Maybe this is becoming monotonous for you guys but when a child gets a new toy he keeps playing with it and that is me with this PITCHf/x stuff. So here is a look at the starting pitchers going in the series this weekend through the eyes of PITCHf/x. Note, only home pitches when the system was turned on have been added to these plots to eliminate calibration issues. I've done some tweaking with my algorithm and hopefully things have improved some.Let's call it a series preview. --- roguejim, who is easily swayed by pretty colors.
Starting with the Reds starters Tom Shearn is throwing Friday. Shearn pitched a few innings of that blowout game last Sunday against the Brewers so Brewer hitters should have seen him before. He is a nice story as a 30 year old getting his first taste of the show. Here is what he offers.
Shrean throws a four seam fastball, a two seamer, a slider, and a curve (with an occasional change mixed in). Shearn lives and dies with his off speed pitches getting very good movement from his slider and exceptional movement from his curve. His four seam fastball offers almost no horizontal break at all which appears to be a very bad thing the more I look at this stuff. His sinker is better but not close to the quality you see in good sinker/slider type pitchers. His curve is almost exactly 12 to 6 as well.
Here is the horizontal break vs. pitch speed. You can see that his sinker has basically the same speed as his four seamer with his change and his slider at basically the same speed. He has a very good speed differential with his off speed stuff.
Same sort of thing with the vertical break. His sinker and four seamer start to break apart here.
Here is Shearn's release point. He has a very much an over the top type release which is what you would expect looking at his 12 to 6 curve. Interestingly, it appears he cheats a little and is throwing that curve from a slightly higher and closer to his body point. Here is a blowup of that plot.
Here you can see the spread of his release point for each of his pitches. His curveball appears to be half a foot different from his fastball which is half a foot different from his slider (and change). I wonder if that difference is enough for batters to pick up on it?
Next is Kirk Saarloos. PITCHf/x has very little data on him so I am going to show his plots without much commentary.
During his most recent callup with the Reds Saarloos has featured his sinker which is a pretty good one (you can see the extra vertical drop he is getting compared to Shearns). My algorithm thinks he threw four changes in ther but it really is hard to tell. With more data we would know for sure.
Saarloos doesn't seem to have too much behind that sinker though with it sitting near 85 MPH.
Maybe those are changes after all.
Saarloos has a much tighter cluster for his release point but because he was mostly throwing sinkers it is hard to tell if his release point is that constant or not.
On to Matt Belisle who should be the most difficult pitcher the crew faces this weekend.
Belisle throws a four seam fastball, curve, change, and a slider.
Belisle's slider is much more of a "hard slider" than Shearns and you can see the difference in speed between his slider and fastball is much less. I don't know why the clustering algorithm puts that group of sliders in with his curves. I am fairly confident those should all be curves. Again, you can see improvement from the algorithm but it still has work to do.
Again you can see the hardness of his slider and how close it comes to his fastball in this plot. Also notice that his curve doesn't have nearly the break Shearn has.
Now that is a repeatable motion. Belisle has basically one release point though again, his curves are mostly located in the upper right. The difference though is measured in inches rather than feet like Shearn.
Ok on to the Brewers and starting with Sheets.
Sheets only has one start at home to look at but basically this sums up Ben perfectly. Fastball 70% of the time, curve ball 25% of the time, change up 5% of the time. Like Shearn, Sheets has a devastating 12 to 6 curve with little horizontal movement at all. His fastball though is offers much more horizontal break breaking in to right handers. I believe that Sheets' ability to generate such movement on his fastball while throwing it in the mid 90's and still have a 12 to 6 curve is much of the reason he has gotten away with basically two pitches as a starter.
This start his fastball was more in the lower 90's than mid 90's and that seems to be something that has fluctuated a lot this year with Ben. It would be interesting to see his stats broken up by how fast his fastball was that day.
Not only does Sheets' curve break a good amount in comparison to the fastball it is breaking almost a foot and a half. That is huge and means if the hitter sees the ball wrong or guess wrong he isn't going to even make contact with the curve.
Like we have see with other pitchers Sheets throws his curve from closer in that his fastball. Here is another blow up picture.
Again his curve seems quite a distance away from the center of his fastball zone. Maybe he is saved because he is throwing a few fastballs closer to his body? Hitters haven't seemed to picked that up so maybe pitchers can get away with a six inch difference is release point.
Suppan is the crafty vet of the group. He throws a fastball, change, cutter, and curve. His cutter has about the opposite break you would expect from a slider but similar downward movement in relation to the fastball. His fastball zone is pretty large and maybe that should be broken up into a four seamer and a two seamer. I couldn't find any mention of this with a quick google search though.
Suppan's change and fastball don't seem to have a great speed differential. His change was supposed to be one of his better if not his best pitch but it is hard to see that on his plot.
Here as well. He just doesn't seem to be getting very much differential with his fastball/cutter/change. Now as I pointed out in other posts, Maddux seems to be the same way so maybe there is something too that.
Suppan seems to have a smeared release as well though his curves are on the other side of his motion. That unknown point up high is a PITCHf/x screw up that I am going to get rid off next time I run the code (which now takes about 5 hours over the whole data set by the way).
When I look at Carlos' plots I get pretty excited about his future. His change appears to mix in with his fastball quite well and his slider and curve both seem to have excellent life to them. His slider isn't a hard slider like Belisle with plenty of difference between that and his fastball. His curve isn't nearly as 12 to 6 as we have seen with some others but still offers quite a bit of vertical break. Lastly, he has the same horizontal movement on his fastball we saw from Sheets.
Carlos has a bit more speed differential on his change than Suppan and his fastball is a tick or two better. His curve really falls off the table with more than a 15 MPH speed difference between that and his fastball.
Again great separation between his pitches. While there isn't too much of a difference in horizontal movement between his fastball and change there is a considerable difference in vertical break. This should help disguise the pitch and produce a lot of ground balls as batters see fastball and swing just a hair over it.
Carlos has a spread in his release point like Suppan has but it appears that he throws every pitch from everywhere on the smear. I wonder if this is might be slightly poor mechanics from Carlos? Hard to tell. In any case there is absolutely no way a hitter will be able to pick up the curve or the change from looking at his release point.
A little more gushing over Villy. Really from looking at these plots it appears he is the total package. He has a decent fastballs that rides in against right handers. He has a great change that should be able to produce a ton of ground balls when he needs them. His slider and curve both have great movement and definitely appear good enough to get a lot of strikeouts when he needs them. The one thing these plots don't show is control which seems to be the last thing Carlos has to master. Obviously, many pitchers have had great stuff but never learned control over it. But Carlos is young and seems to "know how to pitch" already. If he does learn better control look out because he is going to be better than the #4 or #5 starter many have predicted for him