Defensive Doctrine: An Interview with John Dewan

Ad1_89I am posting an interview with John Dewan, author of the new book, "The Fielding Bible", that is currently appearing on our website today. I think you’re going to like it!

More from me next week.


Written by Joe Hamrahi   
Friday, 24 March 2006

Today we are privileged, and I mean that in every sense of the word, to talk to a real innovator and pioneer in the field of baseball statistics. John Dewan of Baseball Info Solutions (BIS) sits down with us today to talk about his new book, The Fielding Bible. Most of you know John from his years of work at Project Scoresheet and Stats, Inc. He is also a good friend and colleague of Bill James.

Joe Hamrahi (JH): Hi John.

John Dewan (JD): Hi, how are you doing?

JH: Great, thanks. Thank you so much for taking time out of your schedule to talk to us about your new book.

JD: Sure, no problem.

JH:  Just to start things off, how long have you been working on the revised fielding zone ratings and the plus/minus system before they were actually published?

JD: Actually, I began working on this during the 2003 season. BIS started in the 2002 season. I was looking at data from the 2002 season and the beginning of 2003, trying to come up with something new. Fielding has always been a special interest of mine. Back when I was with Project Scoresheet, I immediately starting using data to try to digest fielding. In particular I was trying to eliminate the lefty-righty bias. A team that throws a lot of lefties tends to face a lot of righties. Those righties, in turn, will hit two-thirds of their grounders to the left side of the infield. And vice versa. Lefties will hit about two-thirds of their ground balls to the right side.

So when you have a lefty predominant pitching staff, you have a lot of balls hit to the shortstop, and in particular, third base, which tends to inflate their numbers. So when I was at Project Scoresheet that was something we worked on. I actually gave a presentation at SABR regarding the work that we did on that.

Then when I got involved with Stats, I came up with zone ratings. At Stats we kept track of the location of batted balls, and I came up with a system that evaluated that.

Then, at BIS, we concentrated on getting even better hit location information. A pixel on a computer screen was used. We’re now measuring distance from home plate within a foot. We have like 260 vectors along which we measure so we get very good hit location data.

JH: Do you use film and transpose it onto a computer screen?

JD: Yeah, we chart every game off of videotape. And during that charting, we can really pinpoint the hit location and review every play to make sure we’re getting the location right. We can also do that for pitch location.

So for fielding, we’re looking not only at the location, but the velocity of the batted ball…not radar gun velocity. We’re not quite there yet, but you know…slow, medium, hard hit, line drive, pop ups.

JH: It’s eyeballed I guess, correct? And it’s somewhat subjective?

JD: Right. There is a subjective element to it. But I feel so good about it because the results are consistent with what you’d expect. For the players you know are good, the results are consistent. There are some players that fall outside the (constraints) and you can find the problems.

You can look at a guy like Mike Lowell, a Gold Glove third baseman who looks tremendous at third base. And he is tremendous. He is the only “A” rated fielder on bunts in each of the last three seasons on our system. So he’s great at handling bunts, and he is great on fielding balls down the line. But it turns out he is substantially below average on balls hit straight on and balls hit to his left. And he’s so much below average that it really makes him a fairly average third baseman overall in my eyes. I don’t consider him one of the top third baseman, let alone the Gold Glover. It could be all positioning, but I also think it’s a bit more than that.

JH: You bring up a good point. How much does positioning matter in the end? Does it all equal out in the end after 162 games? Or pretty close?

JD: I think positioning is important, but if a guy makes a play or doesn’t make a play, what we’re trying to measure is whether he made the play or not. That’s really what it comes down to. Either they are making the plays or not making the plays. Whether he makes the play because he can run the play down, or he makes the play because he gets a tremendous jump, or whether he makes the play because he is smart enough regarding positioning or his coaching staff is smart enough regarding positioning, or if he has a great arm…whatever the reason they’re making the play, they’re making the play!

If we were able to figure out where each player started out at his position, we could then figure out how much range he had to get to the ball. But that’s just breaking it down further.

JH: And I’m sure this is just one of several steps you will take over the years.

JD: Right, and that’s a good comment you make. This is a process, but we have not hit nirvana. We have advanced the study of fielding and taken some big steps, but there’s more that can be done. We’re thinking about it every day.

JH: Is there a team of people who make the decision on the subjective data…like how hard a ball is hit? Is it one person? And what does BIS do to ensure that there is consistency in judgment across the team?

JD: Yeah, we have a limited number of scouts or charters, whatever you want to call them, that are doing the videotape analysis. They have gone through extensive training on how to chart everything. So, it’s a limited number of people who have received extensive training, and because of that, we think there’s great consistency in the information. It’s not perfect, but we feel the results we are getting are meaningful.

JH: How much of a role did Bill James have with the formulas and the book? Indirectly or directly?

JD: Well, the relative range system is Bill James. He developed that from scratch, and it’s a very good system because it not only allows you to evaluate current players, it lets you go back in history and evaluate, say, all the Gold Glove shortstops. His results are very accurate and what you would expect to see based upon subjective information. For example, Ozzie Smith rates number 1 among shortstops. Everybody knows and everybody feels, and I’m certain everybody is right, that Ozzie Smith is the best shortstop who ever played the game.

So, you like to see a subjective rating system that comes out that way. That’s 100% Bill James.

The plus/minus system is something I invented. What Bill helped with was making it a plus/minus. I had it as a fraction…a lot like my zone ratings. He suggested that I convert it to a plus/minus system. I was like, “Yeah!” So that was his addition to it.

Bill and I had a lot of discussions about fielding. On the zone system, he didn’t like some of the things, and in the book, we changed those things to come up with revised zone ratings.

One of the missions of the book was to look at fielding from as many angles as we could. The revised zone ratings were one way, Bill’s system was another, and then there was the plus/minus system. We also looked at double plays for infielders, handling bunts for corner infielders, and outfielder throwing arms. We wanted to look at from as many ways we could and just advance where we’re at.

JH: Obviously Bill did the analysis on the shortstops so let’s talk a little bit about the Derek Jeter – Adam Everett analysis for a minute. Obviously Derek Jeter’s fielding is a heavily debated topic. I’m in New York (but not a Yankees fan!) so I hear quite often how great a fielder Derek Jeter is. The statistics in the book don’t support the fact that Derek Jeter is such a great shortstop. What do you think the reason is that we hear that Derek Jeter is so good? Are we missing something here? Are the plays he does make just flashier than others?

JD: I do think there is an element that makes him flashier than others. When he comes in on a slow roller and makes a throw on the run, that really looks good. In fact, according to the data, he does alright on softly hit ground balls. I think overall he was a +2 or something on slowly hit ground balls. But the fact is, he doesn’t get to as many ground balls as other shortstops. It’s a combination of his positioning…he plays shallow…and his throwing arm. He has learned to make the most of that weakness. But he doesn’t get to medium or hard hit balls to his left or right. And he’s not getting to those balls because he does play shallow.

He does do a lot well though. As Bill says in his article, you can’t be that great a ballplayer and play as much as he does and contribute as much as he does. But, a lot of great players have flaws. And he has a defensive flaw. You know, he’s smart. He does smart things on the field that aren’t involved with ground balls. That’s to his credit. I don’t evaluate him as one of the worst shortstops, but I don’t consider him one of the best either. But he does do a lot of good things, and he is a leader. That’s very important to a team.

He is also very good on balls hit in the air. Over the last three years he has the highest rating of all shortstops on balls hit in the air. So there is a defensive element where he’s the best! Not just good, but the best!

Bill and I have been studying pop ups, and we think we may be understating the contribution infielders make on balls hit in the air. We’re researching that as we speak. So as we develop this system ever further, we may find that (Jeter) is actually better than what his (+/-) number is right now.

JH: Again, we go back to the evolution process and as time goes by, we can see changes…

JD: Right. We’re not just concerned with if a player is good or bad. We want to know why. And are there factors that we need to consider…

JH: While we’re on the topic of shortstops, there was a bit of an uproar in the Strat-O-Matic community this year about whether Jhonny Peralta should be a “4” or not in the field.

The scouting report on Peralta in the book is not exactly glowing, but it’s not exactly unfavorable either. What is your general opinion on Jhonny Peralta?

JD: We found out a few key things about Jhonny Peralta when we did the book. The first thing was that he was a -14 while playing shortstop. But the other key thing was that he had the highest double play percentage of any shortstop in baseball. So he did a very nice job of turning the double play. Now, I have to be honest with you. I have not personally seen many games with Jhonny Peralta playing shortstop. And that has to be weighed in.

We asked people’s opinions about what they thought about Jhonny and they mostly felt he started out poorly, but settled in as the season went on. If it were purely based on numbers, I’d factor in the -14 with the great double play, and I might have given him a 3. But I don’t have a real problem with the 4 either. It remains to be seen how he does. Can he improve and can he remain consistent on the double play? That statistic has shown both consistency and inconsistency with players over the years.

JH: I know that Hal Richman of Strat-O-Matic has had a big influence on you over the years. Do you guys do any work with Hal Richman on any of these ratings?

JD: Hal is my good friend! Hal is a good man who is passionate about what he does. He wants to make that game the absolute best it can be. There’s no one who does more fielding analysis than Hal Richman. So when I was going to do this book, I contacted him. I said, “Hal, there is nobody that does as much fielding research as you do. Can I get some assistance?”

Hal said there was nothing personally he could do, but he offered the assistance of his number one person who works on fielding for him, and that’s Len Schwartz. Len worked on all the player comments. I got the benefit of all the work that they did to supplement the numbers…all the subjective work. They have their own reading that they do, but they solicit help from various people throughout the industry and various parts of the industry. There are broadcasters involved, scouts involved, former players involved…

They do more research on fielding than anyone I could possibly think of. I wanted to make sure I got their assistance.

JH: To switch gears a little…there’s been a little criticism of your comment that each extra play made is about equal to one-half a run. A lot has been written recently that, in actuality, each extra play is worth three-quarters of a run. Is it that important really? What would you say to the critics who disagree with you?

JD: I would say we still have more research to do with what these numbers can do to measure defense. To get a real good runs prevented number, a lot more has to go into the analysis.

I’m not convinced that the linear weights system (which is where the .75 number comes from) can be used in a one game environment. I’m not convinced it can’t be! I’m open to discussing it. Is it .5? Is it .75? Is it something in between?

For now I am comfortable with the .5.  We have a rule of thumb to work with as a starting point until we do more research. In fact, I put in an email to Tom Tango (Author of The Book: Playing the Percentages in Baseball) about that very thing.

He put together a great book. I read it, and he did a great job. For 20 years I wanted to do a book like that, and these guys did it. In all these years, I wanted to write that book, but never got to do it.

JH: You we’re pretty busy though!

JD: That’s true. It’s so funny. They finally did this work that I wanted to see done, and it’s great.

JH: You obviously are very familiar with Dave Pinto’s fielding system as well. If you had to pinpoint a few things, what would be the main differences between Dave’s system and the plus/minus system?

JD: When I came up with the concept of doing this book, I said I was going to talk to all the people who have fielding systems and research all the different systems out there, and if I had time, put them into the book or relate to them.

I started with UZR by doing research on the web. I invented zone ratings, and someone (Michel Lichtman) came up with ultimate zone ratings. I thought that would be interesting. But honestly, I couldn’t find anything to figure out what it was. I have no idea how he does that system. So I decided to put that on the side burner.

JH: I’ve had a hard time finding that information again recently as well.

JD: Right and Michel Lichtman now works for the Cardinals.

JH: And all that information is proprietary…

JD: Right. So I said ok. David (Pinto) has something. Let me talk to David. So David and I had some back and forth email discussions. Then we had some phone conversations. I discovered the basic premise of his system is very similar to the plus/minus system. There are some things I am doing that he doesn’t do, but the premise is consistent, and I would expect that a lot of his results are similar. Of course there will always be some differences.

JH: One last thing…looking toward the future, are there any things that have come up since the book was published that you may want to look at more closely…positioning was one thing, maybe ballparks…

JD: Ballparks is definitely one thing I’d like to look at. Ballpark adjustments were something I wanted to do, but just didn’t have the time to do it.

JH: Were there any other things that may have cropped up that you thought you really need to include?

JD: Having the corners in was one thing. Oh, the big thing we’re doing this year is “flyners.” Making the distinction between a fly ball and a liner was difficult so we came up with something in between. There will be more things like that.

Our mission is to keep improving the system to make it the best it can be.

JH: Well the book is already a great one, and I highly recommend it to everyone. Thanks again for taking the time today John.

JD: I really appreciate it.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s