FanGraphs’ Jeff Sullivan was wondering about something really insightful, as he so often does: how telling is a team’s 50-game record, really?
In other words, as we approach the 50-game mark in baseball, how well correlated are the results in those first 50 games with the 112 games that follow? Do we know what we know about a team through 50 games, or do we know very little?
In a strongly-encouraged read, Sullivan finds out that a team’s first 50-game record does bear some relationship to the team’s record in the following 112 games, but it’s not all that strong. Much more notably, the team’s 112-game record after the first 50 bears a much stronger correlation with the team’s projected record from the preseason.
Isn’t that pretty incredible, when you think about it? Through 50 games, actual stuff has happened. Guys have been injured. Breakouts and slumps have kicked in. We know a lot about a team! Or, at least, we think we do. Whereas preseason projections have none of that. They have only data. And yet, even after 50 games have already occurred, the preseason projections do a much better job of predicting those latter 112 games than the team’s record in the first 50.
Ideally, Sullivan notes, you’d blend the 50-game results AND the preseason projections to best predict the team’s record in the final 112 games, but he also points out that to do that, you’d have to weight the preseason projections six times the 50-game record (that’s how much more predictive it is! wow).
This should all be fairly heartening to Cubs fans. After all, the team’s preseason projections called for a winning percentage near .600, not the modest .523 winning percentage they’ve accumulated through 44 games. Statistically and historically speaking, you’d expect the Cubs’ games the rest of the way to be closer to the preseason projected winning percentage than what they’ve done so far.
Let’s be really clear, though: projections are always imperfect. They are a best statistical guess of sorts, and they cannot contemplate a total annihilation of a pitching staff by injuries, or a complete offensive crash by a previously productive star. Nothing in this analysis means for certain that the Cubs will suddenly start winning many more games.
But the analysis does suggest that, living in the moment, especially at the start of a season, fans may put too much stock in the first chuck of results they see.
(Note, though, that this is all about a team’s record in the ensuing games. It does NOT address the problem of banked losses, which is my greatest concern for this Cubs team – even if a team’s projections say they should play at a .600 clip in the final 112 games and they do in fact do so, if they’ve already banked 30 losses in the first 50 games, they’re going to be in trouble regardless.)