The three-point-for-win rule in soccer: are there incentives for match fixing?

This work by OLEKSANDR SHEPOTYLO is an example of an unusual application of game theory .

<blockquote>In the middle of the 90's the European soccer body UEFA urged National Soccer Federations to award three points for a win in a match instead of two points, as under previous regulations. Soon, the new system was universally adopted by all countries. The purpose of this change was to discourage playing for a tie, since ties receive only one point each, and to encourage a more attractive, attacking style of play. While there is some evidence of success, the effects are not as big as was probably expected. The potential danger of the new rule is that it penalizes "quality" tied games and encourages teams to collude in order to maximize the expected number of points per game. This problem is especially relevant if teams can strategically interact during long tournaments, potentially leading to corruption and point trading. There is also evidence that the change in rules had heterogeneous effects on top clubs and lesser clubs. While top clubs began attacking more, lesser clubs emphasized defense even after the change in rules. We develop a game theoretic model of strategic interactions between teams in a tournament that predicts that lesser clubs are more likely to collude if the probability of a draw is high or the expected punishment of being caught is low. We further test predictions of the model using a data set of soccer games in the USSR, Ukraine and Italy in 1980-2002. We apply a logit model that predicts the outcome of the current game as a function of the outcome of the previous game, controlling for the quality of each team. We find a statistically significant relationship between the outcomes of two games involving the same teams under the new rule in the Ukraine. However, we do not find such a relationship under the old rule or in the more competitive Italian league.</blockquote>

Read more at www.econ.umd.edu/~shepo...