- A deep dive into the numbers to examine whether recent title teams got off to slow starts, and what it means.
The questions over whether the 2017 Cubs would experience a championship hangover were raised long before the banner commemorating their World Series win had gone up. So far, as expected given the talent on hand, they've shown little sign of it, as they enter Tuesday in first place in the NL Central at 11-8. However, they are already three games off the blistering pace set by last year's champions and they're likely to fall further behind; Chicago would have to win its next 14 games to match the 2016 version's 25-8 start.
In truth, recent history has been unkind to World Series winners. None has repeated as champions since the 1999 Yankees won again in 2000, and in that span, only two other winners—the 2000 Yankees and 2008 Phillies—even returned to the Fall Classic the following year. Just seven out of 16 made the playoffs at all. Particularly with a postseason that can now stretch into November, it seems natural to suggest, as many have, that there's a championship hangover effect.
Finding evidence of that is a different matter, but doing so requires defining what we're looking for. The possibilities about where one might look—a year-after effect on pitcher performance and injury, or on hitters and aging — are endless and potentially fascinating, but lacking the time to pull off a more granular study, we can keep things simple by testing a couple of theories:
• 1) that the defending champions are more likely to start slowly the following year;
• 2) that last year's winners are likely to finish more slowly.
Admittedly, it's easy to project narratives onto each of those theories. Perhaps, for instance, the first one is true because those teams' shortened off-seasons didn't give their players' aching bodies enough time to rest and recover, or their architects were reluctant to break up a winning combination. Or perhaps the second one is true, because the heavier workloads take time to catch up with those players, or their luck finally runs out. We're getting ahead of ourselves, though, and as the great baseball writer Leo Tolstoy once wrote (it's somewhere in the back of Anna Karenina, I swear), happy teams are all alike, but every unhappy team is unhappy in its own way. For the purposes of this quick-and-dirty study, it's best to keep those narratives aside and see what the data says.
As for slow starts, recently, I noted a study published by Derek Carty in Baseball Prospectus' 2012 book Extra Innings (to which I also contributed). Examining the years 1962 to 2011, Carty found that at the 16-game mark, a team's current record reaches a correlation of 0.5 with their final record, which is to say that their year-to-date performance became more predictive than simply assuming they'll finish at .500 (as the largest possible sample of teams inevitably does). Before that 16-game mark, there's more noise than signal in a team's record (to say nothing of the underlying individual performances). At the 20-game mark, a point that most teams will have reached by Tuesday, a team's current record is slightly more predictive than the .500 one. According to Carty's formula—PW% = G/(G+16) * W% + (1-G/(G+16) * 500, where G is the number of games played, W% is current winning percentage and PW% is their final winning percentage—at that point a team's current record would account for 56% of its projection.
In other words, checking to see how the defending champions fared over their first 10 games wouldn't have told us much, but at 20 games, we've at least got a sample size that's somewhat meaningful. As it turns out, at the 20-game mark, the recent defending champions performed quite similarly to how they did in their championship year. The 2000 to '15 World Series winners combined for a .578 winning percentage in the regular season, around a 94-win pace. In the first 20 games of the following season, they posted a combined .569 mark, about a 92-win pace—a falloff, but hardly a dramatic one; without the rounding, it amounts to a 1.46-win slip over the course of a season. Only four of those 16 reigning champs won fewer than half of their first 20 games: the 2015 Giants went 8-12, while the '03 Angels, '07 Cardinals and '14 Red Sox went 9-11. None of those teams made the playoffs, nor did the lone 10-10 team in the set, the 2011 Giants. The cold or lukewarm starts didn't help those clubs, of course, but with more than two-thirds of the clubs in question playing at least .550 ball through the first one-eighth of a season, it's tough to argue that there's a hangover at work.
Stretching the definition of a season's start to 40 games, the data shows that those 16 defending champs cooled off just a bit. Their combined winning percentage at that point, roughly one-quarter of the way through the follow-up season, was .558, just over a 90-win pace, and a total drop relative to the year before of 3.2 wins. Only one of those teams was below .500 (the 2007 Cardinals, at 16-24), but three more (the '03 Angels, '14 Red Sox and '16 Royals) were right at .500, and none of those teams recovered to make the playoffs. Still, if there were hangovers, they were specific to those teams; the other 12 had a combined winning percentage of .576, nearly identical to the set's overall mark of the teams in their championship seasons.
As it turns out, even that .558 winning percentage at the 40-game point is far better than the 16 defending champions' full-season combined winning percentage of .535, suggesting that it's the other end of the season where we should be looking. Indeed, the 16 teams won at just a .506 clip over the final 40 games of the season, with only the 2001 Yankees and the '05 and '08 Red Sox winning more than 22 games; they each went 24-16 and were three of the seven teams to return to the playoffs. Six defending champions were below .500 over those final 40 games, none of whom made the postseason.
It gets worse. The 16 reigning champs played at a .497 clip (a 159-161 record) over their final 20 games. Six of them won either seven or eight games—only one of whom, the 2010 Yankees (8-12), still made the playoffs—while only four won either 12 or 13 games; three of them made the playoffs.
Within the set, there are teams that fit the patterns I suggested above. The 2006 White Sox, who went a very respectable 90-72, lost 12 of their final 20 games while starters Mark Buehrle, Jose Contreras and Jon Garland and closer Bobby Jenks, important pieces from their championship run, were lit for ERAs above 6.00 over those final 20 games. The '14 Red Sox, who finished 71-91 and dropped 25 of their final 40, traded away pitcheres John Lackey, Jon Lester, Jake Peavy and Andrew Miller and infielder Stephen Drew by that point and lost star second baseman Dustin Pedroia to a season-ending within that timeframe. The 2016 Royals, who lost 13 of their last 20, were down reliever Wade Davis and third baseman Mike Moustakas by the end, with catcher Salvador Perez, first baseman Eric Hosmer and leftfielder Alex Gordon skidding in September.
Still, it's a risk to generalize, because not every team conforms to the pattern. The 2012 Cardinals needed a 13-7 closing run to hold onto the second NL wild card spot (interestingly, they're the only defending champion in the two-wild card era even to return to the postseason). And what we don't have is a basis of comparison to distinguish this "hangover effect" from garden-variety regression to the mean, or what Bill James called the Plexiglass Principle: teams that overachieve in one year will tend to underachieve the following year.
Since the wild-card-era playoffs are something of a crapshoot given how rarely the team with the season's highest win total comes out on top, it's worth comparing the follow-up seasons of the champions to those of their fellow playoff teams. As noted the championship teams produced a combined .578 winning percentage in the years that they won it all, and a .535 mark the next year, a 43-point drop, with seven out of 16 (43.8%) making the playoffs. Meanwhile, the other postseason teams from 2000 to '15 actually produced a .580 winning percentage that year and fell off to .548 the following year, a 32-point drop—less than that of the champions. Sixty of those 120 teams (50%) made the playoffs.
By those measures, these post-millennium championship teams fared slightly worse than expected, albeit not by much, and tended to struggle late in the season, though without breaking down the 20- or 40-game closing stretches of those other 120 teams, we can't say to how great a degree relative to their rivals. At least within the narrow parameters of what I've examined, it does seem there's a slight hangover effect at work here.