Q&A with John Dewan
First of all, thanks to all the questions I got from readers. Some really interesting stuff. Secondly, thanks to John for answering my questions.
John Dewan is the author of The Fielding Bible and is one of the leading thinkers on the statistical analysis of fielding.
Dewan has just released his 2007 Fielding Bible Awards, based on the voting of 10 panel members, including Dewan, Bill James and Joe Posnanski. Here are the winners:
FIRST BASE—ALBERT PUJOLS, ST. LOUIS CARDINALS (91 POINTS)
The only repeat winner; his excellent defense is becoming as well-known as his prodigious offense.
SECOND BASE—AARON HILL, TORONTO BLUE JAYS (82)
Edged out the 2006 winner, Orlando Hudson, by two points.
THIRD BASE—PEDRO FELIZ, SAN FRANCISCO GIANTS (89)
A surprising but clear-cut winner with 89 points. He is especially adept at handling bunts.
SHORTSTOP —TROY TULOWITZKI, COLORADO ROCKIES (87)
A rookie wins the award! You can’t say that too often about the Gold Gloves.
LEFT FIELD —ERIC BYRNES, ARIZONA DIAMONDBACKS (85)
Sports Illustrated anointed him the second coming of Pete Rose. He beat Carl Crawford by three points.
CENTER FIELD —ANDRUW JONES, ATLANTA BRAVES (86)
It briefly seemed that Jones’ star was fading, but his range and intimidating throwing arm rebound in 2007.
RIGHT FIELD —ALEX RIOS, TORONTO BLUE JAYS (73)
Last year’s runner-up in right field fills the spot vacated by Ichiro Suzuki, who finished third in center field.
CATCHER—YADIER MOLINA, ST. LOUIS CARDINALS (83)
Threw out 49% of would-be base stealers and upstaged last year’s winner, Pudge Rodriguez.
PITCHER—JOHAN SANTANA, MINNESOTA TWINS (62)
Only 11 base runners attempted steals against this former Cy Young Award winner; five (45%) were gunned down.
So, anyway, here's the Q&A, some of the Qs are from you, some are from me and all the As are from John. Thanks again, John:
Q: Got the list of the Fielding Bible Awards and Aaron Hill was the winner at second base and in the comments, it said Orlando Hudson was second. What about Brandon Phillips? How does he rank in your rankings?
A: Brandon Phillips did very well. He came in third in the voting and had a +11 for the season.
(Editor's note, as a follow-up, I asked him where the +11 ranked Phillips, and he said it was seventh among MLB 2B)
Q: Adam Dunn and Ken Griffey Jr. have been rating very low on your rankings, how much does that hurt the Reds and their pitchers?
A: Dunn was horrible in left field once again (-29 in 2007). Griffey doesn't hurt the team playing right field as much as he did in center. He was a -8 in RF in 2007 after posting -29, -24 and -24 in center field in the previous three years. Between Dunn and Griffey that's -37 in 2007 and that costs the team 20-30 runs more than a couple of average outfielders would have given them.
Q: As a follow-up, have you or anyone else come up with some sort of matrix to see if a player's offensive prowess overcomes their defensive liability? And if so, where do those two players come into play in that kind of situation?
A: If Dunn is giving up about 20 runs defensively, he needs to produce 20 more runs than the average offensive left-fielder to reach average overall. Does he do that? He had the third highest OPS in left-field (page 387 of the Bill James Handbook 2008). I could do a bunch of math to figure this out, but I'm sure it will come out close. Suffice to say, Dunn hits like a first baseman, and plays left field like one too. Griffey was fourth in OPS among NL right-fielders.
All in all, I'd say yes, between the two of them their offense overcomes their lack of defense.
Q: Many of the statistical analyses depend on a guy in the press box charting plays. How much margin of error is there when determining a players' fielding ranking based still on a human judgment? Can this ever be overcome?
A: There is human error in every endeavor. We've found that there is consistency in the charting done by our Video Scouts. Having said that, technology may one day allow for even greater accuracy. Think of it this way. Umpires do a great job of calling balls and strikes. Is it perfect? No way. But it works. Technology may some day replace them, but I doubt that will happen in our lifetime (well, my lifetime anyway).
Q: The Reds went into 2007 talking about how they wanted to be improved defensively, and their errors were down, however, how much -- according to your analysis -- did they improve?
A: Yes, the Reds improved. But that's a bit of a stretch. We publish team plus/minus figures each year in the Hardball Times Baseball Annual. Last year the Reds were at -61 as a team. In this year's book (coming out December 1st), they came in at -59. But their ranking in turning double plays went from 26th to 16th, a significant improvement.
Q: I'm not sure if you've got enough data to analyze him, but what about Jeff Keppinger? Did he do enough to be considered an everyday shortstop and not be a liability?
A: Keppinger was a -6 in 47 games at shortstop. Extrapolate that to a full season and you get a -15 to -20 or so. If you can extrapolate his .332 batting average as well, he'd be great to have as an everyday shortstop. But if he's going to hit .290 with average on-base ability and limited power and speed, he would be a liability. But as you say, there's not enough data and it's too early to tell.
Q: Have any teams approached you about using your services?
A: Baseball Info Solutions provides services to about half of the teams in Major League Baseball, including each of the last four World Champions. We have various products, but defensive information is our specialty.
Q: Are you currently tracking, or are you planning to begin tracking "hang time" for fly balls and "time to get through infield" on ground balls? This, along with the hit location data you already collect, would allow you/us to back-calculate trajectory and would eliminate the need for qualitative, subjective categorizations of hit speed or batted ball type (fly vs. fliner vs. liner). ?
A: We're not using a stop watch, if that's what you mean. But we factor in virtually the same thing by utilizing both the speed (soft, medium, hard) and the type (bunt, fly, liner, fliner, grounder) of batted ball. Can we get more precise at some point? Probably.
Q: Is there any part of fielding statistics that you're working on improving? Will there ever be a way for fielding statistics to tell us as much as batting statistics?
A: We're always working on improving. Since we went to print with the Handbook just two weeks ago, we have already added a new element to the Plus/Minus System. It's the Manny Adjustment. We no longer "punish" outfielders on balls that hit an outfield wall at a height that is physically out of reach. We now treat these balls the same way as home runs by removing them from consideration. Manny Ramirez went from a -38 to a -24, a huge improvement, but still a poor performance. Coco Crisp winds up with the highest plus/minus figure (+26) in center field with the adjustment.
Q: How much of your statistics -- and how much of what you value in measuring fielders -- are indicative of not messing up (routine plays) and how much is the ability above and beyond other players?
A: Messing up a routine play can hurt you just as much as a spectacular play can help you. In our system, if you miss a play that 90% of all other fielders at your position make, that's -.90. on that play. If you make a play that 90% of all other fielders would miss, that's +.90.
Q: Which Reds players, or any, do you consider exceptional defensively?
A: You have a great double play tandem in Brandon Phillips and Alex Gonzalez. I mentioned Phillips already. Gonzalez was slightly above average in plus/minus (+4) in 2007 but he ranked number two in turning double plays among shortstops. In 2006 when he was with Boston, he should have won the American League Gold Glove over Derek Jeter based on our Fielding Bible Awards voting. We crown only one defender as our Fielding Bible Award winner (Adam Everett last year and Troy Tulowitzki this year), but Gonzalez was the highest ranking American Leaguer in our voting. The Handbook shows the entire voting breakdown, including a ranking of all players receiving votes at each position.
Again, thanks so much for your questions and to John and his time.