On Spring Training Game Pace

As the resident FanGraphs Game Pace Cares-A-Lot-About-er, I know it can be weird to talk about pace and watchability, because to prefer a faster pace could be interpreted as not really liking baseball that much. One could argue that a “true fan” cares only about the outs and the score and minds not the speed at which the conclusion is approached. But I know I can’t help what I care about, and I care about things like pace and duration. As such, I’ve gotten to wondering about how spring-training speed matches up with regular-season speed. Are the games this time of year longer? Are they shorter? Are they exactly the same? If there are differences, what might they mean?

So far in 2013 spring training, at this writing, 83 games have been completed. That’s not a whole lot of games, but that’s a meaningful sample of games, and it’s what I’ve examined. In every official MLB.com game box score, the length of the game is displayed toward the bottom. I’ve gone through and compiled all the data and performed some elementary math on it, as you do. As a potentially unwelcome spoiler alert, the results are not astonishing.

One thing to acknowledge first: this time of year, games don’t drag on forever. Games get called either before extra innings, or shortly into extra innings. In the regular season, of course, games go until they’re decided, which is a factor, but it’s a relatively small factor. Now, it was hard to track down the average regular-season game length in 2012, but it seems to have been right around 177 minutes, meaning it was just shy of three hours. This is more or less where it’s hovered for a little while. Some teams, naturally, took longer to play than others, but we care not about the individual data points, and only about the overall picture.

So far, through 83 spring-training games in February 2013, the average game length has been almost exactly 177 minutes. To be more precise, it’s been 176.494 minutes, meaning it’s virtually identical to last year’s average regular-season game length. So there’s our answer. Based on our potentially representative sample, spring-training baseball isn’t faster than regular-season baseball.

But! As Carson Cistulli pointed out earlier Thursday, you have to look at the run environment, and there have been plenty more runs scored on average this spring than last season. See, there’s a difference between game duration and game pace — given two games with identical lengths, one might have more packed into it. Last year, the average game featured 75.8 combined plate appearances. This spring, the average game has featured 78.4 combined plate appearances. The games haven’t been any faster, but the action has gotten along a little faster, to fit more into the same amount of minutes.

The difference is still small. We’re talking about a few seconds, on average, per plate appearance. On one hand, spring-training games don’t have to deal with major network commercial breaks of a mandated, extended length. On the other hand, spring-training games feature a hell of a lot more substitutions, especially this early on. Should those substitutions be made mid-inning, they serve as delays. This time of year there are fewer bells and whistles, but there’s still a lot happening.

I wasn’t sure if the reduced stakes would make people work faster. Because this is spring training, you can consider pretty much all events to have a leverage index of zero. Everyone acknowledges that the games don’t mean anything; people are just trying to get in their swings and their pitches. But, ultimately, habits are habits, and you don’t want to work too differently just because it’s still glorified practice. Maybe pitchers attempt slightly fewer pickoffs, and maybe batters step out of the box or call timeout slightly less often, but hitters with quirks will stick with their quirks, and slow pitchers will keep doing most of the things that cause them to work slowly. One doesn’t abandon his routines in the absence of importance. In every plate appearance, the pitchers and hitters are still trying. Trying now isn’t too unlike trying in May or September.

I did take a quick look at an individual case study, that being the dreadfully glacial Jonathan Papelbon. Here is a Papelbon chart plotting pace against game leverage for 2008-2012:

There is the slightest hint of slowing things down as things become more important. It isn’t pronounced. The other day, Papelbon threw a long and unsuccessful inning of relief against the Tigers. It was actually two-thirds of an inning, and he pitched in the top of the fifth of a game that was broadcast and archived on MLB.tv. I went through and monitored Papelbon’s pace, and I came out with a result of 25.1 seconds. For his career, during the PITCHf/x era, he’s had a pace of 30.9 seconds. So that’s a sign of quickening, but then, it’s based on one game, in February, and Papelbon pitched in the middle innings instead of his customary ninth. So I don’t know how much to read into it, but Papelbon still looked like Papelbon, in terms of process, and not results. He still took his long looks in. He still made his faces.

I suspect most people work just a tiny bit quicker in spring training than they do during the regular season. The games don’t fly by, though, nor should we expect them to. That would be a silly thing to expect, and I don’t know if you’ve learned anything from this exercise. I know I’ve learned that Jonathan Papelbon once had an appearance with a pace north of 45 seconds between pitches. I don’t really care that much for Jonathan Papelbon.