Tendu (November 24, 2003)
Some rather interesting work being done:
http://www.thesunlink.com/redesign/2003-04-27/sports/133739.shtml
http://developer.java.sun.com/developer/technicalArticles/ThirdParty/Tendu/
http://tendu.net/
--posted by TangoTiger at 10:39 AM EDT
Posted 5:05 p.m.,
November 24, 2003
(#1) -
Jim
I would think it's awfully subjective to judge pitch type, location, and speed just from watching games on TV. Camera angles are different, the radar speeds shown on the screen graphic are unreliable, and there would be a lot of biases among the individual reviewers.
On the other hand, this seems like a great application for Questec.
Posted 7:35 p.m.,
November 24, 2003
(#2) -
Ron(e-mail)
Jim's comment is a very good one. The process that we use for hiring only former players who have seen a lot of pitches, for training and monitoring accuracy and consistency is extensive, for precisely Jim's reasons. Our data collectors freeze and rewind each pitch as many times as they need to, to increase our accuracy to as high as we can get it. Nevertheless, it is subjective data that will never be perfect, but that doesn't remove all of it's value. The kind of tendencies that we find can be discovered in this data and small numbers of mistakes are hidden in the volume of data. Therefore, you have worry about it most, when you have small sample size.
Posted 12:36 a.m.,
November 25, 2003
(#3) -
Jim(e-mail)
Ron, I agree that the system still has value despite the subjective element. But it just seems like the manual process your data collectors use would be very tedious and error-prone, in addition to the biases mentioned above.
I'm curious, has your company made any attempts to work with Questec, perhaps to use data obtained from their SuperVision PitchTrax at least for a few games for quality-control purposes? Tendu sounds like the sort of system that would be much more accurate if the data collection process were automated.
Posted 12:43 a.m.,
November 25, 2003
(#4) -
Jim
I just realized that Questec already claims to have many MLB teams as clients and that they also apparently offer tools for statistical analysis. Perhaps Tendu views Questec as a direct competitor.
Posted 7:28 a.m.,
November 25, 2003
(#5) -
studes
(homepage)
FYI, this was written up in Newsweek a month or two ago. I can't find the link. The Tendu guy must have a good PR firm.
Posted 9:20 a.m.,
November 25, 2003
(#6) -
Tangotiger
(homepage)
Studes, see above link. It was written by Alan Schwartz of Baseball America (who also did that Carl Morris article) for Newsweek.
Posted 11:54 a.m.,
November 25, 2003
(#7) -
tangotiger
Ron mentioned the training practice to me, which I will quote in full:
"I only use former players for recording this data, players who have many years of seeing these ptiches. Every player must prove to me his knowledge of pitch movement, pitch grips, pitch follow through, pitching mechanics and prove it via tests before they are hired. Every data collector goes through a one month probation, proving that the hiring tests were not flukes and also proving that they will focus intensely and be dedicated to data accuracy. Roughly half of the guys who start are dropped during training, or more likely, they quit."
Posted 4:47 p.m.,
November 25, 2003
(#8) -
Jim
"Every player must prove to me his knowledge of pitch movement, pitch grips, pitch follow through, pitching mechanics and prove it via tests before they are hired."
OK, but what makes Ron qualified to be the judge? I would expect even the most experienced experts on the subject to have significant differences of opinion on matters that are fairly subjective.
Posted 4:53 p.m.,
November 25, 2003
(#9) -
tangotiger
Well, his quote did say to prove it to him via tests. So, just as umpires are evaluated by their peers, Ron seems to imply that his tests were constructed by professional players (akin to the umpires). I don't see Ron as the judge (evaluator) here, any more than Sandy Alderson is the evaluator of the umpires.
Posted 12:03 a.m.,
November 26, 2003
(#10) -
Jim
I think part of the issue is inherent to the difficulty of judging exactly where a baseball traveling 90 mph passes with respect to an invisible floating 3-dimensional object. MLB umpires have far more experience and the best view of the pitch. Yet they are still notorious for having many different interpretations of the strike zone. I would think Tendu's data collectors, despite having the ability to use replay technology, would still be limited by not being there live and not having the proper camera angle (or even a consistent one from park to park). This is likely a significant hinderance to the accuracy of the system.
Posted 12:15 a.m.,
November 26, 2003
(#11) -
MGL
Two things:
One, it is simply not that important to have anywhere near perfect data input. Two, as one who already does this on my own, using games downloaded from MLB.com's website, I can tell you that after watching thousands of games in my life on TV, it is a relatively simple task to judge pitch selection (if you are ever in doubt, you simply look at the pitch speed). Occasionally a splitter and a changeup are hard to distinguish, but who throws a splitter and a changeup and at around the same speed? Same thing for a curve and slider. Very occasionally, a change and a curve look similar, although a curve is usually thrown much slower. A cutter as opposed to a regular fastball can be problematic, but who cares. Ditto for a 2-seam and 4-seam fastball. As long as you get 90+% of the pitches right (which you should), again, who cares?
Pitch location is ONLY difficult when the batter hits the ball, and even then, you have a pretty good idea with practice. And if you are a Tendu stringer, practice you will get!
As far as pitch speed, you hear about the TV guns being so "inaccurate." As someone who watches more than 200 live games a year, that is BS! And even if it's true, again, who cares? A pitcher's average velocity on each of his pitches will even out even with different guns in different games. And once again, no one is going to care (nor should they) if there is a little "slop."
I have nothing to do with the company. I am only defending them because this "crying" about how hard it is to "string" games is ridiculous. It's not hard! Now if Tendu were not hiring and/or training players properly and consistently, or if they didn't have some sort of quality control system to make sure that stringers were not slacking off and/or faking results, that would be a different story...
Posted 10:27 a.m.,
November 26, 2003
(#12) -
Jim
My guess is that Questec is significantly more expensive than Tendu. Questec is probably more accurate, but I'm guessing they don't yet have the capability for 100% coverage of all MLB games. Aside from each company's marketing claims, does anyone have any inside knowledge of how the client teams are using each technology? Getting back to my original point, it seems like a combination of the two systems would be best, and it would give you some quality control and some idea of the strengths and weaknesses of each system. In fact, it wouldn't surprise me if the Athletics are already doing this.
It's probably a lot like scouting vs. sabermetrics. Knowing how to use both in concert would give you a nice advantage.
Posted 11:13 a.m.,
November 26, 2003
(#13) -
tangotiger
The ideal would be to have a "FoxTrax Puck" that measures the precise location and spin of the ball at all times. Anything that tries to do that is great.
I suppose Questec tries to do that, and I'm sure it would be a great feed into Tendu.
The Tendu scorers apparently, according to one of the articles, spend 12 hours per game recording the various ball locations (strike zone, and field). I have to figure that you'll get SOMETHING of value. It's certainly better than not recording any pitch location whatsoever.
And, if the scorers are consistent, and there's quality control, it must be half-decent. And if you do it for 700,000 pitches, the poor quality that may exist (you still drive a car, and buy a computer, even though some parts break down don't you) will certainly not take away from the incredible benefit that you may gain.
Everything comes with a margin of error. The key is for Tendu, STATS, anyone else to SHOW what that margin of error is. It's the only right thing to do.
Posted 2:29 p.m.,
November 26, 2003
(#14) -
J Cross
A cutter as opposed to a regular fastball can be problematic, but who cares. Ditto for a 2-seam and 4-seam fastball.
Every player must prove to me his knowledge of pitch movement, pitch grips, pitch follow through, pitching mechanics and prove it via tests before they are hired.
yeah, I don't think there's anyway you can see fastball movement on TV but if you slow it down you might be able to see the grip. To distinguish btw the 2-seam and 4-seam fastballs you'd need to see the stitching on the ball which might not be possible. Maybe you just have to know what a pitcher throws.