Can Karstens do it again?

By John Franco

Last season, Jeff Karstens posted an impressive 3.38 ERA in 162.1 innings for the Pirates. He only earned a 9-9 record for his efforts, but his ERA was just outside the National League’s top 10 (Madison Bumgarner finished 10th with a 3.21 ERA).This year, Karstens has posted a 2.08 ERA in 13 spring innings as he continues to inspire hope for a repeat performance the Pirates’ shaky rotation desperately needs.

Karstens’ 2011 season looks less impressive when viewed through advanced metrics – he posted a 4.29 FIP and 4.00 xFIP, two metrics that attempt to show what a pitchers’ ERA should have been. He stranded 77% of his baserunners (the league average is around 70%) and opposing batters hit just .275 on balls in play (the league average is around .300). Put simply – Karstens pitched reasonably well, but he got lucky.

Of course, the reverse was true in 2010 – Karstens posted a 4.92 ERA, but his FIP was 4.82 and his xFIP was a much more reasonable 4.26. Despite an actual improvement of more than 1.5 runs per 9 innings between seasons, Karstens’ expected performance improved by only one-quarter of a run. Advanced metrics aren’t always perfect, and there are pitchers who consistently outperform their xFIP (poster boy: Matt Cain) and pitchers who consistently underperform their xFIP (poster boy: Ricky Nolasco). After two full seasons with the Pirates, Karstens has one year on each side of the fence, so it’s hard to predict where he will fall in 2012.

Looking at Karstens’ performance in 2010 and 2011, his underlying numbers show why his xFIP looked about the same in both years:

Other than a slight improvement in his ground ball rate, Karstens was essentially the same pitcher. Even his arsenal was basically the same (Pitch/FX data courtesy of Fangraphs):

FanGraphs also lists a cutter in Karstens’ arsenal – thrown 1.5% of the time in 2011, but not used in 2010. I found a couple mentions of Karstens throwing an “occasional cutter” including this article from earlier in March. It doesn’t appear to be a focus of Karstens’ efforts, but when your fastball averages 89 miles per hour you have to be willing to try anything. Otherwise, Karstens threw a few less breaking balls and added a few more changeups, but didn’t make any significant changes.

If Karstens didn’t change his arsenal, and didn’t have a noticeable change in his strikeout and walk numbers, does that mean he’s due for a regression back to his ugly 2010 numbers? The last time I looked at Karstens, I pointed out that 20 of his 22 home runs allowed in 2011 were solo shots, and an average number of runs allowed per homer would have inflated his ERA to 3.88. Karstens does benefit from an extremely low walk rate, so he might keep his R/HR rate below average, but 20 out of 22 probably isn’t happening again.

One way to predict what Karstens’ future might hold is to look at similar pitchers. His combination of K/9, BB/9 and GB% rates was fairly unique among MLB pitchers last year. The closest matches were Kyle Lohse (5.3 K/9, 2.0 BB/9, GB% of 41.4) and Mark Buehrle (4.8 K/9, 1.97 BB/9, GB% of 44.9, but a lefty). Buehrle posted a BABIP of  .294, and Lohse posted a stellar .269 mark. In 2010, Doug Fister posted numbers similar to Karstens’ 2011 numbers (4.89 K/9, 1.68 BB/9, GB% of 47.1) and ended up with a .302 BABIP.

So an anecdotal study of pitchers similar to Karstens shows that he might be able to sustain his luck in terms of BABIP, or at least not regress all the way to the .309 mark that he posted in 2010. Replacing Ronny Cedeno with Clint Barmes will certainly help Karstens, as will the continued improvement of Neil Walker at second base. Garrett Jones and Pedro Alvarez won’t do Karstens many favors at the corners, but the Pirates should have a strong defensive outfield. For a more analytical approach, Fangraphs offers a formula for calculating expected BABIP. Using that formula, we can see some amazing consistency from Karstens:

Defensive improvements aside, a few hits one way or the other won’t have as big an impact on Karstens’ season as his ability to keep the ball in the park. He allowed home runs on 12.5% of his fly balls in 2010, then cut that rate to 11.5% in 2011 while also allowing less fly balls. That dropped his HR/9 from an ugly 1.54 to a workable 1.25. He also cut his HR/AB rate against lefties from 4.61% in 2010 to 4.27% in 2011. While that may not seem like much, it’s representative of a larger success: Karstens went from having a 400-point difference in OPS against lefties to a 200-point difference.

I’ve thrown a lot of numbers around in this post, so let me sum things up as simply as I can. Karstens was fundamentally similar in 2010 (when he was unlucky) and 2011 (when he was lucky). In spite of the similar fundamentals, the subtle changes he did make (a few more ground balls, adding a few cutters, handling lefties better) are all positive. None of the projection systems listed on Fangraphs project him to have an ERA below 4.00, which is probably reasonable given all the possible avenues for regression to creep in. Nonetheless, I think Karstens can fool them all again and finish in the 3.70-3.90 range.  If Erik Bedard and AJ Burnett can pitch well enough that Karstens is the Pirates’ 3rd best pitcher with an ERA in that range, I think the Pirates will be happy with their starting rotation.