Prospectus Feature: Th... (05/10)
5 comments have been left for this article. BP Comment Quick Links
| |
|
|
May 12, 2016 Prospectus FeatureRISPy BusinessIf you spend a lot of time behind the wheel, you might have satellite radio. If you spend a lot of time behind the wheel and you’re a baseball fan, you might have SiriusXM so you can listen to home radio broadcasts of games. (This is not a commercial. This is just a statement of facts.) If you spend a lot of time behind the wheel and you’re a baseball fan, though, you can’t listen to games all day, for the simple reason that baseball is not played around the clock. So when there isn’t a game on, you might listen to MLB Network Radio, a SiriusXM station.
I sometimes spend a lot of time behind the wheel, and when there isn’t a game on, I often listen to MLB Network Radio. I like some of the shows better than others. Some have strong elements of sportstalk radio, and like all sportstalk radio, you sometimes hear things that are, well, interesting.
A while ago one of the hosts—not a caller—was talking about Rougned Odor. Odor has had a pretty good start to his season. He’s also been better with runners in scoring position (.360/.407/.600 through Monday) than not (.283/.309/.528). The host said that some batters are consistently better with runners in scoring position (henceforth RISP) than they are otherwise.
Now, I don’t want to get into a fight over clutch hitting. Much smarter people have debated this. Some aren’t buying it. Others think there might be something there. It’s easy to get lost in the fog. So I’m going to sidestep the whole clutch thing and instead look at the simple question: Are some players better at hitting with runners in scoring position?
To check, I looked at every batter with more than 400 plate appearances (a round number that’s convenient not only because it implies a more-or-less regular player, but also because it gives a nice sample size of 200 or so players per season) for every year in the 30-team era, 1998-2015. That gave me 3,831 player-seasons. For each, I checked whether a batter was better than average at batting with RISP.
I had to figure out a batting metric to use. With RISP, the goal is to drive the runner home. That reduces the value of walks and hit batters (unless the bases are loaded) relative to base hits. Extra-base hits drive in virtually all runners, as do singles with a runner on third, while runners on second scored on 59 percent of singles last year. To synthesize all of this, as well as runners scoring on outs (sacrifice flies and groundballs with the infield back), I developed RISP-specific run expectancies and correlated them to familiar measures of batting (batting average, on base percentage, slugging percentage, and OPS). The correlation with slugging percentage was higher than for any other metric, so I checked whether batters have a better slugging percentage with RISP than they do in other situations.
Then, to see whether batters were, in fact, consistently better with RISP than not, I paired consecutive seasons. For example, in 2014, Mike Trout had a .615 slugging percentage with RISP. The major-league average was .386. In 2015, he slugged .693 with RISP compared to an average of .405. So he gets credit for repeating.
Before I show you the data, let’s figure out the ground rules. If there are players who are consistently better with RISP than not, we should see a majority repeat year to year. If it’s a 50/50 shot, like a coin flip, we should see about half of them repeat. If there’s a bias against repeating, less than half will repeat. So how many of the 400+ plate appearance batters who are better than average with runners in scoring position repeat? Quite a few, it turns out:
But wait. This isn’t really right, is it? It’s comparing batters with 400 or plate appearances in a season—by definition, those good enough to play in most of a team’s game—with all batters. It includes the batting performance of Trout and Bryce Harper and Manny Machado, but it doesn’t include the batting performance of Mike Zunino and Pedro Florimon and Jon Lester. That’s a significant omission. Last season, as noted, major-league hitters slugged .405 with RISP. But the 211 players who amassed 400 or more plate appearances slugged .431 with RISP. That’s a big difference. So let’s do an apples-to-apples comparison of slugging percentage with RISP for hitters with 400 or more plate appearances in a season with the overall average slugging percentage with RISP for that group. That’s a different picture:
Hmm. That’s closer to the 50/50 coin flip probability of repeating—well above it some years, but close to or even below it in others. So maybe there aren’t batters who consistently do better with RISP?
Well, maybe. But let’s look at it another way. The guy on the radio didn’t say that there were batters who were consistently above average at batting with RISP. He said there were batters who were consistently better with RISP. Last season, for example, Mark Teixeira slugged a .509 with RISP. That’s good! But had a higher slugging percentage when there weren’t RISPs. On another extreme, his teammate, Didi Gregorius,slugged .361 with RISP, but that was higher than his slugging percentage without RISP. So, do some hitters consistently do better with RISP, regardless of their baseline? Here are the data:
As with the prior tables, there’s been a blip the past couple seasons, but overall, there doesn’t appear to be a trend here. It’s looking a lot like a coin flip. When a player posts a slugging percentage with RISP higher than his slugging percentage without RISP in a season, it’s just about 50/50 whether he repeats the next season. That’s more indicative of randomness than a trend.
Well, maybe not. Let’s give it one last try. True, the percentages are all pretty close to the coin flip result of 50 percent. But they're not exactly 50 percent, and some are higher. So maybe there are just a few players who really do exceed by posting slugging percenages with RISP consistently above their overall averages, but they get lost in the randomness for everybody else.
To check for this, I looked at every batter from 1998-2015 who had at least six consecutive seasons with at least 400 plate appearances. In most cases, those were consecutive years as well, but in some cases they weren’t. Barry Bonds had 400 or more plate appearances every year from 1998 to 2004, but was hurt most of 2005, then returned in 2006 and 2007. In cases like this, I ignored the missing season. So for Bonds, I counted four six-year streaks: 1998-2003, 1999-2004, 2000-2006, and 2001-2007, skipping his 14-game 2005 season in the latter two. I wound up with 1,012 six-year periods. Then I counted the number of times in each period the batter’s slugging percentage with RISP exceeded his overall slugging percentage. The result could be anything from zero to six times, of course. Here’s how it looks:
This seems to lend some credence to the idea that there are a few hitters who really are consistently better with RISP. And there’s a handful that are consistently bad. Most hitters fall in the middle, but some are at the extremes. So maybe there is something there!
But wait. Look at that graph long enough, and you may have a flashback to a probability and statistics course you took in college. That distribution looks something like…What was it again?...Oh yeah.
That’s it. You flip a coin six times, it’ll come up heads six or zero times 1.6 percent of the time, five or one time 9.4 percent of the time…I’ll stop. Do it 1,012 times, and you’ll end up with the red bars above. Yes, there were 21 six-year streaks during which a player slugged higher with RISP than not every year. But that’s almost exactly what we’d expect from a 50/50, coin flip type of environment. In fact, the shape of the graph, with the actual yellow bars longer than the theoretical red bars at on the left side of the graph, representing batters who fail to out-slug with RISP, more often than not, suggests that having a higher slugging percentage with RISP than in other at bats is not only a coin flip, but a coin flip using a biased coin!
Let’s look at those extremes. The 21 streaks of six straight years of slugging percentage with RISP exceeding slugging percentage without RISP (recalling that I’m omitting seasons of fewer than 400 plate appearances) were made by 16 players, some of them pretty unlikely candidates:
The flip side, here are the 12 players who account for the 21 streaks of six straight years with a worse slugging percentage with RISP:
Keep in mind that we’re comparing players’ slugging percentages with runners in scoring position to their slugging percentage in other situations. In some cases on the first list, the slugging percentages were still pretty bad. (Grissom slugged .362 with RISP in 2000, the third year of his streak.) In the second list, some of them are good. (Ortiz slugged over .500 with RISP over the last four years of his streak, but that was still below his slugging percentage without RISP.) But the idea that there are players who are consistently better with RISP than without—well, the evidence just isn’t there.
And the player that got me going on this? The 2016 RISP hero cited on the radio, Rougned Odor, off to a good start in 2016 with RISP, batted .245/.299/.441 with RISP in 2015 and .265/.322/.472 otherwise. Tails.
Rob Mains is an author of Baseball Prospectus. Follow @Cran_Boy
|
The Stash List: Sixth Edition, 2016
There's also one other hidden reason why batting with RISP would be generally higher than non-RISP overall: bias in the skill of the pitcher faced. In theory, a pitcher who has allowed a RISP isn't pitching as well as a pitcher without a RISP, almost by definition. In an aggregate sense, when a batter has a RISP during his AB, he'll be facing a pitcher whose "stuff" isn't as strong, at least on a micro-level (that game, or even just that inning). Put another way, the RISP has revealed some level of skill about the pitcher, and all else equal, you'd want to face pitchers with RISP because they're likely not pitching quite as well.
I'm intentionally ignoring the idea that a reliever could be brought in to face the new batter - the reliever is supposedly an inferior pitcher (that's why he's not starting), but he is being brought in because he's supposed to be better than the original pitcher at that time (that's why the first guy was pulled). It's not clear which direction this would push the RISP argument, but my completely idle, non-scientific theory would be that it would still be a net boost for batters' RISP stats.