Why Doesn't WAR adjust for home field aided players?

[willshad]

I am curious as to why the WAR totals are so high for guys like Larry Walker, Todd Helton, etc, who were obviously only great because of where they played. For instance, Helton in 2004 came up with 8.1 WAR. This is a GREAT total, and in fact more than Mike Piazza had in his historic 1997 season.

What gives? You can't argue that it is the game played factor, because they played about the same number of games. Not only was Piazza a better hitter outright, but he was doing so from the catcher's spot, which should have given him MUCH more offensive WAR. Even more importantly, Helton on the road only hit 11 homers, with 36 RBI, and .991 OPS. In a neutral stadium, he maybe has a typical prime Mark Grace year. Piazza's numbers were actually better on the road than at home, as is the case for every season of his career.

Why can't WAR adjust for this, when it is obvious to anyone looking at the stats, that if Piazza had played half his games at Coors, that his numbers would put Helton to shame? I realize that Helton is a good glove at first, but when including position, I think he is still negative for defensive WAR.

[brett]

I am curious as to why the WAR totals are so high for guys like Larry Walker, Todd Helton, etc, who were obviously only great because of where they played. For instance, Helton in 2004 came up with 8.1 WAR. This is a GREAT total, and in fact more than Mike Piazza had in his historic 1997 season.

What gives? You can't argue that it is the game played factor, because they played about the same number of games. Not only was Piazza a better hitter outright, but he was doing so from the catcher's spot, which should have given him MUCH more offensive WAR. Even more importantly, Helton on the road only hit 11 homers, with 36 RBI, and .991 OPS. In a neutral stadium, he maybe has a typical prime Mark Grace year. Piazza's numbers were actually better on the road than at home, as is the case for every season of his career.

Why can't WAR adjust for this, when it is obvious to anyone looking at the stats, that if Piazza had played half his games at Coors, that his numbers would put Helton to shame? I realize that Helton is a good glove at first, but when including position, I think he is still negative for defensive WAR.

It fully adjusts for home field offensive environment. It doesn't use road rates to estimate home rates, but it does calculate a players value relative to the average value produced by a player with that ballpark. If Helton and Walker had benefitted MORE from their ballpark than an average hitter, they would get more war, but in fact Helton and Walker produced home value exactly in proportion to their road value over their careers. For Helton this is in part because he walked even more on the road than at home so his power and average may take a little more than average cut, but his on-base percentage is better relative to the road than at home. Helton's season is so high because he supposedly saved 18 runs above average at first base (and I don't doubt that he was that good.) Helton gets +18 for defense ansd -10 for position and -2 for defense and +7 for position so Helton actually makes up for more than the positional adjustment with his defense. Piazza also lost .5 war on the bases and Helton was +.1. Piazza hit about 14% better but Helton also played about 5% more to go with the +.3 defense and position war and the +.6 running war.

[willshad]

It fully adjusts for home field offensive environment. It doesn't use road rates to estimate home rates, but it does calculate a players value relative to the average value produced by a player with that ballpark. If Helton and Walker had benefitted MORE from their ballpark than an average hitter, they would get more war, but in fact Helton and Walker produced home value exactly in proportion to their road value over their careers. For Helton this is in part because he walked even more on the road than at home so his power and average may take a little more than average cut, but his on-base percentage is better relative to the road than at home. Helton's season is so high because he supposedly saved 18 runs above average at first base (and I don't doubt that he was that good.) Helton gets +18 for defense ansd -10 for position and -2 for defense and +7 for position so Helton actually makes up for more than the positional adjustment with his defense. Piazza also lost .5 war on the bases and Helton was +.1. Piazza hit about 14% better but Helton also played about 5% more to go with the +.3 defense and position war and the +.6 running war.

It just seems to me that Piazza is being penalized for playing a tougher position, and playing in a tougher home park. Let's put it this way: Let's assume that both guys didn't have to play the field at all, and just were DH, for the same team at the same time. I think in that case Piazza's numbers would blow away Helton's, and nobody could really doubt that.

To me, a below average catcher who is the best hitter in the league, is way more valuable than a good fielding first baseman who is slightly above average hitting for his position. Todd Helton was Mark Grace who happened to play in Coors field. Sometimes trying to be too scientific takes away the human factor, and you overlook things that are obvious.

[brett]

It just seems to me that Piazza is being penalized for playing a tougher position, and playing in a tougher home park. Let's put it this way: Let's assume that both guys didn't have to play the field at all, and just were DH, for the same team at the same time. I think in that case Piazza's numbers would blow away Helton's, and nobody could really doubt that.

To me, a below average catcher who is the best hitter in the league, is way more valuable than a good fielding first baseman who is slightly above average hitting for his position. Todd Helton was Mark Grace who happened to play in Coors field. Sometimes trying to be too scientific takes away the human factor, and you overlook things that are obvious.

Helton may have been similar to Grace in BA and fielding, but Grace averaged 12 home runs and 78 walks per 162 road games. Helton averaged 21.5 home runs and 94 walks per 162 road games.

In 2000, on the road, Helton hit .353 with 31 doubles, 15 home runs, a .441 on-base percentage and a .633 slugging percentage. Given a NORMAL home park we would expect him to have done the following at home: .370, 33 doubles, 16 home runs, .463 on base, .670 slugging for an overall line of about .360, 64 doubles, 31 home runs, .452 on base, .652 slugging and an OPS+ of 183, so he actually did better on the road relative to most hitters than he did at home and at Coors compared to the average. ib

Piazza went .368/.431/.643 on the road in '97. Helton went .353/.441/.633 on the road in 2000 and Piazza got to hit AT Coors on the road and Helton didn't which makes them virtually identical hitters on the road for those seasons.

So comparing Helton to Grace does not hold water unless Piazza was Mark grace playing below average defense at the catcher position. Cutting Helton down doesn't make sense. He almost certainly is hurt slightly in WAR due to his ballpark. I think your contention with WAR and Piazza is that Piazza is rated poorly enough on defense that it cancels out his entire positional offensive edge on an average hitting catcher.

[brett]

It just seems to me that Piazza is being penalized for playing a tougher position, and playing in a tougher home park. Let's put it this way: Let's assume that both guys didn't have to play the field at all, and just were DH, for the same team at the same time. I think in that case Piazza's numbers would blow away Helton's, and nobody could really doubt that.

Helton's 2000 road numbers are virtually identical to Piazza's '97 road numbers!

[JDanger]

1. Piazza had 8.5 WAR in his 1997 season according to B-R and 9.4 according to Fangraphs-- either way that's more than Helton's 2004.

2. A .991 OPS on the road is outstanding, something Mark Grace has never came close to achieving at home at any point in his career.

3. Piazza is not getting penalized for his position, in fact he is rewarded for it in the positional adjustment. He was just not very good at it.

[willshad]

Helton's 2000 road numbers are virtually identical to Piazza's '97 road numbers!

Umm that's ONE season, and not even the season Im discussing here. For their careers, Piazza's road OPS is .960, compared to .866 for Helton. It is only that close because Piazza suffered an early and severe decline due to his position. They are not comparable as hitters. To me, their value would be about equal if their positions were switched, and both played in neutral stadiums. The fact that Helton comes out ahead despite being a worse hitter and playing the easiest position, compared to the hardest, is a huge flaw with the stat. If Piazza had played his entire career at Coors, and had the luxury of playing first base, his raw numbers and rate stats would probably be about the same as Babe Ruth's, with a slightly lower on base percentage.

[willshad]

Helton may have been similar to Grace in BA and fielding, but Grace averaged 12 home runs and 78 walks per 162 road games. Helton averaged 21.5 home runs and 94 walks per 162 road games.

In 2000, on the road, Helton hit .353 with 31 doubles, 15 home runs, a .441 on-base percentage and a .633 slugging percentage. Given a NORMAL home park we would expect him to have done the following at home: .370, 33 doubles, 16 home runs, .463 on base, .670 slugging for an overall line of about .360, 64 doubles, 31 home runs, .452 on base, .652 slugging and an OPS+ of 183, so he actually did better on the road relative to most hitters than he did at home and at Coors compared to the average. ib

Piazza went .368/.431/.643 on the road in '97. Helton went .353/.441/.633 on the road in 2000 and Piazza got to hit AT Coors on the road and Helton didn't which makes them virtually identical hitters on the road for those seasons.

So comparing Helton to Grace does not hold water unless Piazza was Mark grace playing below average defense at the catcher position. Cutting Helton down doesn't make sense. He almost certainly is hurt slightly in WAR due to his ballpark. I think your contention with WAR and Piazza is that Piazza is rated poorly enough on defense that it cancels out his entire positional offensive edge on an average hitting catcher.

OK, Helton hit great on the road in 2000. The season I am talking about, however, is 2004. 2000 was by far Helton's best year on the road, and 1997 was only about Piazza's fifth best season on the road. You compare them as if they were typical of the players' performances.

[brett]

Umm that's ONE season, and not even the season Im discussing here. For their careers, Piazza's road OPS is .960, compared to .866 for Helton. It is only that close because Piazza suffered an early and severe decline due to his position. They are not comparable as hitters. To me, their value would be about equal if their positions were switched, and both played in neutral stadiums. The fact that Helton comes out ahead despite being a worse hitter and playing the easiest position, compared to the hardest, is a huge flaw with the stat. If Piazza had played his entire career at Coors, and had the luxury of playing first base, his raw numbers and rate stats would probably be about the same as Babe Ruth's, with a slightly lower on base percentage.

In '04, Helton's road relative OPS+ was 164 and his overall OPS+ was 165, but that just compares him on the road to the rest of the league on the road. He didn't get to hit at Coors on the road and the rest of the league did which is about a 2.5% edge, so given park effects, his road/park relative OPS+ was 168.

Helton went .326/.446/.544 on the road with 26 doubles, 11 home runs and 59 walks. Again with a normal home park that would give him overall rates of .333/.457/.560 with 54 doubles, 23 home runs and 120 walks. That is closer to Ted Williams than Mark Grace.

The main problem with Piazza's career and comparing his road rates to Helton's is that it has nothing to do with Helton being particularly helped by his park. He is not helped by his park in WAR or OPS+. The problem is that Piazza WAS particularly hurt by his home parks.

Piazza's home/park specific OPS+ was 132 relative to others at home, and his road OPS+ was about 154. That is compared to all players at home and on the road, and adjusting to Piazza's specific home parks. Some of that variation was luck as well so by picking his road numbers which happen to be relatively better you get to take advantage of some chance as well.

[BondsOverBabe]

WAR is just a TERRIBLE stat for evaluating catchers.

[brett]

WAR is just a TERRIBLE stat for evaluating catchers.

You mean bad in comparing them to other players, or comparing catchers to catchers?

War tries to get catchers true value above replacement. It is possible that because catchers don't play as much they simply DON'T produce as much value as other players. It is also possible that WAR misses out on some catcher value in a) giving them some credit for what they do most-catch, and realizing that if a catcher is injured, most teams will be forced to split time between a replacement and a sub replacement level backup, to their replacement level might be too high. Players get about 21 runs or 2.1 WAR per 162 games for the replacement level player who would take their place. I think catchers should get about 29 (somewhere between 27 and 32) runs for replacement level because of this which would give every catcher .6-1.1 extra war per 162 games caught, or using 29 runs, about .8 per 162.) That would give Mike Piazza about 8 more career WAR or bring him up to about 64, which is darn good for only 1900 some games. It would give Bench about 81 war in only around 2100 games.

So maybe its a good stat for evaluating catchers if its used properly, unless you mean its a poor stat for comparing one catcher to another.

[pheasant]

Umm that's ONE season, and not even the season Im discussing here. For their careers, Piazza's road OPS is .960, compared to .866 for Helton. It is only that close because Piazza suffered an early and severe decline due to his position. They are not comparable as hitters. To me, their value would be about equal if their positions were switched, and both played in neutral stadiums. The fact that Helton comes out ahead despite being a worse hitter and playing the easiest position, compared to the hardest, is a huge flaw with the stat. If Piazza had played his entire career at Coors, and had the luxury of playing first base, his raw numbers and rate stats would probably be about the same as Babe Ruth's, with a slightly lower on base percentage.

I definitely agree with you on this one. Piazza during his best 10 year peak hit an amazing .340 on the road while averaging a whopping 42 HRs per 162 games played. I believe Piazza's rate stats would have been even better had he not been ruining his knees by catching during those 10 years. Perhaps Piazza as a Colorado Rocky(not as a catcher though) might have matched Ruth's best 12 year peak, which was from 1920-1931 when he averaged .357/.489/.736.
I.e, I see Piazza as a first baseman with the Rockies having several 50-59 HRs seasons. That alone would skyrocket his walks. Perhaps Ruth's actual numbers are a little lofty. But I see Piazza averaging .360/.450/.680 in a Rockies uniform for a 10 year peak.

[brett]

I definitely agree with you on this one. Piazza during his best 10 year peak hit an amazing .340 on the road while averaging a whopping 42 HRs per 162 games played. I believe Piazza's rate stats would have been even better had he not been ruining his knees by catching during those 10 years. Perhaps Piazza as a Colorado Rocky(not as a catcher though) might have matched Ruth's best 12 year peak, which was from 1920-1931 when he averaged .357/.489/.736.
I.e, I see Piazza as a first baseman with the Rockies having several 50-59 HRs seasons. That alone would skyrocket his walks. Perhaps Ruth's actual numbers are a little lofty. But I see Piazza averaging .360/.450/.680 in a Rockies uniform for a 10 year peak.

That doesn't mean he would "match" Ruth's peak, just his peak numbers. Anyway, its pretty easy to estimate. If you take a players road rates (.340 with 42 HRS per 162), then if you put that player in Coors for half of his games, he would get about a 12% boost to batting average and about a 22% boost to home runs so he would have hypothetically gone .381 with 51-52 home runs a season. That is during the peak offensive period of Coors. The problem once again is that Piazza did not hit as well at home on a relative basis. Of course he should hit worse in lower offensive parks, but he hit about 15% worse at home RELATIVE to other players than he did on the road. If ANY of that is random, rather than due to the uniqueness of his home parks, then we are reading the randomness as real.

If we just take Piazza's overall rates and move him to Coors top offensive environment as a home park, he has a 10 year peak of about .354/.428/.667 with 49 home runs and 157 RBI per 162 games. He still would have had a 155 OPS+ though.

[pheasant]

That doesn't mean he would "match" Ruth's peak, just his peak numbers. Anyway, its pretty easy to estimate. If you take a players road rates (.340 with 42 HRS per 162), then if you put that player in Coors for half of his games, he would get about a 12% boost to batting average and about a 22% boost to home runs so he would have hypothetically gone .381 with 51-52 home runs a season. That is during the peak offensive period of Coors. The problem once again is that Piazza did not hit as well at home on a relative basis. Of course he should hit worse in lower offensive parks, but he hit about 15% worse at home RELATIVE to other players than he did on the road. If ANY of that is random, rather than due to the uniqueness of his home parks, then we are reading the randomness as real.

If we just take Piazza's overall rates and move him to Coors top offensive environment as a home park, he has a 10 year peak of about .354/.428/.667 with 49 home runs and 157 RBI per 162 games. He still would have had a 155 OPS+ though.

That makes sense, Brett. I tend to believe that the best hitters magnified those advantages even more. But I can't prove it. That's just a gut. Thus, we need to go with league averages, which is what you provided.

I appreciate you crunching the numbers.

[brett]

That makes sense, Brett. I tend to believe that the best hitters magnified those advantages even more. But I can't prove it. That's just a gut. Thus, we need to go with league averages, which is what you provided.

I appreciate you crunching the numbers.

But I have demonstrated in other threads that Larry Walker, Todd Helton and Troy Tulowitzki all had slightly less of a boost than the average hitter on a percentage basis. Especially Helton because he walks a lot so he has a baseline of walks on the road and at home that are similar.