|Wednesday, January 29
Updated: March 13, 4:02 PM ET
In terms of history, are Cubs in good hands with Baker?
By Nate Silver
Special to ESPN.com
Chicago loves a winner, and it loves a loser even more. While other cities are full of fair-weather fans, Chicagoans are a devoted lot, and the perennial failures of the Bears, Bulls and Cubs are chronicled in excruciating detail by the city's unrelenting scribes. Chicago will always remember Michael Jordan, but neither has it forgotten Gary Scott, Tim Floyd, or Cade McNown.
It is into this environment that Dusty Baker was introduced this winter. Dusty is optimistic, gregarious, and easy-going -- everything that Chicagoans are not. The scribes did their best to reserve their cynicism, heaping praise on the man, resting the hopes of a city on his broad shoulders. Better to build Dusty up, they must have thought to themselves, and the further we'll have to tear him down. These are the Cubs, after all, and their failure is a matter of destiny.
Or is it? The Cubs, among their other indignities, have suffered through a long string of unremarkable managers. Since Lee Elia was fired in 1986, 14 men have managed the Cubs, and none has since been hired for a full-time managerial job elsewhere. The Cubs have a Hall of Fame player in the prime of his career, an up-and-coming young rotation, a loyal fan base, and a deep farm system; it is easy to see how the manager's job is the most logical target for improvement.
It would seem, for once, that history is on the Cubs' side. Of the 1,266 clubs that have completed a season since 1946, the first full season after the last great war, 237 began the next season with a new manager. Those teams, on the average, improved their win total by just less than three games in the forthcoming season:
Three wins might not seem like a lot, but in the context of an offseason acquisition, it is significant, roughly equal to the difference between Gary Sheffield and Eli Marrero. Forget the Cubs; with gains like that for the taking, why shouldn't every team hire a new manager, or swap them wholesale like Fritz Peterson and Mike Kekich did wives?
If you've come to be familiar with our work at Baseball Prospectus, you'll know that we aren't likely to settle for an answer that simple -- especially when the answer is misleading. The complicating factor in this case is something that Bill James called the Plexiglas Principle. Baseball teams have a very strong tendency to regress to the middle; good teams usually get worse the next season, bad teams usually get better, and large gains or losses in the standings are usually balanced with moves in the opposite direction in the following year.
A bad team, it should come as no surprise, is much more likely to hire a new manager. In fact, as we can see from the following table, the tendency is quite dramatic:
Most teams that hire new managers do improve. But most teams that hire new managers are coming off losing seasons, and because of the Plexiglas Principle, most of them were going to improve to begin with, whether they had Casey Stengel, Terry Bevington, or William Ligue Jr. perched on the first step of the dugout.
An analysis that corrects for the Plexiglas Principle is required. My approach was to place all post-World War II teams into brackets covering 10 wins apiece. I then further divided the teams between those that had changed managers, and those that hadn't. A couple of additional ground rules: a managerial change was considered to have occurred only if it took place during the offseason. Also, as has been done throughout this analysis, all seasons have been prorated to the modern standard of 162 games.
There's a lot of data to work through here, so I'll do it a bit at a time.
Teams in the bottom two brackets, those that had won seventy games or fewer in the previous season, improved by almost exactly the same whether or not there had been a managerial change.
Similarly, the teams at the top of the heap experienced a dropoff in the win column, but ignoring the small sample of 100+ win teams, the dropoff wasn't any greater or lesser with a new manager. That's good news for fans of the Giants, Mariners, and Athletics, but perhaps not such good news for Felipe Alou, Bob Melvin, and Ken Macha. Their teams, though no fault of their own, are likely to win fewer games than they did last season. Surely, an enlightened management team would temper their expectations accordingly, and Alou, Melvin and Macha will each be managing under one of the best administrations in the game. But the business of baseball is competitive, major league front offices are breeding grounds for intraoffice conflicts of interest, and cooler heads don't always prevail; just ask Jack McKeon. The prediction here is that at least one of Alou, Melvin and Macha won't be back in 2004.
The teams in the middle of the pack, everyone between 70 and 90 wins, did appear to derive some benefit from bringing in a new manager, amounting to about a game or two. It would be easy, however, to make too much of this; the difference between the two groups is not statistically significant at a strict degree of confidence. Moreover, even if we have corrected for the distortions created by the Plexiglas Principle, we haven't done anything to ensure that the two groups are equal in other respects. Managerial changes frequently coincide with changes in player personnel; Tim Floyd was a disaster of a basketball coach, but it was the departure of Michael Jordan and Scottie Pippen that sealed his fate, and not Phil Jackson's emblematic Harley ride into the sunset. Certainly, one of the things we rally against at Baseball Prospectus is organizational complacency, and if a managerial change is part of a coherent long-term strategy, we're all for it. But it's the nine guys on the field that make the most difference.
Cubs fans are sure to protest that I haven't given Baker a fair shake. The problem with an analysis like this one is that it doesn't do anything to distinguish Casey Stengel from Terry Bevington, and there's a good argument that Baker shares far more in common with the former than the latter.
Managerial performance is notoriously difficult to evaluate objectively. Baker gets a lot of flack from analysts like us for his strange attachment to veterans like Shawon Dunston, his occasional overuse of starting pitchers, and a pension for the sacrifice bunt that is equaled among contemporary managers only by the man who managed the Cubs entering the 2002 season, Don Baylor. But the first job of a manager in any occupation is, well, to manage people, and there's every indication that Baker does an exemplary job of that. His players love to play for him, he gets them into the lineup, and they respond with performances that far more often exceed expectations than fail to meet them. Baker might do a lot of the little things wrong, but he does all of the big things right.
So let's give Dusty the benefit of the doubt; the Cubs, after all, are undefeated since he took over. Let's assume that he is a manager of Hall of Fame caliber. The Hall of Fame, of course, is itself not a perfectly objective standard of greatness, but the men who were inducted into it as managers share in common success sustained over many seasons, and in some cases, with multiple teams.
There are, by my count, 11 instances since the Black Sox scandal of 1919 in which a Hall of Fame manager changed teams prior to the start of the new season. I'm including only cases in which the manager already had a previous major league gig prior to joining a new team; his rookie season doesn't count.
Since we're dealing with a small sample, we need to introduce a slightly more complicated technique to account for the Plexiglas Principle. The technique, linear regression, allows us to come up with a best guess at the number of games a team is likely to win based solely on their win total in the previous season. For example, a team that won 89 games last year can be expected, all else equal, to win 86 games this year, giving a few games back to the Plexiglas Principle.
The table that follows includes our 11 managerial changes. In each case, I've indicated the number of games the team won in the previous season, our best expectation of its win total in the upcoming season per our regression analysis, and its actual result.
In most cases, the Hall of Fame manager did bring success to his new ballclub. Eight of the 11 teams beat their expected win total, and five of them did so by more than 10 games. The average improvement over the expected win total was nearly seven games, and the improvement over the team's unadjusted win total from the previous season was greater still. It's a small sample, sure, but Cubs fans have found worse excuses to renew their season tickets.
You'll also notice that there is one case that is far out of line with the others; that team, taken over by a Hall of Fame manager, performed about 17 games below its expectation. Its performance was so bad, in fact, that I've protected the identity of its manager, at least for a couple of paragraphs. (Fred Flintstone is not currently believed to have managed a major league team, in spite of occasional reports out of our New York bureau to the contrary).
Our mystery man was a manager with a deserved reputation as a winner. He took over a team with a legacy of losing seasons, but which provided ample reason for optimism on account of its star talent and improving farm system. That team featured a mid-career Hall of Famer in right field, an aging Hall of Famer at first base, and another one just getting his feet wet as a 22-year-old pitcher.
Coming off a 72 win season, the team lost eight of its first nine games, fell deep into the cellar, and never recovered; it finished at 59-103, the worst record in the big leagues by more than seven games.
It is time, Scooby-Doo style, to unveil our masked man. The year was 1966. The manager was Leo Durocher. His team? Ladies and gentlemen, your Chicago Cubs.
You can check out more work from the team of writers of the Baseball Prospectus at baseballprospectus.com. Baseball Prospectus is a registered trademark of Prospectus Entertainment Ventures, LLC.