Ecological Rationality in Expert Decisions: The Case of Penalty Kicks 12 Feb 2025 Cognitive Science Journal Club
Predominance of lab-based approaches in Cognitive Science Critical issues of laboratory research (1)- invariance assumption, context simplification (through control), complexity of cognitive processes The need for (and lack of) field investigations. Not merely applications in the field- the use of field phenomena to generate insights into cognitive processes such as decision-making Cognition in the Field (1) Kingstone, A., Smilek, D., & Eastwood, J. D. (2008). Cognitive ethology: A new approach for studying human cognition. British Journal of Psychology, 99(3), 317-340.
Study of naturalistic, dynamic decisions made 'online' (under time pressure) (2) Different decision agents (coaches, players, etc.), tasks (play-calling, ball allocation, etc.), and contexts (during play, during timeout, etc.) Deriving insights about cognition from complex yet rule-based real-world tasks. Making the Case for Sports Cognition (2) Johnson, J. G. (2006). Cognitive modeling of decision making in sports. Psychology of sport and exercise, 7(6), 631-652.
Clearly defined, familiar rules of the game High stakes decisions with personal consequences under time pressure Expert professionals making the decisions The question of rationality- do experts make 'optimal' decisions on the field? Penalty Shootouts- A Lab in the Field
Multiple biases defined in the literature, e.g. right oriented bias, action bias, sequential bias, etc. Our focus- Action Bias Choosing to act when inaction is more advantageous Predicated on norm theory (3)- deviations from normative behaviour have worse emotional consequences when outcomes are bad. Rationality and Bias in the Penalty Shootout (3) Kahneman, D., & Miller, D. T. (1986). Norm theory: Comparing reality to its alternatives. Psychological Review, 93(2), 136–153. https://doi.org/10.1037/0033-295x.93.2.136
The Question Do Goalkeepers jump more often than they should? Does staying in the middle allow them to save more shots ? Is there an ACTION BIAS ? When action is the norm, negative outcomes due to inaction are judged as worse than identical outcomes due to action (4). Thus, we choose action more often than inaction, even if inaction may lead to better outcomes. But does inaction lead to better outcomes in this case? (4) Ritov, I., & Baron, J. (1992). Status-quo and omission biases. Journal of Risk and Uncertainty, 5(1), 49-61. https://doi.org/10.1007/bf00208786
Bar-Eli et al. (5) Claim: Goalkeepers jump sideways (action bias) despite optimal strategy being staying center. Data: 311 penalty kicks in top leagues. Conclusion: Jumping is suboptimal but driven by social rationality. Previous Research (5) Bar‐Eli, M., Azar, O. H., Ritov, I., Keidar-Levin, Y., & Schein, G. (2007). Action bias among elite soccer goalkeepers: The case of penalty kicks. Journal of Economic Psychology, 28(5), 606–621. https://doi.org/10.1016/j.joep.2006.12.001
Previous Research Jump-direction→ Kick-direction ↓ Right Center Left Total Right 20.6% 2.4% 16.1% 39% Center 10.8% 3.5% 14.3% 29% Left 12.9% 0.3% 18.9% 32% Total 49.3% 6.3% 44.4% 100% Table 1: Distribution of Jumps and Kicks: Bar Eli et al. (2007)
Previous Research Jump-direction→ Kick-direction ↓ Right Center Left Total Right 25.4% 0.0% 0.0% 13.4% Center 3.2% 60.0% 9.8% 13.4% Left 0.0% 0.0% 29.6% 17.4% Total 12.6% 33.3% 14.2% 14.7% Table 2: Chances of Stopping a Penalty Kick: Bar Eli et al. (2007)
Data Transparency- Which Leagues? Which Competitions? Testing Validity of Claims- Does the conclusion hold for experts competing at the highest level? Recent debates- ecological and social rationality (6) The Gap (3) Gigerenzer, G. (2015). Simply rational: decision making in the real world. http://pubman.mpdl.mpg.de/pubman/item/escidoc:2182750
Dataset: 292 on-target penalty kicks from 35 FIFA Men’s World Cup shootouts (1982–2022). Sources: - Public Kaggle dataset (1982–2018). - Independently coded 2022 shootout data by two researchers (92.7% inter-coder agreement). The Present Study
Methods- - Key Variables: - Kicker direction (9 zones). - Goalkeeper jump direction (left/center/right). - Save success. - Analysis: Chi-square tests of independence, descriptive statistics. The Present Study
Jump-direction→ Kick-direction ↓ Right Center Left Total Right 27.4% 3.8% 15.8% 46.2% Center 9.6% 2.4% 7.9% 19.9% Left 12.0% 5.1% 16.8% 33.9% Total 49.0% 11.3% 39.7% 100% Table 3: Distribution of Jumps and Kicks: Present Study The Present Study
Jump-direction→ Kick-direction ↓ Right Center Left Total Right 41.2% 0.0% 0.0% 24.4% Center 17.9% 71.4% 26.1% 27.6% Left 0.0% 0.0% 44.9% 22.2% Total 26.6% 15.2% 24.1% 24.3% Table 4: Chances of Stopping a Penalty Kick: Present Study The Present Study
Insights from the Data Probability distribution of kicker actions (Right, Left, or Center kicks) differs from Bar-Eli et al. (2007) Similar save percentages across all decisions (jumping right, jumping left, staying in the center)- corroborated by chi-square test of independence [χ2 (2, 292) = 1.9041, p = .39] If action and inaction lead to similar outcomes, there are no marginal benefits of increasing the frequency of inaction. Thus, we found no evidence action bias in the present data.
Insights from the Data Is there evidence of ecological rationality? More kicks aimed at the sides than the center [χ2 (2, 292) = 16.078, p < .001] Jumping covers more zones (body coverage during motion). Jumping more aligns with the probability distribution of kicks and is ecologically rational. Goalkeepers optimize decisions under real-world constraints. No need for alternate utility functions; actions align with environmental cues (critique of social rationality in this context).
Limitations and Future Directions Save percentage for keeper staying in the center based on few center observations (33 kicks) Investigations across leagues and tournaments- in game and shootout penalty kicks. Contextual and cognitive variables in goalkeeper decisions (Fatigue, pressure, game state, tournament level, relative skill level- rankings as proxy, etc.) Experts vs Novices- do experts leverage additional/different cues for the decision?