanzamiento tras acierto y tras
Attali (2013) Baloncesto lanzamientos consecutivos Prueba de Mantel-Haenszel
aumenta la error ( χ²=758.7, p<.01). El OR de lanzar después de un
de la NBA (2010-2011).
probabilidad de lanzar acierto fue un 40% mayor que después de un error.
otro.
Muestra: 18 entrenadores
Analizar el tipo de profesionales.
respuesta defensiva y
Variables: Utilizar una Los entrenadores tendieron a incrementar
Csapo et al., la eficacia de ésta ante
Baloncesto defensa individual o con Chi cuadrado significativamente la presión defensiva hacia los
(2015)a (I) jugadores en racha
doble marcaje sobre un jugadores en racha (χ²=26.794, p<.01).
durante una tarea
experimental. jugador tras observar una
secuencia de vídeo.
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Analizar la toma de
decisión (lanzar o no) Muestra: 20 jugadores Los jugadores decidieron lanzar significativamente más
de un jugador con o profesionales. tras observar secuencias de Jordan con rachas que sin
Csapo et al.,
Baloncesto sin racha ante distintas Variables: Lanzar o pasar Chi cuadrado éstas (χ²=13.015, p<.01). Además, esta conducta fue
(2015)a (II)
estrategias defensivas tras observar secuencia de mayor ante defensas individuales que ante defensas
durante una tarea vídeo de Michael Jordan. dobles (χ²=23.123, p<.01).
experimental.
Muestra: 29 universitarios
de Ciencias del Deporte.
La distribución de pases fue similar para el jugador en
Instrumento: tarea racha y para el que no tenía rachas en ambas
Comprobar cómo
experimental (decidir si condiciones.
influyen
pasar a un jugador con
modificaciones de No hubo diferencias significativas entre las dos
acierto racheado o a un
framing condiciones actor vs. observador en el número de pases
jugador con acierto
Raab y (actor/observador; ANOVA al jugador en racha y al que no tenía racha.
aleatorio).
McMahon Voleibol mediante
(2015)(III) instrucciones) en la Variables: Rol del Autocorrelaciones En los análisis secuenciales si se observaron diferencias
conducta de pase participante: observador o entre condiciones: Los participantes en la condición de
durante la actor (mediante actor tendieron a continuar pasando significativamente
visualización de un instrucciones). Número de más al mismo jugador tras haber acertado que los
partido. pases al jugador con racha observadores [F=1.14, p=.07]
positiva o negativa.
Estrategia de pase a un
jugador u otro.
DISCUSIÓN
distintas características que pueda tener esa creencia, y cómo puede ser afectada por
las consecuencias conductuales, observamos que tras años sin ser un objeto de estudio
principal en la investigación sobre hot hand, los cinco últimos años han sido muy
negativas. Hasta la fecha, el autor más productivo en ambas temáticas es Markus Raab
impacto o Simcha Avugos (Israel) con 5 artículos publicados. Además, en esta revisión
una gran variedad de contextos deportivos, las personas tienden a creer en la existencia
89
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
dependencia secuencial en general (para una revisión ver Oskarsson et al., 2009), donde
influenciada por cuatro factores principales: quién o qué genera la secuencia, los
La mayoría de los estudios han tendido a analizar los dos primeros factores
la genere y examinar la precisión de juicios tras observar una secuencia. En este sentido,
Tyszka et al., (2008) comprobaron que las personas muestran creencia en rachas cuando
donde las secuencias son realizadas por personas), mostrando la tendencia contraria (o
‗falacia del jugador‘, Tune, 1964) cuando la secuencia es producida por mecanismos
aleatorios (e.g., ruleta de un casino). De acuerdo con estos resultados, Mathews (2013)
comprobó que los participantes tendían a asociar las secuencias con menos alternancia
entre eventos (más rachas) a habilidades humanas. Esta idea, unida a que las personas
tienen un concepto desvirtuado de qué es una secuencia aleatoria (tienden a creer que en
una secuencia aleatoria los eventos tienen más alternancia de la que estadísticamente
realmente tienen) (para una revisión, ver Hahn y Warren, 2009), muestra que la
lado el estudio de la influencia del contexto en la creencia. Tan sólo Carlson y Shu
90
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
(2007), Gula y Köppen (2009), Castel (2012) y MacMahon et al., (2014) analizaron
factores contextuales, comprobando que a partir del tercer evento positivo surge la
percepción entre los expertos. Profundizar en estos factores contextuales puede aportar
información útil en el futuro así como ayudar a establecer un marco teórico más
modificaciones del contexto, también lo es a las modificaciones del framing, tal como
función del rol aunque, sorprendentemente, los dos grupos percibieron de forma
investigador (e.g., tarea experimental donde se observa a un jugador con rachas y a otro
situ y emitir un juicio predictivo sobre el lanzamiento que está a punto de ejecutar). Si
bien es cierto que los primeros aportan una mayor validez interna y, con ello, una
relación más fiable entre variables dependientes e independientes, hay que resaltar que
al., (2009) destacaron que la mayoría de tareas utilizadas en las investigaciones sobre
deliberada o consciente antes de emitir los juicios. Este mismo hecho lo encontramos en
posible que el propio diseño de las tareas esté potenciando que los participantes emitan
91
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
juicios erróneos debido a que obligan a hacer un uso desadaptativo de los heurísticos
(2009), Raab et al., (2012) y Raab y MacMahon (2015) han aportado mejoras
imágenes. De la misma forma, reducir el tiempo para emitir juicios así como el uso de
los juicios se asemejarán más a los que se emiten en situaciones deportivas reales.
rachas evidencia que existe un largo camino por recorrer en el sentido de identificar los
rachas pues tal como muestran Raab y MacMahon (2015), la creencia de hot hand no es
estable ni lineal. Existe toda una tradición investigadora que vincula los sesgos
deportivo, por ejemplo, métodos utilizados en el ámbito de la falacia del jugador. Claros
ejemplos son los estudios de Gold (1997) donde se analizaron los momentos en los que
92
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
quienes mostraron que los participantes con mayor tendencia al juego, percibían mayor
relacionadas con la creencia en rachas. Sorprende que tras mucho tiempo sin ser una
que la creencia de que un jugador está en racha, ya sea por parte de oponentes,
condiciones ese impacto es positivo o negativo. En este sentido, tal como se describió
y, por ello, se ha asumido que sus consecuencias conductuales serían negativas. Sin
embargo, Raab et al., (2012) se hicieron la siguiente pregunta: ¿Pueden las creencias
recientes (Raab et al., 2012; Csapo et al., 2015a) poniendo en relieve la posibilidad de
comenzado a hablar del heurístico take the hot, el cual ya fue citado por Burns (2004),
creencia sobre la toma decisiones (Attali, 2013; Bocskocsky et al., 2014; Csapo et al.,
93
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
comprobando que los jugadores tienden a realizar lanzamientos más difíciles (mayor
número, a mayor distancia y con mayor presión defensiva) cuando aciertan de forma
conducta de riesgo podría explicar los datos del meta-análisis de Gula et al. (2013), que
parecen demostrar que la probabilidad de fallo aumenta tras haber acertado varias veces
consecutivas.
creencia sobre la toma de decisiones (Burns, 2004; Raab et al., 2012; Csapo y Raab,
2014; Csapo et al, 2015a). Sin embargo, un análisis más exhaustivo indica que estos
forma complementaria a éstos. Por un lado, encontramos un artículo (Burns, 2004) que
muestra implicaciones positivas mediante una simulación por ordenador, con lo cual, la
validez externa de dicho estudio se ve reducida frente a otros más actuales donde se
jugadores eran capaces de detectar rachas y actuar de forma adaptativa mediante una
secuencias de aciertos y otro sin dependencia secuencial. Si bien este estudio es de gran
dependencia entre ensayos (ver, por ejemplo, Gula et al., 2013 o Gilovich et al., 1985).
Los otros dos estudios (Csapo y Raab, 2014; Csapo et al, 2015a) mostraron
aumentando la presión sobre los jugadores tras haber acertado éstos varios lanzamientos
94
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
probabilidad de ser fallados. Sin duda, estas implicaciones positivas a nivel defensivo,
son compatibles y lógicas con las implicaciones negativas para el protagonista, en este
caso el jugador que lanza. De hecho, refuerza la idea de que la conducta de hot hand es
desadaptativa, pues el jugador que lanza podría decidir no lanzar y buscar una opción
menos arriesgada al percibir que la presión defensiva aumenta contra él. Así, los
resultados obtenidos hasta la fecha parecen mostrar que la conducta provocada por la
creencia en rachas dentro del contexto deportivo es, en la mayoría de los casos, una
irracional.
conductuales de a creencia a los resultados reales del juego. Sin embargo, consideramos
las positivas, si es que existen, del hecho de mantener o no mantener la creencia en las
rachas. En otras palabras, ¿es la creencia en rachas un sesgo que debe combatirse
pobres, tan sólo encontramos un artículo (Doron y Gaudreau, 2014) que intenta
procesos predijeran las rachas, si encontraron diferencias entre rachas positivas (mayor
95
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Además, tal como revelan Raab y MacMahon (2015), no sólo la creencia en rachas no
es estable, si no que la conducta provocada por ésta tampoco lo es. Por todo esto, sería
personalidad así como examinar cómo varía dicha conducta en función de variables
no), la localización del partido (local o visitante) o la importancia del mismo (e.g., fase
Por otro lado, salvo algún intento aislado como el de Csapo et al., (2015b) o Suárez-
criterio objetivo para decidir qué es una buena o una mala decisión. En el mejor de los
casos, se ha dispuesto del juicio externo de un experto que valora esas consecuencias,
pero que a su vez está sometido a los mismos sesgos que generan esas decisiones. Por
herramientas que permitan valorar las consecuencias de las decisiones por la ventaja o
desventaja competitiva que ofrecen. Hasta la fecha no existe un método que permita
evaluar de forma objetiva la calidad de las decisiones (Skinner, 2012), y para que pueda
96
REFERENCIAS
of mechanism: The past, present, and future of hot hand research. Thinking &
10.1177/0956797612468452
Avugos, S., Koppen, J., Czienskowski, U., Raab, M., & Bar-Eli, M. (2013). The "hot
Ayton, P., & Fischer, I. (2004). The hot hand fallacy and the gambler's fallacy: Two
10.3758/bf03206327
Blanchard, T. C., Wilke, A., & Hayden, B. Y. (2014). Hot-hand bias in rhesus
Cognition,40(3), 280.
Bar-Eli, M., Avugos, S., & Raab, M. (2006). Twenty years of "hot hand" research:
Review and critique. Psychology of Sport and Exercise, 7(6), 525-553. doi:
10.1016/j.psychsport.2006.03.001
97
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Bocskocsky, A., J. Ezekowitz, & C. Stein (2014). The Hot Hand: A New Approach to
Botella, J., & Gambara, H. (2006). Doing and reporting a meta-analysis. International
Burns, B. D. (2004). Heuristics as beliefs and as behaviors: The adaptiveness of the "hot
10.1016/j.cogpsych.2003.07.003
Burns, B. D., & Corpus, B. (2004). Randomness and inductions from streaks:
"Gambler's fallacy" versus "hot hand". Psychonomic Bulletin & Review, 11(1),
Bruine de Bruin, W., Strough, J., & Parker, A. M. (2014). Getting older isn‘t all that
bad: Better decisions and coping when facing ―sunk costs‖. Psychology and
Carlson, K. A., & Shu, S. B. (2007). The rule of three: How the third event signals the
Cartwright-Hatton, S., Roberts, C., Chitsabesan, P., Fothergill, C., & Harrington, R.
Caruso, E. M., Waytz, A., & Epley, N. (2010). The intentional mind and the hot hand:
98
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Castel, A. D., Rossi, A. D., & McGillivray, S. (2012). Beliefs About the "Hot Hand" in
Basketball Across the Adult Life Span. Psychology and Aging, 27(3), 601-605.
doi: 10.1037/a0026991
Csapo, P., Avugos, S., Raab, M., & Bar-Eli, M. (2015a). The effect of perceived
Csapo, P., Avugos, S., Raab, M., & Bar-Eli, M. (2015b). How should "hot" players in
Csapo, P., & Raab, M. (2014). "Hand down, Man down.'' Analysis of Defensive
10.1123/jsep.2013-0043
Falk & Konold (1997). Making sense of randomness implicit encoding as a basis for
99
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Fernández-Ríos, L., & Buela-Casal, G. (2009). Standards for the preparation and
Gilovich, T., Vallone, R., & Tversky, A. (1985). The hot hand in basketball: On the
Gula, B., & Köppen, J. (2009). Einfluss von Länge und Perfektion einer „Hot-Hand― -
65-70.
Hahn, U., & Warren, P. A. (2009). Perceptions of Randomness: Why Three Heads Are
10.1037/a0015241
Huber, J., Kirchler, M., & Stockl, T. (2010). The hot hand belief and the gambler's
fallacy in investment decisions under risk. Theory and Decision, 68(4), 445-462.
doi: 10.1007/s11238-008-9106-2
10.1037/a0036406
100
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
10.1037/gpr0000026
Koehler, J. J., & Conley, C. A. (2003). The "hot hand" myth in professional basketball.
Köppen, J., & Raab, M. (2012). The hot and cold hand in volleyball: Individual
Landis J.R., & Koch G.G. (1977) The measurement of observer agreement for
Lucas, P. J., Baird, J., Arai, L., Law, C., & Roberts, H. M. (2007). Worked examples of
Marcelino, R., Mesquita, I., & Sampaio, J. (2011). Effects of quality of opposition and
MacMahon, C., Koppen, J., & Raab, M. (2014). The Hot Hand Belief and Framing
Effects. Research Quarterly for Exercise and Sport, 85(3), 341-350. doi:
10.1080/02701367.2014.930089
101
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
on Oppenheimer and Monin (2009). Judgment and Decision Making, 5(2), 133-
137.
Miller, J. B., & Sanjurjo, A. (2014). A cold shower for the hot hand fallacy. IGIER
http://ssrn.com/abstract=2450479. doi:10.2139/ssrn.2450479
Miller, J. B., and Sanjurjo, A. (2015). Surprised by the Gambler's and Hot Hand
Fallacies? A Truth in the Law of Small Numbers. IGIER Working Paper no.552.
de: http://dx.doi.org/10.2139/ssrn.2627354
Newton, P. K., & Aslam, K. (2006). Monte Carlo tennis. Siam Review, 48(4), 722-742.
doi: 10.1137/050640278
Oskarsson, T., Van Boven, L., McClelland, G. H., & Hastie, R. (2009). What‘s next?
Raab, M., Gula, B., & Gigerenzer, G. (2012). The Hot Hand Exists in Volleyball and Is
Raab, M., & MacMahon, C. (2015). Does Framing the Hot Hand Belief Change
102
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Rinott, Y., & Bar-Hillel, M. (2015). Comments on a ―Hot Hand‖ Paper by Miller and
http://pluto.huji.ac.il/~rinott/publications/hothand14.pdf
Sampaio, J., Lago, C., Casais, L., & Leite, N. (2010). Effects of starting score-line,
Scheibehenne, B., Wilke, A., & Todd, P. M. (2011). Expectations of clumpy resources
Strough, J., Karns, T. E., & Schlosnagle, L. (2011). Decision-making heuristics and
biases across the life span. In G. R. SamanezLarkin (Ed.), Decision Making over
Swann, C., Keegan, R. J., Piggott, D., & Crust, L. (2012). A systematic review of the
10.1016/j.psychsport.2012.05.006
Todd, P. M., & Gigerenzer, G. (2007). Environments that make us smart: Ecological
Tversky, A., & Kahneman, D. (1971). Belief in the law of small numbers.
Tversky, A., & Gilovich, T. (1989a). The cold facts about the ‗‗hot hand‘‘ in basketball.
Chance, 2, 16–21.
103
CREENCIA Y CONDUCTA DE HOT HAND EN EL DEPORTE
Tyszka, T., Zielonka, P., Dacey, R., & Sawicki, P. (2008). Perception of randomness
Wilke, A., Scheibehenne, B., Gaissmaier, W., McCanney, P., & Barrett, H. C. (2014).
Xue, G., Juan, C. H., Chang, C. F., Lu, Z. L., & Dong, Q. (2012). Lateral prefrontal
10.1073/pnas.1111927109
Yaari, G., & David, G. (2012). "Hot Hand" on Strike: Bowling Data Indicates
Correlation to Recent Past Results, Not Causality. Plos One, 7(1). doi:
10.1371/journal.pone.0030112
104
ARTÍCULO 2:
Suárez Cadenas, E., Cárdenas, D., Sánchez Delgado, G., & Perales, J.C.
doi:10.2466/25.30.PMS.120v14x
Impact Factor: 0.546; Exp. Psychology’s Rank: 83/85; Quartile: Q4; JIF Percentile: 2.941
THE HIDDEN COST OF COACHING: INTENTIONAL SHOT
ADEQUACY DISCRIMINATION TRAINING IN BASKETBALL
HAMPERS UTILIZATION OF INFORMATIVE INCIDENTAL CUES
1
Departamento de Educación Física y Deportiva, Universidad de Granada
2
Department of Experimental Psychology; Mind, Brain and Behaviour Research Center
(CIMCYC), Universidad de Granada
1
Address correspondence to Ernesto Suárez Cadenas, Faculty of Sport Sciences,
Carretera de Alfacar s/n 18071 Granada, Spain or e-mail ([email protected]).
2
Any aspect of the work covered in this manuscript that has involved human
participants has been conducted with the ethical approval of all relevant bodies, and
such approvals are acknowledged within the manuscript (Boards of the Experimental
Psychology and Physical Education Departments, University of Granada).
107
THE HIDDEN COST OF COACHING
making skill acquisition (Wulf & Prinz, 2001; McNevin, Shea, & Wulf, 2003; Beilock,
Bertenthal, McCoy, & Carr, 2004). Most formal learning models do include parameters
representing stimulus salience and intentional attention to relevant cues, and try to
quantify the effect of such variables on the speed and asymptotic level of learning in
experimental paradigms (Mackintosh, 1975; Pearce & Hall, 1980; Kruschke, 1992). In
procedural terms, selectively pointing the relevant cues out and stressing their
importance by means of instruction can accelerate learning; on the other hand, keeping
cues incidental and/or redirecting the learner‘s attention away from them can slow down
or even completely block their utilization (Ashby & Maddox, 2005; Kim & Rehder,
2011).
The goal of the present paper was to identify the advantages and disadvantages of
between good and bad circumstances to shoot in basketball, in a task simulating the type
contribute to teams‘ efficacy (Oskarsson, Van Boven, McClelland, & Hastie, 2009; Bar-
Eli, Plessner, & Raab, 2011). Hence, coaches are becoming growingly interested in the
conditions that improve these abilities, and a number of technical works addressing this
concern have been recently published (Adkins, Bain, Dreyer, & Starkey, 2007; Sivils,
2009).
adequate for shooting can be modeled as a form of category learning, and thus a cue-
108
THE HIDDEN COST OF COACHING
learning theories (Ashby & Maddox, 2011; Best, Yim, & Sloutsky, 2013). Real,
vicarious, or simulated feedback, in the form of hit/miss, keeping the ball or not after
rebound (in case of miss), or being in a good disposition to defend (in case the rebound
is caught by the opponent team), allows players to gradually discriminate good from
bad situations to throw. Operationally, good situations are those in which the present set
of cues predicts success with a high probability, and bad ones are those in which such
set of cues signals a low probability of success. At the present moment, there is some
consensus among basketball experts and coaches that absence of direct opposition from
alternative better than shooting (i.e., an open pass line to a better situated teammate),
and adequate shooting distance, are the main cues to be taken into account (Wooden &
Nater, 2006; Krause, Meyer, & Meyer, 2008). There is no such consensus however, on
the optimal weight that should be attributed to each cue, or the rules linking these cues
to desirable feedback. Actually, these cues could interact in a non-linear manner, and
their optimal value could vary from individual to individual1 (Araújo, Davids, &
Hristovski, 2006). Nevertheless, there is also some evidence that expert coaches and
players are apt at extracting this kind of high-quality informational cues from the game
Importantly, this type of cue-based learning process naturally occurs in two different
settings: obviously, during the game, but also in extensive technical training sessions
1
As mentioned by an anonymous reviewer, the set of cues that individually determines when shooting is a
good decision is called an ‗affordance‘ from the ecological dynamics perspective. It could be said that the
current study is about how affordances emerge during training, although, in the scope of such training
should be restricted to the specific laboratory task scenario. It cannot be claimed that the study is about
how affordances develop in the pitch: neither the motor response, nor the important cues, nor feedback
(namely, none of the elements that would constitute actual affordances for shooting in basketball) are the
same in the two cases.
109
THE HIDDEN COST OF COACHING
coach and other analysts. As mentioned above, the present paper presents a simulated
task much closer to the latter situation than to the former. In other words, in relation to
the first case, the task can be considered a low-fidelity simulation, whereas, in relation
simulation (see, e.g., Di Stasi, Contreras, Canas, Candido, Maldonado, & Catena, 2010).
Nevertheless, preliminary research shows not only that the use of such cues can be
altered by instructional guidance in a lab setting (Perales, Cárdenas, Piñar, Sánchez, &
Courel, 2011), but also that real technical training producing in-game performance
improvements manifests itself in the form of changes in the way players use those cues
That said, there still is some controversy about the best way to enhance this cue-
here. On the one hand, constraint-based approaches stress the importance of natural
feedback during the game, or in training situations specially designed (constrained) for
the players to detect and utilize the relevant cues (Araújo & Davids, 2009; Travassos,
Araújo, Davids, O'Hara, Leitao, & Cortinhas, 2013). On the other hand, constructivist
and cognition-based approaches stress the importance of technical guidance for the
player to recognize both the relevant cues and the relevant feedback, and to intentionally
elaborate the relationships between them (Koedijker, Oudejans, & Beek, 2007;
Schlapkohl, Hohmann, & Raab, 2012). For the sake of brevity, the authors will not
address the many nuances of these two approaches, their rationales, or variants, nor will
take sides a priori in the controversy. Still, as discussed later, the current results may
110
THE HIDDEN COST OF COACHING
they were presented with series of video frames depicting a real game scenario and were
asked to decide on a trial-by-trial basis whether the player in possession of the ball
should shoot or not. After making each decision, they received feedback on the
correctness of their response (determined on the basis of an artificial rule, applied upon
(incidental) condition, the learners received no guiding instructions, so that they were
asked to try to learn to do better using the feedback for their decisions. The other
(intentional) group was verbally instructed to detect and utilize four of the five cues that
actually determined such feedback, in a way that reproduces how coaches teach their
players in tactical training sessions. The fifth cue (shooting distance) was kept
Hypothesis 1: Cues are not equipotent; i.e., even in similar feedback conditions,
people learn faster about some cues than about others. Perceptual salience,
preparedness, and previous exposure to those cues (see Blair, Watson, & Meier, 2009;
Kruschke & Hullinger, 2010) are among the factors that modulate acquisition.
Hypothesis 2: Intentional guidance towards the relevant cues boosts learning, but it
can also interact with cue properties. For example, in a work by Maldonado, Jiménez,
Herrera, Perales, and Catena (2006), using a contingency-based causal judgment task,
participants rapidly learned to use all relevant cues in an intentional learning condition.
salient cues into their responses, but failed to make use of the least salient ones.
Namely, the incidental condition generated ‗inattentional blindness‘, but only for some
111
THE HIDDEN COST OF COACHING
of the relevant cues. In a similar fashion, in a previous work (Perales, et al., 2011) most
However, learning to utilize rebound proceeded much more rapidly and effectively in
Hypothesis 3: Intentional guidance not only is expected to boost learning about the
cues such guidance is directed toward, but also to interfere with learning about cues not
explicitly mentioned by the instructor. In other words, the benefits of guiding the
useful information. As shown by Kim and Rehder (2011), in a study analyzing eye
changes what features are attended, with knowledge-relevant features being fixated
more often than irrelevant ones‖ (p. 649). Favored (intentional) cues ‗unfairly‘ compete
with non-favored (incidental) cues, in such a way that the latter lose their ability to drive
attention towards certain cues will boost learning to use (at least some of) the cues to
discriminate between adequate and inadequate shots, but also to deter learning about
any cue that had been left incidental in the same context, despite the fact that feedback
could make such a cue objectively discriminative. This could mean that if an instructor
fails to identify some crucial cues, the instruction might end up preventing people from
Beyond testing these predictions, this work is also aimed at providing a new
a two-stage method is proposed. In the first stage all decisions of each individual are
regressed upon the extent of presence of the cues (Opposition, Rebound, Balance,
112
THE HIDDEN COST OF COACHING
Alternative, and Distance). This allows precise estimation of whether or not, and the
degree to which, each participant used such cues when deciding if a situation was
adequate to shoot. In other words, cue- and individual-specific utilization scores are
calculated. Theoretical and practical implications emerged from the second stage;
however, the methodological novelty resides in the first stage. In that regard, the
challenge is to demonstrate that this rationale can be used to compute utilization scores,
METHOD
Participants
Sixty-five first-year sport sciences students (11 women; M age = 18.6 yr.,
SD = 1.3) volunteered in exchange for course credits. Thirty-one were randomly
assigned to the intentional group, and 34 to the incidental group. None had received
knowledgeable in basketball were not considered for the study. Naïve participants were
recruited to prevent, as much as possible, the influence of prior knowledge (apart from
All participants gave written informed consent. In all its aspects, the study was
carried out in accordance with the Declaration of Helsinki and was ethically approved
113
THE HIDDEN COST OF COACHING
the University of Granada, as part of a series of voluntary activities during the course
Measures
1,440 trials, divided into eight (180-trial) daily sessions. In each trial, participants were
asked to decide whether the player in possession of the ball (depicted in a still video
frame) should shoot or not. Trial-by-trial yes/no responses were kept for further
analysis.
Cue presence scores.—All the video frames were assessed by four basketball
point Likert scale, for Opposition, Rebound, Balance, and Alternative in each video
frame (anchors 0: absence and 4: presence to a maximum extent). The presence of each
of the cues in each video frame was computed as the average of the four experts‘
Cue presence ratings were standardized to translate them into a common scale, and
then kept for feedback programming. Correlations of ratings across experts (Table 1)
were acceptable for Opposition and Rebound. Notably, correlations were low for
Alternative and Balance. We will get back to the implications of this low level of
114
THE HIDDEN COST OF COACHING
TABLE 1
Experts Cue
compared Opposition Rebound Balance Alternative
1 vs. 2 .805** .628** .423** .326**
1 vs. 3 .815** .626** .425** .360**
1 vs. 4 .850** .655** .414** .468**
2 vs. 3 .794** .705** .541** .142**
2 vs. 4 .836** .614** .246** .196**
3 vs. 4 .842** .686** .195** .197**
Average .824** .652** .374** .282**
Note.—all correlations statistically significant at p < .01.
Cue utilization scores.—First, for each participant and each 360-trial block of the
task, a binomial logistic regression analysis was run for the actual decisions on the cue-
presence scores. The 360-trial (instead of 180-trial) blocks were used to eliminate
potential effects of the specific set of video frames used in each pair of sessions (each
session containing 180 images). These regression analyses took into account only the
trials in which feedback was not random, i.e., those trials in which feedback was
Second, provided that all cue-presence scores were in the same (standardized) scale,
interpretable as cue utilization scores, and thus were kept as the main dependent
individually, and are interpretable as estimates of the degree to which each participant
takes each cue into account to make decisions (where the score's sign indicates whether
presence of the cue biased decisions towards a yes or a no response). Please note that
the goal of this preliminary analysis was not to test whether each cue significantly
115
THE HIDDEN COST OF COACHING
Procedure
and consisted of eight 180-trial sessions, with each session performed on a different
day. The exact number of trials was determined on projections made from Perales, et
al.‘s (2011) results, to ensure enough learning, and a sufficient number of trials for
further reliable analyses. Each trial contained a video frame extracted from a
professional-level game, with the 10 players visible, and a player in possession of the
ball. That player was oriented towards the basket, but it was not possible to guess
The first session started with 10 practice trials, to familiarize participants with the
task, and to establish the key/response coupling. Different video frames (N=360) were
used. The video frames in Session 1 (n=180) were different from the ones in Session 2
(n=180). The video frames were randomly assigned to these two sessions. A first
randomization process determined the order for the other half. In all cases, mirror
images of the video frames used in Sessions 1–2 were used in Sessions 3–4. Finally,
Sessions 5–8 were identical to Sessions 1–4. Therefore, each video frame was used four
Each participant was provided a pen-drive with an executable file. Participants were
instructed in accordance with the condition they had been assigned to (incidental or
intentional), and told to run the task at home. Such an activity was considered as part of
a larger set of voluntary academic duties, and to strengthen the motivation for proper
116
THE HIDDEN COST OF COACHING
execution, participants were warned that their performance would be inspected and
observed utilization parameters incorporated the feedback provided during the task (see
confidence intervals for cue utilizations in Table 2). In addition, there is extensive
experiments to collect data (Vadillo, Bárcena, & Matute, 2006; Exadaktylos, Espín, &
Brañas-Garza, 2013).
Each trial consisted of a 2500 msec. fixation point, and a 5000 msec. presentation of
the video frame. After the video frame had disappeared from the screen, a question
mark (―?‖) indicated that a decision had to be made on whether a shot opportunity was
adequate or not. Keys Z/M were randomly assigned to yes/no responses. Upon
responding, the decision latency was shown on screen, and a positive/negative sound,
and the words ―Correcto/Incorrecto‖ immediately indicated whether the decision had
been right/wrong.
The two groups differed only in the instructions they received (see translation from
Spanish in the Appendix). The instructions for the incidental group made reference only
to the general goals of the task. Intentional instructions provided definitions of all cues
except Distance, so that participants were biased to take those cues into account.
Shooting Distance was kept incidental in the two groups. Importantly, Distance was
shown an intermediate level of utilization for it. Learners tend to incorporate Distance
into their decision rather straightforwardly, but not as much as they do with Opposition.
Given that the effects on cue utilization are expected to be subtle, choosing Distance as
the key cue to detect some of those effects allowed avoiding both ceiling and floor
effects.
117
THE HIDDEN COST OF COACHING
Selection and classification of stimuli.—427 video frames were extracted from the
Eurobasket 2009 male tournament. Scores for the five cues were used to classify each
opposition > 0.75, the video frame was classified as shooting inadequate. If
opposition < 0.75, shooting adequateness was calculated by using the following
algorithm:
1]
If y > 0.55, then the video frame was classified as adequate for shooting and if
This rule is obviously arbitrary, and was designed to ensure that barely half the
situations were classified as adequate, and the other half inadequate for shooting. In
addition, the rule was intended to reflect the relative importance attributed to them by
professional coaches, in such a way that opposition is the stronger cue, followed by
rebound, and then by the other three cues. Still, the rationality of this rule is not crucial
to the goals. What was needed was a type of feedback provided according to a rule
known a priori. No agreement about such rule exists in the basketball community,
although preliminary work from our laboratory actually shows that computing
Among the 427 original video frames, those 270 frames were selected for which the
correct response was clearest, after applying the equation to determine feedback. The
270 video frames were mixed with 90 images for which feedback values were randomly
yields the full 360 set). This was done to make feedback non-deterministic, and thus
118
THE HIDDEN COST OF COACHING
more similar to feedback in real contexts. Indeed, for that final set, the resulting linear
weight of each cue on shooting-adequateness (cue validities) was –0.40, 0.26, 0.16, –
0.16, and –0.14, for Opposition, Rebound, Balance, Distance, and Alternative,
FIG. 1. Video-frame presented in the task. For this video frame, average expert judge‘s scores on the
presence of cues were 3.00 (Opposition), 2.50 (Rebound), 3.50 (Balance), 2.50 (Alternative), and 1.00
(Distance).
First, to calculate the dependent variable (see ʻ Cue utilization scoresʼ ), binomial
logistic regression analyses were performed for each participant and each 360-trial
block of the task. Subsequently, mixed analyses of variance (ANOVAs) were carried
out on utilization scores separately for each cue (Opposition, Rebound, Balance,
119
THE HIDDEN COST OF COACHING
cues, a one-sample t test was performed for each block and each cue.
The significance threshold for all analyses was p = .05. ηp2 was used to estimate
effect size. All analyses and pre-analyses were run on the IBM SPSS 20 statistical
package.
RESULTS
Figure 2 displays mean utilization scores for each cue, group, and block. A three-way
Group: Incidental, Intentional) yielded significant effects for cue (F4, 232 = 499.14,
MSE = 0.17, p < .01, η2 = 0.90), block (F3, 174 = 2.99, MSE = 0.10,
p = .03, ηp2 = 0.05), the block × cue interaction (F12, 696 = 2.61, MSE = 0.07,
p < .01, ηp2 = 0.04), and cue × group interaction (F4, 232 = 3.52, MSE = 0.17,
p < .01, η2 = 0.06). Neither block x group (p = .16), nor block × cue × group
(p = .14) had statistically significant effects. In other words, utilization scores varied
across cues, and the effect of group on utilization also varied across cues (see details in
Some cues are more difficult to define than others (experts‘ inter-judge reliability for
Opposition and Rebound were higher than for Alternative and Balance). That means
that participants could make a weaker use of such cues, not because they are less
important, but because they are not so salient or discriminable. Effects of cue are thus
interpreted with that caution in mind. Hence, block × group ANOVAs were also
120
THE HIDDEN COST OF COACHING
For Opposition, there was a statistically significant effect of block (F3, 174 = 5.06
MSE = 0.11, p < .01, ηp2 = 0.08. Neither group (F < 1) nor block × group
(p = .15) effects reached significance. Trend analyses of the block effect indicated a
significant linear component (F1, 58 = 7.30, MSE = 0.20, p < .01, ηp2 = 0.11).
Neither the quadratic (p = .21) nor the cubic component (p = .21) were significant.
For Rebound, there was a significant effect of group (F1, 58 = 4.34, MSE = 0.11,
p = .04, ηp2 = 0.07). Neither the effect of block (p = .31) nor the effect of block ×
group (F3, 174 = 2.25, MSE = 0.03, p = .08, ηp2 = 0.04) was significant. As
shown in Fig. 3B, the utilization of Rebound was globally larger for the intentional
group.
For shooting Distance there was a significant effect of group (F1, 58 = 9.14,
MSE = 0.32, p < .01, ηp2 = 0.14), but not of block (F3, 174 = 2.16,
MSE = 0.43, p = .09, ηp2 = 0.04), nor any interaction (F < 1). Importantly, as
shown in Fig. 1E, the utilization of this cue was consistently larger for the incidental
group.
No significant effects were found for Balance and Alternative (Figs. 3C and 3D)
(min. p = .55).
121
THE HIDDEN COST OF COACHING
Opposition utilization
Rebound utilization
-0,60 0,70
-0,80 0,60
-1,00 0,50
-1,20 0,40
0,30
-1,40
0,20
-1,60
0,10
-1,80
0,00
-2,00 Block 1 Block 2 Block 3 Block 4
Balance utilization
-0,15
0,05
-0,20
-0,25 0,00
-0,30
-0,05
-0,35
-0,40
-0,10
-0,45
-0,50 -0,15
Block 1 Block 2 Block 3 Block 4
-0,10
-0,30
Distance utilization
-0,50
-0,70
-0,90
-1,10
-1,30
-1,50
FIG. 2. Mean utilization scores for the five cues across blocks (Panel A: Opposition; Panel B: Rebound;
Panel C: Balance; Panel D: Alternative to shooting; Panel E: Distance). Error bars represent +/-1 standard
errors. Positive values indicate positive association between cue presence and decision to shoot; negative
values indicate inverse association between cue presence and decision to shoot. Please note that, for
122
THE HIDDEN COST OF COACHING
Finally, a one-sample t test was carried out to check whether block-wise cue
utilizations were larger than zero and, therefore, whether participants made a significant
TABLE 2
ONE-SAMPLE t TESTS COMPARING CUE UTILIZATION (AGAINST ZERO) FOR EACH CUE,
BLOCK, AND GROUP
Group
Cue Block Intentional Incidental
P CI (95%) P CI (95%)
Opposition 1 < .01 - 1.41 - 1.10 < .01 - 1.35 - .97
2 < .01 - 1.53 - 1.30 < .01 -1.45 - 1.20
3 < .01 - 1.46 - 1.30 < .01 -1.60 - 1.24
4 < .01 - 1.50 - 1.23 < .01 -1.74 - 1.26
Rebound 1 < .01 .13 .32 .12 - .02 .14
2 < .01 .09 .27 < .01 .04 .18
3 < .01 .09 .26 < .01 .09 .26
4 < .01 .15 .35 < .01 .06 .22
Balance 1 .80 - .06 .08 .76 -.05 .03
2 .83 - .07 .06 .33 -.09 .03
3 .33 - .11 .04 .49 -.09 .04
4 .51 -.14 - .05 .49 -.05 .10
Alternative 1 < .01 - .13 .02 .08 - .13 .01
2 .14 - .16 - .02 .13 - .14 .02
3 .01 - .12 - .02 .01 - .12 - .02
4 .01 - .27 - .13 .05 - .14 .00
Distance 1 < .01 - .29 - .13 < .01 - .49 - .32
2 < .01 - .35 - .22 < .01 - .52 - .34
3 < .01 - .40 - .28 < .01 - .53 - .38
4 < .01 - .36 - .22 < .01 - 1.17 - .27
Note.— CIs refer to the mean values of utilization scores for the five cues, and the four blocks, across
participants in the two groups (displayed in Figure 2).
DISCUSSION
The first hypothesis stated that, in cue-based learning, cues are not equipotent, i.e.,
even in similar feedback conditions, people learn faster about some cues than about
others. As displayed in Table 2, people learn to use (utilization score<>0, p < .05) all
cues except balance, in at least some circumstances. That is, they incorporated the cues
123
THE HIDDEN COST OF COACHING
into their decisions in such a way that utilization scores largely varied across cues
(Opposition > Distance > Alternative > Rebound > Balance), which supports
the hypothesis. Nevertheless, this result can have multiple interpretations. On the one
hand, some cues could be more complex and less salient, and take longer just to be
perceptually discriminated (see, e.g., Pearce, 2002, for a review on the many subtleties
that people use simple cues as Opposition and Distance much earlier and
Alternative. Unfortunately, these cues are not only more complex, their definition is
also less clear-cut (i.e., more ambiguous). As shown in Table 1, inter-judge agreement
on cue presence strongly varied across cues. In terms of agreement, the order of cues
was Opposition > Rebound > Balance > Alternative, which indicates that
presence is more a diffuse property for some cues that for others, which is itself a
relevant result. Still, interestingly, the ordering of cue utilization did not match the inter-
judge agreement ranking. For example, Alternative is the cue for which there was
lowest agreement across judges, but it was incorporated into adequateness decisions
more effectively than Balance and Rebound. This implies that utilization differences
The second hypothesis was more straightforwardly testable. The effect of guidance
was markedly different across cues. In the case of Opposition, attentional guidance did
not make any difference in cue utilization. This effect is similar to that reported by
Maldonado, et al. (2006), in a different task. On the other hand, in our case, there was
no effect of group either for Balance or Alternative, but there was a clear group effect of
124
THE HIDDEN COST OF COACHING
interaction was not significant, the effect of group seemed to vanish (or at least
decrease) with practice. This pattern closely replicates the one reported by Perales, et al.
(2011), and implies that cue properties determine whether learning on them is more or
identify these properties, but it is interesting to note that the effect of guidance did not
strictly depend on the ease of learning to use a cue. As noted above, people learned
more rapidly and efficiently about Alternative than about Rebound, but guidance did not
exert any effect on Alternative utilization (but did on Rebound). This question surely
The third hypothesis stated that favored (intentional) cues would ‗unfairly‘ compete
with non-favored (incidental) cues, in such a way that the latter were expected to lose
their ability to drive attention towards them. Distance actually suffered this deterrent
effect. The intentional group not only utilized this cue less efficiently than the incidental
group, but also failed to reach the same utilization in the long term (after more than
1,080 trials). Intentional instructions almost blocked the utilization of Distance as a cue.
In spite of the limitations imposed by the type of methodology used here, there are
theoretical implications to be drawn from these results, particularly the last one. On the
one hand, the task is simple enough to be explained in cognitive terms; actually, current
general associative learning theories from which the hypotheses were drawn must be
associative learning phenomena (Shanks, 1995). The reported results can be neatly
accounted for in terms of selective attention guidance by intentional instruction, and the
and attention influence each other. On the one hand, attention guides associative
125
THE HIDDEN COST OF COACHING
allocated (Wulf, Gartner, McConnel, & Schwarz, 2002; Roelfsema, Pieter, van Ooyen,
& Watanabe, 2010). Most importantly, when several cues compete for associative
strength, attention can determine what cues are effectively associated to feedback (see
In terms of the type of knowledge generated in this task and the strategies used
during learning, results must be interpreted cautiously. Although it has been sometimes
assumed that processing during incidental tasks is mostly automatic (generating implicit
explicit knowledge; see Bar-Eli, et al., 2011, pp. 84–86 for a discussion on the matter;
assumption can be rather thorny. In tasks like these, it is perfectly possible that learners
in the incidental condition are actively searching for the cues that allow them to predict
feedback, and proceed in a serial, controlled manner. So, the ‗incidental‘ and
‗intentional‘ labels used here refer exclusively to the learning conditions imposed by the
instructor, not to the strategies used by learners, nor the representations resulting from
generated by game and feedback, properly perceiving and using those cues, and
improving the declarative knowledge structures resulting from learning (McPherson &
Kernodle, 2002, 2007). In most cases, this type of training has been provided via video
feedback of the own player‘s performance (García-González, Moreno, Moreno, & Del
Villar, 2012; Koeppen & Raab, 2012; Raab, Gula, & Gigerenzer, 2012). However, the
126
THE HIDDEN COST OF COACHING
current results seem to imply that if an instructor fails to identify some crucial cues,
instruction might end up preventing people from considering and learning about these
unidentified cues. Actually, in the last few years, some authors belonging to the
cognitive tradition have started to become aware of the potential setbacks of basing
tactical training exclusively on guidance from the coach. For example, in García-
González, et al., (2012), the declared aim of the supervisor ―was to guide the analysis of
the game situation through open questioning, but not intervening directly or giving
answers to the proposed questions‖ (p. 3). Still, in what they call ―analysis of the
decision context‖—which was part of the interaction between instructor and learner—
placement, ball placement and direction, and shot executed, i.e., aspects selected
―according to the tactical elements that characterise tennis shots‖ (p. 3). In other words,
The ecological dynamics approach (Araújo, et al., 2006) is undoubtedly more aware
prescription is regarded as a limitation for that kind of selective learning (Araújo &
individuals. However, our task does not allow direct assessment of its predictions, as
most specificities of the game, including motor demands, are absent in simulations (see
Finally, the current approach provides a methodological tool that can be easily
127
THE HIDDEN COST OF COACHING
scores can be used as learning indices, so that training can be monitored, not only by
means of outcome-related data (e.g., individual statistics, coaches‘ judgments), but also
computed from behavior, without the need to collect learners‘ or coaches‘ estimates. In
addition, experts and novices could be compared with regard to the way they make use
of the cues present in the environment. Presumably, utilization scores will dynamically
change as expertise arises, and will tend to correlate highly with other indices of
players‘ and teams‘ performance. Actually, the next objective is to demonstrate that
utilization scores, namely, the extent to which players incorporate certain cues in their
decisions, can significantly predict the final result of the game. Provisional data do
show that shooters whose shot decisions are influenced by the presence of Rebound
have a larger contribution to the final score than those who do not.
Limitations
There are several limitations that must be acknowledged. First, the set of cues
considered in this study is not exhaustive. To date, it is not possible to be sure that such
cues really determine shooting adequateness in a real game. At present, the only place
where such ideas are discussed are in specific coaching technical work written by
practitioners, where they describe what must be done to coach players about these
factors. Still, there is no reason to think that the dynamics of learning for the present
cues are essentially different than the ones for the valid cues in real training. Match
analysis techniques and mixed methodology (Camerino, Castañer, & Anguera, 2012)
advantage.
128
THE HIDDEN COST OF COACHING
deciding about shooting adequateness. In the real game, such decisions are made both
by the shooter (from an insider‘s perspective), and by external observers (e.g., the
coach). It is important to stress that our task does not simulate the kind of decision-
making process executed by the shooter at the moment of shooting. The task is more
similar to the way players and coaches learn to decide if a shot is adequate by
skills are similar to those in operation in shooters (who learn from the feedback of their
own shooting decisions) is still open, and remains as a potentially fruitful line of
research.
The main strengths of this work arise from the novelty of its methodology, the
demonstration that people are able to use complex, non-obvious cues if systematic
between incidental and intentional cues can influence the utilization of ones and the
others.
REFERENCES
Adkins, C., Bain, S., Dreyer, E., & Starkey, R. A. (2007) Basketball drills, plays and
Araújo, D., & Davids, K. (2009) Ecological approaches to cognition and action in sport
and exercise: Ask not only what you do, but where you do it. International
129
THE HIDDEN COST OF COACHING
Araújo, D., Davids, K., & Hristovski, R. (2006) The ecological dynamics of decision
10.1016/j.psychsport.2006.07.002
Ashby, F. G., & Maddox, W. T. (2005) Human category learning. Annual review of
Ashby, F. G., & Maddox, W. T. (2011) Human category learning 2.0. Annals of the
Bar-Eli, M., Plessner, H., & Raab, M. (2011) Judgement, decision-making and success
Beilock, S. L., Bertenthal, B. I., McCoy, A. M., & Carr, T. H. (2004) Haste does not
always make waste: Expertise, direction of attention, and speed versus accuracy
Best, C. A., Yim, H., & Sloutsky, V. M. (2013) The cost of selective attention in
Blair, M. R., Watson, M. R., & Meier, K. M. (2009) Errors, efficiency, and the interplay
10.1016/j.cognition.2009.04.008
Di Stasi, L. L., Contreras, D., Canas, J. J., Candido, A., Maldonado, A., & Catena, A.
10.1016/j.ssci.2010.07.006
Exadaktylos, F., Espin, A. M., & Branas-Garza, P. (2013) Experimental subjects are not
130
THE HIDDEN COST OF COACHING
García-González, L., Araújo, D., Carvalho, J., & Iglesias, D. (2011) An overview of
García-González, L., Moreno, A., Moreno, M. P., Iglesias, D., & Del Villar, F. (2012)
10.2466/30.10.25.pms.115.5.567-580
Krause, J. V., Meyer, D., & Meyer, J. (2008) Basketball skills and drills (3rd ed.).
Kim, S., & Rehder, B. (2011) How prior knowledge affects selective attention during
doi: 10.3758/s13421-010-0050-3
Koedijker, J. M., Oudejans, R. R. D., & Beek, P. J. (2007) Explicit rules and direction
Koeppen, J., & Raab, M. (2012) The Hot and Cold Hand in Volleyball: Individual
131
THE HIDDEN COST OF COACHING
Llorca-Miralles, J., Sánchez-Delgado, G., Piñar, M. I., Cárdenas, D., & Perales, J. C.
Maldonado, A., Jiménez, G., Herrera, A., Perales, J. C., & Catena, A. (2006)
McNevin, N. H., Shea, C. H., & Wulf, G. (2003) Increasing the distance of an external
McPherson, S. L., & Kernodle, M. (2007) Mapping two new points on the tennis
doi: 10.1080/02640410600908035
Oskarsson, A. T., Van Boven, L., McClelland, G. H., & Hastie, R. (2009) What's Next?
doi: 10.1037/a0014821
132
THE HIDDEN COST OF COACHING
Pearce, J. M., & Hall, G. (1980) A model for Pavlovian learning: variations in the
Perales, J. C., Cárdenas, D., Pinar, M. I., Sánchez, G., & Courel, J. (2011) Differential
20(2), 729-745.
Raab, M., Gula, B., & Gigerenzer, G. (2012) The Hot Hand Exists in Volleyball and Is
Raab, M., & Johnson, J. G. (2007) Expertise-based differences in search and option-
Roelfsema, P. R., van Ooyen, A., & Watanabe, T. (2010) Perceptual learning rules
doi: 10.1016/j.tics.2009.11.005
performance outcome and movement patterns for novices and experts in table
10.7352/ijsp2012.43.053
Sivils, K. (2009) Game Strategies and Tactics for Basketball: Bench Coaching for
133
THE HIDDEN COST OF COACHING
Travassos, B., Araújo, D., Davids, K., O'Hara, K., Leitao, J., & Cortinhas, A. (2013)
Travassos, B., Davids, K., Araújo, D., & Esteves, P. T. (2013) Performance analysis in
Vadillo, M. A., Bárcena, R., & Matute, H. (2006) The internet as a research tool in the
Wooden, J., & Nater, S. (2006) John Wooden's UCLA Offense. Champaign, IL: Human
Kinetics.
Wulf, G., Gartner, M., McConnel, N., & Schwarz, A. (2002) Enhancing the learning of
171-182.
Wulf, G., & Prinz, W. (2001) Directing attention to movement effects enhances
134
THE HIDDEN COST OF COACHING
APPENDIX
Instructions for the intentional group (translated from Spanish; instructions for the
incidental group were identical except for the absence of the definitions of the cues and
the rules).
In this task you must learn when shooting is adequate during the game in basketball.
In each trial you will be presented with a 5-second video-frame displaying a 5x5
situation. When the image disappears, a question mark will indicate that you have to
decide whether shooting is adequate or not. The keys to respond yes/no are Z/M [M/Z].
You must use the 5 sec. to mentally reach a decision, but you will not be allowed to
make it during that time. After that interval, you will be given only ¾ of a second to
Once you make a decision, you will be immediately told whether your decision was
right/wrong. Please, try to gradually improve your performance, but also to reduce the
If you have opposition from a defendant, i.e., if a player from the other team can
If your team is well-set for rebound, i.e., whether the players from your team are
If your team is well-set for defensive balance, i.e., whether one or more players
from your team could set the defense against a potential fast break, in case the
shooter misses and your team does not catch the rebound.
135
THE HIDDEN COST OF COACHING
If there is little opposition, take the other factors into account. Rebound, balance
shooting.
136
ARTÍCULO 3:
Suárez-Cadenas, E., Courel, J., Cárdenas, D., & Perales, J.C. (in press).
Impact Factor: 0.584; Multidis. Psychology’s Rank: 93/129; Quartile: Q3; JIF Percentile: 28.295
TOWARDS A DECISION QUALITY MODEL FOR SHOT SELECTION
IN BASKETBALL: AN EXPLORATORY STUDY
1
Department of Physical Education and Sport, Faculty of Sport Sciences, University of
Granada
2
Department of Experimental Psychology; Mind, Brain and Behaviour Research Center
(CIMCYC), Universidad de Granada
Summary.—We take the first steps towards a shot selection quality model in basketball
that incorporates decisional cues that might be predictive, not only of proximal results
(e.g., scoring), but also of distal results (e.g., winning/losing the match). 2976 jump-
shots from 50 Euroleague matches were sampled, following systematic observation
guidelines. The decisional cues under scrutiny were shooting opposition, distance and
lateral angle, disposition to offensive rebound and disposition to defensive balance at
the moment of shooting. A first set of regressions between decisional cues and proximal
results showed higher opposition and distance to decrease the probability of scoring
(OR= .81; p< .001 and OR= .89; p= .013); a better disposition towards rebound to
increase the chances of catching rebound (OR= 1.57; p< .001); and better defensive
balance disposition to decrease the probability of a fast break (OR= 1.27; p< .036). A
second set of regressions between proximal and distal results showed shooting and
offensive rebound effectiveness to predict total points scored (β= .62; p< .001 and β=
.32; p< .001) and game result (winning/losing the game; OR= 1.12; p< .001 and OR=
1.05; p= .021). Finally, an analysis of the impact of decisional cues on distal results
showed a positive relationship between likelihood of winning and average team‘s
disposition to offensive rebound (OR= 1.18; p= 018). These results cast light on the
actual weights (validities) of the different cues involved in predicting outcomes of
shooting decisions. This evidence could help coaches provide objective feedback about
players‘ shooting performance beyond hit percentages.
1
Address correspondence to Ernesto Suárez Cadenas, Faculty of Sport Sciences,
Carretera de Alfacar s/n 18071 Granada, Spain or e-mail ([email protected]).
Any aspect of the work covered in this manuscript has been conducted with the ethical
approval of all relevant bodies (Boards of the Experimental Psychology and Physical
Education Departments, University of Granada).
139
TOWARDS A DECISION QUALITY MODEL
In the last years, sport sciences have experienced a growing interest in decision-
making processes and their relationship with competitive success (Bar-Eli, Plessner &
Raab, 2011). Researchers are working on extracting objective information from natural
competitive environments, in order to identify and modify trainable decisions that could
Sampaio, 2013).
sport. In the more restricted arena of offensive phases of the game, teams‘ tactics, as
well as individual players‘ decisions are oriented towards generating optimal shot
opportunities. Indeed, several studies have addressed the problem of shot selection from
a theoretical point of view (see Goldman & Rao, 2011, for analyses of allocative and
approach). Other attempts have focused on the hot hand phenomenon and its impact on
players‘ decisions (Csapo, Avugos, Raab & Bar-Eli, 2014), or have adopted a
relevance of the abovementioned studies, to date, none of them have addressed the
analysis of shot selection in a more practical manner. For example, it remains unclear
how specific pre-shot decisional cues impact on proximal and delayed outcomes of
shooting. That means that an objective method to determine the decisional quality of a
shot (how well that shot was selected) that considers its effect on both close and distant
In general, previous research has stressed the difficulty and complexity of modelling
the processes underlying shot adequateness judgments and shooting decisions (Skinner,
2012). Nevertheless, research about shot selection has taken into account, almost
140
TOWARDS A DECISION QUALITY MODEL
exclusively, direct shot outcomes (hit/miss), and have neglected more delayed outcomes
that could also contribute to competitive advantage. In the absence of this potentially
valuable research, and with a more applicative aim in mind, experts have tried to
describe the conditions that discriminate between good and bad shots in technical works
2006; Krause, Meyer, & Meyer, 2008). According to these, good shooting decisions
should increase the likelihood, not only of scoring, but also of offensive rebound (in
case of miss), and decrease the likelihood of suffering a subsequent fast break (in case
of no rebound).
notational analysis research has pointed out different performance indicators that
influence game result (e.g., controlling the defensive rebound increases the likelihood of
winning games, García, Ibáñez, Martinez, Leite, & Sampaio, 2014). Some other studies
Delgado, Cárdenas & Perales, 2015) have tried to operationalize decisional cues,
namely, environmental cues that could signal the consequences of shooting decisions:
(1) degree of presence of direct opposition by a defender, (2) shooting distance, (3)
shooting lateral angle, (4) team‘s disposition to offensive rebound, and (5) team‘s
These decisional cues are probabilistic, in the sense that they can portray a heavier or
lighter weight in predicting shot outcomes. For example, we do not know the validity
catching the imminent rebound. Namely, we do not know the parameters of the function
that links a specific quantitative index of team‘s disposition and the likelihood of
141
TOWARDS A DECISION QUALITY MODEL
regression index, after regressing the outcome under consideration (catching or not the
rebound) over its possible predictors (Hastie & Dawes, 2010). Decision-making
al., 2015).
To date, main limitations in the field of decision making in sports are attributable to
two reasons. First, to the prevalence of studies based on experts‘ judgements and their
notion of what is a good decision. For example, Bar-Eli et al., (2011) acknowledged the
ambiguity of what an expert and a correct decision are. And second, research on this
valid methodologies (Farrow & Raab, 2008). The tentative decisional quality model
proposed in this study extracts objective information about decisions quality directly
from the real game and consequently, helps partially surpassing the abovementioned
limitations.
shot selection, and expert coaches and players have been shown to be better than
Araújo, Carvalho, & Iglesias, 2011), there have been no attempts to estimate the
statistical validity of decisional cues to predict both proximal and distal consequences of
quality, understood in terms of the conditions that a shot situation should meet in order
to maximize positive outcomes and minimize negative ones (i.e., validities). Making
shooters (in order to inform better individual decisions), and to the coach and the other
players (in order to set optimal conditions for shooting). In addition, observed changes
142
TOWARDS A DECISION QUALITY MODEL
different levels of possession time pressure, across trials in a hot-hand streak, or under
different degrees of emotional or cognitive load) (see, for example, Csapo et al., 2014 or
Raab, 2002) could allow analyzing the impact of such circumstances on decision-
making processes.
In this study, we take the first steps towards developing a shot decision quality-
assessment model. 2976 jump shots from 50 Euroleague matches were sampled,
al., 2011; Suárez-Cadenas et al., 2015), the decisional cues taken into consideration
Similarly, the proximal outcomes from shooting to be predicted were scoring (or not),
catching (or not) the following rebound, and preventing (or not) a fast break. Distal
outcomes were the amounts of points scored and received, and the final result of the
match (winning/losing). A first set of (logistic) regression analyses was run to explore
the impact of decisional cues on proximal outcomes. A second set of (logistic and
linear) regression analyses was performed to explore the relationships between proximal
and distal outcomes. Finally, regression analyses were carried out to check whether the
effect of decisional cues on proximal outcomes do transfer to distal outcomes. Given the
exploratory nature of this study, we hold no previous hypothesis about the map of
143
TOWARDS A DECISION QUALITY MODEL
METHOD
Sample
The sample consisted of 2976 shots corresponding to 50 matches from the 2012-13
and 2013-14 seasons of the Basketball Euroleague. Only jump shots performed outside
the 3-second-restricted area during offensive phases and with possibility of rebound, for
the two teams in each match, were considered. Shooting fouls, free throws, and off-time
Variables
These variables refer to the degrees of presence of each of the environmental cues
considered relevant to shot outcomes, registered exactly at the moment of shooting (on
all scores were expressed in discrete scales. The procedural details and rationale of how
Opposition was defined as the degree to which an opponent was close enough and
ready to interfere with the shot attempt, measured in a 1 (low) - 4 (very high) quasi-
quantitative scale. Shooting distance was computed as the linear distance from the
vertical feet projection on the floor to the vertical basket projection on the floor, and
then translated into a 1 (close) to 4 (very distant) scale. Lateral angle was first measured
as the angle formed by the end line and the imaginary line from the shooter to the basket
and then translated into a 1 (corner shot) - 4 (frontal shot) scale, regardless of court side
144
TOWARDS A DECISION QUALITY MODEL
offensive players inside the three second restricted area at the moment of shooting and
5 players in data collection). Team’s disposition to defensive balance was defined as the
number of defensive players situated behind the free throw line. Based on the number of
observations, this variable was clustered as 0/1, 2, 3 and 4/5 players (1 - 4 quasi
quantitative scale). Additionally, we also computed mean scores for each decisional cue,
averaging their values across shots for each team in each match. These mean scores
represent the general tendency in each team and each match to shoot with higher or
smaller degrees of opposition, distance, lateral angle, rebound and balance. In other
words, these mean values can be interpreted as match and team-specific shooting
Proximal outcomes
The in-game outcome of each shot was described by using three dichotomous
variables: scoring (yes/no); catching the subsequent offensive rebound (yes/no), in case
of miss, and preventing fast break (yes/no; a fast break was considered to be prevented
if the opponent team took longer than 7 seconds to make an advantage shot, with
defence not yet organised), in case of not catching the rebound. Additionally, we also
possibility of rebound for one team or the other), and defensive balance effectiveness
(proportion of fast breaks prevented over the total number of attacks from the opponent
team following rebound) scores. These effectiveness scores were computed for each
team and each match. Please note that the three outcomes are logically chained: only
rebounds subsequent to the shots included in the analysis, and only attacks resulting
from not catching those rebounds were taken into account for further analyses. This is
145
TOWARDS A DECISION QUALITY MODEL
obviously due to the fact that our interest is to inspect exclusively those variables
Distal outcomes
Match outcomes were defined by using two quantitative variables (points scored and
observation guidelines, performed by two expert technicians specifically trained for this
Variable recording training took place during a four-week long period (20 hours in
total). For all variables, agreement between judges, as measured by Cohen‘s Kappa was
.85 or higher, which qualifies as very strong (Altman, 1991). All data were collected
using the LINCE software, a program specifically designed to measure sport behaviour
Statistical analysis included binary logistic and linear regressions. For logistic
regressions, b-values (B) and Odds Ratio (OR) with their 95% confidence intervals (CI)
were estimated and Nagelkerke‘s R2 was used to assess goodness-of-fit of the models.
regressions, b-values (B) and standardized b-values (β) were estimated and R2 and
predictors were assessed by t tests. The significance threshold for all analyses was
p=.05. Statistical analyses were conducted in IBM SPSS Statistics, Version 20.0 (2011,
146
TOWARDS A DECISION QUALITY MODEL
Analyses1 were carried out in three separate stages (see Table 1). In Stage 1,
proximal results were regressed over decisional cues, according to three logistic binary
regression models.
In Model 1, scoring (yes/no) was regressed over opposition, distance and lateral
angle, across all the shot attempts under consideration (n=2976). In Model 2, performed
upon missed shots (n=1741), catching the rebound (yes/no) was regressed over team‘s
disposition to rebound. Finally, in Model 3, carried out upon all shots for which the
rebound was not caught (n=1059), defensive balance with fast break prevention (yes/no)
was regressed over team‘s disposition to defensive balance. In general, the inclusion of
connections. For example, in the first model, disposition to offensive rebound and
disposition to defensive balance were not included as there is not any evident way in
which they could be connected to scoring. In the same way, opposition, distance and
lateral angle were not included in models 2 and 3 as there are no evident ways in which
they could affect the likelihood of an offensive rebound or preventing a possible fast
break2. (Note, however, that in the last analysis all decisional cues were included in the
same model, as all of them can impact on distal outcomes, i.e., game result).
1
Given that data showed a multilevel structure (i.e., shots nested into teams and teams nested into
matches), intraclass correlations (ICC) for all variables were computed regarding the 50 matches
analysed, in order to determine an adequate analyses strategy (single or multilevel modelling). ICC ranges
varies from 0 to 1, indicating between-level variance. Following Dyer, Hanges and Hall (2005) with
values close to 0 (e.g., .05) multilevel models are difficult or impossible to estimate. In our study ICCs
ranged from <.01 to .02 showing that data variance was almost entirely associated to the individual/single
level. Thus, we decided to examine relationships at the single/individual level.cases.
2
Complementary analyses were performed, including all available decisional cues in Models 2 and 3, in
order to check for not-so-evident cue-proximal outcomes relationships. The decisional cues considered
for analyses were opposition, distance, lateral angle, disposition to rebound and disposition to balance for
alternative Model 2 and 3. It was not possible to formulate an alternative Model 1, as variables recorded
for scoring analyses did not included either disposition to rebound or disposition to balance. These
analyses are consistent with the original models. Fitting indices were Nagelkerke‘s R2= .03; χ2 (5)=
45.99; p< .001, for alternative Model 2, and Nagelkerke‘s R2= .44.; χ2 (5)= 696.44; p< .001, for
alternative Model 2. The predictors added in alternative models (but not present in original ones)
remained far from significance (Min. p= .116).
147
TOWARDS A DECISION QUALITY MODEL
analyses were carried out to linearly regress points scored over shooting and rebound
effectiveness (Model 4), and points received over defensive balance effectiveness
(Model 5). Please note that shooting and rebound effectiveness were computed over
shots and resulting rebound for each team and each match (n=100).
Similarly, a logistic binary regression analysis was carried out to regress the final
and defensive balance effectiveness (Model 6; n=50). In order to avoid the violation of
sample of 50 matches was selected (25 matches from winning teams and 25 losing
Opposition, distance, lateral angle, and disposition to rebound scores (averaged across
all shots under consideration for each team and each match) were used to linearly
predict points scored (Model 7, n=100). Similarly, team‘s disposition towards balance
was used to linearly predict points received (Model 8, n=100). Two final logistic binary
regression analyses were carried out to predict the final result of the game from all mean
decisional cue scores (Model 9, n=50). Note that, in this stage, decisional cue values
were averaged across all the shots under consideration for each match and team. In this
as suggesting that teams who tend to select shots against larger degrees of defenders‘
opposition tend to make less points. In other words, averaged decisional cue values can
be interpreted as alternative team stats with potential links with game results (which is,
148
TOWARDS A DECISION QUALITY MODEL
TABLE 1
SUMMARY OF DECISIONAL CUES, PROXIMAL RESULTS AND DISTAL RESULTS, AND STAGES OF ANALYSES
Opposition
Points scored
Shooting effectiveness
Distance Game result
Lateral Angle
RESULTS
both shooting opposition (i.e., the more opposition, the lower probability to score) and
shooting distance. No significant relation was found for lateral angle. Model 2
offensive rebound and the number of players in the three-second restricted area at the
relationship between fast break avoiding and team‘s disposition to defensive balance.
points scored at the end of the match. A similar relationship was found between rebound
effectiveness and points scored. Model 5 (R2<.01) did not reveal a significant
149
TOWARDS A DECISION QUALITY MODEL
relation between defensive balance effectiveness and game result were found.
Results from stage 3 (model 9) are shown in Figure 4. Models 7 and 8 did not yield
significant relations and are not represented. Finally, Model 9 (accuracy=74%) showed
offensive rebound. There were no significant relations for the rest of the cues.
results from this complementary analyses are shown in Appendix 2. Those extra
analyses showed interesting results, specifically for missed shots. In that submodel, not
150
TOWARDS A DECISION QUALITY MODEL
FIG 1. Logistic regression models (1, 2 and 3) from analysis Stage 1. Depiction of relations between
151
TOWARDS A DECISION QUALITY MODEL
FIG 2. Linear regression models (4 and 5) from analysis Stage 2 (I). Depiction of relations between
152
TOWARDS A DECISION QUALITY MODEL
FIG 3. Logistic regression model (6) from analysis Stage 2 (II). Depiction of relations between proximal
153
TOWARDS A DECISION QUALITY MODEL
FIG 4. Logistic regression model (9) from analysis Stage 3. Depiction of relations between averaged
DISCUSSION
This study was aimed at setting the way to develop a quality model of shooting based
on the relationship between decisional cues (contextual cues present at the moment of
The first two analysis stages showed that decisional cues were significantly linked to
proximal results, and these, in turn, were linked to more distal game results. More
specifically, opposition by a defender and shot distance negatively predicted jump shot
effectiveness, and this significantly predicted points scored and the odds to win the
match. In general, those results are in accordance with studies from match analysis and
performance indices analysis, confirming that losing teams take more shots with high
154
TOWARDS A DECISION QUALITY MODEL
opposition than winning teams, and shot efficacy increases with lower opposition and
closer distances to the basket (Álvarez, Ortega, Gómez & Salado, 2009; Lorenzo,
With regard to offensive rebound, the number of offensive players inside the three-
rebound. Offensive rebound effectiveness, in turn, predicted points scored and game
result. These two pieces of evidence concur with previous studies at showing that
rebound control increases the chances of winning (Csataljay, O‘Donoghue, Hughes, &
Dancs, 2009; Sampaio, Drinkwater, & Leite, 2010). The few studies that have studied
basketball performance analysis (Kubatko, Oliver, Pelton, & Rosenbaum, 2007; Oliver,
2004), and have shown the direct relationship between team‘s performance and
breaks avoidance effectiveness with either points received or game result. In apparent
contradiction with this result, fast breaks have been shown to be important determinants
explanation of this contradiction may rely on the fact that defense disposition at the
moment of shooting determines only a fraction of the total number of fast breaks. Elite
teams average low numbers of fast breaks (approximately 15% of attacks), and among
these, six out of ten do not follow immedately after catching a rebound, but after
155
TOWARDS A DECISION QUALITY MODEL
Importantly, a third analysis stage checked for the relationships between decisional
cues and distal results. Team‘s disposition to offensive rebound, computed as the
rebound score averaged across all shots under consideration for each team and each
match, directly predicts game result. To our knowledge, this is the first time a
relationship between a decisional cue (prior to shot) and a distal result (winning or
losing the game) is reported and quantified. This result is particularly interesting, as
and as a consequence more difficult to be learnt (see Perales et al., 2011; Suárez-
Cadenas et al., 2015). In other words, although disposition to rebound by the shooter‘s
team (at the moment of shooting) has a significant impact on the odds to win the match,
there is no empirical evidence so far that shooters do refrain from shooting when their
al., 2013; Suárez-Cadenas et al., 2015) show that strategy is probably trainable with
Before going any further, and in relation to our previous argument, it is important to
note that we have not considered so far the individual or psychological factors by means
of which decisional cues enter actual players‘ decisions (in case they do). In general, we
can reasonably assume strategic decisions made by the whole team are oriented towards
which they take advantage of the scenarios generated by the whole team. We suspect
that psychological features like impulsivity, resistance to pressure and interference, and
other personality traits can have a particular strong impact on determining whether
shooters incorporate this sort of non-obvious cue into their shooting decisions. The
156
TOWARDS A DECISION QUALITY MODEL
consequences of shooting (e.g., securing rebound) on the match result, and the link
between pre-decisional cues (e.g., team‘s disposition to rebound) on such feedback can
help coaches bring relevant feedback forward, making it more contiguous in time both
Still, future studies must also take individual differences into account. Despite the
fact we have shown that general cue validities can be computed for large samples and a
variety of teams and playing circumstances, they are also likely to differ between
individuals (Araújo, Davids, & Hristovski, 2006). Nevertheless, the same tracking and
the individual level, as far as the athlete is recorded for long enough to have sufficient
consequences resulting from her decisions, and her contribution to her team‘s
performance. In this sense, future studies could include different cues in the analyses
(e.g., an alternative better than shooting) and examining their impact on proximal and
distal results. Similarly, different models could be developed analyzing other shot types,
for example, those taken inside the three seconds restricted area, where different
decisional cues could emerge as relevant for proximal and distal feedback.
157
TOWARDS A DECISION QUALITY MODEL
REFERENCES
Altman, D.G. (1991). Practical Statistics for Medical Research. London: Chapman &
Hall.
Álvarez, A., Ortega, E., Gómez, M. Á., & Salado, J. (2009). Study of the defensive
Araujo, D., Davids, K., & Hristovski, R. (2006). The ecological dynamics of decision
http://dx.doi.org/10.1016/j.psychsport.2006.07.002
Bar-Eli, M., Plessner, H., & Raab, M. (2011). Judgement, decision-making and success
http://dx.doi.org/10.1002/9781119977032
http://dx.doi.org/10.1037/h0047470
Csataljay, G., O‘Donoghue, P., Hughes, M., & Dancs, H. (2009). Performance
Csapo, P., Avugos, S., Raab, M., & Bar-Eli, M. (2014). The effect of perceived
http://dx.doi.org/10.1080/17461391.2014.982205
158
TOWARDS A DECISION QUALITY MODEL
Dyer, N., Hanges, P., & Hall, R. (2005). Applying multilevel confirmatory factor
149-167. http://dx.doi.org/10.1016/j.leaqua.2004.09.009
Farrow, D., & Raab, M. (2008). A recipe for expert decision making. In D. Farrow, J.
Gabín, B., Camerino, O., Anguera, M. T., & Castañer, M. (2012). Lince: multiplatform
4694.
García-González, L., Araújo, D., Carvalho, J., & Iglesias, D. (2011). An overview of
García, J., Ibáñez, S. J., Gómez, M. A., & Sampaio, J. (2014). Basketball Game-related
Garefis, A., Tsitskaris, G., Mexas, K., & Kyriakou, D. (2007). Comparison of the
159
TOWARDS A DECISION QUALITY MODEL
Goldman, M., & Rao, J. M. (2011, March). Allocative and dynamic efficiency in nba
(pp. 4-5).
Hastie, R. & Dawes, R.M. (2010). Rational Choice in an Uncertain World: The
Heuzé, J. P., Raimbault, N., & Fontayne, P. (2006). Relationships between cohesion,
http://dx.doi.org/10.1080/02640410500127736
(Eds.), Intuition in judgment and decision making (pp. 91-105). Mahwah, NJ:
Jiménez-Sánchez, A. C., Buñuel, P. S. L., Ibáñez, S. J., & Lorenzo, A. (2012). The
perception female basketball players who play internationally have about their
Krause, J. V., Meyer, D., & Meyer, J. (2008). Basketball skills and drills (3rd ed.)
Kubatko, J., Oliver, D., Pelton, K., & Rosenbaum, D. T. (2007). A starting point for
http://dx.doi.org/10.2202/1559-0410.1070
160
TOWARDS A DECISION QUALITY MODEL
Lorenzo, A., Gómez, M. Á., Ortega, E., Ibáñez, S. J., & Sampaio, J. (2010). Game
related statistics which discriminate between winning and losing under-16 male
Llorca-Miralles, J., Sánchez-Delgado, G., Piñar, M. I., Cárdenas, D., & Perales, J. C.
http://dx.doi.org/10.1038/ncomms1580
Oliver, D. (2004). Basketball on paper: Rules and tools for performance analysis,
Perales, J. C., Cardenas, D., Pinar, M. I., Sanchez, G., & Courel, J. (2011). Differential
20(2), 729-745.
Ribas, R. L., Navarro, R. M., Tavares, F., & Gómez, M. A. (2011b). Analysis of
161
TOWARDS A DECISION QUALITY MODEL
Skinner, B. (2012). The Problem of Shot Selection in Basketball. PLoS ONE 7(1):
e30776. http://dx.doi.org/10.1371/journal.pone.0030776
Sampaio, J., Drinkwater, E.J. & Leite, N.M. (2010). Effects of season period, team
http://dx.doi.org/10.1080/17461391003699104
Suárez-Cadenas, E., Cárdenas, D., Sánchez-Delgado, G., & Perales, J. C. (2015). The
Wooden, J., & Nater, S. (2006). John Wooden's UCLA offense. Champaign, IL: Human
Kinetics.
162
TOWARDS A DECISION QUALITY MODEL
APPENDIX 1
These variables were registered exactly at the moment of shooting (frame previous to
Opposition. Defined as the degree to which an opponent was close enough and ready
to interfere with the shot attempt (1 – 4). Low opposition (1): the closest opponent from
the shooter was without jumping in a distance further than two steps at the moment of
throwing the ball without jumping and was not able to interfere with the shot (even with
his arm/s raised). Medium opposition (2): the closest opponent from the shooter tried to
interfere the shot through a jump action in a distance further than two steps or without
jumping in a distance between one and two steps (with his arm/s raised or not). High
opposition (3): the closest opponent from the shooter tried to interfere the shot through a
jump action in a distance between one and two steps far or without jumping in a
distance closer than one step (with his arm/s raised). In this case, the opponent clearly
interferes the shot although not modify the natural shooter‘s motion. Very high
opposition (4): the closest opponent from the shooter tried to interfere the shot through a
jump action or without it in a distance closer than one step with his arm/s raised. In this
case, the opponent interferes the shot and modifies the natural shooter‘s motion (e.g.,
Shooting distance. Computed as the linear distance from the vertical feet projection
on the floor to the vertical basket projection on the floor. Continuous distances were
Lateral angle. Measured as the angle formed by the end line and the imaginary line
from the shooter to the basket (1 - 4). (1) Corner shot: 0˚-20˚/161˚-180˚. (2) lateral shot:
163
TOWARDS A DECISION QUALITY MODEL
21˚-40˚/41˚-160˚, (3) semi lateral shot: 41˚-60˚/121˚-140˚, and (4) frontal shot: 61˚-
120˚.
inside the three second restricted area at the moment of shooting (0 – 3; there were no
situated behind the free throw line (on the bassis on the number of observations per
category, this variable was clustered into four values: 0/1, 2, 3, 4/5).
FIG 5. Shooting distance zones. 1 = close shots; 2 = intermediate shots; 3 = distant shots; 4 = very distant
shots.
164
TOWARDS A DECISION QUALITY MODEL
APPENDIX 2
separately for successful and unsuccessful shots. Six complementary regression models were
run trying to predict distal outcomes (game result, point scored and points received) from
decisional cues, separately for successful and unsuccessful shots. Unfortunately, for
successful shots, disposition to rebound and disposition to balance were not recorded and it
Results from those models are quite similar the original ones. For missed shots, model 10
(accuracy=72%), showed that, not only disposition to offensive rebound, but also disposition
to defensive balance, positively impact on game result. There were no significant relations
The fact that disposition to defensive balance predicted distal results only for missed shots
probably deserves further investigation. Please note that at the moment of shooting, shots are
not misses or hits yet. The fact that disposition to rebound in ―to-be-missed‖ shots impact
distant outcomes could be implying that the members of the team (apart from the current
shooter) can preview the result of the shot (a miss, in this case), and are sensitive to the cues
165
TOWARDS A DECISION QUALITY MODEL
FIG 6. Alternative logistic regression model (10) from analysis Stage 3. Depiction of relations between averaged
decisional cues and dichotomous distal outcomes (game result), for unsuccessful shots only
166
Discusión y Conclusiones
167
DISCUSIÓN Y CONCLUSIONES
En esta tesis doctoral hemos analizado la toma de decisión del lanzamiento a canasta
el Artículo 1). Para ello se han desarrollado herramientas que permiten evaluar el aprendizaje
de la toma de decisiones de forma fiable (Artículo 2) y se han obtenido criterios objetivos con
los que poder definir qué es una decisión correcta y así poder comparar la calidad de las
aportaciones podrían ser de utilidad a la hora de aplicar los distintos avances teóricos a la
En primer lugar, se realizó una revisión sistemática sobre la creencia de hot hand y sus
forma evidente que la creencia en rachas tiene un impacto altísimo en la toma de decisiones.
Queda por dilucidar los factores que determinan el signo de ese impacto (beneficioso o
perjudicial, desde el punto de vista competitivo), pues encontramos estudios con resultados
contradictorios. De hecho, tal como se ha discutido a lo largo de esta tesis, es en este punto
donde encontramos una de las mayores limitaciones (que además es generalizable a la mayor
criterios objetivos para examinar la calidad de las decisiones ha impedido comprobar si las
selección del lanzamiento a los resultados de juego, y aunque recientemente se han realizado
169
Discusión y Conclusiones
algunos intentos (e.g., Csapo, Avugos, Raab y Bar-Eli, 2015a) sigue existiendo un vacío en la
literatura, que hemos intentado abordar en este trabajo (a lo largo del Estudio 3).
través de una tarea experimental comprobamos las ventajas y desventajas de las instrucciones
momento de la decisión fue formulada de forma arbitraria por los experimentadores, esa
Este estudio mostró que, con independencia del tipo de instrucciones, los participantes
incorporaron todas las claves a sus decisiones salvo la de Balance defensivo. Además, las
(Oposición > Distancia > Alternativa > Rebote > Balance). La principal ventaja
Rebote, incorporada de forma temprana por el grupo que recibió ese tipo de instrucciones,
algo que no ocurrió en el grupo incidental. Sin embargo, la instrucción intencional también
mostró desventajas, pues bloqueó el uso de la clave Distancia (clave que se mantuvo
incidental para ambos grupos) a corto y largo plazo (después de más de 1080 ensayos). En
términos prácticos, esta evidencia podría implicar que si un entrenador falla a la hora de
identificar las claves decisivas durante el entrenamiento, sus instrucciones podrían impedir el
170
Discusión y Conclusiones
A pesar de que el uso diferencial de las claves (la falta de equipotencia antes comentada)
experimentador), se observó también que claves con idéntica validez predictiva no fueron
aprendidas en la misma medida por los participantes en sus decisiones. Ello demuestra que
una clave, su complejidad, o su mayor o menor conexión evidente con los resultados del tiro,
Cabe resaltar que en este estudio se proporciona una herramienta metodológica que puede
ser de utilidad para analizar el aprendizaje de la toma de decisiones. Nos referimos al cálculo
del grado de utilización de cada una de las claves decisionales, que es, en realidad, un
indicador directo del uso que un participante hace de una clave para tomar sus decisiones y
cómo ese uso cambia a lo largo del proceso de aprendizaje (o lo que es lo mismo, del impacto
de esa clave sobre sus decisiones). Para ello se llevaron a cabo análisis de regresión logística
de las decisiones tomadas por los participantes (Hastie y Dawes, 2011, pp. 47-49), utilizando
como predictores los valores estandarizados de cada una de las claves decisionales para cada
una de las situaciones sobre las que había que decidir. Dicho análisis arroja un parámetro de
Por tanto, este estudio muestra la posibilidad de caracterizar el tiro, ya no sólo desde el
punto de vista de sus resultados (e.g. encestar o no), sino también de las claves decisionales
que determinan ese tiro. Sin embargo, tanto la selección de claves como la valoración de su
expertos, estando sus juicios expuestos a parte de las limitaciones ya comentadas (Bar-Eli et
al., 2011).
171
Discusión y Conclusiones
Estudio 3 se dieron los primeros pasos hacia el desarrollo de un modelo para medir la calidad
defensivo) que predicen no sólo los resultados próximos a la acción del lanzamiento, como
por ejemplo anotar o no anotar, sino también resultados distales como ganar o perder un
partido.
Por un lado, las claves decisionales mostraron una relación significativa con los resultados
próximos, y éstos, a su vez, mostraron una relación significativa con los resultados distales.
probabilidad de coger el rebote ofensivo y, a su vez la eficacia del Rebote ofensivo predijo
(número de jugadores atacantes más cercano al aro propio que los rivales en el momento del
encontró relación entre la eficacia a la hora de evitar el contraataque y los puntos recibidos o
del partido. Este hecho resulta especialmente interesante, en primer lugar, porque hasta donde
sabemos, esta es la primera vez que se reporta y cuantifica una relación entre una clave
172
Discusión y Conclusiones
no lanzar a canasta, no es una clave muy saliente (i.e., con capacidad de captar la atención)
tal como se puede comprobar en el Artículo 2 de esta tesis (ver también Llorca-Miralles, et
al., 2013; Perales, et al., 2011). Y por último, conviene señalar que la disposición al rebote
ofensivo no es una decisión que dependa únicamente del lanzador (como podría ser lanzar
con más o menos oposición), sino que implica un conjunto de decisiones a nivel colectivo.
Con este tercer estudio cerramos un ciclo que comenzaba con la constatación de un
problema en la investigación sobre hot hand en el Artículo 1 de esta tesis. En éste se señalaba
que, hasta la fecha, no se ha podido comprobar si dos o más aciertos encadenados llevan a los
jugadores a hacer una mejor o peor selección del tiro; como mucho se había examinado si los
siguientes tiros son más o menos efectivos en términos de anotación. Sin embargo, el
Artículo 3 muestra claramente que el hecho de que un tiro esté bien o mal seleccionado
depende de factores que van mucho más allá de la efectividad anotadora. En este estudio,
observamos que, por ejemplo, en relación con el resultado final del partido, tirar cuando hay
jugadores atacantes al rebote puede ser incluso más importante que tirar con posición cómoda
Las implicaciones de los estudios desarrollados en esta tesis doctoral son múltiples. Por un
lado, conocer la validez objetiva de las claves decisionales permitirá, por ejemplo, analizar el
fenómeno hot hand desde un punto de vista decisional, comprobando si realmente los
jugadores lanzan de forma más arriesgada al sentirse en racha, y en tal caso, detectando qué
claves decisionales se ven afectadas por ese deterioro decisional. Además, permitirá dar un
paso más allá de los paneles de expertos y obtener un feedback objetivo, basado en el impacto
de las decisiones en el juego real. Por último, el hecho de poder, por un lado, monitorizar el
173
Discusión y Conclusiones
(mediante la herramienta metodológica del Artículo 2) y, por otro, comparar esas decisiones
con un criterio objetivo (modelo decisional del Artículo 3) permitirá detectar patrones en la
toma de decisión en función, por ejemplo, del momento de partido o del marcador parcial
(e.g., Marcelino, Mesquita, y Sampaio, 2011 o Sampaio, Lago, Casais y Leite, 2010). Más
importante aún, facilita una vía para conectar la psicología individual y colectiva con el
Estudios futuros
Los estudios desarrollados en esta tesis nos permitirán abordar problemas concretos que,
hasta la fecha, no nos había sido posible resolver. Nuestro próximo objetivo será comprobar
cómo afecta el acierto o error de varios lanzamientos consecutivos a la calidad decisional del
vincularlas al perfil de impulsividad de los jugadores. Así, podremos analizar si los jugadores
más impulsivos tienden a sufrir un mayor deterioro en la calidad de sus lanzamientos tras
varios aciertos o si, al contrario, los jugadores más reflexivos modifican más el uso de las
claves decisionales durante secuencias con mucho errores. De esta forma podremos
determinar cuáles serían los lanzamientos óptimos para cada jugador (en función de las
de la selección de lanzamiento en las circunstancias en las que ésta se vea más afectada.
De forma paralela a ese estudio, nos proponemos mejorar el modelo de calidad decisional
en varios puntos:
174
Discusión y Conclusiones
como podría ser la clave alternativa mejor que el lanzamiento (ya sea mediante un
pueden ser los que se producen dentro de la zona de tres segundos. Este tipo de
lanzamientos, al ser más cercanos al aro (donde suele haber mayor aglomeración de
jugadores) podrían verse afectados de forma distinta por las claves decisionales ya
analizadas.
decisional para generar una herramienta que pueda resultar útil a los profesionales del
de visión por ordenador capaces de medir las posiciones de los jugadores con una
2009). Si bien estos datos son menos específicos (ofrecen menos información
cuentan con la ventaja de que se adaptarían en tiempo y forma (los datos se generan
175
Discussion and Conclusions
DISCUSSION AND CONCLUSIONS
In this work we have analysed decision-making for shot selection from different
approaches, providing methodological novelties that help to solve the main limitations on the
topic (detected in Article 1). For this we have developed tools to assess the learning of
decision-making in a reliable manner (Article 2) and we have tried to obtain objective criteria
to define what a right decision is and so, to compare the quality of decisions through various
personal and environmental circumstances (Article 3). These contributions could be useful
First, a systematic review on the hot hand belief and its possible behavioural implications
was developed, confirming that people strongly believe in streaks in most sports, although it
is not stable or linear. From a behavioural point of view, the studies show that the influence
of the belief has a huge impact on decisions although we have yet to clarify the factors that
determine the effect of that impact (beneficial or detrimental from a competitive point of
view), since studies showed conflicting results. In fact, as has been discussed throughout this
thesis, it is in this point where we find one of the major limitations (which is also true of most
of the research into decision-making in sport): the lack of objective criteria for examining the
quality of decisions has prevented checking whether the behavioural consequences of hot
hand belief lead to better or worse decisions. To solve this problem, it is necessary to link
decisional cues associated with shot selection to game results. Although recently there have
been some isolated attempts (e.g., Csapo, et al., 2015b), the literature has not addressed this
Before this, in Article 2, we focused on examining the learning of decisional cues related
to shot selection through training. Using an experimental task we checked the advantages and
179
Discussion and Conclusions
This study showed that independently of the instruction, participants learn to use all cues
except Balance, in at least some circumstances. However, cues were not equipotent, that is,
(Opposition > Distance > Alternative > Rebound > Balance). Also, intentional
incidental group. However, a group effect was not found for Opposition, Alternative or
Balance. This pattern implies that cue properties determine whether learning is more or less
However, the intentional instruction also showed disadvantages as it blocked the use of
Distance (cues that remained incidental for both groups). The intentional group not only
utilized this cue less efficiently than the incidental group, but also failed to reach the same
utilization in the long term (after more than 1.080 trials). The current results seem to imply
that if an instructor fails to identify some crucial cues, instruction might end up preventing
Although the differential use of cues depended largely on its actual predictive validity
(programmed by the experimenter), it was also observed that cues with the same predictive
validity were not learned in the same measure by participants in their decisions. This shows
that the particular characteristics of the decisional cues determine how the type of instruction
affects their learning. In other words, properties such as perceptual salience of a cue,
complexity, or more or less obvious connection with the results of the shot, determine the
It is worth noting that this study provides a methodological tool that can be useful to
analyse the learning of decision-making in sport. We are referring to the cue utilization scores
that are interpretable as estimates of the extent to which each participant takes each cue into
account to make decisions (or the impact of each cue on decisions). For this purpose we
180
Discussion and Conclusions
conducted logistic regression analysis of the decisions taken by the participants (Hastie and
Dawes, 2011, pp. 47-49), using as predictors the standardized values of each cue for each
situations on which he had to decide. This analysis yields a parameter of use for each
Therefore, this study shows the possibility of characterizing the shooting, not only from
the point of view of its results (e.g., score or no score), but also from the decisional cues that
determine that shot. However, both the selection of cues and the assessment of its importance
in determining the adequacy of the shot was done by an expert panel and their decisions were
In future studies real-game versions of utilization scores can be used as learning indices,
so that training can be monitored, not only by means of outcome-related data (e.g., individual
statistics, coaches' judgments), but also by means of progressive tuning of decisions to the
statistics directly computed from behaviour, without the need to collect learners' or coaches'
estimates.
These difficulties show the need to check the ecological validity of the decisional cues
involved in the shot selection and, in fact, that was the main aim of the last study of this
thesis. Therefore, the next and final step was to link different decisional cues related to the
shot selection to the actual game results, and thus, verify its ecological validity. In Study 3 we
took the first steps toward developing a model for measuring the quality of the shot selection
in basketball. In this model we incorporate decisional cues from real games (Opposition,
Distance, Laterality, Offensive rebound and Defensive balance) that predict not only the
proximal outcomes to the action of shooting, such as scoring or not scoring, but also distal
181
Discussion and Conclusions
On the one hand, decisional cues were significantly linked to proximal results, and these,
in turn, were linked to more distal game results. More specifically, Opposition by a defender
and shot distance negatively predicted jump shot effectiveness, and this significantly
predicted points scored and the odds to win the match. Also, the number of offensive players
catching the rebound. Offensive rebound effectiveness, in turn, predicted points scored and
game result. In the same way, a team‘s disposition to defensive balance was positively related
between the effectiveness of avoiding fast breaks and either points received or game result.
The relationships between decisional cues and distal results showed that a team‘s
disposition to Offensive Rebound directly predicts game result. To our knowledge, this is the
first time a relationship between a decisional cue (prior to shot) and a distal result (winning or
losing the game) is reported and quantified. This result is particularly interesting, as
as a consequence more difficult to be learnt (see the Article 2 of this thesis or Llorca-
Miralles, et al., 2013; Perales, et al., 2011). Also, it is worth noting that the decision-making
involved in the Offensive Rebound depends not only on the shooter but also on the rest of
This third study closes a cycle that began with the finding of a research problem associated
with the hot hand in Article 1 of this thesis. It was noted that, to date, it was not possible to
verify if two or more consecutive scores led players to make a better or worse shot selection;
at best, it was only possible to examine whether the following shots are more or less
effective. However, Article 3 clearly shows the fact that the shot selection depends on factors
that go far beyond scoring effectiveness. In this study, we observed that, for example, in
relation to the outcome of the game, shooting when there are players in position to get the
182
Discussion and Conclusions
offensive rebound is more important than shooting from a comfortable position (with low
The studies developed in this thesis have multiple implications. On the one hand, knowing
the actual weights of different decisional cues will allow, for example, analyzing the hot hand
manner when they are feeling hotness, and thus, detecting which particular decisional cues
are affected by this decision-making impairment. Moreover, it will allow taking a step
beyond expert panels and get objective feedback based on the actual impact of decisions on
game results. Finally, the possibility of monitoring the learning or the dynamics of the
players‘ decision-making (methodological tool in article 2) on the one side, and the
article 3) on the other side, will help to detect shooting decision patterns depending, for
example, on the match status or the critical game period (e.g., Marcelino, Mesquita &
Sampaio, 2011 or Sampaio, Lago, Casais & Leite, 2010). More importantly, it provides a
way to connect individual and collective psychology with match analysis allowing us to link
Future Studies
The studies developed in this thesis allow us to address specific problems that, so far, we
had not been able to resolve. Our next aim will be to check how the success or failure of
several consecutive shots affects the quality of decision-making, examining the use of
relevant decisional cues (determined in Article 3) along shot sequences. Once we have an
understanding of the dynamics of the cues in ‗hit and miss‘ shot sequences, we will try to link
them to player‘s impulsivity profile. Thus, we can analyse whether more impulsive players
183
Discussion and Conclusions
tend to suffer further deterioration in the quality of their shots after several hits or whether, on
the contrary, the most thoughtful players modify the use of decisional cues during sequences
with a lot mistakes. In this way we can determine what would be the best shots for each
player (depending on utilization trends of decisional cues) and propose specific interventions
for improving the shot selection in the circumstances in which it is most affected.
In parallel with this study, we aim to improve the quality decisional model in several
ways:
1. To increase the ecological validity of the model. This will be done by including
different decisional cues that can affect the shot selection, as there could be a better
dribbling and getting closer to the basket). Also, we will include in the model analysis of
different shots such as shots within the three seconds restricted area. These types of shots,
closer to the ring (where there is usually greater concentration of players) could be affected
2. Enhance the applicability of the model. As discussed throughout this work, one of the
consequences caused by the limitations of research in decision-making in sport has been the
tool that can be useful to professionals in the sport from our decisional quality model. To
achieve this, we intend to replicate the model developed using data from optical tracking data
instead of data from systematic observation. This technology uses computer vision systems
capable of measuring the players‘ positions at a rate of 25 frames per second (Perše, Kristan,
Kovačič, Vučkovič, & Perš, 2009). Although these data are less specific (offer less
qualitative information) than those obtained by systematic observation, they are considered
totally reliable in research (see, for example, Sampaio et al., 2015) and they have the
184
Discussion and Conclusions
advantage of being taken in a time and form (data are generated in real time) relevant to the
185
Referencias/References
187
REFERENCIAS/REFERENCES
Araujo, D., & Davids, K. (2009). Ecological approaches to cognition and action in sport
and exercise: Ask not only what you do, but where you do it. International
Araujo, D., Davids, K., & Hristovski, R. (2006). The ecological dynamics of decision
10.1016/j.psychsport.2006.07.002
Araújo, D., Davids, K., Bennett, S. J., Button, C., & Chapman, G. (2004). Emergence of
sport skills under constraints. Skill acquisition in sport: Research, theory and
practice, 409.
Araujo, D., Davids, K. W., Chow, J. Y., Passos, P., & Raab, M. (2009). The
Ashby, F. G., & Maddox, W. T. (2011). Human category learning 2.0. In M. B. Miller
& A. Kingstone (Eds.), Year in Cognitive Neuroscience (Vol. 1224, pp. 147-
161).
Avugos, S., Koppen, J., Czienskowski, U., Raab, M., & Bar-Eli, M. (2013). The "hot
Bar-Eli, M., Avugos, S., & Raab, M. (2006). Twenty years of "hot hand" research:
Review and critique. Psychology of Sport and Exercise, 7(6), 525-553. doi:
10.1016/j.psychsport.2006.03.001
Bar-Eli, M., Plessner, H., & Raab, M. (2011) Judgement, decision-making and success
189
Referencias/References
Bar-Eli, M., & Raab, M. (2006). Judgment and decision making in sport and exercise:
Rediscovery and new visions. Psychology of Sport and Exercise, 7(6), 519-524.
doi: 10.1016/j.psychsport.2006.07.003
Blair, M. R., Watson, M. R., & Meier, K. M. (2009). Errors, efficiency, and the
doi: 10.1016/j.cognition.2009.04.008
Bocskocsky, A., J. Ezekowitz, & C. Stein (2014). The Hot Hand: A New Approach to
Csapo, P., Avugos, S., Raab, M., & Bar-Eli, M. (2015b). The effect of perceived
http://dx.doi.org/10.1080/17461391.2014.982205
Dickinson, J. (1977). Incidental motor learning. Journal of Motor Behavior, 9(2), 135-
138.
Edwards, W. (1954). The theory of decision making. Psychological Bulletin, 51, 380–
417.
190
Referencias/References
Lessons from Theory and Practice (pp. 137–154). New York, NY: Routledge.
Fernández-Ríos, L., & Buela-Casal, G. (2009). Standards for the preparation and
García-González, L., Araújo, D., Carvalho, J., & Iglesias, D. (2011). An overview of
Garcia-Gonzalez, L., Moreno, A., Gil, A., Moreno, M. P., & Del Villar, F. (2014).
Gigerenzer, G., Todd, P. M. and ABC Research Group. (1999). Simple Heuristics that
psychology, 31-41.
Gilovich, T., Vallone, R., & Tversky, A. (1985). The hot hand in basketball: On the
191
Referencias/References
Griffin, L. A., Mitchell, S. A. and Oslin, J. L. (1997). Teaching Sport Concepts and
Hastie, R. & Dawes, R.M. (2011). Rational Choice in an Uncertain World: The
Koehler, D. J., & Harvey, N. (Eds.). (2008). Blackwell handbook of judgment and
Llorca-Miralles, J., Sanchez-Delgado, G., Pinar, M. I., Cardenas, D., & Perales, J. C.
223-226.
Maddox, W. T., Ashby, F. G., Ing, A. D., & Pickering, A. D. (2004). Disrupting
10.3758/bf03195849
Marcelino, R., Mesquita, I., & Sampaio, J. (2011). Effects of quality of opposition and
192
Referencias/References
Memmert, D., & Roth, K. (2007). The effects of non-specific and specific concepts on
tactical creativity in team ball sports. Journal of Sports Sciences, 25(12), 1423-
Miller, J. B., & Sanjurjo, A. (2014). A cold shower for the hot hand fallacy. IGIER
http://ssrn.com/abstract=2450479. doi:10.2139/ssrn.2450479
Moreno Domínguez, A., Gil Arias, A., Arroyo, M., Perla, M., Villar Álvarez, F. D., &
0785-800.
Perales, J. C., Cardenas, D., Pinar, M. I., Sanchez, G., & Courel, J. (2011). Differential
20(2), 729-745.
Perše M., Kristan M., Kovačič S., Vučkovič G., & Perš J. (2009). A trajectory-based
406-433.
Raab, M., & Johnson, J. G. (2004). Individual differences of action orientation for risk
taking in sports. Research Quarterly for Exercise and Sport, 75(3), 326-336.
193
Referencias/References
Raab, M., & Johnson, J. G. (2007). Expertise-based differences in search and option-
Raab, M., & Johnson, J. G. (2008). Implicit Learning as a Means to Intuitive Decision
Ripoll, H. (ed.). (1991). Information processing and decision making in sport (Special
Sampaio, J., McGarry, T., Calleja-González, J., Sáiz, S. J., i del Alcázar, X. S., &
Sampaio, J., Lago, C., Casais, L., & Leite, N. (2010). Effects of starting score-line,
Shanks, D. R., & Darby, R. J. (1998). Feature- and rule-based generalization in human
Shanks, D. R., Rowland, L. A., & Ranger, M. S. (2005). Attentional load and implicit
194
Referencias/References
Turvey, M. T., & Shaw., R. E. (1995). Toward an ecological physics and a physical
doi: 10.1007/bf00122574
Votsis, E., Tzetzis, G., Hatzitaki, V., & Grouios, V. G. (2009). The effect of implicit
391.
195
Agradecimientos
Gracias a mis directores, José César Perales y David Cárdenas, por todo el trabajo y ayuda.
199