This is a review of the expected rewards in fantasy sports research conducted by Martin B. Haugh and Raghav Singal
Daily Fantasy Sports is an ever-growing industry with millions of users participating each year. DFS covers a wide variety of sports including football, basketball, baseball, soccer, and golf. Each competitor puts together a fantasy team of real world players within a designated league. Typically, there are restraints such as budget; a constraint that real world teams face as well. Additional constraints include positional constraints, as the user is restricted to a certain number of players chosen for each position. The users construct portfolios putting together a team they feel has the best expected outcome of winning. These users can be split into two groups – those who use statistics to form their portfolios and those who do not. The success of a user’s portfolio is determined by the success of the real players in their actual games.
Success is based on the amount of money the user is awarded which depends on the payoff structure. One type of payoff structure is the double-up payoff structure. In this configuration, the top number of players each receives an equal payoff. A second type of payoff structure is the top-heavy one. Within this framework, the amount of the cash payoff increases with the ranking of the portfolio. How can users optimize their decisions when putting a portfolio together? Optimal decisions will ultimately lead to a higher expected reward.
While a user’s success is dependent upon the success of the real world players, it is also dependent on the outcomes of their competitors. Final rankings are based on well users did compared to the hundreds or thousands of competitors who put together teams for the same league. Therefore, optimizing one’s portfolio includes not only choosing the best combination of players but also choosing the combination that will outplay the teams put together by other competitors.
The first step to answering this question is to determine a ranking system for portfolios. This ranking is accomplished by looking at the performance of a user’s portfolio against the performance of their competitors’ portfolios. The portfolios are ranked based on their point’s total and cash payoff.
In order to be more effective in putting teams together a user must have some idea of the team his opponents will put together. Analyzing the teams an opponent has put together on prior occasions will help predict teams they will put together in the future.
Research was conducted using DFS contests on FanDuel during the first 12 weeks of the 2017-18 NFL season. The expected rewards outcomes clearly indicate that portfolios that take into account their predicted competitors’ portfolios perform better than those that do not. How much better they perform of course depends on how accurate the predictions regarding the competitors’ portfolios actually were.
How mature is your team’s analytics program? Take the Sports Analytics Maturity Assessment.
Learn about the Groundbreaking Sports Analytics Model coaches and sports analysts are talking about!
Learn all about Sports Analytics here.