College football and the Vegas line: Deconstruction and arbitrage
The Journal of Prediction Markets · 2016
DOI: 10.5750/jpm.v10i1.1183 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Journal of Prediction Markets, 2016. Verfasst von McDonald Paul Mirabile und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).
Abstract (Originalsprache)
We examine the Vegas line in college football games by employing two separate regression models to deconstruct the Vegas line and actual margin of victory for 4,590 unique contests from the 2005 through 2011 seasons. A comparison of these two models suggests which factors represent a true relationship with the margin of victory and which reflect bettor biases. An additional model of the margin of victory illustrates which factors the Vegas line systematically misrepresents. The authors find a number of factors inadequately priced in the Vegas line that help explain variation in the actual margin of victory. Using a holdout dataset comprised of the 2012 and 2013 seasons we identify the magnitude of any mispricing and opportunities for arbitrage. We exploit this mispricing to develop and evaluate profitable betting strategies. A strategy betting on the top 35% mispriced games yields 55% correct picks and a 2.7% APY. A second strategy in which only the top 8% of mispriced games are bet yields 59% correct picks and a 5.9% APY.
Zitieren
APA
Mirabile, M. P., & Witte, M. D. (2016). College football and the Vegas line: Deconstruction and arbitrage. The Journal of Prediction Markets. https://doi.org/10.5750/jpm.v10i1.1183
Vancouver
Mirabile MP, Witte MD. College football and the Vegas line: Deconstruction and arbitrage. The Journal of Prediction Markets. 2016. doi:10.5750/jpm.v10i1.1183
BibTeX
@article{Mirabile2016,
author = {Mirabile, McDonald Paul and Witte, Mark David},
title = {College football and the Vegas line: Deconstruction and arbitrage},
journal = {The Journal of Prediction Markets},
year = {2016},
doi = {10.5750/jpm.v10i1.1183}
}
EUVIMED