The game is not valuing WAR correctly, or using a formula I am not familiar with.

Pacoheadley · 08-07-2012, 01:08 PM

I have noticed for a while that the game did not seem to be valuing some players correectly with WAR, or using some different formula. I have always had a feeling that, for example, OOTP's WAR undervalued catchers some. I looked at some catchers in my game to test this.

In my last season, Jarrod Saltalamacchia had an OPS+ (I am using OPS+ because it is close enough to real life wRC+) of 89, while only playing at catcher. For the season he had exactly 0.0 VORP in 423 plate appearances, the offensive part of WAR. Looking at Russell Martin in real life, he currently has a wRC+ of 89, in 309 plate appearances. He currently has a WAR of 1.3 though, while playing neutral defense. He also had a wRC+ of 89 in 2010, in 387 plate appearances. His WAR minus defense was 1.7.

I will use another example here, this time with someone who is hitting well. In my last season (593 PA), Brian McCann has a 125 OPS+, and his VORP was 34.4 (3.4 oWAR). In real life, in 2010, he had a wRC+ of 124 (566 PA). The offensive part of his WAR was 5.0, even though he hit about the same in less plate appearances.

I am not sure at the moment if any other positions are valued wrong, but I have a feeling DH is, although I'd need to test it first to find out for sure. These WAR values are just too far off, it is like they are not receiving a positional adjustment at all for playing catcher.

RchW · 08-07-2012, 01:45 PM

My guess is OOTP is reporting oWAR only. Not adjusted for position. I stand to be corrected as it is a pure guess.

Pacoheadley · 08-07-2012, 02:13 PM

Quote:

Originally Posted by RchW

My guess is OOTP is reporting oWAR only. Not adjusted for position. I stand to be corrected as it is a pure guess.

I might have to test that, but I think there is at least a small positional adjustment for some positions. I have noticed shortstops seem to be rated fairly, for example.

olivertheorem · 08-07-2012, 03:21 PM

I think Markus is just using different math than baseball-reference or anyone else, but still using offense+defense (and maybe +baserunning) in some form. I haven't seen any glaring issues with WAR in my games.

Pacoheadley · 08-07-2012, 07:26 PM

Quote:

Originally Posted by olivertheorem

I think Markus is just using different math than baseball-reference or anyone else, but still using offense+defense (and maybe +baserunning) in some form. I haven't seen any glaring issues with WAR in my games.

I just think that, even if it is his own formula, it underrates catchers too much, and overrates DHs too much. Also, if it is his own stat, then it shouldn't be called WAR...

olivertheorem · 08-07-2012, 08:22 PM

I recently finished going through six leagues (1 universe), manually determining award winners and "all-stars" (not the in-game ones, but something I'm doing outside of the game). In any event, the highest WAR for a catcher I found was 8.9. I forgot off-hand how many were over 5.0. I do think that the relatively limited playing time, even for a healthy catcher, tends to depress their WAR totals a bit.

mets8484 · 08-07-2012, 09:34 PM

Different sites calculate WAR differently. For instance, I know baseball reference and fangraphs have different formulas for their WAR. I am sure Marcus has a different formula for his.

JWay · 08-07-2012, 09:39 PM

I'm guessing OOTP catchers get hammered by playing time, where as other websites adjust catchers to something less than 162 for replacement level. It would make sense to assume that a backup catch will have to be used at least for 30 or so games. Where as other positional players should miss no more than 5-10.

That makes the most sense.

Pacoheadley · 08-07-2012, 10:28 PM

Quote:

Originally Posted by JWay

I'm guessing OOTP catchers get hammered by playing time, where as other websites adjust catchers to something less than 162 for replacement level. It would make sense to assume that a backup catch will have to be used at least for 30 or so games. Where as other positional players should miss no more than 5-10.

That makes the most sense.

That doesn't explain why the difference is so big though. The difference in the McCann example was 1.6, in about a full season. The playing time can't be it, because in the game an OPS+ of about 89 seems to be replacement level for catchers, while that is pretty solid in real life. I just wish he would tell us the methodology behind how it is calculated, because it is obviously not the same as either Fangraphs or Baseball Reference.

kingcharlesxii · 08-08-2012, 02:55 PM

Unless I'm completely mistaken, WAR in OOTP is just (VORP/10)+(ZR/10). ZR is the Zone Rating at all positions. If it's not that, it's something very close to that.

jaa36 · 08-08-2012, 03:01 PM

Quote:

Originally Posted by kingcharlesxii

Unless I'm completely mistaken, WAR in OOTP is just (VORP/10)+(ZR/10). ZR is the Zone Rating at all positions. If it's not that, it's something very close to that.

And to my frustration (and frequent bug reports), in some places, it doesn't even incorporate the ZR portion, for example in the all-time and single-season leaderboards.

Pacoheadley · 08-08-2012, 09:15 PM

Quote:

Originally Posted by kingcharlesxii

Unless I'm completely mistaken, WAR in OOTP is just (VORP/10)+(ZR/10). ZR is the Zone Rating at all positions. If it's not that, it's something very close to that.

Which means his VORP formula is not being used correctly.

Markus Heinsohn · 12-19-2012, 03:57 AM

Pitchers use the WAR formula of Fangraphs, and for position players I use (VORP + ZR) / 10.

The Wolf · 12-19-2012, 09:09 AM

And there you have it. Thanks, Markus.

The Wolf · 12-19-2012, 09:42 AM

A discussion of this with definitions is here: http://www.ootpdevelopments.com/boar...-war-ootp.html

Leo_The_Lip · 12-19-2012, 12:37 PM

While the discussion thread will prove interesting and, hopefully, informative, is it really needed? Future versions of OOTP will likely not use WAR as anyone is currently calculating it because there will be new algorithms developed that claim to do the job better and they will. I have no doubt these new measurement systems will be included in the future. For now, this is the way it is, and that should be acceptable to everyone. Me, I'm keeping my eye on Nate Silver. Now that he knows how all of us are going to vote even before anyone is nominated, I'm hoping he goes back to baseball and develops a new system which he names after a more worthy player.

08-07-2012, 01:08 PM	#1
Pacoheadley All Star Reserve Join Date: Feb 2012 Location: Kincheloe, MI Posts: 518	The game is not valuing WAR correctly, or using a formula I am not familiar with. I have noticed for a while that the game did not seem to be valuing some players correectly with WAR, or using some different formula. I have always had a feeling that, for example, OOTP's WAR undervalued catchers some. I looked at some catchers in my game to test this. In my last season, Jarrod Saltalamacchia had an OPS+ (I am using OPS+ because it is close enough to real life wRC+) of 89, while only playing at catcher. For the season he had exactly 0.0 VORP in 423 plate appearances, the offensive part of WAR. Looking at Russell Martin in real life, he currently has a wRC+ of 89, in 309 plate appearances. He currently has a WAR of 1.3 though, while playing neutral defense. He also had a wRC+ of 89 in 2010, in 387 plate appearances. His WAR minus defense was 1.7. I will use another example here, this time with someone who is hitting well. In my last season (593 PA), Brian McCann has a 125 OPS+, and his VORP was 34.4 (3.4 oWAR). In real life, in 2010, he had a wRC+ of 124 (566 PA). The offensive part of his WAR was 5.0, even though he hit about the same in less plate appearances. I am not sure at the moment if any other positions are valued wrong, but I have a feeling DH is, although I'd need to test it first to find out for sure. These WAR values are just too far off, it is like they are not receiving a positional adjustment at all for playing catcher.

08-07-2012, 01:45 PM	#2
RchW Hall Of Famer Join Date: Jul 2004 Location: The big smoke Posts: 15,628	My guess is OOTP is reporting oWAR only. Not adjusted for position. I stand to be corrected as it is a pure guess. __________________ Cheers RichW If you’re looking for a good cause to donate money to please consider a Donation to Parkinson’s Canada. It may help me have a better future and if not me, someone else. Thanks. “Conservatism consists of exactly one proposition …There must be in-groups whom the law protects but does not bind, alongside out-groups whom the law binds but does not protect.” Frank Wilhoit

12-19-2012, 03:57 AM	#13
Markus Heinsohn Developer OOTP Join Date: Dec 2001 Location: Germany Posts: 24,739	Pitchers use the WAR formula of Fangraphs, and for position players I use (VORP + ZR) / 10. __________________ Buy Out of the Park Baseball 25 now! Download OOTP Go 25 for iOS Download OOTP Go 25 for Android

08-07-2012, 03:21 PM	#4
olivertheorem All Star Starter Join Date: Jun 2006 Posts: 1,919	I think Markus is just using different math than baseball-reference or anyone else, but still using offense+defense (and maybe +baserunning) in some form. I haven't seen any glaring issues with WAR in my games.

08-07-2012, 08:22 PM	#6
olivertheorem All Star Starter Join Date: Jun 2006 Posts: 1,919	I recently finished going through six leagues (1 universe), manually determining award winners and "all-stars" (not the in-game ones, but something I'm doing outside of the game). In any event, the highest WAR for a catcher I found was 8.9. I forgot off-hand how many were over 5.0. I do think that the relatively limited playing time, even for a healthy catcher, tends to depress their WAR totals a bit.

08-07-2012, 09:34 PM	#7
mets8484 Bat Boy Join Date: Jun 2009 Posts: 7	Different sites calculate WAR differently. For instance, I know baseball reference and fangraphs have different formulas for their WAR. I am sure Marcus has a different formula for his.

08-07-2012, 09:39 PM	#8
JWay Hall Of Famer Join Date: Dec 2001 Location: Sioux Falls, SD Posts: 5,021 Blog Entries: 3	I'm guessing OOTP catchers get hammered by playing time, where as other websites adjust catchers to something less than 162 for replacement level. It would make sense to assume that a backup catch will have to be used at least for 30 or so games. Where as other positional players should miss no more than 5-10. That makes the most sense.

08-08-2012, 02:55 PM	#10
kingcharlesxii Hall Of Famer Join Date: Mar 2003 Location: Madison, WI Posts: 2,730	Unless I'm completely mistaken, WAR in OOTP is just (VORP/10)+(ZR/10). ZR is the Zone Rating at all positions. If it's not that, it's something very close to that.

12-19-2012, 12:37 PM	#16
Leo_The_Lip All Star Starter Join Date: Nov 2009 Posts: 1,703	While the discussion thread will prove interesting and, hopefully, informative, is it really needed? Future versions of OOTP will likely not use WAR as anyone is currently calculating it because there will be new algorithms developed that claim to do the job better and they will. I have no doubt these new measurement systems will be included in the future. For now, this is the way it is, and that should be acceptable to everyone. Me, I'm keeping my eye on Nate Silver. Now that he knows how all of us are going to vote even before anyone is nominated, I'm hoping he goes back to baseball and develops a new system which he names after a more worthy player.