The home-field paradox: These go to eleven
Abstract
Two baseball teams plan a best-of-seven series, beginning with three games at the home field of one team and followed by up to four at the home field of the other, stopping when one team reaches four wins. The teams are evenly matched, but the home team wins with a fixed probability p > 1/2. Which team is expected to win more often? Which team is expected to win the series? Are these the same? Does it depend on p? Peter Winkler recently described a “home-field paradox” that arises in this problem. We describe this problem and its analysis, and explore a number of computational investigations into it, varying the series length, the value of p, and the sequence of home and away games. Can we design series schedules and p values that exacerbate the home-field paradox? Or ameliorate it? Or make it even more confounding? Can we push this problem to the utmost level, so that we can truly make the home-field paradox go to eleven?
© 2026 John M. Harris, Michael J. Mossinghoff, published by Ludus Association
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.