Question about the NEW WAR

[willshad]

On Baseball-reference I noticed that their WAR totals have changed. I believe that they now do the positional adjustments on the defensive part of WAR. Does this mean that as far as OFFENSIVE WAR is concerned they are all treated as equals? This does not seem to be the case. I see in 1995 that John Valentin led the league in WAR. I'm assuming most of this is due to his position and defensive work. However, he totaled 5.9 OFFENSIVE WAR that season. Frank Thomas had 6.1 offensive WAR..just a slight bit higher even though he was far superior offensively. His OPS was 1.061 compared to Valentin's .931. He also played 10 more games. Shouldn't his offensive WAR be much higher? Valentin did have 20 steals..but do they make up that much difference in hitting?

[Bothrops Atrox]

oWAR and dWAR both include positional adjustments. Thomas's oWAR is only slightly higher because WAR is considering the fact that Thomas was a 1B but Valentin was a SS. BBref clearly says that this is why we can't add oWAR + dWAR together, or you would be double-counting positional value.

They only list defense+ position together and offense + position together to organize/present the data one certain way. The positional values are the exact same as before. Before, they listed offense + position together but defense by itself, which had everybody up in a tizzy.

If you want to know who the better offensive player was regardless of position, just look at Rbat + Rbaser + Rdp, which Thomas leads 55 to 35.

[brett]

On Baseball-reference I noticed that their WAR totals have changed. I believe that they now do the positional adjustments on the defensive part of WAR. Does this mean that as far as OFFENSIVE WAR is concerned they are all treated as equals? This does not seem to be the case. I see in 1995 that John Valentin led the league in WAR. I'm assuming most of this is due to his position and defensive work. However, he totaled 5.9 OFFENSIVE WAR that season. Frank Thomas had 6.1 offensive WAR..just a slight bit higher even though he was far superior offensively. His OPS was 1.061 compared to Valentin's .931. He also played 10 more games. Shouldn't his offensive WAR be much higher? Valentin did have 20 steals..but do they make up that much difference in hitting?

It is kind of funny how they did it. Owar effectively shows a player offensively versus his position, but Dwar shows a player's defensive value against ALL players given defense at the position AND the positional adjustment. Furthermore, Owar is versus a replacement level hitter, and assuming that a replacement level is only below average as a hitter, while Dwar is versus an average major league level plaer. Adding them together will not give total war, but the quickest way to get Owar is just to subtract the positional adjustment from Owar (Matt, wouldn't that work?).

Some WAR systems make a replacement level player below average on offense AND defense. As long as you know what you are getting it doesn't really matter.

[Bothrops Atrox]

Adding them together will not give total war, but the quickest way to get Owar is just to subtract the positional adjustment from Owar (Matt, wouldn't that work?).

.

Sure, if you don't mind the positional adjustment being there. We have this debate on every "best hitters ever" polls. Somebody brings up Piazza and says "we have to consider the fact that he is a catcher" and somebody responds "but I am just looking at the pure hitting numbers." Depends on what you are looking for.

[brett]

Sure, if you don't mind the positional adjustment being there. We have this debate on every "best hitters ever" polls. Somebody brings up Piazza and says "we have to consider the fact that he is a catcher" and somebody responds "but I am just looking at the pure hitting numbers." Depends on what you are looking for.

What I mean is to take the OWAR and then remove the positional adjustment, but then again that's done in runs so you divide it by 10. So if OWar is 5 and positional adjustment is 7 runs, take away .7 war and you get a position neutral 4.3

[Second Base Coach]

If you had Piazza close to his prime, maybe signing him to four of his best seasons with the Mets, and you with the organization in some capacity... which of the following statements would best sum up your feelings about the situation:

We GET to use Piazza as the catcher in our line-up.

We are FORCED to use Piazza as the catcher in our line-up.

How you answer that will say a lot of your take on defensive metrics and wether or not Piazza comes to the plate as a catcher or just another hitter in your mind.

[DoubleX]

Maybe I'm missing something, or perhaps everything is still being updated, but the new WAR figures at baseball-reference don't add up.

Take Derek Jeter, for example. Before the revamp, I believe his career WAR was slightly over 70 and included something like -14 for defense. Now his career WAR is 67.9, but his offensive WAR is 89.1 and his defensive WAR is -8.0, which should equal 81.1. So how could his career WAR be lower than it previously was, but his offensive and defensive aggregate now be much higher?

Or let's look at Mike Piazza. Before the revamp, I believe his career WAR was just over 59, it's now at 56.1, but his offensive and defensive splits show 63.2 and 1.0, respectively, which should equal 64.2.

Even if you look at individual years it doesn't add up. For instance, in 1997, Mike Piazza's overall WAR is listed at 8.5, yet his offensive and defense WAR totals were 8.7 and 0.5, respectively, which should be 9.2.

So am I missing something here?

[brett]

Maybe I'm missing something, or perhaps everything is still being updated, but the new WAR figures at baseball-reference don't add up.

Take Derek Jeter, for example. Before the revamp, I believe his career WAR was slightly over 70 and included something like -14 for defense. Now his career WAR is 67.9, but his offensive WAR is 89.1 and his defensive WAR is -8.0, which should equal 81.1. So how could his career WAR be lower than it previously was, but his offensive and defensive aggregate now be much higher?

Or let's look at Mike Piazza. Before the revamp, I believe his career WAR was just over 59, it's now at 56.1, but his offensive and defensive splits show 63.2 and 1.0, respectively, which should equal 64.2.

Even if you look at individual years it doesn't add up. For instance, in 1997, Mike Piazza's overall WAR is listed at 8.5, yet his offensive and defense WAR totals were 8.7 and 0.5, respectively, which should be 9.2.

So am I missing something here?

Here's the deal: for offensive WAR they put the positional value in with the players offensive value above replacement. This tells you how much better he hit than a replacement hitter at the same position. For defense they put defensive WAR above average and also add the positional value so that we get a ranking of defensive players not based on the most dominant at their position, but the best fielder with a positional adjustment, so a shortsop who is +5 runs a year is going to be as good as a first baseman who is +18 runs a year because SS has a +7 adjustment and first base about -10. So when you add the two together you get the positional adjustment two times. Again, the REASON for showing this info is that people want to rate all time fielders with consideration of their position, but its worth seeing how good a hitter is compared to others at his position to see how much value he brings to a team. The change is in putting the positional adjusment into defensive WAR because people like to see that. If you add offensive WAR and Defensive war then you get total war plus an extra second counting of positional value.

[brett]

If you had Piazza close to his prime, maybe signing him to four of his best seasons with the Mets, and you with the organization in some capacity... which of the following statements would best sum up your feelings about the situation:

We GET to use Piazza as the catcher in our line-up.

We are FORCED to use Piazza as the catcher in our line-up.

How you answer that will say a lot of your take on defensive metrics and wether or not Piazza comes to the plate as a catcher or just another hitter in your mind.

Piazza's positional adjustment is +72 runs. (this is really an estimate of how much worse catchers hit on average than other players) His defense at his position is -63 relative to all catchers. The overall benefit of getting his offense from a catcher just barely outdoes the defensive cost he had as a catcher if you trust the defensive metric. He would have to have been a near gold glove first baseman through his career if he hit the same and played as long as he did to be as valuable. He could have been a little below average defensive third baseman. Still, I'd bet he would have played 15% longer and hit somewhat better had he not been a catcher.

[DoubleX]

Here's the deal: for offensive WAR they put the positional value in with the players offensive value above replacement. This tells you how much better he hit than a replacement hitter at the same position. For defense they put defensive WAR above average and also add the positional value so that we get a ranking of defensive players not based on the most dominant at their position, but the best fielder with a positional adjustment, so a shortsop who is +5 runs a year is going to be as good as a first baseman who is +18 runs a year because SS has a +7 adjustment and first base about -10. So when you add the two together you get the positional adjustment two times. Again, the REASON for showing this info is that people want to rate all time fielders with consideration of their position, but its worth seeing how good a hitter is compared to others at his position to see how much value he brings to a team. The change is in putting the positional adjusment into defensive WAR because people like to see that. If you add offensive WAR and Defensive war then you get total war plus an extra second counting of positional value.

That's all kind of confusing and means you have to look at WAR and oWAR and dWAR in different contexts.

[brett]

That's all kind of confusing and means you have to look at WAR and oWAR and dWAR in different contexts.

Yes offensive WAR and total war are both effectively bookkeeping methods, they tell a team how a player should affect winning and losing. DWar is simply a way of rating defensive players.

[dominik]

I hate the new dWAR with positional adjustment. now the two WAR don't add up anymore since of course position is not counted twice. was much more easy before.

In overall WAR this does make no difference but I still think positional adjustment should be done on offense and not defense.

[dominik]

Maybe I'm missing something, or perhaps everything is still being updated, but the new WAR figures at baseball-reference don't add up.

Take Derek Jeter, for example. Before the revamp, I believe his career WAR was slightly over 70 and included something like -14 for defense. Now his career WAR is 67.9, but his offensive WAR is 89.1 and his defensive WAR is -8.0, which should equal 81.1. So how could his career WAR be lower than it previously was, but his offensive and defensive aggregate now be much higher?

Or let's look at Mike Piazza. Before the revamp, I believe his career WAR was just over 59, it's now at 56.1, but his offensive and defensive splits show 63.2 and 1.0, respectively, which should equal 64.2.

Even if you look at individual years it doesn't add up. For instance, in 1997, Mike Piazza's overall WAR is listed at 8.5, yet his offensive and defense WAR totals were 8.7 and 0.5, respectively, which should be 9.2.

So am I missing something here?

this is new. the old WAR did add up but the new dWAR considers position too.

[brett]

I hate the new dWAR with positional adjustment. now the two WAR don't add up anymore since of course position is not counted twice. was much more easy before.

In overall WAR this does make no difference but I still think positional adjustment should be done on offense and not defense.

Its just information though. You can take total WAR and subtract OWar to get the old DWar.

[Bothrops Atrox]

Its just information though. You can take total WAR and subtract OWar to get the old DWar.

Right - do whatever you want with the data. I never look at dWAR and oWAR separately, personally.