Despite growing interest in quantifying and modeling of the scoring dynamics within professional sports, relatively little is know about what patterns or principles, if any, cut across different sports. In this talk, we introduce a novel generative model of scoring dynamics within a game, and apply it to comprehensive data on 1.2 million scoring events across 10 seasons of four professional and college team sports and 1 billion scoring events across 10 million competitions from the popular online game Halo Reach.
Across these sports, we find a set of common patterns in scoring tempo (when scoring events occur) and scoring balance (how often a team wins an event). We then show that a model based on these patterns effectively reproduces the observed patterns in all five systems and makes accurate real-time forecasts of game outcomes. These results highlight the central role that luck plays throughout a game, and point to simple underlying principles governing the dynamics of competition. We close with a few brief comments on what these results may say about more complex types of social competition, including economic and political conflict.