No Theo Epstein? Also, where did Jim Hendry add up?

Shawn, if you re-ran Cashman's numbers just accounting for what's happened since he was given "total control" (i.e., 2006-2009), how does that change things?

awesome work.

I think that the problem with this system of evaluation as per Theo Epstein (and Cashman making the 'worst' list) is that it doesn't do a good job of taking into account the non-linear value of adding the marginal wins for teams that are already contending and just need to add a few wins to be great. This site above all others stresses the difference in player value relative to the team signing them, so there is a bias in this measurement against teams with high payrolls that can afford to overpay for a few extra wins, even if they aren't maximizing value. Is it better to win 80 games with a $50 million payroll or 100 games + a title with a $200 million dollar payroll? As mentioned in a recent article, Theo managed to get a lot of value out of his non auction-market players, which suggests that he would be one of the better GMs, but he was also able to get away with overpaying the corpse of Mike Lowell and paying Julio Lugo to play for another team, which kills him in this valuation but not in real life. Is there some way to adjust the metric to take this into account?

Can you add team names for the GM lists over the first two thirds? I know current GMs but not the ones from the start of the decade.

I think there should be some way to account for the fact that three of the top five (I'll be generous) organizations play in the same division. The Yankees and Red Sox and to a lesser extent the Rays are penalized for facing each other 19 times a year.

Does this account for say, a city with two teams? For example, in Chicago, you would be out of your mind to think the White Sox would generate the same amount of revenue as the Cubs if both teams win 100 games.

Steve Phillips just can't catch a break this offseason, can he?

I don't know how you would work it into the metrics, but it seems like a flaw that postseason success (or even appearances) gets zero consideration here. Also, while Phillips, Thrift and Bavasi were horrible, I have to think that any objective measure would say that Lamar and Littlefield are 1-2 in some order. They were disasters.

This is really interesting stuff.

Can you comment more about why this system is so kind to Sabean, who many Giants fans appear to believe is the worst GM on the planet?

As a long-time Mariners fan, I have to disagree with you a bit ElAngelo. Bavasi's suckitude goes well beyond any objective, quantifiable amount, and into metaphysical realms that mortals are not meant to understand, and seek to do so at their own peril.

Interesting stuff.

I'm not sure it's right to base how well the GM did using entirely third order wins with no weight on actual wins. For example, if Mark Shapiro insists on keeping Eric Wedge as his manager despite the fact that Wedge has knack for delivering fewer actual wins than expected, doesn't that eventually reflect on Shapiro too?

It does seem like the model was not work well with the Yankees, who are an extreme case. Given that they made the playoffs 9 out of 10 years in the decade, I'd be hesitant to call them the second least efficient team of the decade.

I can't figure out how this system ranks GMs at all. It's purely based on revenue vs production of the TEAM. There's no way that former Astros GM Tim Purpura would not be on a bottom-10 list of worst GM's of the decade. He was responsible for fun things like:
--trading for Jason Jennings
--signing Woody Williams
--having an opening day lineup featuring stalwarts Willy Taveras, Adam Everett, and Jason Lane,
--giving a full-time gig to Mike Lamb,
--paying Carlos Lee a contract that was probably 130% of what he was worth to the rest of the market,
--letting Carlos Beltran get away, and
--presiding over the complete collapse of the farm system that was third in 2002 to dead last by the time he was ousted.

But because the Astros made the World Series in 2005 riding on the arms of no-brainer-signees Roger Clemens and Andy Pettitte, along with a host of Gerry Hunsicker assembled pieces (Roy Oswalt, Lance Berkman, Craig Biggio, Jeff Bagwell, Brad Lidge), and remained competitive for the next couple of years with the same predecessor-built club, Purpura gets a pass.

For shame!

Why, for cumulative team rankings, are only 9 teams at 1.00 or below. Is this logical? Does this mean that most GMs are actually somewhat good at their jobs? Am I completely missing some part of the methodology?

I had to read your article twice to understand it, and, unfortunately, I think it paints an innacurate (or at best, incomplete) picture of rating General Managers.

For 1, it is not always the job of a General Manager to win the most games he can for the price paid. Some GM's are told to win at any cost. Sometimes that means going out and spending money and assembling a relatively expensive roster. Unless that team wins a huge number of games, he's not going to do well in the evaluation you listed. If that team does poorly, well, that means that he chose poorly as to where the money went.

On the other hand, some GM's are told to strip cost and attempt to rebuild the organization. Eschewing current wins for future promise. These teams will often have low payrolls, but also poor results. If the team gets lucky with their prospects (or they are just very good at it), their team might become successfull before those players get expensive. Also, your formula says nothing about trades, or farm systems, or the like. GM's are responsible for all of that.

I guess one thing that's bothering me about this, as interesting and entertaining as it is, is the use of third-order wins. I understand the desire to correct for bad luck and stiff competition, but GMs aren't paid to build teams that finish first in the third-order standings, they're paid to win.

Furthermore, it's quite interesting, at least to me as I chew on Cashman's spot here, that over the past decade the Yankees ranked third in the majors in third-order delta (D3), the amount by which they exceeded their third-order record. They led the majors in both third-order wins (932.5) and actual wins (965), but the margin of their lead over the #2 team, the Red Sox, increases from 10 (W3) to 45 (W0). Data is here in a Google Doc.

Granted that the margin between 93 and 96 wins in a given isn't huge, but looking back on Nate Silver's BBTN graph of the marginal win curve (see here, that appears to be the backside of the peak portion; beyond 96 things level off considerably. So I guess what I want to know is how the team rankings at the end look if you use real wins instead of theoretical ones.

Yeah this type of thing is fun to read but very simplistic in it's measurement.

Also, I think this piece should be entitled best and worse GM in terms of efficiency- or something of the like. To just use this as the criteria for best GM seems silly.

As a Mets fan, I completely agree with their #30 ranking. ANY Mets fan will tell you they have the most incompentent ownership as well as front office. Like I said before, they are run by idiots and chickenshits.

And one of the biggest lessons here is, if you spend big, you'd damn well better win big.

Lou Piniella is even more justified in asking where Steve Phillips got off criticizing his handling of Fukudome. Phillips has proven to be quite a rascal and a poor GM.

It saddens me to see Duquette's 2001 team there, but I understand it. That was a great team that never took the field together for a single-game. Nomar went down with a non-baseball injury before the season started, Pedro went down, Jason Varitek broke his elbow in a fluke play, Frank Castillo went down, Rich Garces went down, Manny Ramirez went down, Carl Everett went down - it was a disaster.

And you can talk about training staffs and contingency plans all you want - if the 27 year-old defending batting champ and three-time All Star goes down at the beginning of Spring Training with an injury suffered off-site and unrelated to baseball, there aren't many contingency plans that deal with that effectively.

It seems to me that your theoretical link between wins and revenue fails to account for what we might call "enthusiasm stickiness."

The Rays had lousy attendance and revenues in 2007 when they massively overachieved relative to budget. The Red Sox in 2006 and the Yankees in 2008 had excellent attendance (both teams having collapsed in September after being in the race all year) and therefore, high revenues, in spite of finishing out of the playoffs.

It seems to me that the marketplace cuts these GM's a bit more of a break in terms of the primary goal of any team- turning a profit- than does this metric.

Given all the negative feedback here, i'd like to respond with some gratitude: i thought this article was really sweet.

This comes up with some pretty counter-intuitive results which people don't dig that much, but once you remember to interpret the results as a metric instead of GM grade, these counter-intuitive results are fun to see. I'm a big fan of seeing an objective take on the efficiency of certain teams, and it really does help you take a second look at the Sabeans and Epsteins. And re: Cashman-- we all hear so much about the Yankee's financial hegemony, but articles like this do a great job of putting it into perspective. Understanding that Big George had a lot to do with this can then help clarify our mental impression of Cashman's work.

No this system doesn't present a perfect measure for GMs (actually, i think the biggest problem has rarely been touched on with all these complaints -- the fact that one of the GMs biggest jobs is to ensure future success by strengthening the farm system, which can be completely unrelated to big-league efficiency, especially for teams with high GM-turnover), but pray tell me this wasn't an engaging, enlightening article. Kudos Shawn.

It seems in these results that payroll (as a proxy for market size) is negatively correlated to ranking. This kind of makes sense - a team with a large payroll has a bigger downside than upside and vice-versa. In finance its referred to as convexity. Simply put, a team which spends less has less to lose!

What is included in a team's revenue?

If you look at the past decade, the Jays have been the class of the AL East but got so little out of it, partially because being efficient is only good if you can spend on the same rough scale as your opponents. If your opponents spend twice as much as you, then being 40% more efficient isn't enough to close the gap.

Jays decade PER 1.16
Rays decade PER 1.13
Red Sox decade PER 1.01
Orioles decade PER 0.92
Yankees decade PER 0.87

If you look at the teams who averaged at least 10% better than average you get 11 teams. 2 in the AL East and 9 elsewhere. The 2 in the AL East made the playoffs just 1 time (Rays in 2008). The other 9 teams made the playoffs 33 times and each team made it at least 2 times.

So that is 3.5 playoff trips per decade of 10%+ baseball if you play outside the AL East and 0.5 playoff trips per decade of 10%+ baseball if you play in the AL East. In terms of fairness, something is wrong with the current model and the Rays, Jays, and Orioles are suffering!