Did
the 2009 New York Yankees have the most efficient payroll in baseball?
According
to some recent research we've done, they did indeed.
Just not in the conventional sense. If the above assertion struck you as ridiculous, it's because you think of payroll efficiency in the same way as most sabermetricians. Current studies that focus upon how well teams spend their money do so on a large scale: they almost always examine the relationship between how much the team spends and how much the team wins.
This is an intuitive and highly intelligent way of looking at the issue, but it's not the only way we can do it. Daniel Adler's recent research, the subject of a Wall Street Journal piece, showed that the relationship between salary and wins isn't as clear as we'd like it to be. The problem is that we're not moving out of this "Dollars per Win" framework. There are many ways of looking at how intelligently a team spends its money. Some new methodology can't hurt, right?
I decided to examine payroll efficiency by looking on a more microscopic level - individual players. The new statistic, calculated with an ingenious formula devised by Dan Yamins, measures payroll distribution on a given team. It holds total wins and total payroll constant from team to team, and instead simply asks, "are the best players on the team making the most money, and vice versa?" I had no idea before I conducted the study if distribution was connected to winning. After all, if a minimum-salary player performed very well, that certainly wouldn't be a bad thing, but would it be reflected that way in the study.
The
Yankees topped the 2009 rankings with an Inefficiency Rating (IR) of just .11,
meaning that their salary distribution contained only 11% of the error of their
worst possible salary distribution. Their archrival Boston Red Sox finished
right behind them at .16.
But
efficient distribution comes in all market sizes. The Florida Marlins and Minnesota Twins, owners of the lowest and 7th lowest payrolls in baseball, were next in the rankings.
So, does intelligent distribution lead to more wins? Here's where it get's interesting: among the 15
highest-spending teams, IR was not effective at predicting a given team’s
winning percentage. However, for the 15 lowest teams, IR was significant. For every percentage point decrease in the index,
those teams won another 1.1 games over the course of the season.
The main conclusion of our study was thus that while teams with large payroll can absorb
the hit of bad contracts, small-market teams have to be more discerning in whom
they sign.
There
are two ways to interpret that deduction. Perhaps it shows yet another way in
which teams that don’t spend as much are at a disadvantage. It’s also possible that the data shows that big spenders are
crucial – if there were no teams with large payrolls, then all bad contracts would be a death sentence, since no team would
have a buffer to absorb them through trades.
Which
is right? With debate brewing over a salary cap, the commissioner’s office will
soon have to decide.
The team-by-team results are below. For future analysis on this issue, be sure to bookmark our blog.
31%