This offseason will be a busy one for Danny Ainge. Not only will he need to pray that the lottery balls fall his way in May, that he ends up with a franchise player out of the June lottery, and that Theoâ€™s contract can be traded for a nice veteran, but Ainge will also have to worry about extending the contracts of some key Celtics players. Jefferson and West could have their contracts extended, but Ainge will probably take a wait and see approach to Allen and Telfair. Gomes will also be in the last year of his rookie contract, and the Celtics risk losing him the following summer if they donâ€™t extend him as well. The question is less about â€œwho to extendâ€ and more about â€œhow much is the player worthâ€. As much as we like the young players on the team, salary cap realities mean that the owners will not be able to keep all of them around.
Letâ€™s focus on the issue of â€œhow much is the player worthâ€. One logical approach to putting a value on a player is to compare the playerâ€™s performance to other players in the league and then benchmark his salary to those of similar players. One easy way to compare players is to look at statistics. The problem with using this approach is that players anticipate it, and so in their â€œcontract yearâ€ (the final year of their old contract) players will attempt to pad their stats so that they look good in comparison to other players. Conventional wisdom holds that players are more productive in contract years and less productive in non-contract years. Is this true? In what follows I seek to test this idea.
I gathered a list of NBA players who signed new contracts during the summer of 2006. This information came from Patricia Benderâ€™s website (http://www.eskimo.com/~pbender/index.html -- I thank fellow Celticsblog member petula for pointing me to the site). These players could either be free agents who signed a new contract (such as Ben Wallace) or players currently under contract who signed an extension (such as LeBron James). From this list I excluded players who had not played in the NBA for the last two years and players who signed minimum contracts (most of the players who signed for minimum contracts were rookies, so they would have been excluded anyway). The sample reduction left me with 48 players. I then compiled points per game (PPG) and rebounds per game (RPG) for each player for the years 2005-2006 and 2006-2007 (as of March 26, 2007). The averages are below (numbers are rounded):
Notice that the averages drop in both categories after the contract is signed. The differences are not statistically significant, but for the sake of argument, the numbers imply that, on average, a player plays about 10% â€œbetterâ€ in contract years.
This analysis is obviously a very rough first cut. A more detailed analysis would cover many years so as to incorporate any trend in a playerâ€™s productivity and to incorporate contracts signed over multiple time periods. One would also want to control for the playerâ€™s position, age, total contract amount, and any other number of observable characteristics. One would also be curious about other outcomes such as assists per game, steals per game and blocks per game.
What do these numbers mean for the Celtics? Let me put on my cynical hat for a moment and assume that Jefferson and West are fully aware that they could get extensions this summer and are playing better because this is a potential contract year for them. Jefferson is currently averaging 16 and 11; the analysis above implies that if he extends his contract over the summer he could average 14 and 10 next year. West is currently averaging 12 and 3; next year he could average 10 and 2.5. I do not actually believe this will happen, but the point is that there is some evidence that players improve productivity in contract years, and so we (and more importantly the Celtics management!) want to keep this in mind when thinking about how much to pay players.