Maddon vs. Gardy: How much of an improvement would Maddon be?

[alexlegge]

Maddon vs. Gardy: Comparing Performance of Two Managers

 

With the news of Joe Maddon’s opt-out, Twins fans everywhere are clamoring for the front office to make him a serious offer. Maddon has earned a reputation as one of the greatest minds in the game, a savvy and sabermetrically-aware manager capable of getting a lot of production out of his players.

Just how much of an improvement would Maddon be over Gardy? It is widely thought that he will command a salary near Mike Scioscia’s $5 million. This would be a big step up from the $2 million Gardy will be earning for the last year of his contract, but certainly not a bank-breaker by MLB standards.

Since 2006, both Gardy and Maddon have been at the helm of several playoff teams with little postseason success. Factor in Gardy’s 2002-2005 seasons, and you generally have more of the same. Both managers have also received praise from baseball writers. Maddon was named American League Manager of the Year in both 2011 and 2008. Gardy won the award in 2010 and finished runner-up in 2006, 2008, and 2009. Both men are generally seen as “player’s managers” who run loose and comfortable clubhouses.

A major difference between Gardy and Maddon is their usage of sabermetrics and advanced stats. Gardy has publicly mocked them, while Maddon has embraced them. For instance, Maddon revolutionized defensive shifting. Gardy, meanwhile, has been consistently hesitant to platoon players, demonstrating a non-acknowledgement of numbers.

 

Unfortunately, it is difficult to quantify managerial performance. From a fan’s perspective, we can only judge based on what we see on the field and in the papers. Here, I’ve attempted to quantify (albeit crudely) the managerial performance of Gardy and Maddon. I’ve used a few simple metrics to clump managerial performance into two broad categories:

1. Making the most out of every run, as evidenced by team performance in 1-run games and relative to Pythagorean record. This is largely related to in-game managerial decisions and luck. Team confidence also factors into this; a team that is confident in their ability to win should, in theory, win more close games. Presumably, a manager can affect this.
2. Making the most out of cheap players, as evidenced by money spent per win. This reflects several things – lineup decisions, clubhouse rapport, defensive positioning, and front-office performance. The extent to which a manager affects this is unknown and undoubtedly varies from team to team.

 

Making the Most Out of Every Run
Here are comparisons of Gardy (red) vs. Maddon (blue) in Pythagorean difference (Pythagorean wins – Actual wins) and 1-run games from 2006-2014:

Over the course of 2006-2014, Gardy has a Pythagorean record of +1 overall, while Maddon’s is -4. In that span, their 1-run records are virtually identical: Gardy is 221-211 (.512), while Maddon is 215-202 (.516).

What do these numbers mean? Basically, they say two things. First of all, both managers have fielded teams that are, on average, slightly better in close games than their opponents. In that sense, both Gardy and Maddon seem to be above-average managers. However, neither manager is capable of making extraordinary in-game decisions that lead to far more wins than expected based on runs. If Maddon is substantially better than Gardy, it doesn’t show up in these metrics.

 

Making the Most Out of Cheap Players
While neither Gardy nor Maddon has had the opportunity to work with gargantuan payrolls, there is a clear difference here: from 2006-2014, the Twins won 709 games for a total payroll of $723.05 million ($1.02 million per win), while the Rays won 754 games for $479.65 million ($0.64 million per win). This is a striking difference. Furthermore, it is completely in-line with Maddon’s reputation as someone who can do a lot with a low payroll.

But a huge question remains – just how much does a manager actually affect wins per dollars spent? Does a great manager make a difference? Or is this more closely linked to front office decisions? In an effort to examine this, I compared 2006-2014 year-by-year money spent per win of four teams:
1. Twins (red)
2. Rays (blue)
3. Mets (orange; a team with changing managers and a front office with a poor reputation)
4. A’s (green; a team with changing managers and a front office with a great reputation)
Here are the results:

What’s interesting here is that the A’s have been almost as efficient with their money per win as the Rays, despite a revolving door of managers. And this doesn’t even include Billy Beane’s first 8 years as the A’s GM, when he originally championed the “Moneyball” philosophy.

The Twins were part of the “efficient” group until 2010-2011. Two things happened around this time: 1) Joe Mauer started earning a lot more money, and 2) Bill Smith’s craptastic decisions reached a peak, trading Wilson Ramos for the overpriced Matt Capps and trading the underpriced J.J. Hardy for nobody noteworthy. Since then, the Twins have been struggling to spend money efficiently, especially with Mauer’s clunker contract still looming large.
The Mets have been relatively inefficient spenders from 2006-2014, although they have heading in the right direction more recently. Like the Twins, they fired a GM during this time - Omar Minaya in 2010.

How much is Gardy to blame for the Twins inefficiency over the last four years, and how much should Maddon be credited with the Rays efficiency? These figures suggest that front office changes are far more meaningful in this regard – the consistently clever A’s and Rays front offices have kept their teams efficient despite drastically different managerial histories. Meanwhile, the Twins and Mets efficiency has fluctuated more with front office changes than managerial changes.

 

Take-Home Points
Overall, Maddon hasn’t really outperformed Gardy in any in-game metrics that could confer an advantage per run scored. And although Maddon probably had some impact on the Rays’ efficiency per dollar spent, front offices appear to be far more influential than managers.

Bottom line: it would be wicked cool to welcome Joe Maddon as the Twins’ next manger, and he’d certainly sell more tickets. But in comparison to the front office, his impact would be relatively small.

[glunn]

I applaud the effort that went into this and don't dispute the statistical analysis.  However, it seems to me that the charts may not reflect intangibles, such as getting the most out of each player and attracting better free agents.  

 

I would also enjoy seeing a chart of how many wins Maddon would have had relative to Gardy if the Rays payroll had been as high as the Twins payroll.  Such chart would require a lot of subjective assumptions, but the process could be illuminating.

[Linus]

Don't know about the methodology but I agree with the conclusion. Managers just don't make that much difference - players do. Spend the money on a pitcher.

[big dog]

Interesting stuff, thanks.  One concern I have about the Pythagorean projections is that it refers to how run differential translates into wins.  I would think that a really good manager, someone who concentrates on matchups, uses the hit and run efficiently, etc., would also be able to generate more runs out of a given set of players.  If a manager generates more runs, they ought to have more wins, even if they slightly underperform from a wins/run differential expectations.

 

If you look at two above-average managers, the difference might come down to whose personal style best fits the ballclub.  The wrong manager can wreck the whole thing, of course (see Red Sox recent history).  All that said, though, as long as you have a good manager I don't think it's easy to quantify the value of the best, as you point out.

 

That makes it easy to say we should spend the money on a stud pitcher instead, but managers almost never have Tommy John surgery and they don't cost a quarter of the payroll.  So my conclusion is I'm glad I'm not a GM, I guess.

[jay]

I would think that a really good manager, someone who concentrates on matchups, uses the hit and run efficiently, etc., would also be able to generate more runs out of a given set of players. If a manager generates more runs, they ought to have more wins, even if they slightly underperform from a wins/run differential expectations.

Shouldn't that same manager be able to use those skills to get the runs needed to scratch out a few more close wins than his counterparts? I don't see how we can say situational decision making improves runs but doesn't improve situational runs.

[gunnarthor]

I applaud the effort that went into this and don't dispute the statistical analysis.  However, it seems to me that the charts may not reflect intangibles, such as getting the most out of each player and attracting better free agents.  

 

 

What free agents has Maddon attracted?  I'm not sure that's a good measure anyway but a lot of guys that played under Gardy have wanted to come back to him after they've left. 

 

As for the "most out of each player" thing, I've mentioned before that the hardballtimes and WSJ did a study on best managers a few years back using three criteria: one run records, overachieving/underachieving pyth w/l records and players performance before and after leaving said manager.  Gardy was #1. 

 

In any event, I'm still leaning Molitor b/c I think he has a better baseball mind than Maddon and he also has a good relationship with Buxton and Sano.  I kind of worry that Maddon is opting out to get paid before his value as a manager declines. 

[nicksaviking]

Great analysis.  I agree that managers don't generally have a huge impact on the W/L record.  Now general managers....

 

Still, there is something about new managers that is likely intangible.  Why else would Buck Scholwalter, Terry Francona or Bob Melvin almost immediately turn around teams that the previous managers failed with? 

[big dog]

Shouldn't that same manager be able to use those skills to get the runs needed to scratch out a few more close wins than his counterparts? I don't see how we can say situational decision making improves runs but doesn't improve situational runs.

I didn't try to say it doesn't, though it probably reads that way.  My point is that it might not, and there are reasons to think that they might not be perfectly correlated.  For example, if they are really good at executing/moving players ahead a base because of coaching, but I don't call it at the right time quite so often, we might get more runs but somewhat fewer wins than the Pythag projection.  Or maybe I'm really good at platooning, which isn't situational on a day-to-day basis.

 

The Pythag projection doesn't usually show huge variations anyway, so I'm inclined to think that more runs (based on expected runs absent the manager) is important as its own measure.  I have no belief that could be estimated with any degree of accuracy, though.

[Dman]

I really liked the article but was surprised by the result.  Although I have never been a huge believer in managers making a ton of difference I guess I did buy in to the Maddon Mystique.  I was surprised the numbers didn't match the hype.  I still think Maddon would be a good manager for the Twins but I don't know why he would want to come to Minnesota over the Cubs or several other clubs.  

[tobi0040]

Here is an example of getting more out of cheap players via platooning them. Here is a look at the Rays 2013 outfielders, a utility guy, and backup catcher:

 

Matt Joyce   .747 OPS

Luke Scott    .741 OPS

Delmon Young  .780 OPS

David Dejesus .741 OPS

Desmond Jennings .748 OPS

Jose Lobaton  .714 OPS (.590 OPS in 2014 with Washington)

Sean Rodriguez   .714 OPS

[tobi0040]

 I kind of worry that Maddon is opting out to get paid before his value as a manager declines. 

 

I agree that is what he is doing.  He is at peak value and the pipeline appears to be thin in Tampa Bay.  But I still think he is the best guy out there so I want him in MN.

[Willihammer]

Yeah but Maddon has thick rimmed glasses.

[TheLeviathan]

Yeah, I've always tended to think the role of the manager is overblown.  It's more symbolic than anything and far more prone to destroying a team's chances than turning water into wine.  We shouldn't expect anything radically different with a new manager, the choice is more about a symbolic changing of the guard.

[Seth Stohs]

I think it's pretty close to a wash myself over time.  When Gardy had talent, his teams won a lot of games and outperformed the Pythag. When the Rays weren't great, Maddon didn't win either, but as they accumulated a lot more talent, he became a much better manager.

 

Could the platoons, shifts, etc., win a game here or there, probably. I can't imagine that the change would mean more than 1-2 wins a season.

 

Now, that said, the Twins should be able to possibly jump from 70 wins this year to hopefully 75, and maybe even 78... Whoever the manager is will probably look pretty good if that's the case. 

[tobi0040]

Yeah, I've always tended to think the role of the manager is overblown.  It's more symbolic than anything and far more prone to destroying a team's chances than turning water into wine.  We shouldn't expect anything radically different with a new manager, the choice is more about a symbolic changing of the guard.

 

I say about 3 games is the difference between the best and worst manager.

 

I just think about the difference between certain guys facing righties and lefties, as well as how many hits pull hitters lose to the shift.  I remember Mauer from April-June this year it seemed like drilled the ball right at the OF's a ton.   Being on the cutting edge there has to play a role. 

 

But Tony LaRussa was not going to bring the Astros or Twins to the playoffs this year

[Brandon]

I think Molitor is a great bench coach.  But I don't know how good he would be as a manager.  I do think Maddon is a good manager.  I think he has all of the things that make Gardenhire a good manager and he is open to advanced stats and ideas to improve the teams odd of winning as opposed to Gardenhire.  So my preference is Maddon.  though I wouldn't mind the Twins bringing Gardenhire back.

[Linus]

I say about 3 games is the difference between the best and worst manager.

 

I just think about the difference between certain guys facing righties and lefties, as well as how many hits pull hitters lose to the shift.  I remember Mauer from April-June this year it seemed like drilled the ball right at the OF's a ton.   Being on the cutting edge there has to play a role. 

 

But Tony LaRussa was not going to bring the Astros or Twins to the playoffs this year

Except it was hardly cutting edge.  Every team (every manager) was playing Mauer that way.  Most managers are very knowledgeable which is why they tend to cancel each other out, IMO.  I think the biggest factor is when you get a guy that is a real jerk, then they can do some damage ala Bobby Valentine.

[jay]

Could the platoons, shifts, etc., win a game here or there, probably. I can't imagine that the change would mean more than 1-2 wins a season.

 

If that's accurate and you could determine which managers are on the high end of that, wouldn't that make them easily worth $5-10M/season?

[tobi0040]

Except it was hardly cutting edge.  Every team (every manager) was playing Mauer that way.  Most managers are very knowledgeable which is why they tend to cancel each other out, IMO.  I think the biggest factor is when you get a guy that is a real jerk, then they can do some damage ala Bobby Valentine.

 

Fair point. I am using Gardy as my starting point.  The league was 5 years ahead of him

[Willihammer]

If that's accurate and you could determine which managers are on the high end of that, wouldn't that make them easily worth $5-10M/season?

I don't think managers can be considered free agents in the same way players can. For starters, how many wins is a replacement level manager worth? How would you even measure? Pyth. W-L doesn't seem to be very enlightening. Does anyone really think Maddon is worth -4 wins compared to a replacement manager over his tenure? Other issues are that managers don't face the same age-related decline - they can manage into their 70s. This leads to an overabundance of candidates for only 30 jobs, certainly that skews the economics a long way from $7m/WAR.

[markos]

I don't follow the Rays day in and day out so I might be missing things, but I've never gotten the impression that Madden is some kind of extraordinary manager. He has received a lot of support from sabermetrically inclined writers and analysts, but I have yet to read a convincing, detailed statistical account as to what exactly he does that separates him from the rest of the managers in baseball. I feel like he gets the benefit of the doubt since he had worked with THE sabermetric front-office for so many years, and if they stick with him then he must be good.

 

Undoubtedly, there are things he does well. I think he has been very good at maximizing his roster by using platoons and positioning to put his players in the best position to succeed. Additionally, he seems to be able to foster a great clubhouse culture. However, in other ways he is still fairly standard. His intentional walk numbers are league average, as are his sacrifice bunt numbers. Also, I haven't notice that his bullpen usage is particularly unconventional. 

 

I am definitely open to having my mind changed if others have strong arguments in Madden's favor. But considering that it sounds like he wants to be one of the top-paid managers in the game, I'm not sure he is worth it.

[nicksaviking]

 

But Tony LaRussa was not going to bring the Astros or Twins to the playoffs this year

 

I don't know.  The WS games in KC cost a fortune, but LaRussa is still rich enough to splurge for 25 of them even if they couldn't get a group discount.

[jay]

I don't think managers can be considered free agents in the same way players can. For starters, how many wins is a replacement level manager worth? How would you even measure? Pyth. W-L doesn't seem to be very enlightening. Does anyone really think Maddon is worth -4 wins compared to a replacement manager over his tenure? Other issues are that managers don't face the same age-related decline - they can manage into their 70s. This leads to an overabundance of candidates for only 30 jobs, certainly that skews the economics a long way from $7m/WAR.

 

Maddon might be the first instance of a highly desireable, free-agent like manager.  I agree with you on the complications of measurement, but the number of organizations that have reportedly reached out to him would seem to indicate that he's viewed as significantly better than replacement level -- which should translate to wins in some way or another.  There is an overabundance of candidates, but very few managers garner the reputation of being worth additional wins. 

 

If some team thinks he is worth 2 more wins than the other candidates available, why shouldn't he be worth $14M if the alternative is to go after those 2 wins on the free agent player market for the same price?

[Brock Beauchamp]

I think it's safe to say that Maddon is worth anywhere between -162 and 162 games* a season.

 

*not including postseason play

[dxpavelka]

I applaud the effort that went into this and don't dispute the statistical analysis.  However, it seems to me that the charts may not reflect intangibles, such as getting the most out of each player and attracting better free agents.  

 

I would also enjoy seeing a chart of how many wins Maddon would have had relative to Gardy if the Rays payroll had been as high as the Twins payroll.  Such chart would require a lot of subjective assumptions, but the process could be illuminating.

Funny how we mention that the charts above don't reflect intangibles and yet one of the biggest arguments people make for Maddon being and improvement over Gardy is that he "embraces advanced analytics" which, interestingly enough, don't reflect intangibles.

[tobi0040]

Maddon might be the first instance of a highly desireable, free-agent like manager.  I agree with you on the complications of measurement, but the number of organizations that have reportedly reached out to him would seem to indicate that he's viewed as significantly better than replacement level -- which should translate to wins in some way or another.  There is an overabundance of candidates, but very few managers garner the reputation of being worth additional wins. 

 

If some team thinks he is worth 2 more wins than the other candidates available, why shouldn't he be worth $14M if the alternative is to go after those 2 wins on the free agent player market for the same price?

 

$7M per WAR seems a tad high with several nuances.  Such as age, overpaying for career years, potential, injuries, etc.  All impact the players contract.

 

I looked up the four largest contracts (I think, it was quick look).  I excluded Tanaka because he did not have an MLB track record.  

 

Cano  - roughly $3M per WAR at 24M a year.

Choo - roughly 6.3M per WAR at 18M a year

Ellsbury -  roughly $4M per WAR at $21M a year

Mccan  - roughly $6.5M per WAR at 17M a year

 

Mid-tier guys

 

Arroyo -   6.7M per WAR

Byrd  -  5.3M per WAR

Feldman -  16M per WAR

[Willihammer]

The exact value of a Win in free agency isn't at issue. Its in the 6-7m range, according to fangraphs last study. The question remains how do you determine what is a baseline replacement manager worth and how many more is Maddon above (or below) that level? We have no idea of what Win-value Maddon presents to teams looking for a new manager outside of the rumors that about 28/30 teams are interested in him right now.

[Otto von Ballpark]

Interesting post, thanks.

 

What about playoffs?  I don't mean to rehash bad memories for Twins fans, but Maddon equaled Gardy's career playoff victory total by his 8th career playoff game.  13-17 overall isn't great, but it's way better than 6-21.

 

Obviously we aren't terribly close to returning to the playoffs, but it might be nice to know our manager isn't just a good "long haul" manager but can also hang in the October elimination tournaments, should the opportunity (hopefully) arise.

 

Here is another way to compare them: if my math is right, Gardy's teams are 25-38 against Maddon's head-to-head.  Obviously the last 4 years have been terrible for the Twins, so that skews things toward the Rays (who only had 2 similarly bad years under Maddon), but even in 2006-2010 the Twins barely edged the Rays 18-16, despite the Rays having two 96+ loss seasons in that time.  Gardy's success against Maddon is almost entirely based on Maddon's first year in Tampa (2006, when the Twins won the series 6-1) -- the Twins never won a season series with the Rays after that.

[Mike Sixel]

If we assume teams are not stupid...........managers must have a greater effect than we think, if the cheap ones get $2MM per year (meaning they are worth quite a bit), and the great ones will be getting $5MM per year (meaning teams think great managers are worth quite a bit more, more than the average player makes, I believe). Around 25% of players made $5MM last year......

 

http://www.usatoday.com/sports/mlb/salaries/2014/player/all/

 

or, our first assumption is wrong......

[gunnarthor]

Interesting post, thanks.

 

What about playoffs?  I don't mean to rehash bad memories for Twins fans, but Maddon equaled Gardy's career playoff victory total by his 8th career playoff game.  13-17 overall isn't great, but it's way better than 6-21.

 

 

Well, if you want to go there, the Twins opponents in the post season haven't been just good teams - they've generally been monsters.  Gardy's opponents averaged 101 wins each year, to the Twins 92.  That's a huge difference.  The only series Gardy lost where the Twins were the better/favored team was the 06 A's.  In contrast, Maddon had a 96 win team knocked out in the first round to a 90 win team and his playoff teams avg 94 wins while their opponents avg 93.