Silva’s Pitching
So, Carlos Silva last night – that was pretty impressive, no? 7 innings, 3 hits, 1 run, 3 walks, 5 strikeouts against a pretty good line-up made up of a lot of lefty hitters. Certainly, the M’s will take more of that kind of performance from all of their pitchers. But how did Silva do it? Let’s take a look at his approach last night through the eyes of the Pitch f/x data.
On the evening, Silva tossed 116 pitches in total. Here’s the breakdown of pitch types, percentage of each pitch, and average velocity.
Fastballs: 70 thrown, 60% of all pitches, 89.9 MPH starting velocity
Sliders: 26 thrown, 22% of all pitches, 82.8 MPH starting velocity
Changeups: 20 thrown, 17% of all pitches, 80.1 MPH starting velocity
If you compare that to Silva’s Player Card from last year or his pitch type data from Fangraphs, you’ll notice that he threw more sliders than usual last night. Last year, he threw about 8.7% sliders. Had he followed that percentage last night, he would have thrown 10 sliders the entire game. He threw 26.
The extra sliders came mostly in lieu of his fastball – he only threw about 60% fastballs last night, significantly down from his career norms. Silva has pretty impeccable command of his fastball, so it’s not overly surprising then that he would miss the plate more often on a night when he was throwing more sliders and less fastballs. But facing a LH heavy line-up, I’d actually have expected him to throw less sliders than usual, considering that it’s not a great pitch against LHB. So, did he attack lefties with the slider, or was it something else? Let’s take a look at the pitches based on batter handedness.
Vs LHB
Fastballs: 51 thrown, 66% of pitches thrown
Sliders: 6 thrown, 8% of pitches thrown
Changeups: 20 thrown, 26% of pitches thrown
Vs RHB
Fastballs: 18 thrown, 46% of pitches thrown
Sliders: 20 thrown, 51% of pitches thrown
Changeups: 1 thrown, 3% of pitches thrown
Against lefties, he actually pitched pretty much like he always does – 2/3 fastballs, then mostly change-ups and an occasional slider when going off-speed. The pitch selection last night against southpaws matches what he did last year almost to a tee.
Against right-handers, however, he threw actually more sliders than fastballs. When Kinsler, Young, and Laird stepped up to the plate, he was going after them with a ton of sliders, and it worked pretty well – those three were a combined 1 for 9 against him. Interestingly enough, he only threw one change-up in the 39 pitches he tossed at right-handers. Last year, one of every seven pitches he threw right-handed hitters were change-ups. One in 39 seems to be a pretty deliberate move away from that. We’ve seen Felix relegate his change to mostly LHB, so I’m curious if there’s an organizational tendency to not throw change-ups to same handed hitters. Something to keep an eye on going forward.
Anyway, in looking at this, last night starts to make a little bit more sense. The low GB% (37.5%) and higher BB% (4.3 BB/G) and K% (7.3 K/G) rates all fit with the idea that Silva intentionally threw more breaking balls and less fastballs than a normal Carlos Silva start. His fastball is the pitch that he has the most command over, and it’s also the pitch that gets him the most groundballs. By eschewing the fastball somewhat in lieu of more sliders, we’d expect him to throw more pitches out of the strike zone and give up a few more balls in the air. That’s exactly what happened.
So, while a quick look at the box score might lead us to think that Silva had below average command last night, I’d argue that’s not really the case – he made a conscious choice to try to induce less contact than usual by varying his normal pitch selection against right-handed batters, and the plan certainly worked, as we saw from the results.
Also, interestingly, no sign of the famous split-finger fastball that he talked about working in during the second half of 2007. Larry Stone asked him after the game and he admitted he didn’t throw any last night. So, don’t go ascribing his performance to a new fourth pitch. He was still a three pitch guy, but he just used those three pitches a little differently than usual.
Dave, do you think this is Silva pitching to match the ballpark? Throwing more sliders results in more fly balls, but by only throwing them to right-handed hitters, the Safeco effect on righties helps him out (even if Ibanez does not).
Interesting theory – maybe. But, if that was the idea, it didn’t really work out as planned. In the nine at-bats against righties, he got two flyballs, one popout, two groundballs, one lineout, and three strikeouts. Three balls in the air, but only two got to the outfield, and one of them got down for extra bases.
I wonder if it has to do with Johjima’s game calling.
I have no idea how much a catcher really plays in calling the game, but if it is something we see in the whole rotation, maybe that has something to do with it? Just a though, tell me if I am wrong.
I was disappointed with the non-appearance of the splitter. I want to see it.
Looking at GameDay, Silva used the slider a lot when he was ahead in the count and with two strikes (although he only got one third strike with it). Maybe his intent with the slider is to get more strikeouts against right-handed hitters, knowing that, unlike some ballparks, Safeco will help him with the consequence (a fly ball) if the batter makes contact.
My guess is that, if he starts throwing the splitter, it’s going to come at the expense of the change-up. With the way the change-up was working last night, there wouldn’t have been much of a benefit to tossing the splitter out there.
While watching the game I really thought his control was pretty good. Every pitcher has a few that don’t go the right way, but it looked like Silva was putting the ball where he wanted to with ease.
Granted Silva is pitching to contact just like Washburn. I feel a lot better with a pitcher getting ground ball outs and working down in the zone than I do with Washburn on the hill.
So, he pitched the way he always does to Lefties, but did much better against them than usual.
Small sample size?
Well, whatever it was, I hope we get to see more of it this year.
xxtinynickxx,
I thought that as well. It didn’t really look like many pitches every “got away from him”. His command on the slider might not be what it is on the FB, but it still looked pretty dang good.
JMHawkins,
it’s not really a sample size issue, he was displaying a different approach then he use to. The results may be a sample size thing, but this post was less about his results and more about what he was actually throwing to get them.
That’s the best news about Silva that I’ve heard all year. An ability to change approach like that and match up strengths and weaknesses appropriately? Awesome. More, please.
I like the approach of using the change-up less against same-handed hitters. That makes sense. Good changeups have a screwball effect, breaking the opposite direction.
Based on Silva’s outing last night, has your opinion of him changed at all? Or do you need to see more of the big fella?
Last night didn’t change my opinion at all, no. I’ve tried to avoid talking about the results in this, because I found the approach to be the interesting story, but if we were to look at the results in a “does this change what we think about Silva” mindset, it’d go something like this:
1. The different pitching approach helped retire three right-handed batters who, let’s be honest, aren’t much of a weapon against RHPs. Young, Kinsler, and Laird combined for a .680ish OPS away from The Ballpark in Arlington last year, and almost all of Kinsler’s value came from mashing lefties. Facing a righty, in Safeco Field, you don’t really expect those three to do much anyway. They were pretty much neutralized no matter what he threw them.
2. His approach against lefties was basically the same as always. The results were obviously terrific, but were essentially based on a 1 for 11 batting average on balls in play. He didn’t show a new weapon that will help neutralize LHBs or show that the sink on his fastball is back to when he used to be a real GB machine. He only struck out two of the 17 LHBs he faced, pretty much right in line with his career averages. The success against lefties was entirely built on getting outs on balls in play, and we all know by now that’s not sustainable.
So no, my opinion of his expected performance this year hasn’t changed at all. I just found his style of pitching against RHBs last night interesting, and thus, the post.
One game shouldn’t change your opinion of any player.
Darn Dave beating me to the punch!
Well, I’d edit that comment slightly to say one game’s results shouldn’t change your opinion of any player. But if Silva comes out next start throwing nothing but curveballs and splitters, well, my opinion of him is going to change, because he’ll be demonstrating new skills that I’m currently assuming he doesn’t have.
From a statistical standpoint, one data point doesn’t matter. From a scouting standpoint, though, you can learn something in one game that has some predictive ability if it’s radical enough.
it’s not really a sample size issue, he was displaying a different approach then he use to. The results may be a sample size thing, but this post was less about his results and more about what he was actually throwing to get them.
Uh, Jar, this is what Dave said:
Hence, my question, “So, he pitched the way he always does to Lefties, but did much better against htem than usual.”
He was pretty devastating to Righties with he expanded use of the slider. His K rate last night was 36% vs RH, four times his career K rate of 9.8%. Vs LH he was normal on his K-rate (9% vs career 9.8% – identical vs both sides for his career).
I’d say he got lucky against the lefties, but just might have a new and better approach vs righties this year.
Well, I’m always skeptical of people “figuring it out”, especially older players.
So for me, personally, I’d still want to see more than one game’s worth of scouting-based information before I re-evaluate a guy.
But I definitely see your point; I’m just a little anal-retentive that way.
JMHawkins,
I see what your saying. I didn’t realize the sample size comment was in reference to the comment regarding his results against lefties last night. I thought it was more general. Sorry about that.
Any chance this is Stottlemyre’s influence? I noticed that even though all the normal talking heads were urging Felix to throw more FBs, he also threw a lot of breaking balls. Maybe a combo of pitching coach and scouting report?
Regarding his repertoire, Silva talked a lot about improving and throwing more “cutters” this year. Are we assuming those pitches (as far as pitch f/x is concerned) were mixed in with his fastballs, his curveballs, or did he not throw any last night?
I’m glad I’m not the only one who noticed. I mentioned to my buddy that Silva seemed to pitch differently and he agreed that this type of approach is preferable.
If Silva was throwing any cutters last night, they weren’t different from his normal fastball in any meaningful way. There was one pitch that pitch f/x labeled a cut fastball, as it moved in on a lefty quite a bit more than his normal fastball, but it also had a start speed of 83.4 MPH, so I think it was a slider, honestly. His slowest fastball besides that pitch was 87.1 MPH, so I’d bet that cut fastball was mislabeled by gameday.
All his other fastballs had similar horizontal and vertical movement, so if he was throwing a cutter, it’s not much different than his regular fastball. And, if that’s the case, than it doesn’t really matter what you call it – hitters hit movement, not labels.
Dave, how much difference in break will the pitch f/x show between a four-seamer and two-seamer?
Sure, I get that. I didn’t know much about his slider prior to this year, so I wasn’t sure if what was being labeled a slider last night was actually his cutter, or whether there was enough variety in what was being labeled fastball to think that there were some cutters in there. Sounds like there was maybe 1 cutter mixed in there.
Bigger picture, doesn’t sound like his cutter is some major new weapon for him that will change him as a pitcher in a significant way like the splitter did for Putz.
I’m not changing my opinion on Silva.. as I still think he’s an average -at best- starter… but one thing is for sure.
I’m extremely happy with the way he performed. I know it’s one game.. but it is far superior to having to watch Weaver or HoRam pitch on this team. I’m a lot more confident in Silva than I would be in Dickey or Wang, too.
Only one game? Yes… but it’s a win.. and that’s all that matters.
For what it’s worth: Silva is quoted after the game in the AP coverage as saying the ball moved more in the “colder, heavier air” here vs Arizona.
As I said in the game thread, if that’s true, watch out this summer — especially in away games, and especially against these same Rangers.
Except that it will be humid in Texas instead of dry.
I think your math might be a little off.
On top, you have 76FB/26SL/20CH
Totals on the bottom are 69/26/21
Not that this changes anything, necessarily.
Great post! This is why I love this site! I can get info here I can’t get anywhere else.
Back at comment #20 the question was asked whether Stotlemeyer’s influence may have been in play last night; it was the same question I would have asked. Is it possible that the pitchers are getting better advice than they have in the past and that an improvement in pitch selection is the direct result?
Given what we know about the organizational fondness for establishing the fastball and considering that Stottlemyre’s public comments have been for pitching staff to throw more fastballs, I’d call that unlikely.
Honestly, I was impressed with Carlos last night. Do I need to see more? I think so. Let’s just hope his next outing goes as well.
Except that it will be humid in Texas instead of dry.
Humid air is less dense than dry air. That makes it harder to get a ball to break. Cold air is much denser than warm air (and in fact will contain less water vapor than warm air at the same relative humidity). The temperature is a much greater factor than humidity in this respect.
To the extent humidity depresses the long ball it’s because damp balls are less elastic and damp bats don’t transfer energy as well. The ball doesn’t go as far because it gets less initial velocity, not because humid air slows it down. That’s why at Coors they put the balls in the humidor to supress home runs (and why pitchers who rely on ball movement still complain about pitching there). All things being equal — with humidity-controlled dry balls and bats — a day with humid air should (slightly) increase home runs, though in reality the difference in air resistance between atmospheres with high and low water vapor content doesn’t amount to much.
(I did these calcs once but can’t find them at the moment, so I leave them as an exercise for the reader. Relevant equations can be found here.)
Dang. I started reading this hoping that there was some insight to say that last night was the start of a new skill from Silva that would make last nights result more sustainable. Here’s to hoping he can find a way to justify the shiny new contract he got . . . it will prove interesting to see if the back half of the rotation will actually be good this year or if we will end up with our version of “Spahn and Sain then pray for rain” . . .
It’s at least encouraging to see him mixing his pitches well and commanding his fastball. Hopefully he can keep that up this season, and hopefully it’s a sign that he’s smart enough to adapt…
Thanks for the analysis, Dave, but do keep in mind that “less” is used when referring to something that’s not physically measured/counted. “Fewer” is the appropriate term when referring to, for example, the number of fastballs or sliders thrown in a game. 🙂
I think it’s still possible that Silva has a new pitch, or at least has a variation on an old pitch. On the Mariner home page, I saw an article about Silva’s new cutter that he says breaks up and in on a left hand hitter and away from a right hand hitter.
Maybe the cutter was registered as a slider and accounted for the increase in the amount of sliders thrown to the right hand batters.
In any case, thanks for the great post, very interesting!
If his cutter moves like a slider, then it doesn’t matter what Silva calls it, it’s effectively a slider.. Hitters hit movement, not labels.