John Dewan is perhaps best-known as one of the founders of STATS, Inc. many years ago. (He is little-known for hiring Rob Neyer, also many years ago, and generally been forgiven for that rare misjudgment.) Today, Dewan heads Baseball Info Solutions and has just published and co-authored the third edition of The Fielding Bible, loaded with fielding stats and original analysis.
Recently, I sat down with John in Mesa, Arizona and pestered him with questions about measuring defense ...
These Questions 3: John Dewan
Rob: John, have you ever given any thought to coming up with a method for estimating runs saved for the old-time players? Because we really don’t have that, or at least we don’t have anything that’s gained any real credibility compared to the modern metrics. It just seems to me we don’t have a great feel for how good a fielder Tris Speaker was, or even Willie Mays and Mickey Mantle.
John: Yeah, there’s just stuff that can’t be done. I remember when I first got my hands on Project Scoresheet data, what I was most interested in was lefty-righty match-ups, ground-ball/fly-ball tendencies; you know, lefty pitchers face more right-handed hitters, who hit more grounders to the left side. And traditional range factors wouldn’t take that into consideration. So what I did back then was count the number of innings. Well, you can estimate that, and you can estimate lineup constructions, how many righties and lefties they’re going to face. Which would allow you, going back, to get a better idea about infielders and their range. But you’ve got to take into account ground-ball pitchers and fly-ball pitchers, and you can estimate that from the stats as well.
So there are things that could be done, and some people have done some things. But I’m not as familiar with them. So if someone would challenge me to do that ... Are you challenging me, Rob?
Rob: Well, I just think if you’re going to call your book “The Fielding Bible”, you just need to go back and, you know, get Willie Mays in there. Somehow.
Anyway, here’s another question that always comes to mind ... The work in The Fielding Bible is fantastic, and I trust your results. And I think the Fielding Bible Awards do an excellent job of identifying the best fielders every year. But there’s also Ultimate Zone Rating, which I think does a fine job. I’m not as sold on the method used at Baseball-Reference.com, but that’s probably due to lack of familiarity more than anything.
I just don’t know that we need all of them, or that we should have to wonder which method does the best job of measuring the number of runs a fielder saves in a season. Shouldn’t it be possible to look at the various defensive systems and figure out which ones correlate best with run prevention? Shouldn’t you be able to add pitching runs and defensive runs and see which method comes closest to predicting actual runs allowed?
Shouldn’t someone run that study? And why not you? And if your method comes out on top, you can say, “Sure, UZR is great but mine is better. So when you’re writing about fielding, you should just use mine.”
John: Yeah, but what I don’t want to do is disparage anybody. I don’t like doing that. I don’t like it when anybody disparages me, and I don’t want to do it to anybody else. But we’ve come up with this technique where we’ve rated our system, based on even/odd-year correlation. The higher that correlation is, the more dependable that stat is; the more consistent players are, from year to year. And it’s a good technique.
I didn’t invent this technique; I think it was someone at The Hardball Times. And I thought, wow, that’s a good technique: comparing a player’s even years to his odd years. And we did it over six years, so you’ve got three years in your sample size on each side. You’re looking for consistency of a stat, and we found there is consistency with our system. We found the correlation for Defensive Runs Saved was 0.59, and we could look at other systems and see how well they do, and in theory the highest correlation should be the best.
Maybe this isn’t the best technique, but it is a technique.
Rob: Okay, now I’m going to pin you down ... Who are the five most overrated fielders and the five most underrated fielders in the Fielding Bible Era?
John: Here are most overrated ...
1. Derek Jeter -- 5 Gold Gloves, 0 deserved (article in book)
2. Matt Kemp -- 2 Gold Gloves, 0 deserved (article in book)
3. Torii Hunter -- 7 Gold Gloves in nine-year Fielding Bible era, first three deserved. After the first Fielding Bible we rated him the best center fielder in baseball. After that he dropped off becoming an average center fielder. Gold Glove voters still voted for him for four more years.
4. David Wright -- 2 Gold Gloves, 1 deserved. Wright has dropped off badly since he had two good defensive seasons in 2007 and 2008. Runs saved totals in 2009, 2010 and 2011: -14, -14, -6.
5. Nate McLouth -- 1 Gold Glove, 0 deserved. We estimate that he cost the Pirates 23 runs defensively compared to an average center fielder in 2008, his Gold Glove year. In the last two years, he has cost the Braves 28 runs in center field, that in less than a half season of center-field play in each year.
And for most under-rated ...
Albert Pujols should have won the last seven NL Gold Gloves, but won only two. Adam Everett should have won three Gold Gloves, won none. Brendan Ryan should have won Gold Gloves in each of the last two years, one in each league. Brett Gardner and Austin Jackson both should have won Gold Gloves in each of the last two years.
Rob: Thanks so much, John. Best of luck with the book, and everything else.
See More:
