Assessing win strength in MLB win prediction models
Morgan Allen, * · 2025
Read Assessing win strength in MLB win prediction models by Morgan Allen, * (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-02815.txt.
Description
Morgan Allen and Paul Savala test whether baseball win forecasts also capture how strongly a team will win. Earlier machine learning studies predicted the winner of Major League Baseball games. Their 2025 paper trains a broad set of such models on one shared data set, then checks the link between forecast probability and the final score gap.
The common models did show the expected link: higher win probability went with bigger wins. The pair then asked whether the forecasts could guide run-line betting, a wager on the margin. With careful strategies, the forecasts yielded positive returns. Used naively, though, the same models lost money fast. The study is a caution on turning decent predictions into bets without a plan for risk.
Catalog record
| CREATOR | Morgan Allen, * |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-02815