What You Need to Know:... (04/16)
the archives are now free.
25 comments have been left for this article. BP Comment Quick Links
| |
|
|
April 16, 2014 Overthinking ItDoes Baseball Have a Pace Problem?
On June 13, 2012, in a close but otherwise unmemorable game at Great American Ball Park between the Reds and Indians, Joey Votto and Derek Lowe reminded us what baseball is missing:
That’s Votto’s third-inning plate appearance, a six-pitch walk. The PA was notable not because of the outcome—Votto led the National League in walks that season, despite playing in only 111 games, so a free pass was predictable—but because of how little time it took and the way Votto approached each offering. The slugger barely adjusted his stance or his position in the batter’s box from pitch to pitch, moving his feet only when Lowe came inside on 2-2. He showed no compulsion to do a lap around the plate, no need to adjust his batting gloves. And somehow, he survived. Lowe was one of the game’s quickest workers in 2012. We know that not just from watching him, but by virtue of PITCHf/x, which began to tag each pitch with a timestamp in 2010. By comparing each pitch’s timestamp to the previous one—and using only consecutive pitches, so that the time between isn’t counted if a pick-off throw breaks up the action—we can determine the time it took to deliver. A stopwatch would serve the same purpose for a single pitch or plate appearance, but a database allows us to dig deeper, calculating the average pace for any pitcher or hitter in a given season, as well as for the league as a whole. Lowe’s average time between pitches in 2012 was 16.92 seconds. The league’s as a whole was 21.05, which means that a 100-pitch start by Lowe that season would have included almost seven minutes less standing around than one by a typical pitcher. Votto ranked toward the middle of the pack in 2012 time between pitches, so in this case, he was probably just trying to disrupt Lowe’s timing by matching (or exceeding) the opposing starter’s pace for a single plate appearance. In doing so, he gave us a glimpse of the batter-pitcher confrontation stripped down to the essentials, without any of the filler that normally gives us plenty of time to stare at our seconds screens during breaks in the on-field action. Unfortunately, there’s more of that filler than there was even 3-4 years ago ago. Inter-pitch pace took a sizeable hit between 2011 and 2012 and has slowed further in the season-plus since.
The average time between pitches last season was 0.85 seconds longer than it was in 2010, the first season for which we have the same data. That doesn’t sound like much, and from inning to inning, it isn’t, but consider how many pitches are thrown in a given season: 709, 917 in the 2013 regular season alone, according to Major League Baseball Advanced Media. Distribute those over 2431 regular-season games, and you get 292.03 pitches per game. If the average time between pitches last season (and so far in 2014) was 0.85 seconds longer than it was in 2010, we’re talking about over four minutes extra per game, relative to 2010, for the same amount of action (unless this counts as action). If you watch all of your team’s games, you’re looking at a commitment of over 11 extra hours per season, for which you receive nothing but the baseball equivalent of dead air. As the chart above indicates, the average length of games did increase by almost 10 minutes from 2010 to 2013, and it’s up another three minutes so far this season, likely thanks to expanded instant replay review. Even before the new replay rules, April game lengths and times between pitches tended to be very slightly longer than they were over the course of a full season, but not by so much that we should worry much about significantly overestimating the recent spike. We should acknowledge that the increase in time between pitches isn’t necessarily attributable to lollygagging alone. It could be a result of some confounding factor that has also an effect on pace. For example, Dave Allen discovered a few years ago that pace tends to slow down A) in two-strike counts, and B) when a reliever is pitching (even after accounting for the fact that relievers are more likely to pitch in two-strike counts). He also found that the presence of runners on base slows things down considerably. Those observations hold true today, as the following chart of time between pitches broken down by pitcher role and base state (empty or men on) reveals.
Similarly, Jeremy Greenhouse found that pace tends to slow down as the hitter sees more pitches in a plate appearance. So can counts and relief usage explain baseball’s slowing pace? In the table below, “RP%” is the percentage of pitches thrown by relievers, and “Count Depth” is the average pitch number in a plate appearance (an 0-0 pitch being “1,” a 1-1 pitch being “3,” a 2-2 pitching being “5,” and so on).
Count depth hasn’t increased, and the rise in reliever percentage is slight—probably not enough to account for the added time between pitches, especially given that league-wide on-base percentage has fallen by 7-8 points since 2010, which should translate to more faster-paced, “bases empty” situations. Whatever the reason for the slowdowns, both between pitches and overall, it’s something that the next commissioner should probably pay more attention to. Not for the sake of hopelessly addicted dependents like you and me, necessarily, since we’d still watch baseball if the games took twice as long, but because blowing by the three-hour mark has to make the sport a harder sell to undecided viewers who are browsing for the best experience. It would be one thing if more time meant more action, but we’re still doling out 27 outs over nine innings—we’re just making it harder to get to the good stuff. While it’s true that there are more pitches thrown today than ever before, there are also more pitching changes, more commercials, and more vamping between pitches. For a variety of monetary and strategic reasons, it’s tough to target the first two sources of slow pace, which leaves regulating pitcher pace as the likeliest—if least effective—solution. As long as the broadcast bubble refuses to burst, baseball’s financial future seems safe, but even precious event programming might lose some luster with advertisers and potential broadcast partners if the audience shrinks. Forcing players to be a little more like Lowe and Votto by observing existing rues (for real, this time) might help turn the tide. And that way, we won't have to talk abut lopping off a couple innings. Before we go: Because we need to know whom to credit/blame for saving/costing us precious time that we could be using to watch other baseball games, here are two tables listing the 10 slowest and 10 fastest hitters and pitchers from 2013 (min. 1000 consecutive pitches faced for hitters and 500 thrown for pitchers). You’ll notice that there’s a wider range between the fastest and slowest pitchers than there is between the fastest and slowest hitters, which reflects the fact that pitchers—the guys who get to hold the ball—have more control over pace. (Unsurprisingly, the pitcher leaders are all starters, and the pitcher trailers are all relievers.) We’ll make all of this data available via our sortable stat reports soon. *Update* Make that right now, for both batters and pitchers. Hitters
Pitchers
*Update* By reader request, the slowest starters and the fastest relievers from 2013:
Slowest Starters
Fastest Relievers
Thanks to Rob McQuown and Harry Pavlidis for research assistance.
Ben Lindbergh is an author of Baseball Prospectus. Follow @benlindbergh
|
Deep Impact: Week Two
At least regarding hitters, I think you need to control for men on/not on base. For instance, how much of Philips' sluggishness is due to Votto so often being on ahead of him?
Yes, you'd need to try to control for other factors if you wanted to completely isolate the player's context-neutral slowness.