Baseball Prospectus - Chat with Clay Davenport

Chat: Clay Davenport

Welcome to Baseball Prospectus' Wednesday April 13, 2005 7:00 PM ET chat session with Clay Davenport.

Clay Davenport is an author of Baseball Prospectus.

Clay Davenport: Hello, everyone, welcome to the Baseball Prospectus chat. I'm Clay, and I'll be your answer man for the evening.

Evan (Vancouver, BC): When will we start seeing 2005 EqA reports?

Clay Davenport:
Quite a few of you, understandably, have questions about the stat reports tonight, so let me try to fill you in.

Basically, we have a new provider for the raw stats upon which all of our other stats are built. There have been some problems, most of which are now resolved, and we hope to have them all running very soon.

As for the EqA report, specifically, I figured out how to get around a big problem today, and that should be available tonight or tomorrow on the site. Brian Roberts and Edgardo alfonzo are the early leaders.

Alex (Houston): Who has a brighter future: Lane or Burke?

Clay Davenport: As much as I've loved Jason Lane over the last several years, and as happy I am he's getting a chance to play (finally!), I have to go with Burke. The difference in offense shouldn't be all that great, and the position and defense (plus three years) puts him over the top. IMO, of course.

Alex (Houston): BP 2005 says that Biggio’s “continued employment by th Astros is sinecure, rather than a good baseball decision.” There’s little question that as an outfielder, Biggio sucked. But isn’t your analysis a little harsh? He had 18 win shares—59th most in the NL. Heck, as the worst defensive outfielder in MLB, he had more Win Shares than all but 7 NL second basemen. He is an average or slightly above average 2b, all skills considered. He is most certainly not a detriment. Isn’t it time for BP to admit that they have grossly overstated the argument against Biggio’s continued employment?

Clay Davenport: Is Biggio still good enough to play in the majors? Yes. Should he be doing it for Houston, who has cheaper and better options in the aforementioned Lane and Burke for Biggio's positions? No. In St Louis, sure, but not Houston, not now.

TGisriel (Baltimore): Clay: Sorry I missed your book signing in Baltimore this year. It was a bad day for me. 1. Should the O's hold on to both Julio and Ryan, or should one of them be used as trade bait? Which would you trade and what would you seek? 2.What do you think of Cabrera and Bedard? Is their future bright or are the O's fooling themselves?

Clay Davenport: 1) It always depends on what you can get in trade; it would be either aprospect deal or a pitcher-for-pitcher deal, in all likelihood, because the Oriole offense doesn't have a gaping, easy-to-improve hole in it, and you don't want to do the trade for a better backup catcher. So, barring injuries, I'd hold.

2) I like Bedard a lot, always have, and haven't been so high on cabrera. My only worry on Bedard is more injury.

DePoe (San Francisco, CA): In the words of Mr. Beane: if Hee-Sop Choi can hit, why doesnt he hit?

Clay Davenport: Damn it, Jim, I'm a statistician, not a batting coach.

Seriously, confidence is anoverlooked area, mostly because we don't have access to psych reports and because it takes a lot of confidence just to get ther ein the first place, so you're dealing with differences among people who are already above the curve. I'm sorry to say that my best guess is mental.

mistermoy (minneapolis): A few questions ... 1. Strat-o-matic or APBA? 2. Clint Barmes? Is he for real? 3. Richard Hidalgo or Scott Posednik? in mixed league head-to-head. Thanks.

Clay Davenport: Strat, all the way.

Clint Barmes currently has an eqa of .305. He's never had a full or partial season above .250, so I say no, he's not.

Probably Hidalgo, unless your league counts steals as something really special. Although given that I've got an AL team that punted HR and RBI - such an un-BP thing to do, but my holdover personnel helped dictate that - I'd be taking the Pod.

bobstrab (DE): How much confidence do you have in your defensive ratings? In the book, it seems that the defensive ratings and the author's comments diverge quite often, sometimes wildly. This does not seem to happen to the same degree with offensive ratings.

Clay Davenport: Clearly, I do have, and you should have, a higher respect for the offensive accuracy than the defensive. The input data is better, we actually know what the answer is in terms of runs scored, and a lot more work has been done both by me and others to figure it out. The longer the time period involved, the more confidence I have in the fielding numbers, but I've too many one-year flukes to have full confidence in a single season rating.

Bill Johnson (New Mexico): Subject: catchers' defense and Nichols' Law. (BTW, where is Sherri Nichols these days? Haven't seen anything from her in ages.) Do you think the contention that perceived defense and offense are inversely related is still true? There seem to be a lot of recent counterexamples -- Pudge, Varitek, etc. Maybe the historically heavy-hitting catchers with bad defensive reps (Piazza, Simmons) really *were* bad at defense?

Clay Davenport: Last I heard she had become a big fan of the San Jose Sharks, so she might not have been too happy this winter.

There have always been big counter-examples to NLoCD, i.e., bench, Carter, Campanella. With some things, like basestealing, the results are pretty easy for anyone to figure out. Other aspects, like "calling a game", which I-Rod has been criticized on, still work for the perceptions. The Law isn't a universal; it tends to work better around th edges, for the backups and the starters on second-division clubs, who aren't in the spotlight enough to really show off.

Andrew (Chapel Hill, NC): What minor league (ie PCL, IL) comes closest to the majors in terms of level of competition?

Clay Davenport: The IL tends to rate a little bit higher than the PCL, but I have the feeling that it has something to do with the uniformity of the league. I think the range in performances froms ea level to high mountains makes the concept of "league average" more problematic in the PCL, creating a slight bias that works against the PCL in my rankings. Because, really, the pendulum should alternate between the two, since they have the same role in relation to the majors. Its soemthing I've tried to work out but as yet haven't.

jdietz (Portland OR): Hey, A question about records. Ichiro currently has a 20 game hitting streak going back to last year. Can the consecutive games hitting streak be broken over the course of two seasons, or would he need to hit in 57 games all in one season?

Clay Davenport: The wording of the record books requires it to be in one season, but if he managed to reach 57 over two season there would be some sort of recognition, and an interminable and ultimately pointless argument about who "really" holds the record. Personally, I don't see the end of one season and the beginning of the next as a magic eraser that wipes the slate clean, and have no problem with recognizing a consecutive-X streak across a season boundary.

jwb0581 (Dunbar, WV): Clay - gotta love BP; you guys do a great job. What route do most General Managers take to get their jobs? What types of educational degrees do you need? Do you see MLB teams looking at statistically-minded Harvard or Yale grads more in the future?

Clay Davenport: Not being a GM, it is hard to say. I think the job is transitioning from something that former players worked up to, to a CEO-type position that requires some serious financial and business acumen. Although I sincerely hope that MBAs don't become a requirement.

jjapha01 (Washington D.C.): Eric Hinske is off to a quick start. Is that all it is, or is there some reason to believe he's going to fulfill his potential this year?

Clay Davenport: 1) He's taken a lot of criticism for his defense in the past, and the move to first base may clear some of that from his head. 2) His wrist may hve finally healed. 3) His current EqA is .392, so unless he's got superballs in his bat it's safe to say he's going to cool off. But I do think we're gong to see a season more like 2002 than '03 or '04.

Alex R. (New York, NY): Why have the VORP reports from past years disappeared? When will this year's package be unleashed?

Clay Davenport: To be honest, we've been so caught up with getting the current stat reports working that I hadn't noticed they were gone. It is a bug related to the new setup; the files are still there, the links just aren't pointing correctly. The past eqa reports, for instance, are at statistics/eqa2001.html, and t pst rvorp reports have names like statistics/vorp_player_by_pos2002.html. Those will be fixed very soon.

Josh (Boulder, CO): Hey Clay, what do you think of Jack McKeon's use of his pitchers so far this year? Three complete games in a week, with Willis, Beckett and Burnett all landing right around 100 pitches. Is McKeon going to Dusty Baker these guys into injury, or did I miss out on a season's full of CG bonus points in my league?

Clay Davenport: The trick is "all landing right around 100 pitches". They were all on cruise control, they weren't struggling near the end - there was no reason to pull them. So long as its not a precursor to "If they can throw 100 now, they can throw 140 in August, and pitch themselves out of bases-loaded jams in the eighth", I don't see cause for concern.

Greg (Boulder, CO): Did you ever imagine a dozen years ago when you came up with EQA & EQR that it would basically spawn the entire Baseball Prospectus empire?

Clay Davenport: A dozen? Try 20. No, I had no idea that it would becomne as big as it has, but then I had no idea that I'd be be joined by so many hard-working guys to help build it. I'll take some credit for a few ideas at the beginning, but a lot of people have done a lot more than me to make BP what it is.

James (Chicago): Has anyone done a systematic study of home/road splits for teams that play in high-altitude parks in the minors? If playing at high altitude really does have a negative effect on the Rockies' hitters, one would expect this to occur at all high altitude parks, not just Coors. Looking at high altitude parks in the minors could help resolve this issue.

Clay Davenport: You don't really have that many parks that are in Denver's league, strictly speaking - I think Colorado Springs is the only one that's really at the same level, with the other parks people think of in the high-offense PCL being in the 200-3000 ft height range. Its been a while since I actually looked up the heights of the parks (NOT the airport that serves the region, which is what most tables of city elevation use), so I may be way off on that one. Mexico City, in that league. The problem is data availability; major league data is a whole lot easier to get a hold of than the minors.

Greg (Boulder, CO): While you've been analyzing defensive statistics for your Fielding Runs stats, what has most surprised you?

Clay Davenport: That's a tough queston to answer. Probably how sensitive the results are to a small change in assumptions. Since we don't have absolute data for most of the teams about things like the g/f ratio from the ptchers, that has to be inferred, and it is hard for me to believe how drastically what looks like a small change in the process can affect the results.

Marc Normandin (Dracut, MA): I just wrote an article on my website using WARP3 to determine Tim Raines worth for the Hall of Fame versus some similar players in there...I was wondering how you feel about Raines being in, especially since one of your statistics says its a good idea

Clay Davenport: I have a procedure that I've built to create my own personal Hall of Fame, based mainly on WARP-3 and on the Hall's established rules. I intend to write about that when we get closer to Hall induction time in late July, but yeah, I put Raines in on his first ballot in 2007.

Andy (Raleigh): How many games do you think it will take to get a fairly accurate reading on how RFK is playing this year? I know it played a certain way the last time it was used for baseball, but with the different mix of parks in 2005, I'd think it's park factor would change quite a bit.

Clay Davenport: From the way I've seen parks change in in-season, we should have a really good read on RFK's park factor about when the new stadium opens. At least a full season, as some parks can be really different between the hot (June-July-August) and cooler (Apr-May Sept) parts of the schedule. I still expect DC to be a pitchers par.

Anthony (Long Island): I've been looking at BRAR, PRAR, and FRAR on the team DT cards, and it looks like the offense/defense split is 30/70. This seemed confusing. Why might that be?

Clay Davenport: Short answer: I designed it that way.

Longer: the idea that a run on offense is equal to a run on defense is true at the absolute level, but not necessarily at the replacement level. The difference between average and really bad is larger on defense than on offense - in the extreme case, it is infinite on defense but limited to "score zero" on offense. In order for me to balance the offensive, defensive, and pitchign contributions of players in a way that seems appropriate, I need more rar on the defensive side of the ledger. I think that I have a fair balance of the three, but I do have to consider the possibility that I've built the system to give the answers that I wanted, rather than letting the facts fall where they made. Unless the fact is that they really are that balanced, in which case I'm right to do so. I'm going to stop talking about fair and balanced now, before someone at Fox decides to sue me.

Bill Johnson (New Mexico): Just FYI ref. James of Chicago's Q, Albuquerque is at almost exactly the same elevation as Denver. Minors ball here is weird, weird, weird too...

Clay Davenport: I stand corrected. I didn't recall them being quite that high, but the PFs certainly suggest otherwise.

To follow-up on that point: the systems that we use to "correct" stats back to a theoretical average work best when the necessary adjustment is small. When you have parks like Albuquerque and C-Springs on the one hand, New Orleans and Tacoma on the other, you are making large adjustments - large enough that the players themselves recognize the differences and perhaps alter their game to compensate. Players whose skills are unusally susceptible to those changes aren't going to be affected the "average" amount. We have to remember that the stats are a sample of the true ability, and the extreme environments increase our uncertainty about whether the stats are a representative sample. Bottom line: performances from extreme environments should be more prone to analytical error, making them less trustworthy.

Greg Wilson (Boulder, CO): Any thoughts on WARP using league-average defense as its baseline, rather than replacement level? I know you're tying it to the Cleveland Spiders, but the baseline just seems too low. Studies seem to show that replacements, on the whole, play defense at an average level. How about a separate stat that combines BRAR with FRAA?

Clay Davenport: Not to be too tongue-in-cheel, but my thoughts are that I should have used a different phrase than "replacement", since that has a well-established meaning that I am bucking. Be nice to find soemthing else that started with R, so I don't to change the acronym...anyway, the problem with doingthat is that you then have to create a "position adjustment", which is a concept that viscerally agitates me. The fielding RAR concept was designed in part to get around that; the position adjustment should be defined by the value they produce in the field, not at the plate, althugh in an intelligently runs system they should be approximately equal.

Cris E (St Paul, MN): What's your best guess at Rick Ankiel's future? Do the track records of others with this sort of problem shed any consistent light on this?

Clay Davenport: There really aren't that many players who have had a similar situation. The best example I can think of, a successful pitcher who gave up the mound and came back as a position player, was Joe Wood, 90 years ago. I've got his minor league hitting stats, and if he had been an outfielder, I would have projected him to be able to hit .260-.270 eqa in the majors based on his best seasons, so he has established some ability to hit well enough to make the majors. He still has to go out and do it, and I'd say the dds are against him, but not overwhelmingly so.

beanpj (Washington DC): At what point in the season does BP plan to roll out the Playoff Odds reports?

Clay Davenport: I had intended to run it out at the start, but various problems intervened. That said, I am going to go to work on it right now, and I hope I can turn it out better than before by the weekend. Well, Monday is more realistic.

Clay Davenport: Thanks for all the questions, folks, and my apologies for the ones I didn't get to.