Midseason Managerial Changes

With the recent dispatching of Arizona manager Bob Melvin increasing the debate of whether or not Clint Hurdle should be subjected to the same fate, I became inspired to take a look at what effect the act of replacing a manager midseason generally tends to have on a team. The results were generally consistent, and while they don't reveal everything about the situation, they do serve as some kind of insight.

Before posting the results, I should point out that the players are still, in my eyes, far more responsible for a team's success than a manager is. The actual effect of the managerial changes is not necessarily revealed here, but as I said, the results still have patterns.

So, here we go.

There have been a total of 12 midseason managerial changes since the 2004 season:
2004: 2
2005: 3
2006: 0
2007: 3
2008: 4

2008 Mariners: John McClaren 25-47 .374
2008 Mariners: Jim Riggleman 36-54 .400 Improvement: .026

2008 Blue Jays: John Gibbons 35-39 .472
2008 Blue Jays: Cito Gaston 51-37 .580 Improvement: .108

2008 Brewers: Ned Yost 83-67 .553
2008 Brewers: Dale Sveum 7-5 .583 Improvement: .030

2008 Mets: Willie Randolph 34-35 .493
2008 Mets: Jerry Manuel 55-38 .591 Improvement: .098

2007 Orioles: Sam Perlozzo 29-40 .420
2007 Orioles: Dave Trembley 40-53 .430 Improvement: .010

2007 Reds: Jerry Narron 31-51 .378
2007 Reds: Pete Mackanin 41-39 .513 Improvement: .135

2007 Astros: Phil Garner 58-73 .443
2007 Astros: Cecil Cooper 15-16 .484 Improvement: .041

2005 Royals: Tony Pena Sr. 8-25 .242
2005 Royals: Bob Schaefer 48-81 .372 Improvement: .130

2005 Reds: Dave Miley 27-43 .386
2005 Reds: Jerry Narron 46-46 .500 Improvement: .114

2005 Pirates: Lloyd McClendon 55-81 .404
2005 Pirates: Pete Mackanin 12-14 .462 Improvement: .058

2004 Diamondbacks: Bob Brenly 29-50 .367
2004 Diamondbacks: Al Pedrique 22-61 .265 Improvement: -.102

2004 Astros: Jimy Williams 44-44 .500
2004 Astros: Phil Garner 48-26 .649 Improvement: .149

Here are the important notes:
Contrary to what I (thought I had) discovered earlier, all but one of these changes led to an improvement (even if a small one) in that team's winning percentage. The largest improvement was Phil Garner, whose 2004 winning percentage was 15% larger than Jimy Williams' (ironically, Garner would later be a midseason firing victim himself).

The only negative switch was the Diamondbacks' Bob Brenly, who had a 10% larger winning percentage than his replacement, Al Pedrique. Interesting that the Diamondbacks, who have had the worst midseason firing history in the past 5 year period, would be so quick to jump on squishing out Bob Melvin.

So, initially, this suggests that the managerial change can bring nothing but good for a team. But hold your horses.

Of these 12 cases, only two times did the team in question make it to the playoffs. Both of these cases, the 2004 Astros and the 2008 Brewers, the team was at or above .500 when the manager was replaced. The 2008 Brewers are a unique case too, because their firing took place when only two weeks of play remained in the season. Therefore, the best performance of any of these teams is interestingly the one who kept their manager the longest.

Remember, this data only suggests results, not confirms them. It must be kept in mind that each of these firings had a different reason. Ned Yost was fired after a good season, when the Brewers' final month performance was below expectations, while it is widely suggested that Willie Randolph was fired due to his controversial relationship with the Mets' Front Office.

So, assuming the pattern continues, what does this mean for us? Presumably, it means that the Rockies are more likely to perform better than they have so far (which is something that could also happen under Hurdle anyway, this data does not account for so-called bad managers who got to keep their jobs, and whether or not they turned their performance around.) It also means that more likely than not, just as I have believed, that the Rockies' performance will not change dramatically. Even though it is likely to change for the positive, it is unlikely that this will be a successful season, even if we pull the trigger.

The average improvement in these cases of midseason managerial change between 2004-2008 is .067, or about 6-7% improvement. It has reached as high as 15%, and been as low as a 10% decrease in productivity. It's a wide and unpredictable range.

Somebody pointed out in another thread on the subject of firing Hurdle: what have we got to lose? Well, if this data means anything at all, it really is not a lot. But we also do not have a lot to gain either. Unfortunately, while this data is interesting, I don't see how it succeeds in enlightening us too deeply on the decision. For example, Hurdle's performance from late May 2007-September 2007 is a far better revival of a bad team than ANY of the managerial changes have been. That said, how much of that was the players? How much of it was luck? Too many factors to determine. We certainly can't expect it to happen again, but we can't expect it to happen with a managerial change either. Even with the improvement, the bad teams tend to stay bad teams.

Anybody have any thoughts?