Filter posts by...
Ball_Tracking
Friday, November 02, 2007
Yes, but they see the same number as other hitters in pitcher’s counts. Great stuff. I love his grouping of hitters, and that he controls for the identity of the pitcher.
I don’t agree with his choice of what constitutes a hitter’s count though. A 1-0 and a 2-1 are very similar, and they should be pooled together. Joe had the 2-1, but not the 1-0. The 3-2 is also a hitter’s count, which Joe counts as neutral. As the “pass through” chart shows, if the batter has more balls than strikes, it’s a hitter’s count. The only question is whether to treat the 2-2 and 0-1 counts as neutral or pitcher’s count (but they need to be pooled together). Joe includes these two as pitcher’s counts, which I’m fine with. Otherwise, clear pitcher’s counts are 0-2 and 1-2.
Nonetheless, great work.
Thursday, November 01, 2007
A fantastic look by John Walsh on the transformation of Pedro. It’s really tough for us to forecast pitchers, because we have no idea was is actual change in approach and what is simply sampling. Pitchers have so much variability (speed, spin, trajectory, release) as compared to batters (timing, hardness, plane of swing), that pitch by pitch data (scouting data really, in an objective fashion) becomes critical. If I were running MLB, I’d hire 30 John Walshes, Dan Foxes, Joe Sheehans, Mike Fasts, et al.
What was interesting to me is how consistent Pedro is with his spin effects and speeds on his pitches, other than the curve. (John calls it “movement”, but it’s really gravity-less movement. It’s much clearer to think of those numbers as how much spin Pedro is putting on the ball.) Basically, Pedro is able to alter the speed of his pitch a few miles an hour and impart a small but effective change in his spin on the ball so that the batter likely has no idea what the heck his pitches are doing. (John doesn’t show his release points, but I wouldn’t be surprised if they were all fairly consistent.)
I also think we’re going to be at the point that it doesn’t make sense to call his pitches “fastballs”. A Jamie Moyer fastball would be a Randy Johnson changeup. I would much prefer a nomenclature based on all pitchers, rather than to himself. The nomenclature of the pitch can probably be represented by the spin of the ball. So, you can have a 87 mph fastball and a 98 mph fastball, if we agree that a fastball is thrown with a certain amount and direction of spin. So, the “movement” (i.e. spin) numbers that John is showing is what decides what kind of pitch is thrown, irrespective of the speed. The speed is simply a qualifier to the spin of the pitch.
Friday, October 12, 2007
By , 11:54 PM
First of all, another great article and analysis by Dan Fox at BP.
As an aside, he had a great quote at the end of the article (which has nothing to do with the pitch data). People ask me all the time, “Who do you think is going to win the (insert award/series/etc.)?” Like who do you think is going to win the Indians/BoSox ALCS? Even after painfully analyzing the series for several hours, my pat (and factually correct) answer is, “I have no (bleeping) idea!” To borrow a phrase (again) from Bill James, “I am an analyst, not an oracle.” I can tell them the percentage chance of winning I think each team has (based on my model and my analysis), but I cannot tell them “who is going to win (obviously).”
Now let’s say that I had Boston at 65% (which I don’t). If they win, was I “right?” If they lose, was I wrong? What about if I had Boston at 51% (which I do) and they win? Right or wrong? Heck if I know the answers to those questions. I DO know, for example, that if I were a perfect modeler and knew the exact percentages in all baseball series or even every one of the 2,430 games during the regular season, and all my “rights” were when my favored team wins and all my “wrongs” were when my favored team lost, I would be “wrong” a heck of a lot!
As I also like to say, “If a good - no a great - analyst isn’t wrong a heck of a lot, he is probably cheating.”
Anyway, here is what Dan wrote about the difference between probabilities (which is all an analyst can do) and predictions (which are silly and meaningless for an analysts to make):
A subtler but related point in this vein is that some seem to think the models used to discuss events are necessarily predictions and therefore take a “told you so” approach when the end result seems improbable according to the model. But probabilities are not predictions, and so in addition to the fact that the models used to generate the probabilities are incomplete, even events that are unlikely do in fact happen. Only if you could replay the event hundreds or thousands of times could you say with confidence that the model is not useful.
Back to the pitch f/x data…
Read More
Wednesday, September 19, 2007
The future is arriving. The pinnacle of sabermetrics is the convergence of performance analysis and scouting. Mariano Rivera is a righthanded pitcher, and for most RHP, they love RHH. Mo’s splits shows fairly clearly that he loves LHH (OPS of .526 compared to the RHH of .606). That is, he shows reverse platoon splits. If you knew nothing about Mariano, you’d have to regress those splits toward the mean of all RHH platoon splits, even though something tells you that you shouldn’t. But, John Walsh shows us something special: that virtually all of Mariano’s pitches move as if they were thrown by a LH fastball pitcher. (I think it may be necessary to split his pitch data between LHH, RHH.) Now, all of a sudden, we don’t regress his performance stats to the typical RH pitcher, or the typical RH fastball pitcher, but rather to a RH pitcher who relies almost exclusively on a slider, or on a LH fastball pitcher. You see, we’ve always been using the hand the pitcher throws with as a proxy for how the pitch moves. We don’t need to do this any more. We can actually decide based on how the ball actually moves.
Another thing that I found fascinating was:
The good change-ups that I see tend to dive near the plate, they ‘fall off the table’, as they say. I don’t think they have the vertical movement of a fastball.” Well, you’re right. And me, too—we’re both right. The change does “fall off the table,” but that is thanks to gravity and not to any particular trickery applied by the hurler—the change simply takes longer to get to the plate, allowing gravity to do its handiwork for a bit longer.
Which begs the question as to what the horizontal and vertical movement actually represent in reality. From the sounds of it, it represents the movement without the effect of gravity (comparing the movement of the ball with spin and no spin). But, since the time component is of enormous influence here, I submit that it’s silly to use the spin v no spin as the baseline. Perhaps I’m not understanding something crucial here.
I think the baseline should be one where you choose whatever backspin you want (no spin currently) that forces the ball to travel in a straightline. That is, the backspin balances out against gravity. The break of the fastball and changeup is then compared agaisnt this baseline. If this happens, will it result in vertical breaks that we’d expect?
Wednesday, September 12, 2007
Hat tip Kevin.
Some great stuff being done, this time with clustering pitches to figure out what pitchers are throwing. See also links at the in the comments on that page from the equally impressive Mike Fast. Seriously, there’s a good dozen bright bulbs out there working their hearts out, bringing the work to the masses. It feels great to sit back and watch this unfold.
Friday, September 07, 2007
Want to see how and where balls have been put in play against Paul Byrd when he has two strikes on a left-handed batter? Look no further.
Tuesday, September 04, 2007
Reliability results of the release point of PITCHfx data.
Thursday, August 23, 2007
The PITCHf/x work continues, with Joe Sheehan showing us some plot points (overhead, and side). I said this in response to something else:
What I like to do is say: “How many feet away will it take to travel 0.20 seconds to home plate?” If the human reaction time is around 0.20 seconds, this will tell us exactly where the ball is when the batter has to make his decision. This will also handle the issue about throwing the ball from OF at 200 MPH. The question is always: when does the batter have to decide. I think it’s alot clearer to tell a fan that a batter can decide on a Wakefield pitch when it only 20 feet away and when a Zumaya pitch is 35 feet away, then giving out numbers like .387 and .554 (or whatever… all number for illustration only). If it’s not .20, but .15 or .25, so be it, use those numbers.
Dan Fox also checks in. And you too can be an analyst by following the tutorial of Dr. Nathan. And Mike Fast is keeping track of it all.
Saturday, July 21, 2007
From 1958-1963, here are the pitches per batter of the regular Dodger pitchers:
Read More
Friday, July 13, 2007
Add yet another researcher to my hero list.
This from Joe P. Sheehan is pretty cool too.
Wednesday, July 11, 2007
Yet another in a long-series of great research pieces by John Walsh.
He points out to some obvious data quality issues like:
Read More
Tuesday, June 26, 2007
When you establish the true talent level of a player (or create a forecast, which is essentially the same thing), you want to know if he is healthy or not. And, if he’s not healthy, how unhealthy, and how persistent is his illness. So, you try to infer such things. If a player plays 159, 160, 154, 159, 161, 112 games, his OPS+ at any point in that stretch bottomed at 133, and the player is 27 years old at that point, we infer an injury. We don’t need to look more closely at the situation, though it could have been the case of someone even better usurping his playing time. You always have a certain uncertainty level. And if the player is 37 instead of 27, we may be more inclined to infer a longer rehab period. But, we still don’t know what kind of injury because the data doesn’t tell us much more.
John Walsh shows us the data for Curt Schilling. Now, we don’t need to infer if his performance was about balls falling in for a hit, or whether his true talent level was marketdly different. We remove that uncertainty level with the data. Depending on the nature of the illness, we’ll be able to either discount the data from this performance more, or place a greater premium on it. We’re always looking for the establishment of a new talent level, as opposed to randomness creating noise around the data. It’s data like this that we need.
And for MLB teams that are not doing this.... are you kidding me? What Walsh, Fox, Beamer, Sheehan, Appleman, et al are doing is the cutting edge of sabermetrics, the point where performance and scouting converge. This is the pot of gold that is being prospected.
***
Further research would go into the “mix” of pitches, and the “strategy” of pitches, based on the game state (inning, score, base, out) conditions… i.e., Leverage Index.
***
The data itself also has a certain amount of uncertainty, as can be easily seen with David Wells having a bunch of pitches being released from the wrong side of the mound (four feet from where it should be).
Wednesday, June 06, 2007
When all the talk about the “gyroball” came out, I always thought of three things: tell me
- how much fast the ball moves,
- how much the ball breaks (horizontal, vertical), and
- whether the pitch has topspin, backspin or knuckles.
That’s really what we care about. Calling it a split-fingered fastball, a cutter, a slider, a fastball, or what have you is really irrelevant. Along comes John Walsh with some great research on just that topic. Rather than relying on an observer to determine if a pitch is a sinker, why not simply infer it (or more accurately, classify it) based on what the ball actually does. Walsh, Appleman, Fox, Sheehan, et al: keep it up guys!
Thursday, May 17, 2007
This seems to show that he does. Take one part fantastic resource (Gameday) and one part dedicated researcher (Dan Fox), and you get a fascinating chart. This benefit seems to apply to LHH (click on the LHH chart for the blowup).
Remember that the strike zone would be 8.5 inches (0.7 feet) from the center point (0), and the ball itself is 2.9 inches in diameter, so 11.4 inches (0.95 feet) on either side is the strike zone, which is why Dan has it laid out as he does. His strike zone includes the extra space outside the plate itself. So, even given that, Maddux is going beyond that point. He’s getting an extra 2 to 6 inches beyond that! (He doesn’t get the benefit against RHH.)
However, I have my doubts. There are plenty of foul balls and balls put in play that are in the supposed outside of the strike zone. Is it that the batter himself is expanding the strike zone because he figures that the umpire is as well? So he might as well go for it? The whole chart looks skewed.
Look at the -0.3 feet point, right in the center. That filled-in diamond is a called ball! There are also plenty of called balls at the +0.6 feet point (though that might mean that some umps are not giving him those outside pitches, not even the extra space beyond the plate).
I understand that Cory at MLB.com says they take great care in the entire calibration process. But, I’m still on the fence here. I think we need to look at these things game-by-game, park-by-park, pitcher-by-pitcher, and ump-by-ump. I’d like to see researchers present the results of pitches at the +0.60 to +1.05 feet and the -0.60 to -1.05 feet ranges (i.e., the potential questionable calls). What percentages of the pitches there are called balls v called strikes? What percentage are swining-and-miss, foul, or in-play?
UPDATE:
Then again, when I look at Maddux’s splits:
http://www.baseball-reference.com/pi/psplit.cgi?n1=maddugr01&year=2007
He’s killing LHH: .188/.217/.247
So, maybe LHH are expanding their strike zones against Maddux, and MLB.com has it marked just right?
Wednesday, May 09, 2007
Great news guys. Someone from MLB.com will be dropping by our blog at 11AM ET, Friday, to answer our questions about Gameday.
So, use this thread to hash out all your questions, technical or otherwise, and then come back on Friday as our MLB.com rep answers them, and you can ask any followups, or new questions.
Monday, May 07, 2007
Finally, someone does the roll-up-your-sleeves work, to warn us of not to be too trusting of MLB.com’s GameDay. Fantastic stuff.
Thursday, April 05, 2007
Is this cool or what? It’s by Joe Sheehan, but not THAT Joe Sheehan. If he keeps doing work like this, the BP Sheehan will be introduced as “not THAT Joe Sheehan”.
Thursday, January 04, 2007
I asked MLB.com about getting stopwatches for their scorers. Here is what the Director of Stats had to say:
Read More
from Amazon
Recent comments
Older comments
Page 19 of 342 pages « First < 17 18 19 20 21 > Last »Complete Archive – By Category
Complete Archive – By Date