When things aren’t going the way as expected, it is usually a good time to change something. During the baseball season, that often means the manager has to go. Dan Graulich takes a look at the effect of changing managers in Bye Bye Birdie Tebbetts.
Teams will frequently fire their manager during the season when the team is struggling. The main reason being that it’s easier to fire one man as opposed to the many players not meeting expectations. Team management hopes that the players will respond to the leadership change and make improvements. But does this really work? Let’s take a look at previous instances of mid-season manager changes.
In order to obtain a sufficient sample, these results contain mid-season changes from 1980-2014 (2015 was skipped since the final results aren’t in). In an effort to keep at least a minimum amount of statistical significance, a change was counted only if the manager that started the year and the manager that finished the season were at the helm for at least 30 games that season. One team, the 2010 Baltimore Orioles, actually had 3 managers perform their duties for 30 games or more, so they were counted twice. There were a total of 103 instances of managers being replaced over that time frame. There were a few situations where the manager wasn’t fired, but there was a health issue (like Dick Howser’s brain tumor in 1986 for the Royals) that necessitated a coaching change. These were included because a team could have been affected just as much by a replacement even if it wasn’t a firing. Any interim managers, and their records, were ignored (usually just one or two games).
Of those 103 occurrences, the average change to the new manager increased the team’s winning percentage by .043, from .433 to .476. In terms of a 162 game schedule, this increase would take a .500 team at 81-81 to an 88 win team, so the difference could be considered significant. 75 of these teams improved with a managerial change (including the 2010 Orioles both times).
However, only 10 of the 103 changes resulted in teams making the playoffs. Those ten teams were better than the overall average before the firings (.433 all teams’ winning percentage vs. .486 for the playoff teams), but they also had a significantly bigger increase after (.476 for all teams vs .612 for playoff teams). Some of these teams are pretty familiar to baseball fans:
2003 Florida Marlins
1982 Milwaukee Brewers
Their season began with a disappointing record of 23-24 under Buck Rodgers, then finished 72-43 under Harvey Kuenn while leading the league in runs and home runs, they became known as “Harvey’s Wallbangers.” They lost in the World Series to the St. Louis Cardinals.
1988 Boston Red Sox
Started 43-42 under John McNamara, but finished 46-31 under Joe “No, the Other One” Morgan, including winning 12 in a row and 19 of 20 immediately after he was hired. They lost in the ALCS to the Bash Brothers’ Oakland Athletics.
There were 28 instances where the team had a worse record after making the managerial change. Of those 28 teams, 4 of them lost more than 100 points of winning percentage from the original manager. The 2004 Arizona Diamondbacks who went from awful under Bob Brenly (29-50) to putrid under Al Pedrique (22-61). The 1992 Texas Rangers were over .500 under Bobby Valentine (45-41), but collapsed under Toby Harrah (32-44). The 1988 New York Yankees were well over .500 with Billy Martin at the helm (40-28, the best pre-change mark of any team) and were a sub-.500 team under Lou Piniella (45-48). Worst of all, the 2001 Boston Red Sox were somehow solid under Jimy Williams (65-53) and then crashed and burned with Joe Kerrigan (17-26) at the helm.
The biggest improvements included 6 teams that had 200 point increases in winning percentage, led by the 1989 Toronto Blue Jays, who started out 12-24 (.333) under Jimy Williams, then Cito Gaston took over and the Jays went 77-49 (.611) to win the AL East before losing in the ALCS to the juggernaut Oakland A’s. The only other playoff team from that top group was the 2009 Colorado Rockies, who started out 18-28 under Clint Hurdle and finished 74-42 under Jim Tracy, winning a wild card berth and losing in the NLDS to the eventual pennant-winning Philadelphia Phillies.
Bad teams did not necessarily get worse. Of the 55 teams that began with winning percentages below the average of .433, only 4 had worse records under their new skippers. Four of the 10 playoff teams were in this group as well including the 2003 World Champion Marlins. Half of the teams (24 of 48) that started with better than average winning percentages actually did get worse for the rest of the season, a bad sign for decent teams making changes.
The main reason for more teams having bigger increases than the worst teams going backwards is simple: a team is not going to fire a successful manager unless there is an unusual set of circumstances or a case of blatant stupidity. Looking at all of the data leads us to believe that teams generally will improve with a managerial change, but the odds of making the playoffs are long. Additionally, there is no way to know if the improvement is just a regression to the mean from an underperforming team or some direct influence of the new manager. It becomes more advantageous to change managers with a worse record. The question becomes is it worth making the change? The data says this is a catch-22. The worse off you start, the better your chances to make improvements. However, the worse you start, the further you are away from the playoffs, and that means a lot more ground to make up. If you have a better team – closer to or above .500 – your odds of making the playoffs improve, but the odds of collapse are just as likely. Do you roll the dice and take your chances? That’s why general managers get paid the big bucks, to make the tough decisions like this.