Tuesday, July 16, 2013
Sabermetric discoveries
Jack Moore penned a pretty good "history of sabermetrics". Along many other threads, one of them was about some key sabermetric discoveries or inventions.
In my view, the key discovery was the run expectancy 24 base-out matrix that Pete Palmer popularized. It is the basis of every strategy decision.
From Bill James, there were a few:
- I would say it was replacement level. It was actually? kind of benign at the time. He used something like it when he'd ask "what's the chance of a .400 win% player putting up a two-year record of 27-21, or 20-11", etc. He'd rank players along those lines. He kind of forgot about that. Then he brought it up in the more common form in comparing Clemens / Mattingly (as well as Guidry / Rice). And then promptly forgot about that as well. He never did that next step that Clay and others did.
- Bill also had the fielding spectrum, and he kind of put a numeric system around it. And Win Shares tried to do something with it, but I think the key work there was done on our old blog, which I think is what really drove the creation of WAR. And Fangraphs (and then BR.com) popularized it.
- Around that same time, he also discovered DER, which is nothing more than 1 minus BABIP. He focused it more on the team level, than the pitcher level. And he never took that next step that Voros did.
Beyond that, something that is totally different from Pete and Bill is Regression Toward the Mean. This is something that I first heard about from MGL, at the old Baseball Boards, with his constant refrain of "regression, regression, regression" (his take on the real estate location term). And its use litters The Book. And I think this is one of the most important "discovery" (or practical use anyway) since Pete and Bill's work. RTTM is best represented by Marcel, which really kept alot of forecasting systems humble.
Anyway, so then Jack asked, since that first-wave of sabermetrics (Pete and Bill), and the second-wave of sabermetrics (mostly inspired by Pete and Bill, culminating with WAR), where are we in the third-wave:
there has been all of one radical, game-changing sabermetric discovery: the notion of catchers impacting the game with pitch-framing. Baseball Prospectus's Mike Fast—now an employee of the Houston Astros—was among the first to publish work suggesting a large spread in this talent among catchers.
And I thought that was interesting. The catcher-framing is really very similar to MGL's UZR (which was also a necessity in terms of popularizing WAR), that we're talking about a zone-based system, and you have your adjustments based on pitchers and batter hand and what not. The key difference really being the quality of the data. But as Max Marchi has shown, you can even get something similar to Mike (given a large enough sample) without even knowing the pitch location, and simply relying on the binary ball/strike calls. (Hence, given a large enough sample, WOWY might be as good as UZR.)
But, what else has there been post-WAR, say since 2009-onwards?
Recent comments
Older comments
Page 1 of 152 pages 1 2 3 > Last ›Complete Archive – By Category
Complete Archive – By Date
FORUM TOPICS
Jul 12 15:22 MarcelsApr 16 14:31 Pitch Count Estimators
Mar 12 16:30 Appendix to THE BOOK - THE GORY DETAILS
Jan 29 09:41 NFL Overtime Idea
Jan 22 14:48 Weighting Years for NFL Player Projections
Jan 21 09:18 positional runs in pythagenpat
Oct 20 15:57 DRS: FG vs. BB-Ref
Apr 12 09:43 What if baseball was like survivor? You are eliminated ...
Nov 24 09:57 Win Attribution to offense, pitching, and fielding at the game level (prototype method)
Jul 13 10:20 How to watch great past games without spoilers