A Player Selection Heuristic for a Sports League Draft

Sports leagues conduct new player entry drafts in which franchises select, in a pre-determined order, players to complement their existing rosters. We model the decision-making process of a single sports franchise during a player selection draft. The basic premise of our model is that a team selects a particular player based on a combination of the player's estimated value, the value of the other players currently available, and the team's need at each position. We first conceptualize a sports league draft using a stochastic dynamic program. However, this formulation is not directly solvable for practical-sized problems due to the overwhelming computational complexity. Therefore, we introduce additional assumptions and restrictions that result in a tractable deterministic dynamic program. We implement the model within a spreadsheet-based decision support system that allows the user to compute solutions under a variety of conditions. To benchmark our approach, we perform computational comparisons against several competing draft strategies in a series of simulated fantasy football drafts for the 2005 season. With perfect information regarding opposing teams' selections, our drafting strategy dominates these competing strategies. With imperfect information, there are draft instances in which our method is not guaranteed to dominate an alternate strategy; however, our drafting strategy outperforms the competing strategies on average and is more robust on the instances tested. Furthermore, we demonstrate that the decision-maker can incorporate information regarding the drafting behavior of opposing teams to improve the performance of our method.