Abstract
Kelly staking has been proven to maximize long-term bankroll growth of bettors with positive expected yield (profitable bettors). However, it demands for an estimation of the true probabilities for each event. Thus, many sport tipsters opt for simpler flat (unit-loss) or unit-win staking plans. We analyze under which assumptions these strategies correspond to the Kelly method and propose a different staking plan, unit-impact, under the hypothesis that it fits better with Kelly’s. We test our predictions using data of professional tipsters from the betting database pyckio.com. Results show empirical support for our hypothesis.
Get full access to this article
View all access options for this article.
