Major league baseball division standings,sports journalists' predictions and player salaries

This paper analyzes the accuracy of predictions of division baseball winners and rankings for the 1982 to 1990 seasons. The forecasts in Sport, Sports Illustrated and the New York Times are more accurate than purely random forecasts except at predicting the National League winners. The journalists are better at predicting division winners, but not division rankings, than a naive model that assumes this season's rankings, will be the same as last seasons's. Combining the forecasts in the three publications does not result in greater forecasting accuracy. A model that uses the relative average salaries of teams significantly outperforms the journalists at predicting division rankings and, when allowance is made for ‘George Steinbrenner effect’, also at predicting division winners.