Applications and Implementation: A Decision Theory Example in Football
C. T. L. Janssen and T. E. Daniel
Paul Morbitzer & Benjamin Taylor
ESI 6358 – Decision Analysis
Fall 2013
6 November 2013
ABSTRACT
In the 1967 Harvard-Cornell football game, Harvard was ahead 14-0 late in the game when
Cornell scored two touchdowns. On both occasions, Cornell tried and failed on the
two-point conversion attempt and lost the game 14-12. Postgame arguments were divided
on the merits of Cornell’s strategy. For this frequently occurring scenario in college
football, we derive a maximum expected utility decision rule for the decision of kicking
versus running/passing based on the relative utility of a win, a tie, or a loss and the
probabilities of success with a kick as opposed to a run/pass.
Subject Areas: Decision Analysis, Risk and Uncertainty, Statistical Decision Theory, and
Utility Theory.
Applications and Implementation: A Decision Theory Example in Football
In sports and in competitions, decisions are made continuously throughout the entirety of
the event. Players are making decisions every second on the field and coaches are making
decisions about plays or substitutions. Every decision contributes to the outcome of the
game and some have more impact than others. The common phrase, “A bird in the hand is
worth two in the bush” is tested in this journal article relating to a football game between
Cornell and Harvard, and making decisions to have the highest probability of gaining the
most utility at the outcome of the game.
The article cites a 1967 Harvard-Cornell football game where Cornell is losing 14-0 late in
