When You Change Horses

Well, it’s official, Dave Trembley is out as the Orioles’ manager. And Juan Samuel is in. The Orioles have been terrible in 2010, winning just 15 games out of 54. To say their offense has struggled would be an understatement. At 3.33 runs/game, they are dead last in the American League, and have scored 10% fewer runs than their closest competitor for the bottom. Their run prevention has also been terrible, leaving them 12th in the AL in runs allowed/game. Their pitchers have allowed the most homers in the American League and the most hits. They have the second fewest strikeouts in the league. And their allowed batting average on balls in play is a high .306 (10th in the AL). Mostly, this last problem is due to some atrocious defense being played in the Emerald City. Nick Markakis already has cost the team seven runs on defense. Ty Wigginton, miscast while filling in for Brian Roberts at 2B, has allowed an extra 9 runs in. Corey Patterson, Miguel Tejada, Garrett Atkins, and the statuesque Nolan Reimold have also been defensive disasters. Only Caesar Izturis and Julio Lugo have been assets on the defensive side. And the development of the teams’ future, Adam Jones and Matt Wieters, seems to have stalled under Trembley’s watch. So, really, Trembley’s exit is probably a good thing.

But what about the strategy of firing the manager in general? Since the Royals shuffled Trey Hillman out for Ned Yost earlier this year, The Common Man has wondered whether changing horses in midstream is generally a good idea. So he went back and looked at every team that switched managers midseason from 1901 until today, to see what kind of differences a new manager might make. What he discovered is below.

There were 338 cases (including Trembley) of a club changing managers at least once during the season. You won’t be shocked to know that, taken as a whole, managers who start the season with clubs, but are replaced, lose more games than they win. At 7210-8715, they have a winning percentage of .453. That works out to a 73-89 record over 162 games. The performance of their replacements, however, might surprise you. Second, third, and even fourth skippers have a combined record of 5355-6023, a winning percentage of .471, which works out to a 76-86 record over 162 games. In 223 of these 338 team-seasons, the subsequent managers performed better than the man they replaced. In retrospect, however, what should we have expected? Teams that fire their managers tend to be underperforming, and extremely bad squads have a tendency to regress to the mean. Really, in light of this, a three win swing is not all that significant. Indeed, it’s not at all clear that firing the manager has any impact on the performance of teams, as a macro-level strategy.

But on the micro level, with individual teams and individual manager in individual situations, perhaps it does make a difference. It is up to the general manager to make that determination, and to hire the right man to right the ship, if that’s the ultimate goal. Not just to find an interim manager, but a better manager to steer the club’s roster for the long term.

It should be noted that teams change managers for several reasons. The skipper may have lost the clubhouse, be tactically inept, have the wrong mindset to motivate a particular group of players, or simply be losing a lot of games (and since you can’t fire the players…). Or it could be some combination of reasons. And not every managerial change is designed to win more games during the season in question. For instance, the Angels fired Cookie Rojas in 1988 with eight games to play, and a 75-79 record. The team had no hope of catching the Oakland A’s, and eight games wouldn’t have been enough time to really get fans excited about 1989, but the club simply couldn’t stomach Rojas anymore. Rojas was replaced by Moose Stubing, who went 0-8 at the helm (Stubing holds the record for most games at the helm without a win in the 20th and 21st centuries). Likewise, Dave Bristol was fired by Braves owner Ted Turner in 1977, with a record of 60-100. Turner quickly installed himself as the manager for a game, before turning the team over to Vern Benson for game 162.

The first midseason managerial change of the 20th century occurred in 1902, when Horace Fogel’s Giants stumbled out of the gate. Fogel had never played in the Major Leagues, and had previous experience managing the Indianapolis Hoosiers in 1887. Between jobs, he was a sports writer. Somehow, he convinced Giants owner Andrew Freedman to give him the job, not that Freedman would admit it. In March, the club released news that Fogel had been hired as an “associate manager,” but also announced “that Fogel had not signed to manage the New Yorks.” The Pittburgh Press reported, “that Fogel had called on Freedman and informed him that he had several promising young players under contract, whom he could sign for the New York club if they were wanted. Fogel was told to go ahead, but was not given any reason to believe that he would be associated or connected in any way with the management of the team.” Apparently, that’s not what Fogel had been telling others, “Fogel certainly informed several reliable persons on Thursday last that he had signed to manage the New York team. There were even rumors heard of a considerable lump of advance money paid to the Philadelphia man.” Eventually, Fogel was officially installed as the team’s official skipper, and lasted all of 41 games. Sam Crane wrote in the August, “Fogel is the worst that ever happened as a manager. It was he who disrupted the New York team his spring and made Mathewson dissatisfied with his job.” Indeed, in his brief tenure, Fogel had tried to make Christy Mathewson a first baseman. Second baseman Heine Smith took over for Fogel and managed to do even worse, 5-27. Only then did the Giants strike gold, hiring a certain John McGraw, who would hold the job for the next 31 years.


Fogel apparently stayed on the payroll for at least half the season, and was called the “Chief of Scouts” as late as June 24, when “Horace even hinted in Brooklyn something about having the great Lajoie sign to a New York club contract before many days, but Horace is very apt to ‘talk through his Panama’ when his imagination gets working overtime, so we will have to pass the Lajoie story up until more results are shown than the breeze from Fogel’s ‘trap….’ Now is the time for him to get out and hustle—earn his princely salary by getting Lajoie even if he has to kidnap the slugging Frenchman.”


Some teams do indeed send their manager packing to, hopefully, inspire their club. In 2004, the Astros fired Jimy Williams after 88 games (and a 44-44 record) and installed Phil Garner. Under Garner, Houston went 48-26 down the stretch and made it to the NLCS. Last year, famously, the Rockies fired Clint Hurdle after 46 games (18-26) and brought in Jim Tracy (74-42) and surged to the Wild Card.

But for every great success, there are huge disappointments. On August 15, 2001, Jimy Williams managed the Red Sox to a 6-2 loss against the Seattle Mariners. Williams’ Sox had lost 6 of their last 7, and general manager Dan Duquette got out his hook, despite the Sox sitting in 2nd place, just 5 games back of the Yankees and two back of the A’s for the AL Wild Card. Williams was replaced by pitching coach Joe Kerrigan, who immediately had problems. Nomar Garciaparra and Carl Everett both went down with injuries. First base, second base, shortstop, and leftfield all became carousels of personnel. Manny Ramirez, who had been limited to DH, lost a lot of time to injury. Kerrigan went 17-36 down the stretch and the Red Sox finished just 3 games over .500. Now, Williams probably wouldn’t have made the difference, as the Everett and Garciaparra situations were probably unavoidable, and the A’s surged, winning 28 of their final 32 games. But Kerrigan’s tenure had to be a big disappointment, as he was not retained after the season and hasn’t managed since.

In another disappointing managerial swap, Billy Martin helped the Yankees jump out to a 40-28 record in 1988. The team was in first place until June 20, when it lost the second of four straight contests. Swept by Detroit at Tiger Stadium, the team fell 2.5 back of their opponents. More importantly, Martin seemed to be self-destructing. In May, Billy had been ejected in Texas for arguing with an umpire, and went to a topless bar after the game. There, he got in a fight with another patron and ended up needing 40 stitches to reattach his left ear. Later that May, general manager Lou Piniella resigned because he couldn’t work with Martin. The next day, Martin threw dirt on Dale Scott. According to the New York Times, his use of the pitching staff was getting on his team’s nerves, “Dave Righetti complained he did not receive regular work. Cecilo Guante has privately told teammates he has been used too much. Tommy John, the 45-year-old left-hander, was used three times in a five-day period. Charles Hudson, after throwing only 10 pitches in the bullpen, gave up a grand slam last Sunday on '’a batting-practice fastball.’ Some pitchers have expressed dissatisfaction with a lack of attention from the coaching staff.” Finally, after a dispute (which Martin aired with the New York press) with new GM Bob Quinn over whether Don Slaught was ready to come off the DL, Martin was sent home. He was immediately replaced by his former GM, Lou Piniella, but Piniella scuffled. For all the trouble he caused, Billy did win ballgames. Lou went 45-48 the rest of the way and was fired after the season when the Yanks finished in 5th.

Also, not all managerial changes are the owner’s choice. Bobby Winkles had managed the A’s to first place in the AL West in 1978 with a 24-15 record, but actually quit rather than continue to work for owner Charlie Finley. He told the AP later that year, “it just boiled down to a difference in philosophy on how one human being treats another.” Under Jack McKeon (who Winkles had replaced the previous season), the club went 45-78 and finished 6th. Winkles never managed again.

In 1946, Joe McCarthy pulled a similar move, resigning from the Yankees (who were 22-13 and 5 games back of the Red Sox). McCarthy cited “ill health,” but there were rumors that he had been feuding with new boss Larry MacPhail. Whether this is true or not, McCarthy apparently handled his resignation with class and dignity, meeting with his team and throwing his full support behind new player-manager Bill Dickey, “Dickey is the finest, most loyal fellow on the ball club….Give him a chance and I’m sure he’ll make good.” Dickey had just come back from World War II, and didn’t really have his legs back. He hit just .261/.357/.366 as the backup catcher, but probably took too much time away from starter Aaron Robinson (.297/.388/.506) and young phenom Yogi Berra (.364/.391/.682). Dickey’s record, 57-48, was ok, but the team slumped from the heights it had reached under McCarthy. The Reading Eagle reported that Dickey was “as quiet and efficient as a manager as he had been as a player, [and] lacked the fire and spark that MacPhail had hoped he would engener into the Yankee team.” Dickey allegedly asked MacPhail for his release and told him that he “had taken enough as pilot of the club and that he didn’t want to be considered for the job in 1947.” Johnny Neun managed the last 14 games for the Bombers, and Dickey would never manage again. The Yankees finished 3rd, 17 games behind the Red Sox.

McCarthy holds the distinction of having the highest winning percentage of any manager who managed more than 10 games and didn’t finish out the season. Paul Richards, who had a 91-54 record for the White Sox in 1954, comes in a close second. Richards quit the club late in September, with the team 13.5 games behind the Indians, taking a job as the new general manager and field manager of the Baltimore Orioles. Richards’ deal included a bump in salary, additional control as the GM, and three guaranteed years. Marty Marion took over for Richards, and went 3-6, but would win 91 games the next year.Richards helped to improve the Orioles, getting them out of the second division by 1960. Richards would also leave the Orioles before the end of 1961 to become the first GM for the Houston Colt .45s.

The other end of the spectrum, managers fired after a slow start, is much more robust. Cal Ripken Sr in 1988 and Phil Garner in 2002 were fired by the Orioles and Tigers respectively after starting 0-6. Their replacements, Frank Robinson (54-101) and Luis Pujols (55-100) didn’t do a lot better. Malachi Kittridge had the worst start of anyone who managed more than 10 games, going 1-16 for the Senators in 1904. Patsy Donovan (37-97) took over and was equally unable to help the club improve. On the other hand, Bill Norman started 2-15 for the Tigers in 1959 (ironically, Norman had helped the club finish strong in ’58 after replacing Jack Tighe). He was replaced by Jimmy Dykes, who went 74-63 the rest of the way before earning the boot himself in ’60 after starting 44-52.

Eighteen clubs have made it to the postseason with more than one man at the helm. The Cubs, Astros, Dodgers, Brewers, and Yankees have each done it twice. Of these teams, nine have made it to the World Series (though only four from the division era), and two (the 1978 Yankees and 2003 Marlins) have won it all. The Yanks, of course, replaced the contentious Billy Martin with Bob Lemon and cruised the rest of the way. The Marlins dumped bane-of-young-pitchers Jeff Torborg and hired Jack McKeon.

Of the 30 clubs, the Giants have gone the longest without replacing a manager midseason. In 1985, they replaced Jim Davenport with Roger Craig for the final 18 games, then handed the reigns to Craig for the next seven years. Craig begat Dusty Baker, who helmed the club for ten seasons. Then Felipe Alou managed for four years before being replaced by Bruce Bochy. The Twins (Tom Kelly and Ron Gardenhire) and Oakland A’s (Tony LaRussa, Art Howe, Ken Macha, and Bob Geren) each haven’t flipped managers midseason since 1986. This suggests a strong and consistent organizational philosophies of not rocking the boat and not making rash decisions. These clubs have consistently hired good managers (each of the managers on the list above have made the postseason, except Geren) and have not seen the benefit of making a move just to make a move. Indeed, that would seem to minimize team flexibility moving forward, since a new manager’s success may pressure the team to extend a manager who’s not ideal for that team.

The Dodgers, not surprisingly, have been the least likely club to make an in-season change over the course of the franchise. Wilbert Robertson, Leo Durocher, Walt Alston, and Tommy LaSorda were all long-serving and respected skippers, and had enough respect not to be brushed aside midseason, for the most part. Alston, in failing health, bowed out for Lasorda in ’76, and Lasorda did likewise for Bill Russell in 1998. In 1947, with Durocher suspended for the year, Clyde Sukeforth managed Brooklyn for two games (and won both), but did not want the job. Burt Shotton took over the rest of the way, and guided the club to the World Series. The Browns/Orioles (7) and the Senators/Twins (9) also have been shy about pulling the trigger midseason. Of the original 16 teams, the Indians and Cubs lead the way with 22 in-season changes each in 110 total seasons.

So is it a good strategy to change horses midstream, as it were? Again, on a macro level, there is no evidence that it’s significantly helpful. And, indeed, it can undermine a team’s flexibility to search for the ideal candidate the next offseason. That said, this is entirely team specific. General managers who feel their team needs a different skipper to get back in the race may, indeed, want to make this move. But if the GM is responsible for hiring the wrong manager in the first place, there is certainly no guarantee that they will choose the correct replacement. If a team is out of contention, it seems silly to jettison a manager, who you will still have to pay, for an extra three wins a year (particularly if a manager hasn’t lost the clubhouse). Thus, while a potentially good strategy, it is probably overused.

(note: The spreadsheet TCM created of mid-season managerial changes can be found here.