Comments on: Valuing Pat Burrell

By: David Foy

David Foy — Fri, 19 Dec 2008 06:17:43 +0000

Thank you David, for the awesome info. Much appreciated.

By: David Appelman

David Appelman — Thu, 18 Dec 2008 22:27:24 +0000

Here are the basics behind UZR:

I don’t think mgl is using exactly the same methodology and has since made improvements but this is the basic idea.

Sean Smith’s (Chone) Total Zone is based off batted ball data, so fly balls, groundball, line drives You can find more information here:

http://lanaheimangelfan.blogspot.com/2008/11/totalzone-2008-shortstops.html#links

David Pinto’s PMR you can read more about here:

http://www.baseballmusings.com/archives/008270.php

Dewan’s +/- can be read about in the Fielding Bible. I don’t have a link for this.

The reason we use UZR is because mgl agreed to help and I believe that UZR is at least on par with PMR and the +/- system, if not better. Total Zone is very good at what it does, but it just doesn’t have the same level of detail as the others.

By: Eric Seidman

Eric Seidman — Thu, 18 Dec 2008 22:15:18 +0000

Don, could be, but who knows? David, UZR was, until now, not available for any website. MGL has a licensing agreement with the St. Louis Cardinals and couldn’t reproduce it. I cannot speak for Chone or PMR, but both UZR and Dewan compare players to others at their position. So, if Burrell is -14 runs via UZR, it is compared to other left-fielders, thus why positional adjustments are important. Dewan’s system counts the number of plays made above an average fielder at the same position.

By: don

don — Thu, 18 Dec 2008 22:02:35 +0000

However, for the purpose of future projection, should we assume that Pat’s hitting will decline somewhat if he’s hitting off the bench, since he will be adjusting to a position from which it’s harder to hit? That may be what NickP is asking.

By: David Foy

David Foy — Thu, 18 Dec 2008 21:52:02 +0000

Hey can Eric or Dave or anyone on the fangraphs staff explain the difference between Chone, UZR, PMR, and Dewan’s. Also why does fangraphs use UZR rather than any of the other three? If this already been explained than i am sorry but I would like to know something about the differences in methodology (if there is any) between the 4 systems.

By: Eric Seidman

Eric Seidman — Thu, 18 Dec 2008 20:42:42 +0000

Nick, the point isn’t whether or not he has, as the positional adjustments are not concrete or 100% accurate. If Grady Sizemore suddenly went to LF, and we were doing his projection, we would dock him -7.5 runs per 162 games, or prorate it to his playing time. We wouldn’t keep the +2.5 from CF even though he technically never played LF before. The same applies to Burrell. He may not have played DH before, but the adjustment is only used to judge value. Burrell as a DH with the -22.5 + 5.0 is better than Burrell the LF. The +5 added back is to deal with the hypothetical of it being harder to hit off the bench, and it isn’t reserved strictly for those who have played the position in the past.

By: NickP

NickP — Thu, 18 Dec 2008 18:21:01 +0000

As I understand it, DHs get a +5 run bonus because it’s harder to hit as a DH.

Pat has never hit as a DH, so he shouldn’t get that bonus.

Obviously you can do it however you want. But using Tango’s postional adjustments (as you have in the past), I’d give back the 5.

By: Eric Seidman

Eric Seidman — Thu, 18 Dec 2008 17:44:19 +0000

Nick, he would get the +5 runs even without DHing, since there is no rule stating we cannot estimate his value at a certain position. The DH is actually -17.5, but Tango likes to show it in two steps, the -22.5 and then the +5 for difficulty. The only reason he does that is to show the separate steps. Otherwise, DH would literally just show up as -17.5.

By: NickP

NickP — Thu, 18 Dec 2008 16:44:34 +0000

DH gets a -22.5 penalty.

Then, you add +5 to that because it’s harder to DH. However, Burrell has never been a DH, so he wouldn’t get that boost.

That makes the move almost break even. Can he play first?