Tuesday, Justin Verlander and the Tigers took on the Cardinals in Detroit. Verlander faced Matt Holliday with the bases empty in the first inning, and threw him three fastballs at 92 miles per hour. Verlander faced Holliday again with the bases empty in the fourth inning, and threw him a fastball at 92 miles per hour. Verlander faced Holliday with runners on the corners in the sixth inning, and threw him fastballs at 97, 97, and 99 miles per hour. Verlander faced Holliday with the bases loaded in the seventh inning, and threw him fastballs at 98, 99, 100, and 100 miles per hour.
Some pitchers will say that they don't bother pitching to the situation -- that they're always just focused on retiring the batter at the plate. The evidence above suggests that Verlander was pitching to the situation on Tuesday. So does the evidence below.
From ESPN Stats/Info: Verlander's average fastball was 4.7 mph faster when there were runners on base as compared to bases empty Tuesday.
— Buster Olney (@Buster_ESPN) June 20, 2012
By diving into the PITCHf/x, I'm getting a difference of 3.2 miles per hour, not 4.7 miles per hour. That's smaller. But that's still seemingly significant, and something to investigate.
I got to wondering whether Verlander has been doing this all season long. That's easy enough to check, given access to a PITCHf/x database. I found that in 2012, Verlander's average fastball has clocked in at 93.4 miles per hour with the bases empty. When there's been at least one runner on, though, that average jumps to 95.7 miles per hour, for a difference of 2.3 miles per hour. Without performing any tricks, that sure as hell strikes me as being statistically significant.
So how long has Verlander been doing this? Has Verlander always been doing this? I can go only as far back as 2009, as I don't think PITCHf/x tracked baserunner states in 2008, but here's the data:
Year | Empty | On | Difference |
2009 | 94.9 | 96.6 | 1.7 |
2010 | 94.9 | 96.2 | 1.3 |
2011 | 94.4 | 96.0 | 1.6 |
2012 | 93.4 | 95.7 | 2.3 |
In all four years, we see a considerably faster average fastball with a runner or runners on base. The biggest separation has been in this season, which could be chance, or evidence that Verlander is learning how to pitch even better, or something else. We'll have to re-visit the 2012 data when the 2012 season is over.
All right, so we've established that Verlander's fastball has gone up with men on base. That doesn't do us any good if we don't know what other pitchers do. I can't easily check every pitcher in baseball, but I selected a few guys with top fastballs and looked at their 2012 separations. In no particular order:
Stephen Strasburg: -1.3 miles per hour with runner(s) on
Clayton Kershaw: +0.6
David Price: +1.0
Max Scherzer: +0.9
Josh Johnson: -0.2
How about whole teams?
Cincinnati Reds, June: +0.6 miles per hour with runner(s) on
Detroit Tigers, June: +0.3
This isn't a complete check, but Verlander does seem to ramp up his fastball more than average in run-scoring situations. Perhaps well more than average, as we haven't established a real league-average. It's possible that Verlander has the widest velocity separation in baseball, but I'm not going to declare any conclusions on that.
So we've established that Verlander throws a harder average fastball when there's at least one runner on base. The important next question, then, is: Does it matter? Intuitively, you'd think that a harder fastball would be a better fastball, and that Verlander would have good success with men on. The 2009-2012 numbers might surprise you.
Since 2009, Verlander's allowed a .211 batting average and a .323 slugging percentage with the bases empty, against .233 and .359 with runners on. If you care about batting average on balls in play -- BABIP -- Verlander's come in at .269 and .298, respectively. His strikeout rate has dipped just a touch, from 25.7 percent to 25.3 percent, but his walk rate has climbed, although we'd expect that as Verlander would presumably be pitching more carefully.
One expects that pitchers will perform worse with men on than with the bases empty, but Verlander's numbers don't suggest a hotter fastball. They suggest that he's just been pitching like himself. Considering that Verlander has shown his biggest velocity gap in 2012, there might be something there if you squint. This year, his strikeout rate has been a little higher with men on. This year, his isolated slugging percentage has been lower with men on. But the samples are still too small, and Verlander's BABIP has climbed.
Over the past four years, Justin Verlander has saved many of his fastest fastballs for more dangerous situations. Many people would commend him for that, saying that he knows how to pace himself. Somewhat surprisingly, there's only very limited evidence that this has worked to his benefit. It might be that Verlander has sacrificed too much command when increasing his speed. It might be that Verlander's fastball is already super effective in the low- to mid-90s. Whatever the case, we've confirmed one thing, but we haven't been able to confirm that it really matters.