Pitcher Health and Team Wins
Pitcher attrition is on everyone’s mind after the Mariners committed the GDP of a small country to right-handed pitcher/demigod Felix Hernandez. Two articles today examined pitcher injury from different angles.
The first is a statistical look at risk factors for pitcher injuries. You’ve probably all heard of the “Verducci Effect” wherein increasing the innings for a young pitcher portends ineffectiveness or injury, but this hypothesis turns out to be (oft-repeated) bunk.* Instead, the biggest predictor of injury is, erm, *injury*. That is, a trip to the DL in year 1 makes a pitcher much, much more likely to see the DL again in year 2. That’s not exactly a mind-blowing conclusion, particularly for fans of a team that employed Erik Bedard and Franklin Gutierrez at the same time. But the extent of that correlation is much stronger than I would’ve guessed. 43.7% of pitchers who were on the DL in one season went on the DL in the following season – but among those hurlers who WEREN’T injured in one year, less than 5% landed on the DL the next year. Call it the Erik Bedard principle: pitcher injuries cluster over time.
The pattern extends beyond just one year, of course. If a pitcher had any injury in some season (even if it didn’t require a DL trip), 55.6% of them had another injury 2 seasons later. But if those who made it through the season unscathed, only 2.7% were hurt two years later. 2.7%! That’s incredible, given what we normally think about the random, merciless arm of fate, dispassionately doling out labrum tears and tendon snaps to workhorses and DL-regulars alike and putting Frank Jobe’s descendents through college in perpetuity. You can probably guess where I’m going with this: the fact that Felix has been so durable in recent years really does have an impact on his chances of staying healthy for a while. It’s not a guarantee, and no one is counting on him to be completely healthy for seven years, but it’s the kind of data the M’s may have looked at before deciding that a contract of this magnitude and length made sense. That, and the fact that Felix is amazing.
The flip side is that the M’s have a few question marks in their rotation behind El Cartelua. Joe Saunders, Erasmo Ramirez and Hisashi Iwakuma have all made DL trips in the past two years (OK, Iwakuma’s took place in Japan and thus wasn’t an MLB DL trip per se). Extending that to any injury, and not just DL trips, picks up Blake Beavan as well, making four out of five. That’s somewhat worrying, though the impact is probably not as severe for the M’s considering the state of their high-minors pitching depth and based on what four-fifths of the rotation cost to acquire.
The second piece is Dave’s great article at Fangraphs about the White Sox ability to outperform their PECOTA projection.** One big piece of the puzzle appears to be, once again, pitcher injuries. Looking at DL databases going back 10 years, the White Sox have been the best in baseball at keeping their pitchers healthy. This means that they’ve needed fewer replacement-level fill-ins, and it’s allowed them to better the projection systems’ estimate of their playing time. Dave’s back-of-the-envelope estimate is that this, coupled with pitching coach Don Cooper’s excellent work with his pitchers, has added 2-3 wins *per year* to the Sox. They haven’t always been a great – or even good – team, but it really is shocking how few legitimately bad pitchers have suited up for Chicago, and how perfectly Mariners it is that Seattle managed to allow one of them to throw a perfect game against them.
Keeping pitchers healthy, or at least healthier than the competition, has meant a lot to the White Sox. It’s also something the M’s front office talked about when Jack Zduriencik was hired in 2009. So how’ve they done? Fairly well, actually. The M’s lost three players to the DL in 2012 – Ramirez, Charlie Furbush and George Sherrill. Ramirez and Furbush were on for relatively minor things, and pitched effectively after returning. Sherrill played a grand total of 5 minutes in an M’s uniform, and, technically, his second stint in Seattle ended before spring training did. In 2011, the M’s had six members of the 40-man hit the DL, the same number they had in 2011. Sure, this doesn’t prove much, as they had Erik Bedard in 2010-11 and not in 2012, but they lost very few starts to the DL in 2012, and it’s possible that the organization’s coaching and strength/conditioning/flexibility group has had a hand in that. The cynic would argue that the M’s were simply lucky in 2012, and that regression to the mean (along with the injury history stuff I mentioned earlier) indicates that the M’s are likely to fare worse than they did last year. But as the White Sox example shows, you can’t simply assume that every team regresses towards league average. There is skill mixed in with luck,*** and good teams are working very hard to identify and apply the skill portion of this equation.
* – in fact, Carleton’s model showed that a large increase in batters faced was correlated with a *lower* risk of injury in the next two seasons.
** – For what it’s worth, the M’s PECOTA projection is the most optimistic I’ve seen, pegging them just barely under .500 and within the margin of error not only of the Athletics, but of the Rangers as well.
*** – Holy crap, Texas. Kansas City’s been famous for poor handling of pitchers, but seriously, Texas’ 10 year history is mind-blowing. I understand that Thoracic Outlet Syndrome could be credibly renamed Texas Rangers’ disease. Also, as good as the White Sox have been, they stumbled in 2012, losing six pitchers to the DL, including three starters – one for the year (Danks) and one twice (Floyd).