The Value of GM's

Measuring the value of a GM in baseball is pretty much impossible. There are simply too many factors to evaluate. How much scouting and sabermetric infrastructure is the ownership willing to provide? How good are the medical doctors at both evaluating players as potential injury risks, and keeping the team's players healthy? How much should we attribute the success and failure of a team to an individual GM as opposed to other members of his front office team? These questions, and many more, should be why you should read everything I'm about to say with a grain of salt.

Despite the significant difficulty in actually evaluating a GM's value, I thought it would be interesting to try to devise a basic metric for doing so. This started out as a project for taking a look at Kevin Towers as a GM over the course of his career, but has taken on more significance given the firing of Ned Colletti and the joining of hands between Wall Street Wonder Andrew Friedman and the Demon Nemesis Los Angeles Dodgers. I think most of us are agreed that Friedman joining the Dodgers is not good news. But just how bad is it? While the analysis below won't be able to completely answer the question, hopefully it can shed some light on Friedman's potential impact on the Dodgers moving forward.

Methodology

"You play to win the game" - Herm Edwards

That's really what it all boils down to. Good GM's win a lot of games. Bad GM's don't.

Except what we really care about when evaluating GM's is how many games they win above and beyond what they are expected to win. So how do we calculate that?

It seems to me pretty intuitive that money plays a huge factor in how many wins you are expected to have as a team. Money might not be able to buy you happiness, but it sure can buy you a whole lot of wins in baseball. So what I did was take USAToday's team payrolls from 1995 to 2014 and determine the mean and standard deviation for payroll each year. This is because baseball is a zero-sum game with a finite number of wins that teams fight over. The nominal amount of money spent by a team in any given year is not that relevant. What we want to know is the amount of money spent by a team relative to the league as a whole.

Then for any given GM, I would find out how many standard deviations away from the mean his team payroll was each year and run that through a regression formula designed by Benjamin Morris, which predicts how many wins a team should have based on its standard deviation away from mean payroll. Find the difference between actual wins that year, and voila, you now have a very basic metric with which to measure how a GM and his team underperformed or overperformed in any given year, relative to the whole league and normalizing for team payroll.

If you then take those individual seasons and aggregate them over the course of a career, you can find the average wins added (or detracted) a GM provides. As a very poignant example, here is Kevin Towers's career below:

As you can see, over the course of his career, teams with Kevin Towers at the helm have won 1.28 games above payroll expectation per year. That's certainly not great, but it's also not terrible (hey, there are certainly worse GM's out there). However, it should be noted that two of Towers's three best performing years came in the initial year he became GM of a team. To the extent that people do not believe any GM deserves full credit or blame for the first year or two they take control of an organization, Towers's career looks significantly worse if we discount 1996 and 2011.

Results of GM's Around the League

Here are the results for the rest of the GM's in baseball. I only included GM's with at least five years of experience running a team, as an arbitrary cutoff for when a sample size becomes large enough to measure a GM's true talent. I left off a few GM's because either I was lazy in manually finding the years they managed different teams (Sandy Alderson) or I just really don't care that much about them as individuals or their franchises (Doug Melvin and Brian Cashman). Finally, I added Byrnes, our other most recent GM, who has had a fairly unsuccessful career to date as well.

Overrated GM's

Dombrowski - He has a star by his name, because I only looked at his Detroit years. I was just too lazy to compile the data from his years with the Expos and Marlins, though he's been with Detroit long enough that I doubt the Expos and Marlins track records would change my opinion of him much. He's hurt quite a bit by his first two years with the Tigers between 2002-2003, though even taking out those years only bumps him up to 1.40 Wins Above Expectation, aka the Towers-Colletti zone.

Zduriencik - Thought it was rather poetic that two of the most overrated GM's in the past decade of baseball have spawned from the two opposite spectrums of the baseball world. Whereas Dombrowski is widely respected among old-school baseball folks as one of the best GM's in the business, Zduriencik was wildly hyped by Dave Cameron and many in the sabermetric crowd as being a savior of the Mariners franchise. Hasn't exactly quite turned out that way to say the least.

Color Me Confused

Epstein - It both pains me and surprises me to see my alum only manage 0.84 Wins Above Expectation over the course of his largely lauded career. With Epstein, it's really been a tale of two cities. While in Boston, Epstein generated 5.14 Wins Above Expectation, which would have placed him 4th overall on this list. Since joining the Cubs in Chicago however, Epstein has been worth an astounding -12.05 Wins Above Expectation per year. 2014 showed some promise, but I suspect Epstein's long-term legacy will be determined by what he can do in Chicago the next two-three years, and whether he can find the same success he did in Boston.

Sabean - Not sure whether it pains me more to see Epstein with a mediocre Wins Above Expectation or to see Sabean with a pretty decent Wins Above Expectation. People who have followed my comments on the Snakepit the past few years know that Sabean has oftentimes been the butt of my jokes regarding trades he's made in the past. But wins don't lie, and Sabean has been a very effective GM over the course of his career. He may not be the most analytically sound in terms of decision-making, but there's no denying the success the Giants have had in scouting and developing players. Just goes to show that there is no one way to win in baseball. Different strategies may lead to the same result, but ultimately, you just want to be better than everyone else at the strategy you choose to implement.

Why Do You Not Run The DBacks

Beane - Is there anything else left to say about this guy? He's basically a god-tier hero when it comes to baseball GM's and the Wins Above Expectations bear that out over the course of his career. And apparently he looks like Brad Pitt. What a beast.

Mozeliak - This is such a curious case, because Mozeliak was promoted from Assistant GM to GM after the Cardinals fired Jocketty, who as you can see from above, is a pretty good GM in his own right. It's odd enough that you would be willing to take the risk on firing a proven 3.75 Wins Above Expectation GM. It's even more odd when you hire an unknown who has been supporting the previous GM as an AGM for the past few years. And yet the Cardinals sucked out and somehow managed to improve from an above-average GM to an elite calibre GM. Maybe the Cardinals are just lucky. Maybe the ownership there has a really good eye for GM talent. Or maybe there's just something in the water in St. Louis. Whatever it is, I sure hope Tony La Russa has brought it over to the desert.

Friedman v. Colletti

In conclusion, I will leave the 'Pit with my thoughts on Friedman joining the Dodgers. The data does not look pretty. While with the Rays, Friedman has managed to produce a whopping 8.51 Wins Above Expectation. Meanwhile, the unterrifying Colletti has been just a mediocre 1.72 Wins Above Expectation over his career.

Given the Dodgers's current payroll relative to the league, Benjamin Morris's regression formula expects 95.5 wins per season from them. Do I expect the Friedman Dodgers to win 104 games every season? Of course not. As people have mentioned, the more money you have the more susceptible you are to moral hazard problems. In addition, when you already are expected to win a lot of games due to your high payroll, there is an increasing marginal difficulty with winning additional games.

However, it is not difficult for me to imagine at all a Friedman Dodgers team to be as efficient as the Epstein Red Sox were, or the Daniels Rangers are. "Only" being worth 4-5 extra Wins Above Expectation would still mean a Dodgers team averaging 100 wins per season.

The sale of the Dodgers to Guggenheim was a terrible blow to our yearly chances of making the playoffs. However, the silver lining I always told myself was that Colletti was a mediocre GM and that the Dodgers will certainly have down years where we would be able to compete. Unfortunately, my worst nightmare has arrived. Instead of Colletti, we now have a very capable GM guiding the Dodgers. Even in down years, it seems reasonable to think that the Dodgers will still win at least 88-90 games. And that is a terribly scary thought indeed.