A thought or two about Carlos Marmol

I didn’t make it a secret that I thought the Cubs 3-year extension given to Carlos Marmol last offseason wasn’t a wise decision. Despite that, I made it clear that Marmol was not being paid like an elite reliever and was worth the contract at the time it was signed. This was true even if you consider the discount that arbitration eligible players receive. A second year eligible player is paid roughly 60% of his free agent value and then it increases to 80% his final year of eligibility. Marmol’s contract included those two year and one additional year. It was not a bad contract.

It was, however, an unnecessary one. The Cubs were unlikely to contend in 2011. For that matter, they were also unlikely to contend in 2012 and maybe even 2013. The team also had club control of Marmol through 2012. While teams can and do get discounts when offering 3-year contracts, I was skeptical that it was a good idea. First, they had him for two more years. There was significantly less risk in going year to year with him the next couple of seasons. For a team that wasn’t likely to contend, saving money seemed like the best option. Plus, there’s the unreliability of Carlos Marmol. His walk rate is historically awful. He has a great strikeout rate and as a result his K/BB or the better K-BB% is still good. He had survived on low BABIP’s and a remarkably low HR/FB rate. Neither of these were likely to continue.

I know some Cubs fans thought there was something special about Marmol. Batters made such weak contact against him that he could continue to have BABIP’s in the .250s. He was so difficult to make solid contact against there was no reason he couldn’t sustain a ridiculously low HR/FB rate. That’s what they said.

There may be some truth to that, but last season Marmol’s BABIP was about league average. After posting BABIP’s of .259, .169, and .253 in 2007 through 2009, his BABIP jumped to .293 a year ago and .295 in 2011. Even if he was a pitcher who could be expected to produce weak contact, there was little to no reason to think he could sustain the absurdly low BABIP he had from 2007-2009. We know players regress toward the league average. We had some reason to believe that Marmol’s true talent level BABIP was lower than league average, but nowhere near as much as it had been.

The same was true with his HR/FB rate. In 2007 it was 3.9%. Then it jumped to 9.9%, which is just under league average. Then it fell all the way to 2.6% and 1.6% in 2009 and 2010. This year it was 7.5%. We expected it would have been higher than it ended up being and we can expect it will be higher moving forward. There’s reason to believe his true talent HR/FB rate is not the league average of 10.5%, but it’s not 1.6% like it was in 2010 and it’s not 2.6% like it was in 2008 either. It’s not even as low as the 7.5% he posted this year. Realistically it’s much closer to the 9.9% figure in 2008 than it is to this past season.

Marmol’s numbers have been all over the charts. His ERA in 2007 was 1.43, which was in large part thanks to the typically low BABIP of .259, a low 3.9% HR/FB rate, but also an incredible 91.0% left on base percentage. HIs ERA jumped to 2.68 in 2008 and then again to 3.41 in 2009. In 2010 it fell to 2.55 before jumping way up to 4.01 this season. Marmol had an ERA much lower than FIP from 2007 through 2009. These years are the ones in which he had very low BABIP’s and even lower HR/FB rates. So it’s not surprising. But in 2010 that changed. His BABIP jumped and his FIP was half a run lower than his ERA. This was largely because of the silly 1.6% HR/FB rate. This season his ERA was again, half a run higher than his FIP. Considering the small samples for a reliever in any single season, or for that matter the span of several seasons for a reliever, that’s not too surprising, but I noticed it awhile back.

Marmol’s fWAR was 1.7 and 1.3 in 2007 and 2008. It fell to .6 the following year before jumping way up to 3.0 a year ago and back down to .8 this year. In 2007 and 2007 Marmol was consistently good. Even great from a runs allowed perspective, but just really good from a DIPS (Defensive Independent Pitching Statistics) perspective. The difference between those two year and the last 3 is the walk rate. In 2007 and 2008 it was only a below average 4.54 and 4.23. In 2008 it jumped to nearly 8 per 9 (7.91). It obviously fell the following season, but only to 6.0. It almost certainly had to fall again this year, but only to 5.8.

Despite pitching in lower leverage in 2007 and 2008, he was worth nearly as much in fWAR (3.0) as he’s been since (4.5). The walk rate went from passable to ridiculously bad. His ERA and FIP jumped with it. I don’t know if any pitch f/x guys have looked at Marmol or not, but it seems to me something changed considerably after the 2008 season. As a starter in 2006 it was 6.9 per 9, but then back in the Iowa rotation in early 2007 it was a very good 2.63. Nobody ever expected Marmol to have even average control. His minor league numbers gave little indication that was possible, but I don’t think anyone thought they’d be quite as bad as they’ve been since 2009. I wonder what, if anything, has changed.

He’s throwing a similar number of strikes. 61.5% of his pitches in 2007 were strikes followed by 61.6%, 59.0%, 61.3% and 61.1%. Aside from 2009 when he walked almost 8 per 9, nothing stands out there. If he matched his career high in strike percentage each season (61.56%) he’d have thrown 37 more strikes in 2009. That’s significant, but we already knew that considering his 59% strike percentage. However, it’s only 10 additional strikes in 2010 and 2011 combined.

In 2007 and 2008, Marmol posted a combined 2.13 ERA and allowed 2.36 runs per 9. Since then he’s posted a 3.31 ERA and allowed 3.39 runs per 9. Despite the walks playing a significant role in this and being a significant part of FIP, his FIP is a bit lower (2.92) than it was in 2007 through 2008. This is in large part the result of his crazy low HR/FB rate. There is literally no reason whatsoever to think that Marmol has pitched better the last 3 years than the 2 years prior to that. It would be foolish to suggest it.

I don’t know what to think about Marmol’s future and I’m not sure anyone else does either. He’s certain not headed in the right direction at the moment, but if he could just somehow figure out how to lower that walk rate back to where it was in 2007 and 2008, he’d be an outstanding reliever. There may be some explanation as to the increase in his walk rate in the Plate Discipline numbers Fangraphs publishes, but not nearly enough to make up the entire difference. It’s a lot easier to say how good he could be if his walk rate could drop to 2007-2008 levels, but it’s also not likely. But it is the one thing that stands in his way between being a good reliever to being a great one. It’s just too bad that he’d have to improve so much to realize that potential.

Going forward, the Cubs are probably better off using Sean Marshall as the closer. Marmol may have some trade value, though the Cubs wouldn’t get much in return. Over the last 2-3 years, Marshall has been the better reliever. He’s cheaper, more reliable and quite probably the superior pitcher at this point in their careers.

I doubt the Cubs give up on Marmol though. I think we’ll see him for at least one more year. He has the potential to be great and the potential to be bad. He’ll probably be somewhere in between.