Evaluating the Free Agent Class of 2002 & the Effect of Contract Years, by Don Leypoldt

             “Of course he is playing well. It’s his contract year!” Sports fans often bandy about phrases with similar sentiments. Conventional thinking says that an impending free agent plays better in the last year of his contract, in the hopes of making himself more attractive to bidding teams. But is that really the case? If so, logic says that there should be a fall off in his production the first year he signs a new contract, since the player over-performed the previous year. As MLB free agent bidding starts this month, will the new teams overpay for talent based on inflated past performances? Or is the notion of a “contract year” a myth?

             To test the validity of the hypothesis “Contract years inflate performance,” I compared the 2002 and 2003 performances of the 153 major leaguers who were in the free agent class of 2002. I began by collecting the 2002 and 2003 performance, team and contract statistics of every player eligible for free agency at the end of 2002 (all figures from USA Today). I then divided these players into three categories: hitters, starting pitchers and relievers. I next reduced the collective performance of all players in each category into a single, full-season stat line to provide a fair basis for comparison.

             I also analyzed two key subsets of the data:

             1.          Length of contract. Since stars tend to get long term contracts while average players tend to get one-year deals, I compared the performance change of players signing long term deals with those who signed one year contracts.

             2.          Type of contract: those who signed major league contracts vs. those who signed minor league deals.

             I found that while players in all three categories exhibited significant drops in appearances, the performance of hitters and starting pitchers remained remarkably constant from 2002 to 2003. Only relievers experienced a performance drop off.

             Specifically, free agent hitters had 10% fewer at-bats in 2003 compared to 2002. Starting pitchers threw 27% fewer innings; relievers pitched 29% fewer. But as the tables below show, when projected over a full season (500 AB for hitters, 180 IP for starters), their collective performance was strikingly similar:

WHIP = Walks and Hits per Inning Pitched

             While starters’ winning percentage increased significantly, one reason could be that the free agents joined better teams. Nine of the 31 free agent starters pitched for playoff teams in 2003, and many other free agents pitched for Seattle, St. Louis and other contenders. Since their collective ERA actually increased from 2002 to 2003, the better winning percentage is unlikely to be a function of the pitchers’ own performance. The ERA of free agent relievers jumped even more:

With the exception of games played, there is no evidence of a contract year for hitters. For pitchers, there is a slightly stronger case, although the performance changes are not statistically overwhelming.

Subsets:

             As a whole, there is little evidence that players over-perform in their contract year. But not all free agents are created equal. Big stars sign rich, multi-year deals. Regular players and benchwarmers tend to sign smaller one-year or minor league deals. If the hypothesis is correct, a star player is more likely to over-perform in his contract year than a regular player, because the stakes are higher. More bidders will find him attractive if he performs, and that will drive up his price.

             In fact, players who signed a one year or a minor league deal had a 3 point increase in their batting average and no change in their OPS (on base plus slugging) average from 2002 to 2003. Those who signed multi-year deals saw their batting average decrease by an average of 12 points and a decrease in their OPS of 26 points. Here’s the data, with the bottom pair of columns representing players who signed multi-year contracts:

             The same holds true for pitchers. Pitchers who signed a minor league deal or a one year deal had a tiny drop in WHIP and lowered their ERA by .22 points from 2002 to 2003. Those who signed multi-year deals saw their ERA rise by nearly an entire run and had their WHIP increase by .16. The winning percentage of pitchers who signed minor league or one year deals increased 84 points from 2002 to 2003, while those who signed multi-year deals saw their winning percentages fall 72 points. Statistics for a projected 180-inning season are shown below. As similar effects were noted among both starters and relievers, the tables combine the two groups for the sake of space. Once again, pitchers who signed minor-league contracts or one-year contracts are listed first, those who signed multi-year contracts follow. Although the pitchers who signed long-term contracts had been significantly superior in 2002, the performance of the two groups converged in 2003 to the point that it becomes difficult to tell them apart:

             Among both hitters and pitchers, those who signed multi-year contracts exhibited sharp drops in performance, those who signed one-year or minor-league contracts did not. This supports the hypothesis that a star player tends to over-perform in the last year of his contract. It will be interesting to see how Miguel Tejada, Bartolo Colon, Gary Sheffield and others play this out as they take the field with new contracts in 2004.

Copyright © 2003 Don Leypoldt. All rights reserved. Originally published in the Fall 2003 issue of Outside the Lines, the SABR Business of Baseball Committee newsletter.