Carlos Silva Learns

Among early season performances, there is one that is the most shocking.

Carlos Silva, vs left-handed batters:

2008: .348/.381/.555, 2.19 BB/9, 4.37 K/9, 44.5% GB%, 14.6% HR/FB%, .355 BABIP
2009: .380/.436/.718, 3.94 BB/9, 2.25 K/9, 48.5% GB%, 21.1% HR/FB%, .359 BABIP

2010: .083/.081/.083, 0.00 BB/9, 3.97 K/9, 36.7% GB%, 0.0% HR/FB%, .100 BABIP

Silva, who throughout his entire career has struggled mightily with left-handed hitters, has held them to just three singles in 37 plate appearances in his first four starts. And all three of those singles came in his last start. In his first three appearances, he was perfect against LHBs, as they went 0 for 22 against him.

His line against right-handers isn’t all that much different than it has been in the past, even in his last two seasons. Nearly the entirety of the success he’s had to date can be credited to how well he’s gotten lefties out, which is just something he’s never been able to do before.

So, naturally, the first thing I did was take a look at his pitch selection. Silva’s lived primarily off of his two-seam fastball for most of his career, which is why he’s posted such large platoon splits. The pitch works against righties, but not against lefties.

Sure enough, Silva has finally decided to abandon his fastball-only approach to pitching. He’s thrown his sinker just 56.5% of the time (compared to 83.1% last year), and has replaced with his change-up, which he’s now thrown 30.7% of the time.

The change-up has the smallest platoon split of any pitch in baseball, so it would make sense that Silva relying more heavily on it would fare better against southpaws (and, at the same time, see a decrease in his GB%, which he has). To dig further, I asked resident pitch f/x guru Dave Allen to look at Silva’s pitch usage by handedness, and he found that Silva is throwing his change-up 40 percent of the time to LHBs this year, and produced this neat little graph to demonstrate how effective it has been.

He is pounding the down and away corner with change-ups and getting easy outs off of it. His change-up has been +7.7 runs through four starts, according to our pitch type linear weights, making it the most effective change-up in baseball to date.

Now, obviously, lefties won’t post a .100 BABIP against Silva all season, so there’s inevitable regression coming. But it does look like he’s finally learned that he can’t just attack them with two-seam fastballs and hope for the best. If he keeps pounding lefties with his change-up, he might actually stick in the Cubs rotation this year.

Print This Post

Dave is a co-founder of and contributes to the Wall Street Journal.

25 Responses to “Carlos Silva Learns”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Bryz says:

    The change-up has the lowest platoon split among any pitch? Then can you explain why I’ve gone through my entire childhood being told that change-ups area only useful against opposite-handed hitters? (I believe you more than these people in the past, since you likely have data to back up your argument. I’m just curious why they might have argued against you).

    Vote -1 Vote +1

    • Joser says:

      Did they tell you “only useful” or “most useful”? Because if you had other pitches that were better, you should be throwing them to same-handed hitters. But those pitches were likely to be worse than your changeup when thrown to opposite-handed batters. The changeup makes opposite-handed hitters no better against you than same-handed hitters (facing that pitch), which is an asset; but unless it is your best pitch (or you have another pitch that shows similarly low platoon splits) you should be throwing something else to same-handed pitchers.

      Vote -1 Vote +1

    • Dave says:

      Bryz, these are just numbers pulled out of my ass, but here’s a hypothetical platoon split:

      OPS against a RHP
      vs LHB: .800
      vs RHB: .600

      vs LHB: .600
      vs RHB: .700

      vs LHB: .800
      vs RHB: .600

      You have to work off your fastball, so you have to throw it to lefties to set up your change-up. You don’t have to throw a change-up to righties. The curve or slider are much better secondary pitches against them, so you opt for them instead.

      It’s not that change-ups are useless against righties. It’s just that, against them, there are better options, so most pitchers don’t throw it too often.

      Vote -1 Vote +1

    • Bryz says:

      I suppose my “only useful” comment wasn’t exactly clear. What I heard from TV analysts* was that if you were a RHP, your change-up should only be used as a weapon against lefthanded hitters. Using it against righthanders would be bad, because the pitch “broke into the hitter’s swing” or something like that.

      * These guys are starting to lose my approval. I feel like the play-by-play announcers are guys that don’t do their homework on certain players, and the color guys are all former players/coaches that are convinced that everything they know is correct.

      @ Dave: Yes, I did know that pitchers would use a curve/slider against same-sided hitters rather than a change-up because the breaking pitches were more effective. Thanks for the explanation with the ass data.

      Vote -1 Vote +1

      • Ken says:

        I’ve never understood that either, and I was a pitcher for most of my young life. I don’t understand, especially if you only have 2-3 pitches, why you would take one of your pitches away just because of the handedness of the batter? It seems to me like you’re unfairly handicapping yourself based of some archaic notion that certain batters would hit that pitch better based off of what dominant hand they use to bat with. It’s like people who tell me that “lefties have natural movement on their fastball”. It just seems to me like something old baseball people started and people have echoed without checking the facts first.

        Vote -1 Vote +1

  2. jackweiland says:

    Cuz they’re dumb?

    Vote -1 Vote +1

  3. jackweiland says:

    But seriously, it’s probably because someone made an assumption and didn’t bother to look into things like statistics, facts, etc. and proceeded to spread the word as if it were gospel.

    Kinda like how we thought batting average was the most useful hitting metric for, ya know, a hundred years.

    Vote -1 Vote +1

  4. John says:

    Really nice analysis. It really goes beyond the usual historical data people base their projections on. Historical data is great, but it doesn’t take into account that players can and do change their approach and, every once in awhile, it turns projections on their ear. Seeing Silva go after hitters with 3 decent pitches this year, I was really at a loss to explain why he’d been so awful at times in the past, there are many pitchers who have lesser stuff and less command than Silva does, yet none were as bad as Silva the past 2 years. This excellent analysis really goes a long way toward clearing that up. The change-up has argually been Silva’s best pitch this year and it’s a mystery to me why he’s never used it in the past, particularly against lefties. There is no way Silva is going to continue to pitch as well as he has overall in April, but I don’t think it’s too much to ask that he can consistently put up lines like the one he did vs. Washington. And if that’s the case, the Cubs will have a pretty reliable starter on their hands — and whoulda thunk that would have been possible just 2 short months ago?

    Vote -1 Vote +1

  5. Joser says:

    To the extent that I trust any handedness park factors, it would appear Wrigley is almost a mirror-image of Safeco (especially in the spring and fall, when the atmospherics are more similar). So whereas LH batters facing Silva in Seattle likely inflated his HR/FB rate, they’re probably getting suppressed about as much so far in Chicago. Of course Silva has only pitched two games in Wrigley; one of the others was at Citi Field which is getting a rep as a pitcher’s park, leaving only his first start at Great American as one where the park wouldn’t have helped him.

    Which doesn’t change or detract at all from what Mr Cameron wrote, but it does suggest that while that sparkling 0% HR/FB factor (vs LH bats) won’t last all year, we should expect it to be notably lower vs his rate with the Mariners (thus making him a superficially better pitcher, over and above the change in leagues) even if he hadn’t made the adjustments detailed in this post.

    Vote -1 Vote +1

  6. poodski says:

    Silva’s changeup has looked so incredible this year.

    I dont get how its been worth -22 runs up until this year. Now all of a sudden better than Johan?

    I don’t know what happened, as I have never really seen him pitch before this year, but man that changeup has looked almost unhittable.

    Vote -1 Vote +1

  7. Rudy says:

    I was told once by a guy that played in the minors that lefty pitchers don’t throw changeups to lefties because they see it well. Any truth to that?

    Vote -1 Vote +1

  8. Llewdor says:

    Why couldn’t he have learned this before?

    Actually, the bigger question is why anyone bothered to employ him after those last two years in Seattle.

    Vote -1 Vote +1

    • John says:

      I was thinking along the same lines. Who knows? Maybe Seattle just gave up on him and the Cubs looked at him with a fresh pair of eyes. That he was throwing so many fastballs to lefties is just absurd. It’s odd that no one picked up on that until now.

      Vote -1 Vote +1

    • poodski says:

      You mean after the 30 innings of 2009?

      In 2008 he still had a 4.64 xFIP and was worth 1.5 WAR.

      In 2009 yes he was terrible but was also hurt. People talk about sample size when talking about most players but when it comes to Silva its like sample size and advanced statistics just go flying out the window.

      Vote -1 Vote +1

      • John says:

        I agree. I’ve read my share of people bending over backwards to make his projections look gloomy. Take for instance, the Fans projection on this site that has him with a .352 BABIP which would be a career high, even higher than his 2008 outlier. I think this is why I enjoy this kind of analysis and others like it, it incorporates sound baseball knowledge in addition to good statistical theory.

        Vote -1 Vote +1

  9. Rudy says:

    Cubs made it out well in the Bradley deal if Silva keeps this up.

    Vote -1 Vote +1

  10. Bradley says:

    I honestly don’t think we can say his new approach is paying off. I think it’s a fluctuation, or perhaps a function of game theory as hitters haven’t adjusted to his new approach.

    Frankly, Silva is surrender a metric TON of flyballs right now. Add to that his above-noted BABIP luck, and he’s poised for regression. What happens when his career-low HR/FB% catches up to his career-high FB%? I think we will notice a more Silva-esque platoon split at that point.

    It’s like Poodski pointed out: “I dont get how its been worth -22 runs up until this year. Now all of a sudden better than Johan?”

    Vote -1 Vote +1

    • John says:

      There’s no question Silva’s going to regress. The Cubs know it, Silva knows it…we all know it. No one thinks he’s going pitch this way all year. The only real question is to what point will Silva regress? Just because he’s gotten by with some luck, it doesn’t mean he’s going to regress to his dreadful days in Seattle…where he was a victim of a career high .347 BABIP in 2008. A new approach, a new league, better health, a better defense than the Bavasi-built Mariners…all of this points toward better days than the ones he had with Seattle. We’re not asking Silva to be an ace, just to be a solid starter. This is a guy who as recently as 2007 was a 3.3 WAR pitcher. If he can even approach that I think most Cub fans will gladly take it.

      Vote -1 Vote +1

      • Ian says:

        Do the Cubs know it? I think the fans do but does Lou think Silva will regress? I don’t know. After moving Zambrano to the bullpen I’m just not sure what anyone is thinking there anymore…

        Vote -1 Vote +1

      • John says:

        I seriously doubt the Cubs really expect Silva to have a sub 2.00 ERA, a sub. 200 BABIP, zero hrs vs lefties, and a quality start everytime out for the rest of the year. Of course they expect regression. That’s just plain nonsense. What they don’t expect is that he’ll regress to the way he pitched in Seattle. There’s a difference. People put up BABIP numbers and FB rate numbers, etc. and automatically assume that when Silva inevitably regresses, it means he’ll regress to the pitcher he was in Seattle. As if there is nowhere else he can regress in the vast space from where he is now to where he was in Seattle!! He has plenty of room to regress and still be a quality starter…that is what the Cubs are banking on. It isn’t like Z is an ace, and he’s probably overrated based on last years numbers. At any rate, it isn’t whether Zambrano is a better starter than Silva, it’s whether the downgrade from Z to Silva is less than the upgrade to the bullpen – and if that’s the case, it makes the Cubs better overall. And the whole point is to make your team better, isn’t it? And if Silva turns out to be the pitcher he was in Seattle, you just put Z in the rotation again. So far, for one start, it worked out fine. Silva pitched 7 strong innings then Z pitched 1 2/3 scoreless frames to get the game to Marmol…Cubs win. And, for one game at least, I don’t think anyone cared that Z was an 18M reliever.

        Vote -1 Vote +1

      • Bradley says:

        That’s a good point, John. I don’t mean to imply that others aren’t expecting Silva to regress — though I do fear, like Ian, that the Cubs are underestimating what that regression will look like. Also, there is something to be said for the comparative advantage gained by moving Z to compensate for a weakness elsewhere on the roster, but — as is illustrated on Cubs Stats — I don’t honestly think the bullpen is weak, just misused (see John Grabow).

        Vote -1 Vote +1

    • Jason B says:

      Is a metric ton more or less than a metric shitload? Someone used that measure in another post a couple of days back. Either way, I like that we’re using the metric system.

      Vote -1 Vote +1

  11. Mike Z says:

    The confidence that Silva has from his good start and better pitch selection hopefully mitiagate the inevitable regression and the league catching up to him. He’s throwing better pitches, walking noone again and success breeds success. Maybe there is a reason Rothschild has been the Cub’s pitching coach for 9 seasons with Silva using this effective pitching pattern and keeping the walks down. Maybe it’s the Maddux effect as the entire rotation has been effective this year (everyone but Zambrano with an ERA under 2.78). Maybe its the year of the rotund right handed latino innings eaters with Silva and Livan Hernandez’s success and neither will regress ;)

    Hernandez: 2010 ERA = 0.87; AVG Fastball = 84.3 mph; K/BB = 1.25
    Silva: 2010 ERA = 1.73; AVG Fastball = 89.5 mph; K/BB = 5.19

    Vote -1 Vote +1

  12. Mike Z says:

    Should have included “Metric Ton” somewhere in that post about Hernandez and Silva. Looking at those two I’ve talked myself into thinking that Hernandez is the most shocking success story and that Silva may be sustainable. Not at the current level of course, but do more than just stick in the rotation and be an effective starter the rest of the year.

    By the way, everybody still think the genius in Seattle got the better end of the Bradley/Silva swap. Good riddance to that headache.

    Vote -1 Vote +1

  13. wobatus says:

    So, the Cubs may actually have gotten something for Bradley after all. Silva has had 3 3 WAR seasons in the past. With an improvement he may make a nice back end rotation piece, even with inevitable regression.

    I don’t make as much of the league switch. Zito switched and rotted in the N.L. And it doesn’t seem like facing A.L. West teams will be all that tough this year. Except Silva wouldn’t have gotten to pitch against the M’s.

    The M’s are learning a bit about subtraction by addition with Milton, but it is early yet.

    Vote -1 Vote +1