April 29, 2010
Early Attendance Patterns
We're only 3
(Note: these numbers adjust for the different number of home games each team has played, by comparing each team's total against their total through this many games last year. See the B-Ref link in the next paragraph-that should explain it a lot better than I just did. Also, these figures are as of yesterday afternoon.)
Baseball Reference has a pretty cool tool that breaks all of this down, both by team and for MLB as a whole. Right now, 17 of the 30 teams are down on a per-game basis, and a number of clubs are hitting multi-year lows for April. Amazingly, the Indians-they of the 455 consecutive sellouts from 1995-2001-were last in baseball until just last night, when the Blue Jays managed to retake the bottom spot despite having the Red Sox in town. The Marlins, A's, and Pirates round out the bottom five, while the Mets, Marlins, Padres, Jays, and Rays have seen the biggest per-game declines.
The question is, how important are these numbers? If a team's attendance is down in April, does that mean they're doomed for the rest of the year? And how important are the other months, relative to April? Should MLB be worried?
The easiest way to look at it is to find the r-squared coefficient* between each team's April attendance growth (or decline), and its full-year result. Note that I'm using the teams' per-game averages as compared to their previous April, instead of their raw attendance total, since using the raw number would only tell us that yes, the Dodgers are likely to continue drawing more than the Marlins.
*(The r-squared coefficient tells you how good one thing is at predicting another. One means that they move together in perfect harmony, while zero means there is no connection.)
I took every team in the wild-card era (since we know that the expanded playoffs have had a significant impact on late-season attendance) and compared its April attendance trend to its full-season result. The resulting r-squared: .43. That's not particularly high or particularly low; it just kind of is. There's certainly plenty of predictive value, but it's far from a given. Looking at it another way, 74 percent of teams that saw greater attendance in April were up for the entire season, while only 28 percent of those that declined saw gains for the full year.
In other words: Yes, April is a pretty good forward indicator, but you're not totally out of hope if you don't hit the ground running.
How does that compare to other months? Here are the r-squareds when we look at each month independent of the others:
That's pretty clear: May through August are all centered around .6, while April and September are closer to .4. Part of that has to do with the fact that there are, on average, fewer games in April and September, depending on when the season begins and ends, and attendance is lower in those months anyway, due to the colder weather and school being in session. Taken together, they are simply weighted less, which helps explain the lower correlation.
(Also, while I haven't taken the time to actually prove this, I would guess that April attendance is largely based on pre-season expectations, and most teams have their fates sealed by September, so the middle months better represent the team's performance for most of that season. Nevertheless...)
So if your team's attendance is down, at what point does it go from being "unlikely" to "nearly impossible" for them to turn it around? Here are the probabilities, after each month, for a team with negative attendance growth to actually come out positive for the season:
The big inflection points actually come in June and September. June makes a bit of intuitive sense: teams that aren't playing well at the midway point tend to start trading away pieces and looking toward the following season. But I'm not sure I've totally figured out September-my best guess is simply that there are a bunch of teams right on the borderline going in, and they just run out of time when the season ends.
Regardless, for the teams that are lagging right now, I would take all of this as pretty encouraging news. Having a slow April certainly isn't good, but it hasn't been a death sentence, either. If nothing else, there's still plenty of time to turn it around, and a bunch of teams in that position have done so in recent years.
For MLB as a whole, one would think they're really going to be sweating it out through May and June. Remember something: just being flat from 2009 really isn't a very positive result, given how horrible last year was (down over 6 percent, all told). Many others and I had figured they'd be slightly up this year, given the improving economy and the Twins opening their new stadium. Having attendance fall again, on top of last year's brutal-but-expected losses, would be a really tough blow.
I still think they have a pretty good shot at turning it around, but honestly, that's just a gut feeling. Some of the teams that have declined from last season make sense (e.g. the Mets, coming off a disappointing year in a new stadium), but the Angels? The Mariners? Really? I don't think it's the weather, as we didn't have a single rainout until this past weekend, and the weather was actually pretty warm in places like New York, Minnesota, and Chicago for the first couple weeks of the season. Most teams also either cut or held ticket prices, so that shouldn't be a huge factor.
Whatever it is, it would behoove MLB to figure it out soon. April attendance isn't make or break, but May and June very well could be.