Research comparing Kelly Criterion-based strategies to flat betting highlights key differences in risk, return, and long-term performance. The Kelly Criterion is widely recognized for its ability to maximize long-term growth rates, significantly outperforming flat betting in theoretical models and controlled environments (Browne, n.d.; Dinis et al., 2020). However, its practical application carries risks, as short-term variance can lead to large fluctuations in capital, making it unsuitable for risk-averse bettors. In contrast, flat betting offers stability at the cost of slower capital growth, making it a more conservative approach.
Studies have demonstrated that Kelly-based strategies incorporating multibets tend to generate higher returns compared to single-game betting approaches, leveraging statistical advantages across multiple events (Grant & Buchen, 2013). However, one of the key limitations of the Kelly Criterion is its reliance on precise probability estimates. Parameter uncertainty can lead to suboptimal bet sizing, prompting the development of “shrunken” Kelly approaches that incorporate adjustments to improve out-of-sample performance and mitigate risk (Baker & McHale, 2013). These modifications help refine Kelly’s application in real-world betting scenarios where probability estimates are inherently uncertain.
Further research has explored the relationship between the Kelly Criterion and other risk management strategies. In certain financial contexts, it has been found to be nearly equivalent to Vince’s optimal f, a strategy focused on maximizing capital growth while managing drawdown risks (Wu et al., 2015). Additionally, researchers have examined extensions of Kelly-based strategies for non-mutually exclusive bets, which are particularly relevant in horse racing and other multi-outcome betting markets (Jacot & Mochkovitch, 2023). Overall, while the Kelly Criterion remains a powerful tool for optimal bet sizing, practical adaptations and risk-mitigation strategies can enhance its effectiveness across different betting environments (Haigh, 2000; Dinis et al., 2020).
Overall, while the Kelly Criterion remains a powerful tool for optimal bet sizing, practical adaptations and risk-mitigation strategies can enhance its effectiveness across different betting environments
Summary of: Haigh, 2000 and Dinis et al 2020
Anecdote
Have a story to share? Write to us at research@bettingresearch.org if you have a related, personal experience you would like to see placed here and share with the community.
Articles Cited
- “Mu-En Wu, Chia-Hung Wang, W. Chung, R. Tso, I-Hsuan Yang (2015): An Empirical Comparison between Kelly Criterion and Vince’s Optimal F, https://doi.org/10.1109/SmartCity.2015.166
- The paper compares the Kelly Criterion and Vince’s optimal f for determining optimal bet ratios in a momentum trading strategy on the Taiwan Weighted Index Futures, and finds that the two methods produce very similar results.”
- “Andrew R. Grant, P. Buchen (2013): A Comparison of Simultaneous Kelly Betting Strategies, https://doi.org/10.5750/JGBE.V6I2.579
- The paper compares the performance of different Kelly betting strategies, including those that use multibets versus those that only bet on single game outcomes, using a simulation model based on the Dirichlet distribution and empirical odds data from the 2007-08 English Premier League season, and finds that the Kelly strategy using multibets outperforms the single bet strategy.”
- “J. Haigh (2000): Focus on Sport The Kelly criterion and bet comparisons in spread betting, https://doi.org/10.1111/1467-9884.00251
- The paper provides a summary of the Kelly strategy for optimal betting, its application to spread betting, and methods for evaluating and comparing interlocking spread bets on sporting events.”
- “S. Browne (-): CAN YOU DO BETTER THAN KELLY IN THE SHORT RUN ?, –
- This paper analyzes the short-term performance of the Kelly criterion strategy compared to a dynamic, time-dependent strategy that optimizes the probability of achieving a specific target value or outperforming another strategy by a given deadline.”
- “L. Dinis, J. Unterberger, D. Lacoste (2020): Phase transitions in optimal betting strategies, https://doi.org/10.1209/0295-5075/131/60005
- The paper presents a study of optimal betting strategies that maximize the long-term growth rate of capital while keeping the risk (fluctuations) at a low level, and analyzes the trade-off between the average growth rate and the risk, finding an analog of a phase transition between two optimal strategies, one with risk and one without.”
- “L. Dinis, J. Unterberger, D. Lacoste (2020): Phase transitions in optimal strategies for gambling, https://doi.org/10.1209/0295-5075/131/60005
- The paper presents a study of optimal betting strategies in gambling models, finding that there is a phase transition between a null strategy and a mixed strategy that balances the trade-off between average growth rate and risk, and deriving general bounds on this trade-off similar to thermodynamic uncertainty relations.”
- “R. Baker, Ian G. McHale (2013): Optimal Betting Under Parameter Uncertainty: Improving the Kelly Criterion, https://doi.org/10.1287/deca.2013.0271
- The paper presents a method to improve the out-of-sample performance of the Kelly betting criterion by shrinking the size of the bet in the presence of parameter uncertainty.”
- “Benjamin P. Jacot, Paul V. Mochkovitch (2023): Kelly criterion and fractional Kelly strategy for non-mutually exclusive bets, https://doi.org/10.1515/jqas-2020-0122
- The paper examines how the Kelly criterion can be applied to non-mutually exclusive bets, which are common in horse racing where multiple types of bets are available for a single race, and presents a new formulation of the fractional Kelly strategy for this type of scenario.”
Insufficient Detail?
At times it is difficult to answer the question as there are not enough relevant published journal articles to relate. It could be that the topic is niche, there’s a significant edge (and researchers prefer not to publish), there is no edge or simply no one has thought to investigate.
