Do Manager Firings Really Change Team Trajectories?

“Throw the bums out!” is a rich American tradition. While most used in the context of the messiness of that whole democracy thing, it’s also applicable to sports. When you’re a fan, especially a passionate one, and things are going horribly wrong for your favorite team, there’s a real sense of wanting the perpetrators of these crimes against excellence to be figuratively carted out in tumbrels and to meet their makers like Danton or Robespierre. And heads do roll in baseball when things are going badly, because someone has to take responsibility for a team’s crapitude, and it’s not going to be the team’s owner. Most often, it’s someone public-facing, as fans will not be appeased by the firing of some relatively anonymous staffer in operations. Since general managers and team presidents get first priority to hold the axe (but not always), and individual coaches don’t usually have wide-enough authority to take responsibility for the whole team, that leaves managers as the common sin eaters.

The moment of catharsis happens, and lo and behold, teams play a lot better, vindicating the demise of the ex-manager. It certainly feels that way, and it’s not the craziest idea in the world to think that there’s something to it. While you would expect teams in the midst of a spate of sucking to be underplaying their talent level rather than overplaying it, when you drive by an accident with a car that’s been unfortunately integrated into a telephone pole, it’s also quite likely that the driver had something to do with it.

Two managers have already been fired this season, after their large-payroll teams with championship aspirations got off to awful starts. Surprisingly, Alex Cora was first to go, as Red Sox chief baseball officer Craig Breslow canned not just Cora, but also anyone on the coaching staff considered to be one of Cora’s guys, on Saturday night after the team started the season 10-17. Then, on Tuesday, the Phillies fired Rob Thomson after they began the year 9-19, a woeful start that included a 10-game losing streak. Four years ago, Thomson became one of the most successful midseason replacement managers ever, as he steered a sinking Phillies ship back from a 22-29 start all the way to the World Series. Mets manager Carlos Mendoza can’t be comfortable about his job security right now, despite the team’s insistence that his job is safe.

(Editor’s note from Matt: As I edit this piece on Thursday afternoon, I keep checking to make sure the Mets haven’t fired Mendoza after yet another brutal loss.)

The point is, the first step for a struggling team that was supposed to be a contender is to get rid of its manager. A fresh start with a new skipper, though, doesn’t automatically mean things are going to get better. So, what’s a realistic expectation for a team that replaces the manager in-season?

There have certainly been occasions when teams made stunning comebacks immediately after replacing their manager. I already mentioned how the Phillies won the NL pennant after firing Joe Girardi. But there’s also the 2004 Astros, who were 44-44 when they replaced Jimy Williams with Phil Garner; they went 48-26 from that point on, made the playoffs, and advanced to the NLCS, where they lost to the Cardinals in seven games. (The next year was the Brad Lidge–Albert Pujols home run year, and that was now more than 20 years ago and I feel really old.) But most of the time, things don’t turn out that well. We just have a tendency to remember the outliers.

But how to measure the effect of a midseason firing, if one exists? I went through every midseason managerial firing/hiring going back to 2004 (so that there’s projection data, but we’ll get to that a little later). I intentionally excluded April and September firings in order to make sure both the incumbents and the replacements got more than a handful of games to influence their teams. After all, it wouldn’t be quite fair to expect, say, Freddie Benavides to meaningfully change the trajectory of the 2024 Reds after David Bell was gratuitously fired with just five games left. I also have not included the games for guys who only filled in for a game or two.

For each of the 40 managerial firing/hiring transactions that have come during the season since the start of 2004, I tallied up the team’s record under the fired skipper and its record under the replacement.

(There were two other midseason managerial moves that were not due to a firing. On July 1, 2007, Mike Hargrove unexpectedly resigned as Mariners manager; at the time, Seattle had won eight straight games and was in second place. He didn’t give a reason publicly, though according to his SABR bio, “almost immediately speculation arose that he was forced out because of his increasingly toxic relationship with the team’s star, Ichiro Suzuki.” It’s worth mentioning that his contract was up at the end of the season. Bench coach John McLaren took over for Hargrove. Then, last June, Ray Montgomery replaced Ron Washington as the manager of the Angels when Washington was placed on medical leave. Soon after, Washington underwent quadruple bypass heart surgery.)

On a fundamental level, the teams did play better ball after firing their manager. The teams had a .414 winning percentage when their managers were fired, and the replacements accumulated a .467 winning percentage the rest of the way. Thirty two of the 40 teams saw their record improve. The 2010 Orioles even did it twice, with their .278 winning percentage under Dave Trembley improving to .331 during Juan Samuel’s 51-game stint as skipper; they then played nearly .600 ball after Buck Showalter took over in August.

But how much of this is the manager’s doing, and how much of this is underperforming teams simply regressing toward the mean? After all, Thomson was fired with the Phillies on pace for a 52-110 season, and I don’t think anybody on this planet honestly believes they would have lost 110 games if he had kept his job. You could find a TaB-branded soda machine that hasn’t been plugged in since 1989 from an abandoned Amoco station in Dubuque, Iowa, make that the Phillies manager, and the team would improve on Thomson’s record — and it would have little to do with the magic of saccharin and caramel color.

To get an idea to an extent of the good vibes from new managers, I went back to the point of firing and got rest-of-season projections from the moment the new managers took over. While projections are no doubt going to have significant error bars, they also care little about the good feelings of a new manager, and if managers are changing something fundamental about the roster, you’d expect those teams to exceed their projections as a group. From 2014 on, I used the FanGraphs projected records, since FanGraphs is conveniently updated every morning. For 2004-2013, I ran a ZiPS projected standings for the date, certainly a more time-consuming task, but at least I have the projection data to do it!

Across 3,061 games managed, the new managers won 1.5 fewer games than expected. This is a statistically insignificant difference, of course. Incidentally, this can be taken as yet another data point in favor of the notion that projections aren’t too slow in lowering the projections for underperforming teams. But that’s a piece for another day!

Does this mean that new managers don’t matter? That would be a preposterous conclusion, one too heavily drawn from the information available. It’s certainly quite possible that if these managers had been doing an especially poor job, their teams would’ve continued to fall short of expectations. Taking a team that’s usually playing miserably and at least getting it back on track is a good thing.

But what this does mean is that, at least based on the last couple of decades, new managers shouldn’t fundamentally change how we look at a team’s performance. If the new managers, Don Mattingly and the Non-Diamondback Chad Tracy, get their teams playing like they should, they deserve credit for that. But for those of you expecting sorcery, I’m sorry we live in this thoroughly unmagical world.