The PITCHf/x system is provided by Sportvision, but the pitch ID's come from MLBAM. They are working on it, and they're doing some very cool stuff that I hope to report on in the next few weeks. Here's what Gameday gives you vs. what I pull from the data, for today's game (4/22 v NYM). There is one oddball pitch, you can find it on MLB.tv on the 400k stream just after 12:30 (go back a minute to 11:30 to see the whole at bat). It's the last pitch of the Pagan strike out, and it is tough to figure. 
Above, you can see my ID's. The CH (change) and FA (fastball) differ quite a bit in speed - he has a great strange change, since the movement is so close. The curve is the one with the big vertical drop (this is spin movement, in inches, not location etc) and sweep to it. The sliders are hanging out where they always are.
Now, Gameday gives us something different - see below. FA and CU are the same, they call the sliders "FC" - cut fastball, I think - and they stick a slider (or FC in their book) in the CH group. Oddly, they call the one random pitch a Slider. Go figure.
updated
These are curveballs averaged by game. each dot is the average spin movement for all curves thrown during that appearance. I have not adjusted these for park differences, so there are some wacky ones. 
Tuesday, April 22, 2008
Lilly Sidebar - Pitch ID's
Subscribe to:
Post Comments (Atom)
14 comments:
How does the break on his curves compare to last season?
2007 (253)
70.77 mph
-4.3 inches horizontal spin break
(that's in on a righty)
-9.11 inches vertical spin break
(that's drop beyond gravity and trajectory etc)
2008 (32)
68.74 mph (gameday/sportvision adjusted speed about 2mph I hear)
-4.6 horizontal
-9.76 vertical
I wouldn't call them statistically different, so I guess we can't say he's lost anything on it.
I'm anticipating your next question .... accuracy?
a chart does a better job - some of thsoe numbers are unadjusted for park differences, so see the post for a new chart
Actually, accuracy isn't much of a concern for me when we're working with such a small sample. I assume some inaccuracies in all of this.
Good to see his curve, at least to this point, is still breaking like it did in the past. Have you checked to see if they've been called a strike at a lower rate this season?
Also, that FC pitch (cut fastball?), averages only 82.25 mph. Isn't a 4+ mph drop a little much for a cutter? Mariano Rivera's fastball over the last 4 years is thrown at 93.6 mph and the cutter at 93.1 mph. Is he an exception? I always thought the cutter was thrown quite close to fastball speed.
what they call "FC" I call a slider.
I would think so, Harry. I just don't think a cutter loses 4 mph. So he threw 22 sliders then, is that right?
Looks like a slider to me. I've gotten to go through and watch some of today's game. Being able to watch archived games for free is AWESOME!
Yep, I got 22 sliders + the "XX" pitch he sat Pagan down with
Gotta go with a slider on that pitch to Pagan. It was either a slider or a pitch we didn't know Lilly throws.
by accuracy, I mean location. Plate location.
Yes! Ever since 1060 went away, I needed more graphs in my life! Power to the graph!
Wildly unrelated, but Harry, could you look at this graph and tell me if this looks even close to right to you?
http://i164.photobucket.com/albums/u30/cwyers/hill_pitches_001.png
Something's up with the spin directions - they should fall b'twn 90 and 270. Did you use Mike Fast's formula to arrive at those numbers?
Here's the formula I used:
= (ATAN((O2+32.174)/N2) * 180/3.14) + (IF(N2<0,270,90))
The ATAN part of it is from Fast (via SoSH's wiki); the IF part of the formula is based upon something Jini said. If I remove it, the values still aren't within the band.
Post a Comment