There's very little evidence that momentum means anything in baseball ... until October, when suddenly it can seem to mean almost everything.
I'm on record, many times, saying there's no such thing as momentum in baseball. Or as Earl Weaver put it, "Momentum? Momentum is the next day's starting pitcher."
I have long believed that, and I still believe it.
Well, except for this: Momentum -- or something that looks very much like it -- does seem to play a big role in October. The numbers are just overwhelming. Last week, when the Tigers lost the first three games of the World Series, Al Yellon wrote about what's happened to other teams that did that. Those numbers are dramatic, and of course Game 4 added yet another data point.
Just for fun, I tossed in the League Championship Series, too. Those numbers aren't as dramatic ... but they suggest the same thing: Teams play differently when one of them has a 3-0 lead in a best-of-seven postseason series.
There have now been 33 best-of-seven series in which one team won three games while the other team didn't win any (which includes a couple of tied games).
If all the teams were equally skilled, we would expect the winning teams to win around half of the next games, maybe 16 or 17 of 33. But let's instead assume -- because it's probably true -- that the teams that win three games were probably, you know, the better teams. In that case, we would expect the winning teams to win around 55 percent of the next games: around 18. Which is an overly precise prediction, so let's say 16-20 of 33.
They've won 27 of 33.
The results are even more striking in the World Series, with the leading team going 21-3 in the next game ... and then 3-0 in the game after that. All things being equal, we would expect the losing team to win 12 of the next games, and in six World Series to win another; that is, even with one team taking a 3-0 lead in the World Series, we would expect six of those World Series to last at least six games. Instead, none have.
But the results haven't been as lopsided in the League Championship Series. There have been nine best-of-seven LCS's wherein one team took the first three games. Six of those resulted in sweeps. But in two of them, the losing team won Games 4 and 5 before falling in Game 6. And in in one incredibly memorable series, the losing team somehow won the next four games to become (and remains) the only team in major-league history to overcome a 3-0 deficit in a best-of-seven series.
Maybe the stakes in the League Championship Series just aren't as high, which changes the psychological equation. Or maybe nine League Championship Series just aren't enough to tell us much. The more data the better, which is why I prefer to lump all best-of-seven series together.
And you gotta admit, 27-6 is pretty impressive.
Could that be chance? Yes, it could be. So could 28-5, or 33-0. But 27-6 is some standard deviations from what we would expect, and points me to the possibility that yes, for a few weeks in October momentum really does exist.
So for the moment, let's assume that it does. What would account for it? It's often said, after a team completes what seems an unlikely comeback, that the players were loose, figuring that they didn't have anything to lose. But of course most comebacks don't actually happen, because they are unlikely. In fact, we might guess that teams just one game from elimination -- and knowing they have to win not just one game, but four straight games, are the opposite of loose.
Conversely, it's also possible that teams with a 3-0 lead in a best-of-seven series are particularly loose. Nobody except maybe the starting pitcher feels much pressure; everybody else knows that even if they lose this one, they'll have plenty more chances to win.
Maybe there's something in one of Malcolm Gladwell's books about the psychology of losing; if so, I haven't been able to find it. But I wonder if there really is a difference between the regular season, when losing just means you'll fight again tomorrow, and Game 4 of a postseason series, when losing means you're dead.