Fluke watch: Carlos Marmolby Josh Smolow
February 28, 2011
Carlos Marmol's 2010 performance was an outlier of epic proportions. He accumulated 38 saves, but did so in a manner never seen before (or, well, in recent memory), striking out 42 percent of batters (15.99 strikeouts per nine innnings (K/9)) while walking 15 percent of batters (6.03 walks per nine innings (BB/9)).
The end result was this: only 43.3 percent of at-bats against Marmol resulted in a ball being put in play, by far the lowest mark in the league. The No. 2 in this statistic, Billy Wagner, still had at-bats ending with balls put in play 53.6 percent of the time.
For fantasy purposes, this made Marmol the ultimate wild card. He'd get you saves, sure, and rack up a ton of striekouts. On the other hand, there could be weeks where he'd simply blow up a team's walk rate. (Mind you, if you're not counting pitcher walks, then Marmol's your guy!) Meanwhile, while Marmol's whiff rate increased in 2010, his walk rate actually dropped from a catastrophic 2009 season.
So the question is, are these improvements real? Can you count on Marmol to continue his extreme ways and help you as a fantasy closer?
Marmol's odd Z-Swing rate
As detailed here, batters have swung at Marmol's pitches in the strike zone at the lowest rate of anyone in the majors both of the last two years. (This rate is referred to as the Z-Swing rate). I suspect, as detailed in that article, this is a cause of Marmol's strikeout and walk rates being so high, because batters take pitches they'd ordinarily be able to put into play, resulting in deeper counts, which frequently turn into walks and strikeouts.
Marmol has had the lowest Z-Swing rate for two years running, at least, and in general there is a high consistency in pitchers' Z-Swing rates from year to year, so we can expect this to continue. Thus, we would expect the underlying numbers of Carlos Marmol will be exaggerated next year; we shouldn't be surprised to see high punchout and free pass rates yet again.
Marmol throws two pitches. First, he throws a fastball that doesn't have great movement but has a great average velocity of 94.1 mph. Then, more frequently, Marmol will rely upon his 83.7 mph slider, which gets great horizontal movement in the other direction of the fastball.
Interestingly, Marmol pitches backwards in a sense: whereas most pitchers concentrate more on the fastball as they get into higher-ball counts (to avoid walks), Marmol starts to drop his fastball in favor of his slider instead (except on 3-0 counts).
Marmol does possess a very good ability to hit the strike zone with the slider, and seemingly does not possess a great ability to hit the zone with his fastball, so this doesn't hurt him as much as it might other pitchers. Still it only exacerbates his high strikeout and walk numbers.
Will Marmol's strikeout rate fall next year?
Immediately upon looking at Marmol's pitches, we see a clear change from 2009 to 2010. His slider's whiff rates (percentage of pitches that batters swung at that they missed completely) against left- and right-handed batters have gone up dramatically, as you can see in Table 1 below:
|Pitch Type||Batter Handedness||Year||Whiff Rate|
The question of course is, is this change sustainable? The answer appears to be both yes and no.
Against left-handed batters, there is a clear change in how the slider is used. The pitch was thrown more inside and low to left-handers and was less often over the plate than it was in 2009. In addition, Marmol clearly started to use the slider more frequently in 2010 against left-handed batters, to the point where the pitch was used more often than the fastball.
In 2009, Marmol's slider use against left-handed batters was more normal—he'd use the pitch less in high-ball counts. But, as stated before, Marmol pitched extremely backwards to left-handers in 2010. So another potential cause for the higher whiff rate simply could be that Marmol was using the pitch in 2010 against left-handed batters in situations where batters would normally see fastballs, whereas in 2009, batters would not be as caught off guard by the slider in these counts.
These factors would seem to give us reason to believe that Marmol's whiff-rate increase against left-handed batters is sustainable, and thus so is his strikeout rate against these batters, as these factors are highly correlated.
Unfortunately, against right-handed batters there's no real reason to explain the increase in whiff rate. There's very little change in where Marmol locates the slider in the strike zone from 2009 to 2010 against these batters.
Similarly, there's very little change in how often Marmol used the slider in each count against righties; whereas Marmol only started pitching "backwards" against left-handers in 2010 (this isn't totally true, there's clearly some backwards tendencies in 2009, but it just wasn't as extreme), he was already doing so against right-handers in 2009.
Thus, one would expect Marmol's whiff rate, and therefore his strikeout rate, against right-handed batters to drop next year. That's not a great result.
Marmol's walk rate dropped from 2009 to 2010, to an only slightly less insanely high 6.03 per nine innings in '10 from an utterly insane walk rate of 7.9 per nine innings in '09. Is this sustainable?
Well, some signs point to no. Oddly enough, Marmol's rate at hitting a wide strike zone dropped significantly on his slider in 2010. Similarly, against left-handed batters, the fastball's rate of hitting the zone also dropped, though the rate actually went up against right-handed batters. This would make us expect more walks, not fewer.
However, lower in-strike-zone rates don't necessarily result in higher walk totals, as you can see by the fact that Derek Lowe, the man with the lowest rate of hitting the strike zone, manages to always have a good walk rate.
In Marmol's case, we do see here that a large portion of his zone-rate's decrease occurs in 0-2, 1-2, and 2-2 counts. In fact, in 2010 on 2-0 and 2-1 counts, Marmol was actually better at hitting the strike zone than the year before.
Meanwhile, Marmol's seemingly lesser ability to hit the strike zone was heavily countered by the fact that the percentage of pitches out of the strike zone that opposing batters swung at (O-Swing) rose from 19.8 percent in 2009 to 25.0 percent in 2010, a huge difference.
O-Swing rates tend to be rather consistent from year to year, so I'd expect this change to hold up next year. As a result, I'd suspect that Marmol's "lower" walk rate in 2010 is sustainable and will continue next year.
Normally, Marmol would be a riskier option for a fantasy player as a closer; after all, with his high walk rate, he could start the season on a walk bonanza and lose his closing job, depriving him of most of his worth.
However, fantasy owners can breathe a sigh of relief on this account because the Cubs just signed Marmol to a three-year deal. There's no way they'd make a snap judgment after that and remove Marmol from the closing job before his natural averages can balance out and get his ERA back to normal if he starts out with a rocky start.
That said, he's still not someone I'm confident putting a good prediction on. As you can see from the above, I expect his strikeout rate to drop a little and his walk rate to stay roughly the same, making him a worse option than he was last year, but still a good pitcher.
My confidence in either of these projections, of course, is low due to how unique Marmol is, but essentially I'd still warn potential fantasy owners not to expect as much from Marmol as owners received from him last year. After all, his season last year was historic, and, thus, it's probably not a hard prediction to expect some regression from such a player.
Josh Smolow wishes that one day pitching coaches would seem to actually, you know, do something. You can contact him on twitter under the name garik16.
<< Return to Article