Analysing human prediction in nondeterministic events : biases and implications in a betting environment
Files
ParraGabarron_28921800_2023.pdf
Closed access - Adobe PDF
- 2.29 MB
ParraGabarron_28921800_2023_Appendix3.zip
Closed access - Unknown
- 2.49 MB
Details
- Supervisors
- Faculty
- Degree label
- Abstract
- This master thesis investigates estimation quality in human probability judgements concerning nondeterministic events. The primary objective is to comprehend these estimations and examine whether generic misestimations occur within specific contexts. Statistical analysis tools as hypothesis tests are employed using odds data obtained from betting websites. By facing historical odds with the associated outcomes of events, a comparison is made between the expected and actual results, leading to conclusions regarding the quality of estimation. The findings reveal that within the context of betting, estimators exhibit a central tendency bias, which manifests as a tendency to assign less extreme probabilities than those reflected in reality. This knowledge enables individuals to either adopt a betting strategy for potential gains or refine estimation practices to avoid such errors. This paper presents a testing method to assess the quality of human estimation in nondeterministic events. The proposed method is a well-defined mathematical technique that offers valuable insights into the known biases prevalent in psychology. Finally, the findings contribute to enabling improved estimation practices and enhanced decision-making processes.