Heat Check: Holistic Performance and the Hot Hand Fallacy
Open Access
Author:
Diehl, Sean
Area of Honors:
Management Information Systems
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Robert Alexander Novack, Thesis Supervisor John C Spychalski, Thesis Honors Advisor
Keywords:
Hot Hand Fallacy Conditional Probability Sports Analytics Performance Psychology Basketball Analytics Flow State
Abstract:
This paper aims to reframe the discussion and analysis of the “Hot Hand Fallacy” by incorporating performance psychology into the background assumptions used for conditional analysis of a made field goal. Typical research into the “Hot Hand Fallacy” performs streak analysis and conditional probability analysis on one metric only: field goals. By incorporating the psychological concept of the flow state and broadening the evaluation of a player’s performance to encompass a more holistic depiction of their “hotness” or “coldness,” this paper performs conditional probability analysis on league-wide field goal percentages. The study found that players who outperform their expected statistics in points, rebounds, assists, steals, and blocks at a level higher than league average can be expected to have a higher field goal percentage on their next attempt. The study also found that through only the use of a holistic metric of past performance relative to expected performance, a binary classification model sees a significant increase in accuracy.
