Changing horses in midstream

At the beginning of the season, we THT-ites made our annual predictions. One of the questions asked was “Which manager do you think will be the first to lose his job?” Questions involving managers always interest me, after all I did write a book called Evaluating Baseball’s Managers, 1876-2008 due out next month. (Its release date got pushed back due to circumstances beyond anyone’s control. Rats!)

Anyhow, I had a different answer to the managerial firings question than anyone else here at THT. I said no manager would lose his job during the season. I knew I went out on a limb with that answer. Odds are, someone would get fired, but I knew as recently as 2006 a full season went by without any midseason managerial changes, so it could happen.

Regardless, I was badly wrong. Four teams—Arizona, Colorado, Washington, and Houston—ended the season with a manager who wasn’t calling the shots on Opening Day. (And that doesn’t include Cleveland, which announced just before the season’s end that Eric Wedge was gone.)

That brings up a question: How many teams should you expect to fire their managers in midseason? Was 2009’s four unusual? Furthermore, how often does an entire season go by without any skipper getting scratched? For that matter, what’s the largest number of midseason changes ever?

Plan of action

To answer the above questions and affiliated ones, I want to look at baseball history from 1892-present to see when managers lost their jobs in midseason. Why stop at 1892? Mostly because, based on my research on managers, it was around this time that the position came into its own. Many (though by no means all) managers before them were more like contemporary traveling secretaries than modern managers.

Second, from 1876 to 1891, entire teams went bust midseason, creating an extra layer of instability. Once the American Association folded after 1892, the business of baseball has been much more functional, which means managers are more likely to be hired/fired based on their own merits rather than team-wide problems.

Now I have to figure what qualifies as a managerial dumping. Simply put, anytime a team has more than one man manage them in a season, it’s in. It doesn’t matter if the change was caused by firing, quitting, retirement, death, illness, or whatever. Since I don’t know the ins and outs of every single managerial switch in history, I have to cast a wide net

Furthermore, I’ll count all teams, even if it’s only a matter of one game. The 1973 Red Sox, for instance, had Eddie Kasko at the helm for the entire season, except for one. Hey, that’s all it takes; they’re in the group. That said, if a team makes numerous managerial changes in the course of the year, I’ll only count them once.

In short, I’m counting the number of teams each season that switched managers and am not worrying about the hows/whys of what caused the switch, because I’m not able to know that.

With that in mind, I can approach the questions asked at the outset.

How unusual was 2006?

First, how odd was it for no MLB teams to switch their managers in midseason, as happened in 2006?

Well, it turns out this has happened exactly nine times. Every team stuck with its Opening Day skipper in 1893, 1901, 1903, 1920, 1926, 1931, 1942, 2000, and 2006. So yeah, it happens—but only twice in my father’s lifetime.

Had I known this back in March, I wouldn’t have predicted no teams would fire their managers. Yeah, I knew most years had at least one departure, but if I’d known the chances of it happening were about three percent, I would’ve hazarded a guess.

That said, looking at the list of non-firing years, there are clear trends across time. I’ll look into that in more detail in a little bit.

How odd are four firings in one year?

Four managers lost out in 2009, far more than I would guessed. Am I right in thinking that’s high or is that just a sign of my own obliviousness?

Well, let’s see. In 2008, four teams made midseason managerial changes: Seattle, Toronto, Milwaukee, and the Mets. In 2007, exactly four did it as well. For that matter, precisely four teams did it in 2005 and 2004. Well, guess I can’t say it’s too unusual in the current game for four teams to oust their skippers in midseason. Yeah, 2006 sure was the out-of-sync season.

A Hardball Times Update
Goodbye for now.

Admittedly, the recent run of exactly four changes per year is a fluke. It’s only happened 19 times in MLB history. Prior to 2004, you have to go back to 1996. Then again, one reason MLB doesn’t usually have four teams dump their skippers in midseason is because they often have more than four changes per year. In the last 50 seasons, four or more teams changed skippers during the campaign 37 different times.

The most in one season?

Clearly, midseason swaps are more common than I would’ve hazily/lazily guessed. So what’s the highest number of teams to do this in one campaign?

Nine. That’s quite a bit. It happened twice: in 1988 and 1991.

In both cases, there was an oddity inflating the numbers. In 1988, MLB Commissioner Bartlett Giamatti suspended Pete Rose for a month, and as a result the Reds officially had two different managers that year. In 1991 Cito Gaston missed a month due to back surgery.

That said, even without Rose and Gaston, 1988 and 1991 top them all, for no other year had more than seven firings.

Some overlap exists between the 1988 and 1991 dumpings. John McNamara was fired in both seasons—in 1988 by the Red Sox and three years later by the Indians. In 1991, Buck Rodgers lost his job in Montreal in May but took over the job in California for the last fourth of the season. Four of the six teams in the NL East jettisoned their Opening Day skippers in 1991. The Phillies, Orioles, and Angels all fired a manager in both 1988 and 1991.

There were also some noteworthy firings in 1988. Both Billy Martin and Hall of Famer Dick Williams had their careers end with midseason axings. The Red Sox had their “Morgan miracle” that year, as the team won 19 of their first 20 games under new skipper Joe Morgan, propelling them to the division crown. At the other end of the spectrum, at the beginning of the year the Orioles fired manager Cal Ripken six games into an eventual AL record 21-game losing streak.

Actually, there’s another way of looking at this. Nine teams changing managers during the year is the record, but then again there were 26 teams in 1988 and 1991. What happens if you look at the percentage of teams switching skippers instead of raw numbers?

Well, but with that approach, the top year was the first year under study: 1892, when five of MLB’s dozen teams removed skippers. That’s not a very satisfying answer because it’s on the edge of the study. (For various other reasons, I usually consider 1893 the big turn from pre-modern to modern baseball.) I should note that season was something of a fluke, as baseball wouldn’t have five midseason changes until 1938.

In the 20th century, the percentage winner for most midseason switches was 1961, when seven of the game’s 18 teams did it. That included the College of Coaches-era Cubs. My favorite change was the 1961 Indians, as Mel Harder managed exactly one game for the team. I’m sure there’s a story there.

Ups and downs over time

As noted already, there have been ebbs and flows with firing managers in midseason. For example, in the 1910s, 14 teams changed managers in midseason, but it happened 57 times in the 1980s. Expansion doesn’t explain all the difference.

Let’s look at it a new way: take all my information and divide it up by decade. Take all the teams that changed managers in midseason and divide by the total number of teams in that decade. That will give some ideas of how this practice has changed over time. Here are the results (the 1890s only counts 1892-onward):

Decade	Changes	All teams    %
1890s	21	 96	21.88%
1900s	17	152	11.18%
1910s	19	176	10.80%
1920s	15	160	 9.38%
1930s	23	160	14.38%
1940s	21	160	13.13%
1950s	25	160	15.63%
1960s	47	198	23.74%
1970s	50	246	20.33%
1980s	57	260	21.92%
1990s	35	278	12.59%
2000s	34	300	11.33%

Something happened in the 1960s, didn’t it? There’s a very modest uptick in the 1950s, but that’s largely due to the end of the decade. From 1957 to 1959, teams changed managers 11 times, which is 23 percent of the times it could’ve happened. From 1950-56, it happened only 12.5 percent of the time, about the rate it happened in previous decades. Dating the shift exactly is never possible, but it looks like around 1960 something happened.

That’s an interesting point to focus on, because several unusual things did happen in those years that teams hadn’t seen happen before. In the 1960s, the Tigers and Indians did something unprecedented: they traded managers, Joe Gordon for Jimmy Dykes. That same year, the Cubs did something even stranger, trading their manager (Charlie Grimm) with their broadcaster (Lou Boudreau). The next year the Cubs did something far more unusual, creating the College of Coaches.

A theme exists in those maneuverings: Managers were being treated in ways they had never been used before. Teams were treating them more like cogs in a franchise-wide machine than had ever been the case. The overall trend toward firing managers in midseason fits this pattern.

OK, so it fits a pattern. That doesn’t really address the question, though. Why were managers being treated like this if they hadn’t been previously? In other words, what’s causing the shift?

I think I know. In his book on baseball managers, Bill James noted a change occurred in the role of manager at the mid-century. From the 1920s-onward, team’s general managers were responsible for the farm system, but managers were usually given authority over picking veterans. At the mid-century, this shifted. There is no precise date or event you can point to. Shifts like these are gradual, occurring at various places at different times.

Still, it’s worth noting managers became easier to fire in midseason once their position within the franchise eroded. When they were master architects, fewer than one in 10 lost their jobs in midseason. As cogs it was one in four.

James approximated the change in authority occurring a bit before managers began losing their jobs en masse midseason, but that strikes me as reasonable. Once the situation changes, people don’t always recognize its full implications instantaneously. There is some lag time where old practices survive, at least in part due to inertia. Why be more ready to fire a manager in midseason if you haven’t before?

This delay can also be because the front office just wasn’t in the habit of firing managers during the year. For instance, when Clark Griffith called the shots in Washington, the Senators never fired a manager midseason. He came on as manager in 1912, took over the franchise outright, and when he died in October 1955 the team hadn’t dumped someone during the season. In 1957, the team broke its streak.

But once Griffith and his guard even passed on or adjusted, they just needed to be shaken up enough to realize how fungible managers now were. And 1960 served as just that sort of year. The Cubs’ manager-broadcaster swap came at the beginning of the year. By the season’s end, six teams changed their managers, a new record. The Cleveland-Detroit trade was part of that, but it’s also worth noting Baltimore manager Paul Richards, one of the best regarded skippers of his generation, voluntarily left the dugout in September to become GM for the fledgling Astros. 1960 was the year teams became starkly aware of the importance of managers in the new environment. The next year seven teams switched managers in midseason.

The evolution of managers also can explain why the 1890s were higher. As I noted up top, previously managers were often little more than traveling secretaries, whose main job was to keep the team’s finances in order on a day-to-day basis. (Given how often teams went under back then, that was needed.) While a new wave of managers emerged in the 1890s, it doesn’t mean the scenario changed immediately. These sort of changes never occur like thunderclaps.

This is all very nice, but it leaves me with one big problem. All the logic and arguments used above indicate that in recent decades managers have thus greatly increased their importance. After all, midseason firings are barely more common now than they were in the heyday of John McGraw and Connie Mack. Heck, despite having twice as many teams now, MLB has overseen two times when no teams fired managers in midseason. Does this mean managers are increasing their power?

I doubt it. If anything, I think something quite the opposite is occurring. Managers might not be fired in midseason as often because they ain’t worth firing in midseason. If you want to shake things up in midseason, try to make a deadline deal. (Without bothering to do any research, it seems like deadline deals have gained more prominence in recent years.)

I really don’t have a fully satisfactory explanation for the recent downturn in midseason firings, but it’s real. Not so real as to make my preseason prediction worth a fart, but teams are less likely to swap horses in midstream now than they used to be.

References & Resources
My source for this was Baseball-Reference.com’s team pages.

The Bill James Guide to Managers came in handy as well.


6 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
David P. Stokes
14 years ago

I realize that there are limitations in the data, but if we really want to study the issue in detail, I think we’s need to make an effort to take temporary changes, such as when a manager misses a few games due to illness, suspension, or the like.  After all, that sort of situation doesn’t really address the issue of teams making a conscious decision to replace the manager.  Of course, if the manager dies during the season, or retires unexpected because of major health problems, that’s also not a case of the team making a conscious decision to change managers, either—but it’s still a lot different than someone missing a couple games to attend their mother’s funeral.

Gilbert
14 years ago

The Clark Griffith illustration is probably a good one.  In the days when pitchers used to count attendance in their off-days, the budget just could not stand paying off the contract of a manager to not manage.  To a Depression survivor that would be like throwing out a light bulb that flickered occasionally but hadn’t burned out.
Once the manager’s salary became a small enough percentage of the total budget it became more palatable to eat a sunk cost, particularly if the person the manager could not get along with was a highly-paid star.

Chris J.
14 years ago

David,

I kept that in because I had two possible scenarios, neither of which I liked.  First, keep it in and have minor stuff screw it up as you note.  Second, dump it when I know it doesn’t quite fit the parameters.  While the second scenario sounds more reasonable, it has its own problems.  Namely, I’m more sure why guys left for illness or suspension in recent years than I am in years past.  Since I’m trying to look at/compare across eras, I wanted some consistency.  I figure the odd reasons for leaving shouldn’t be that different across the years.

I should note, it’s sometimes hard to tell when a manager was thrown out and when he left for other reasons.  On the one extreme you have something like the 1937 or 1939 A’s when Connie Mack left on extended scouting trips.  But it’s not always that clear cut.  There are occassions when someone like Pinky Higgins or Jack McKeon will be rehired by the same team that fired him just shortly before. It isn’t entirely clear that they were fired or something else looking just at the data.  Actually, in those two situations I know what happened, but there are other occassions with more obscure managers it’s hard to tell.

Basically, I wanted to be consistent when comparing across eras.

kds
14 years ago

Another factor that may be worth addressing is player-managers. I think there would be hesitation to fire a player as a manager, unless he was traded.  And of course the player-managers were usually stars.  So the prevalence of player-managers would affect the overall firing rate.  See, for instance, Lou Boudreau Bill Veeck and the ‘48 Indians.

Chris J.
14 years ago

kds,

That’s a good point.  I’m sure that plays a role – though the big upsurge in midseason firings didn’t come until a ways after player-managers fell out of vogue.

Robert Haymond
14 years ago

You mentioned Connie Mack )I can still see him in my mind’s eye in the early day of baseball broadcasting in the dugout holding the scorecard, waving it like a fan all dressed up with a tie on) and I wanted to say that he was the manager of the Philadephia Athletics for many many years and, as bad as the team was, couldn’t be fired without his own sayso as he was also the owner!  Of course, this is an obscure point in terms of your own study.  Just a winsome thought.