Bang for the Buck

Doug Pappas, the late, great former chair of SABR's "Business of Baseball" committee, created a formula for determining the marginal cost of a win, in terms of payroll dollars spent, for a given MLB team. He started by determining the minimum payroll required to field a team of "replacement level" players (essentially, what it costs to fill a 40-man roster at league minimum) and estimated the number of games that such a team could expect to win (Pappas set this at 49 games). The marginal cost per win was then the ratio of payroll dollars above minimum to wins above minimum.

Using a minimum payroll of $15.2 million (literally just the league minimum of $380,000*40) and Pappas' benchmark of 49 wins, here is how the numbers break down for 2007.

As you can tell, I've highlighted the playoff teams and made some other modifications to the raw numbers. For one thing, I adjusted the Red Sox and Yankees numbers to reflect the Competitive Balance or "Luxury" tax, and just for grins I estimated the Yankees payroll minus Roger Clemens (assuming Clemens' starts would have been given to a "replacement level" player like Ian Kennedy; replacement level is in scare-quotes for obvious reasons).

A few things are readily apparent here. The Yankees spent an awful lot of money on 94 wins and a first-round playoff elimination. A fate made more embarrasing by the fact that the team that knocked the Yanks out shared the best record in baseball with the Red Sox at a third of what it cost John Henry.

Even though the Yankees look like the biggest wastrels in baseball, they at least got something for the money. Not so for the Orioles, White Sox and especially the Giants. On the bright side, Bonds' home run chase easily brought in enough additional revenue to offset his $15.5 million salary and 5.8 WARP1 performance, especially when you consider that Bonds only charged the Giants about $3 million for each of the wins he contributed, well under their regular marginal costs.

Things are even more interesting at the top of the chart. The Florida franchises once again justify their continued existence with efficiency instead of wins. The Marlins in particular demonstrate their excellent understanding of both the Wins Curve and the Success Cycle, and one begins to understand how a team could collect two World Championships in ten years despite an organizational ethos that kills a little part of the baseball's soul each year.

The efficiency of the Indians and Diamondbacks is just one of the many reasons I think each is the best run franchise in its league (more on that in a future post).

The biggest macro lesson to be learned is that the market for wins is still incredibly inefficient, though some teams are getting better at exploiting this inefficiency. Here's the scatter-dot regression of average wins versus payroll from 2002-2006, courtesy of Dan Fox at BP:



So over the five years preceding last season, payroll explained about 43% of a team's success. Compare that to the regression I did with this year's data:

Last year, only 21% of a team's fate was determined by its financial resources. Good time to be a Rays fan, huh?