Lately - especially in light of yesterday's performance - there's been a lot of talk about Daisuke Matsuzaka's ability to strand runners and get out of jams. To carry at 2.90 ERA with those kinds of peripherals is no small feat, and his .225 BABIP with men on base is something to behold.
At first glance it just seems like he's getting lucky, but Sox fans and Matsuzaka supporters will tell you that there's something else going on besides luck, that Matsuzaka really is in control of those at bats where he has to buckle down. What they'll tell you is that Matsuzaka throws so many pitches, and that he's so good at mixing them around, that he keeps batters completely off balance, thereby preventing them from having the kind of success against him you'd expect. And that's his secret. That's how he defeats DIPS theory. Pitch variation. He doesn't always, say, throw fastballs in fastball counts, and that's what allows him to have so much success.
People will tell you this with such conviction that you almost want to believe they're right. Me, though, I'm just not seeing it.
There are two things, I think that work against that argument:
(1) little proof that pitch variation helps with BABIP
(2) little proof that pitch variation helps Daisuke Matsuzaka with BABIP
As far as #1 is concerned, I played around with Fangraphs and pulled out pitch information for 337 individual pitcher seasons since 2005. Once exported to Excel, I looked at every pitcher and marked down how many pitches he threw at least 2% of the time. I then grouped them by the resulting number and looked to see if this had any kind of effect on BABIP.
| # of pitches | BABIP | # of seasons |
| 2 | .307 | 1 |
| 3 | .299 | 91 |
| 4 | .300 | 154 |
| 5 | .300 | 80 |
| 6 | .309 | 11 |
There's no significant difference in there. The majority of pitchers work with between 3-5 pitches, and regardless of which way they go, in the end it gets them the same results, at least as far as BABIP is concerned. This isn't a perfect way to run this study (we don't have enough PITCHf/x information to go on), but it should get us close, and so generally speaking, this doesn't look like a satisfactory explanation for why Matsuzaka has had so much success. I'm reminded of Padre Josh Banks, who's known for his eight-pitch repertoire. He posted a league-average BABIP.
As far as #2 is concerned - and I think this is the bigger point - if Matsuzaka's pitch variation helps him so much with runners on base, why doesn't it help him when the bases are empty? This year he posted a .289 BABIP with the bases empty and a .225 BABIP with runners on. Last year it was .322 and .272. Overall his BABIP over two seasons is .289, which is in the ballpark of where you'd look for an average pitcher to be. So where's the magic? You can say that he buckles down when he gets in a jam, but what does that really mean? If he has the ability to induce weaker contact when he has to, why doesn't he just use that ability all the time, instead of waiting until he gets himself in trouble? It doesn't make sense.
Daisuke Matsuzaka may very well pitch better with men on base than he does with the bases empty. Over two seasons, his K/BB in the former situation is 2.26 versus 1.92 in the latter situation, so who knows, maybe that's meaningful. Maybe the pitch variation really does help him keep some sort of an edge over the hitter. But I just don't see any reason to believe that he has some fantastic ability to get out of trouble by generating weaker balls in play. The only indication that he does is two seasons' worth of splits, and two seasons of splits can tell you pretty much anything.
Daisuke Matsuzaka is a good pitcher. He's a legitimate #2 who's going to help the Red Sox for a long long time. But his performance with men on base isn't sustainable. That ERA is going to get the living crap regressed out of it.