Payrolls don’t help winning?

Wes Helms thinks Larry Beinfest could do a whole lot more with the Yankees cash:

"“If he uses the same philosophy he does with the low payroll I think it would be pretty amazing,” Helms said. “It might be one of the best teams ever.”…“I’ve always said it’s not about the payroll,” Helms said. “It’s about the chemistry of the team. I’ve always been a big believer if you get a lot of superstars in one room it might cause more conflict than it does help the team.”"

No offense Wes, but these statements are a bit contradictory. If “it’s not about the payroll,” how would having a whole lot of payroll help Larry Beinfest make a better team? In addition, we’ve all heard too much about the mystical quality of “chemistry” to know that all of that stuff is essentially luck. Sure, if players feel good about themselves, maybe they’ll play better. But they’re just as likely to play better if they suddenly get hot, or if they were actually good to begin with, or because it’s July. The fluctuations are no different most likely.

Tom D’Angelo, the author who wrote the blog post, also did some cherry-picking of his numbers, stating that the Marlins and Yankees have won the same number of postseason series since 2004 and one more World Series since 2000. First, he should really just pick 2000 if he was planning on making the World Series argument; the way he did it made the cherry-picking more obvious.

Here’s a look at the Marlins and Yankees payrolls, records, run differentials, and dollars per marginal win from 2000-2008:

Florida Marlins

Record

Run Diff.

Payroll

Dollars/Marginal Win

2000

79-82

-66

$25,864,697

$834,345

2001

76-86

-2

$35,504,167

$1,268006

2002

79-83

-64

$41,979,917

$1,354191

2003

91-71

+59

$48,368,298

$1,124844

2004

83-79

+18

$42,118,042

$1,203373

2005

83-79

-15

$60,408,834

$1,725967

2006

78-84

-14

$14,998,500

$499,950

2007

71-91

-101

$30,507,000

$1,326391

2008

84-77

+3

$21,811,500

$605,875

New York Yankees

Record

Run Diff.

Payroll

Dollars/Marginal Win

2000

87-74

+57

$113,365,877

$2,906817

2001

95-65

+91

$109,791,893

$2,335998

2002

103-58

+200

$125,928,583

$2,289611

2003

101-61

+161

$149,710,995

$2,824736

2004

101-61

+89

$182,835,513

$3,449727

2005

95-67

+97

$208,306,817

$4,432060

2006

97-65

+163

$194,663,079

$3,972716

2007

94-68

+191

$189,639,045

$4,122588

2008

89-73

+62

$209,081,577

$5,099551

Pardon my table editing, I’m still not up to snuff with this system. Also, data provided by USA Today’s Salary Database and of course the immeasurably excellent Baseball-Reference. Marginal wins were determined by taking away the expected win total of a team of replacement-level players, as calculated here by Tom Tango

The tables are to be expected of course. They certainly show payroll having a definite effect on wins. They also show that the Yankees are overpaying compared to the Marlins for their wins. It’d be nice to see this sort of quickie exercise done for all teams over the last 10 years or so, as I’d imagine Billy Beane would look quite a bit better under this analysis. Still, can we say that Larry Beinfest, who began running operations for the Marlins in 2002, could run the Yankees; payroll as well as he did the Marlins’ payroll? Undoubtedly no. There’s only so much talent in the league and so many games to win that it would silly to think Beinfest could manage $1.12 million per marginal win every year with that kind of payroll. I’d say he’s probably doing a better job than Brian Cashman is for the Yankees, but considering the Yanks sizable coffers and the limited talent pool of players, Cashman is just working in a different spending environment than the Beinfest. Could he be focusing more on the farm system and overseas? Sure, and that may trim those high marginal dollars down a bit. But it would never approach the levels that low-market teams can and have to attain simply because the Yankees can’t flood the market with their money, as the talent pool isn’t large enough for that.

I also think this quickie survey uses the wrong statistic for measuring success. I have the run differential up there, and I think a better evaluation of economic efficiency would be to use dollars per marginal run scored, as run differential is a far better estimator of a team’s success. From there, the usual 10 runs/win conversion can be used to determine a more accurate win value. I’ll be working a little bit on this project and post some results here.