Fair Odds for Noisy Probabilities
We suggest that one individual holds multiple degrees of belief about an outcome, given the evidence. We then investigate the implications of such noisy probabilities for a buyer and a seller of binary options and find the odds agreed upon to ensure zero-expectation betting, differ from those consistent with the relative frequency of outcomes. More precisely, the buyer and the seller agree to odds that are higher (lower) than the reciprocal of their averaged unbiased probabilities when this average indicates the outcome is more (less) likely to occur than chance. The favorite-longshot bias thereby emerges to establish the foundation of an equitable market. As corollaries, our work suggests the old-established way of revealing someone's degree of belief through wagers may be more problematic than previously thought, and implies that betting markets cannot generally promise to support rational decisions.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
The Law of Total Odds
The law of total probability may be deployed in binary classification exercises to estimate the unconditional class probabilities if the class proportions in the training set are not representative of the population clas…
Binary ClassificationGeneral ClassificationEquivalence and Divergence of Bayesian Log-Odds and Dempster's Combination Rule for 2D Occupancy Grids
We introduce a pignistic-transform-based methodology for fair comparison of Bayesian log-odds and Dempster's combination rule in occupancy grid mapping, matching per-observation decision probabilities to isolate the fusi…
Paradoxes in Fair Machine Learning
Equalized odds is a statistical notion of fairness in machine learning that ensures that classification algorithms do not discriminate against protected groups. We extend equalized odds to the setting of cardinality-cons…
BIG-bench Machine LearningFairnessGeneral ClassificationFairBalance: How to Achieve Equalized Odds With Data Pre-processing
This research seeks to benefit the software engineering society by providing a simple yet effective pre-processing approach to achieve equalized odds fairness in machine learning software. Fairness issues have attracted …
BIG-bench Machine LearningFairnessEqualised Odds is not Equal Individual Odds: Post-processing for Group and Individual Fairness
Group fairness is achieved by equalising prediction distributions between protected sub-populations; individual fairness requires treating similar individuals alike. These two objectives, however, are incompatible when a…
Fairness