Impactful Midseason Managerial Changes
Posted by Andy on September 27, 2010
A reader named Brian Klaff sent me an email about the biggest turnarounds by team that change managers mid-season. I thought would be interesting to everybody. Follows is an excerpt of his email, which shows that the 2010 Orioles and Cubs could join the list of biggest turnarounds in history.
Brian says:
I did the research on this myself using Baseball-Reference’s Franchise Encyclopedia for each team. Hope I didn’t miss anything! This certainly could lead to the bigger question of whether midseason managerial changes are actually beneficial overall.
Top 10 midseason improvements since 1900 after a managerial change (not including changes in the first or last 20 games of the season):
#1: 1989 Blue Jays 12-24 (.333) under Jimy Williams 77-49 (.626) under Cito Gaston Midseason improvement: +.293
#2: 1940 Cardinals 15-29 (.341) under Ray Blades and Mike Gonzalez 69-40 (.633) under Billy Southworth Midseason improvement: +.292
#3: 1912 Indians 54-71 (.470) under Harry Davis 21-7 (.750) under Joe Birmingham Midseason improvement: +.280
#4: 1999 Angels 51-82 (.383) under Terry Collins 19-10 (.655) under Joe Maddon Midseason improvement: +.272
#5: 2009 Rockies 18-28 (.391) under Clint Hurdle 74-42 (.638) under Jim Tracy Midseason improvement: +.247
#6: 1988 Padres 16-30 (.348) under Larry Bowa 67-48 (.583) under Jack McKeon Midseason improvement: +.235
#7: 1925 Cardinals 13-25 (.342) under Branch Rickey 64-51 (.556) under Rogers Hornsby Midseason improvement: +.214
#8: 1980 Twins 54-71 (.432) under Gene Mauch 23-13 (.639) under Johnny Goryl Midseason improvement: +.207
#9: 2002 Rockies 6-16 (.2727) under Buddy Bell 67-73 (.4786) under Clint Hurdle Midseason improvement: +.2059
#10: 1969 Angels 11-28 (.2821) under Bill Rigney 60-63 (.4878) under Lefty Phillips Midseason improvement: +.2057
Now here’s the really interesting relevance part: This year there are two teams that might join this list. The Orioles could actually break the record if they win 6 of their final 7 games.
2010 Orioles 32-73 (.305) under Dave Trembley and Juan Samuel 29-21 (.580) (so far) under Buck Showalter Midseason improvement: +.275
2010 Cubs 51-74 (.408) under Lou Piniella 19-11 (.633) (so far) under Mike Quade Midseason improvement: +.225
September 27th, 2010 at 12:33 pm
Interesting stuff, Andy and Brian. Please, excuse me for asking a follow-up question to some good work. Could you report the average improvement for all managerial changes?
September 27th, 2010 at 12:54 pm
I was sort of surprised that Harvey's Wallbangers didn't make this list; but then, there isn't a statistic for "Ulcers Avoided".
September 27th, 2010 at 1:10 pm
I've actually found myself rootin' for the Orioles a lot of days lately... they're exciting again. Maybe the PIrates should pull a mid-season manager change in 2011.
September 27th, 2010 at 1:17 pm
@Frank
They don't make the list because Buck Rodgers got fired at 23-24. Harvey Kuenn did go 72-43 though, a rather impressive turnaround.
September 27th, 2010 at 1:38 pm
The 2003 Marlins was a famous one.
Jeff Torborg (16-22) and Jack McKeon (75-49)
September 27th, 2010 at 1:44 pm
I had completely forgotten that Joe Maddon managed before Tampa Bay hired him.
September 27th, 2010 at 1:59 pm
JC, that's a much bigger project. As an O's fan, Buck inspired me to do this much research into the best turnarounds. But what it's made me question is the efficacy of midseason managerial changes in general. There have been incredibly few such changes that have resulted in a real turnaround, whether in terms of winning percentage or, even rarer, in terms of postseason berths.
And what's even more stark is the fact that the few guys who have been able to engineer these turnarounds have, by and large, also been fired midseason too. Guys like McKeon and Torborg and Hurdle have all turned around a team or two, only to also be on the firing end as well.
I intend to take a closer look at the overall efficacy of the midseason switch. I'm pretty sure that the results will show that it's hit or miss.
As for Harvey's Wallbangers and some of the other teams that I expected to see on this list, it seems that it's especially hard to make the list if you were even an average team before the change, because you'd have to play consistently .700 baseball thereafter.
The Orioles this year had the advantage of being truly horrific before Showalter arrived, getting Brian Roberts back at the same moment, and they made their change late enough in the season that they didn't need to maintain their improved play for as long as many other such teams had to.
What's sad to me is that they probably could have broken this record had they not blown their last 4 games (3 to the Blue Jays). At that point, they were only needing to go 6-5 in their final 11 games to break the record, which, given how they had been playing, seemed a piece of cake. It's almost as if Cito Gaston was protecting his record by sweeping the O's this week.
September 27th, 2010 at 3:50 pm
Are you going by just winning percentage? For sheer impact, Bob Lemon replacing Billy Martin in 1978 has to be up there. Without that change and its calming effect, the Yankees don't make that 1978 comeback.
September 27th, 2010 at 3:59 pm
Have any mid-season replacements been clearly WORSE than the previous manager?
September 27th, 2010 at 5:02 pm
It would be interesting to see the results going forward. Obviously, the larger the window, the less you are likely to see manager impact (as much as anything can tell us manager impact) because there are sure to be many other variables. But I'd be curious to see this data along with the new coach's record the next season and the previous coach's record the previous season (especially if this was broken down by month, indicating some sort of downward spiral he and the club were one). Regardless, interesting work. Thanks to both!
BCK, I do have to say I am intrigued by the name, seeing as how I am also a Brian with an online moniker similar to yours, though mine is not my initials.
September 27th, 2010 at 5:47 pm
@9 - Joe Kerrigan replacing Jimy Williams as Red Sox manager in 2001 comes to mind, though I don't remember the exact managerial record for either one (I remember the Sox performing worse under Kerrigan than Williams, however).
September 27th, 2010 at 7:06 pm
This is interesting, but I'm not sure it really tells us anything. In many of these cases, we're dealing with a pretty small sample of games on either side.
Take a look at this:
22-31 (.415)
71-32 (.689)
Difference of .274. Which, according to this list, is the fourth best ever.
Of course, the first number is the Phillies this year from May 29 to July 21. The other number is the rest of the season. Are we to believe that Charlie Manuel was a horrible manager for most of June and July, but a genius the rest of the year?
September 27th, 2010 at 7:56 pm
The 1912 Indians should be #1 on this list, not #3; 54-71 is only .432, so the improvement is .318. Also, if you recognize the Federal League, Buffalo was 15-30. .333, in 1915 when Harry Lord took over and went 58-48, .551, which is +.218, good enough to make the list.
September 27th, 2010 at 8:51 pm
Tim L-
That is EXACTLY what we are supposed to make of it.
September 27th, 2010 at 9:19 pm
@7
BCK, gotta react to "had they not blown their last 4 games (3 to the Blue Jays)". The lone international ML franchise won those games fair and square. The Jays have owned the O's in recent times. The Orioles are the sole reason Toronto doesn't argue more strenously for re-alignment of divisions.
September 27th, 2010 at 10:43 pm
Some more teams of note. The 1966 Yankees were 4-16, .200 under Johnny Keane, then 66-73, .475 under Ralph Houk. That's +275, but I suppose the "not including changes in the first or last 20 games of the season" proviso applies, even though the change was just after, not in, the first 20 games of the season.
The 1965 Athletics at +.2048 are just barely off the list, also the 1952 Phillies, .2039.
Now let me make the case for the 1959 Tigers. They were playing .118 ball when Jimmy Dykes took over, .540 after. It's not on the list because that .118 was 2-15, not making the 20-game cutoff. But you know how they add extra hitless at-bats to a batter's record if he has the highest batting average in the league but not enough plate appearances to qualify? Well, add three wins to Bill Norman's record, and he's 5-15, .250, and we have a +.290, well up there on the list. Or, if that 20 is really 21, make it 6-15, and there's still a +.254, quite notable.
September 27th, 2010 at 11:17 pm
Jimy Williams apparently was a good guy to follow. Phil Garner led the Astros to a .649 winning pct in 2004 after Williams took Houston to a .500 record through the first 88 games.
September 27th, 2010 at 11:27 pm
@15.... I agree with what you're saying, but Toronto just got swept in Baltimore 2 weeks ago. So I wouldn't say the Jays have owed them. Especially since that was the only series played with Showalter as the manager, which is the current incarnation of the team. Additionally, the Orioles literally blew a lead in game two of that series (the most recent one in Toronto) as well as a lead (albeit a small early lead) in the first of the 4 games mentioned by BCK.
Although, there's no doubt that Toronto won them fair and square.
September 28th, 2010 at 1:28 am
With tonight's win, Buck Showalter now has 30 wins in 51 games (30-21/.588). Two more wins, and his team will match the 32 win total over the previous 105 games.
I'm actually less impressed by what Cito Gaston did with the Blue Jays than what Showalter has done with the O's. When Gaston took over, the Jays had only play 36 games. In baseball, that's too small of a sample size. Good teams can play poorly over a stretch of 30 to 40 games. Showalter took over a team that had already played 105 games. They clearly had established their level of poor play that could not be excused by being a poor month or so.
In fact, many teams on that list had only played around 40 or so games before the manager was dismissed. On the other side, there are managers who took over later in the season than Buck, limiting their improved results to 30-35 games, such as the 1980 Twins or the 1999 Angels.
For the list to be meaningful, I'd say the dismissed manager and the incoming manager will have had to have managed at least 50 games. Quade and the Cubs from this year wouldn't make the list. Too small a sample size.
It appears what Showalter has done this year is quite unique.
September 28th, 2010 at 5:15 am
If you are going to up the stakes to Managers fired and Manager hired need min 50 games all 10 fail . Does this mean the O's make the WS by default.
September 28th, 2010 at 8:23 am
@8,
My first instinct was Bob Lemon from 1978 also, until I noticed the list was strictly by winning % differential.
When Lemon took over, the Yanks were 52-43 10.5 games out. They finished 100-63 and won the A.L. East and W.S.
September 28th, 2010 at 9:25 am
@19 >>With tonight's win, Buck Showalter now has 30 wins in 51 games (30-21/.588). Two more wins, and his team will match the 32 win total over the previous 105 games.
Yes! Assuming they win those two, this must be the record for getting to the same win total. I'd also like to see what these managers did the following year (along with the list of worst records after a change).
September 28th, 2010 at 10:17 am
Here are their performances the following season:
1989 Blue Jays
Total: 89-73 (.549)
Next year: 86-76 (.531), 2nd place (of 7)
Next year improvement: -.018
1940 Cardinals
Total: 84-69 (.549)
Next year: 97-56 (.634), 2nd place (of 8)
Next year improvement: +.085
1912 Indians
Total: 75-78 (.490)
Next year: 86-66 (.566), 3rd place (of 8)
Next year improvement: +.076
1999 Angels
Total: 70-92 (.432)
Next year: 82-80 (.506), 3rd place (of 4)
Next year improvement: +.074
2009 Rockies
Total: 92-70 (.568)
Next year (so far): 83-73 (.532), 3rd place (of 5)
Next year improvement (so far): -.036
1988 Padres
Total: 83-78 (.516)
Next year: 89-73 (.549), 2nd place (of 6)
Next year improvement: +.033
1925 Cardinals
Total: 77-76 (.503)
Next year: 89-65 (.568), 1st place (of 8), won WS
Next year improvement: +.065
1980 Twins
Total: 77-84 (.478)
Next year: 41-68 (.376), 7th place (of 7)
Next year improvement: -.102
2002 Rockies
Total: 73-89 (.451)
Next year: 74-88 (.457), 4th place (of 5)
Next year improvement: +.006
1969 Angels
Total: 71-89 (.444)
Next year: 86-76 (.531), 3rd place (of 6)
Next year improvement: +.087
Among this small sample size, there's an average improvement of about 0.027 to the winning percentage from one year to the next.
But I think what's more relevant in such a small sample size is that among those teams that improved (7 of the 10), 5 of the 7 improved between 0.065 and 0.087.
It wouldn't surprise me in the slightest to see the Orioles improve this much next season, from approximately 0.400 in 2010 (about 65 wins) to about 0.480 in 2011 (about 78 wins). Especially when you consider that most people were expecting the Oriole sto win about 75-77 games before this season, it seems reasonable for them to finally live up to that.
September 28th, 2010 at 1:52 pm
OK, here are the 10 worst instances of teams that got significantly worse after changing managers midseason (not including changes in the first or last 20 games):
#1: 1902 Giants
Under Horace Fogel: 18-23 (.439)
Under Heine Smith (who was subsequently replaced): 5-27 (.156)
Difference: -.283
#2: 1978 Athletics
Under Bobby Winkles: 24-15 (.615)
Under Jack McKeon: 45-78 (.366)
Difference: -.249
#3: 1937 Reds
Under Chuck Dressen: 51-78 (.395)
Under Bobby Wallace: 5-20 (.200)
Difference: -.195
#4: 1987 Cubs
Under Gene Michael: 68-68 (.500)
Under Frank Lucchesi: 8-17 (.320)
Difference: -.180
#5: 1969 Athletics
Under Alvin Dark: 52-69 (.4298)
Under Luke Appling: 10-30 (.250)
Difference: -.1798
#6: 1902 Orioles
Under John McGraw: 26-31 (.456)
Under Wilbert Robinson: 24-57 (.296)
Difference: -.160
#7: 2001 Red Sox
Under Jimy Williams: 65-53 (.551)
Under Joe Kerrigan: 17-26 (.395)
Difference: -.156
#8: 1908 Highlanders
Under Clark Griffith: 24-32 (.429)
Under Kid Elberfeld: 27-71 (.276)
Difference: -.153
#9: 1966 Indians
Under Birdie Tebbetts: 66-57 (.537)
Under George Strickland: 15-24 (.385)
Difference: -.152
#10: 1977 Athletics
Under Jack McKeon: 26-27 (.491)
Under Bobby Winkles: 37-71 (.342)
Difference: -.149
I love that the 1977 A's went from average under McKeon to terrible under Winkles, but kept Winkles as manager the next season. They were doing quite well under Winkles in '78, but Charlie Finley (even crazier than normal at the end of his tenure) went back to McKeon, and the team tanked again.
I found only 8 other instances in which a team was at least .100 in winning percentage worse under a midseason managerial replacement:
1909 Naps
1948 Phillies
1971 Indians
1975 Braves
1981 Yankees
1988 Yankees
1996 Angels
2004 D-backs
Also, it's worth mentioning 2 other teams, although the situations were slightly different.
1988 Reds - Pete Rose was suspended 30 days for pushing Dave Pallone, and the Reds were -0.116 under his replacement, Tommy Helms
1999 Astros - Larry Dierker missed 4 weeks after brain surgery, and the Astros were -0.149 under his replacement, Matt Galante
I haven't run all the numbers yet to see whether a midseason change is, on average, a good or bad move. But in general, it looks like most midseason managerial changes have little impact one way or the other. You can't usually get a leopard to change its spots.
But there are definitely more examples of teams improving significantly than of teams regressing significantly, which is the allure. If you haven't got anything to lose, it's worth a shot.
September 28th, 2010 at 2:51 pm
[...] Impactful Midseason Managerial Changes (Baseball Reference). Check out Jack McKeon and the ‘88 Padres. [...]
September 28th, 2010 at 4:03 pm
Curiously, the same day as this piece, I did a report on mid-season managerial changes over the past decade, looking at the records of a team the season before the change, before and after the switch, and the season after.
In the season prior to a firing, the 34 examples had an average W% of .480. That dropped to .441 before the change, and improved only fractionally to .457 afterward. The year beyond that, it improved, but again only slightly, to .465. It all seems more evidence that a manager generally has very little impact on a team's results.
September 28th, 2010 at 9:27 pm
@18
Thomas, good post about the Jays' record against the O's in the pre-Showalter era vs. currently. Buck has definitely injected something tangible into an under-preforming Orioles roster this year. Point conceded.
@23 @24
BCK, thanks for the searches. It will take me a while to digest the significance of the info you've uncovered.
@26
But, Jim, you would have to concede that Showalter's short-term impact on the Orioles is an exception!
September 29th, 2010 at 2:26 am
@20, Hubert -- If you are going to up the stakes to Managers fired and Manager hired need min 50 games all 10 fail. Does this mean the O's make the WS by default.
----------------------
That's right. They all fail, except for Showalter. What we'd get, though, if we ran a new list with 50-game minimums is an entirely new list, which would have less spectacular percentage swings, but one that I think would have a bit more meaning in determining the impact the new manager has had on the team.
Ummm, no World Series for the O's...yet
September 29th, 2010 at 3:46 am
The master of the midseason turnaround?
Steve O'Neill
Check this out:
1935 Indians
Under Walter Johnson: 46-48 (.489)
Under Steve O'Neill: 36-23 (.610)
Difference: +.121
1950 Red Sox
Under Joe McCarthy: 31-28 (.525)
Under Steve O'Neill: 63-32 (.663)
Difference: +.138
1952 Phillies
Under Eddie Sawyer: 28-35 (.444)
Under Steve O'Neill: 59-32 (.648)
Difference: +.204
Total, 3 teams
Under other managers: 105-111 (.486)
Under O'Neill: 158-87 (.645)
Difference: +.159
Three teams, three spectacular improvements. And all three would fall under the revised guidelines of 50 games in both directions (before and after the change). This guy was, for all intents and purposes, a midseason miracle worker.
For those of you who don't know, O'Neill is one of only 2 managers to manage in at least 5 (even partial)seasons and never have a losing record. The other is McCarthy. He won the World Series with the 1945 Tigers, but never made the postseason other than that in 14 seasons total of managing (he came in 2nd with the Tigers in '44, '46, and '47). His overall winning percentage of .559 is 15th among skippers who managed at least 1000 games in the majors and 8th among those who managed 1000 games exclusively after the 1900.
September 29th, 2010 at 3:50 am
Sorry, let me qualify my last statement about O'Neill...
I shoudl have said he was 10th among managers who managed at least 1000 games after 1900.
The "exclusively" qualifier was misleading, as it eliminated McGraw and Clarke, both of whom managed mostly in the 1900's, but who I had ruled out by virtue of a couple years of managing pre-1900.
September 30th, 2010 at 12:29 pm
percentages seem a little mis-leading for the smaller samples. I probably would have gone with games above/below .500