SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
WAR and More (...what is it good for)
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 17:36:13 GMT -5
Read that link: “add in a positional adjustment, a snall adjustment for their league...” etc. For pitchers: “FIP... adjusted for park, scaled to how many innings the pitcher threw.” That is not context free. Obviously Degrom is great. But the WAR number seems meaningless in any real sense. Even you wrote it isn’t “literally” wins... so what are they? Metaphors? Symbols? On the other hand, Degrom may be WASTED — so “worthless” in that sense... in that another pitcher of far less talent might do no worse. If Bartolo Colon took his starts, could the Mets still be 11-14? Well, it is about their overall winning percentage, so... maybe. So on a strictly literal sense, it seems illogical to suggest the Mets see 8 wins of value from Degrom. If it means a different team would, well, who knows? He’d probably have a great record on the Sox. Maybe worse on the Padres. Again, your link: “WAR is trying to answer the time-honored question: how valuable is each player TO HIS TEAM.” (My caps). So is DeGrom worth 8 extra wins to his team? Certainly not in any literal way. Then.... how? That is a different question from how great is he. But is this just a semantic issue, then? I'm partially with you, if the issue is just that "Wins Above Replacement" seems to promise a little more than the stat actually delivers. But that doesn't mean the stat itself is useless. But what if it were called "Goodness Units" - a measure of how good a player has been, calculated the exact same way as WAR but without the misleading implications regarding wins. "Right now, Mookie Betts has 7.8 fGU, 0.4 behind league leader Jose Ramirez at 8.2." It seems like that would answer all your criticisms, no? Would answer some, except I’d still be left wondering what “goodness” translates to. It also doesn’t address the problem with “goodness” as a unit of exchange: if it is pegged to position, season, “replacement” — it varies wildly in a way that makes Mookie’s “goodness” rating at right field a number on a different scale from Ramirez at 3B. But I like “goodness” because it makes it more obviously abstract. “Wins” makes it sound like it measures something real, which it doesn’t. As for the bad 20-game winner vs., say, DeGrom: look, we ALL agree I’m sure that DeGrom is having a great year and the Mets suck. And likely that there are exceedingly mediocre pitchers who put up big numbers on great teams. But trying to quantify the difference strikes me as more deceptive than people who use WAR as a defining stat would make it seem. For example: maybe it was Colby Lewis on the Rangers earlier referenced? 2015, he is 17-9, despite a 4.66 ERA. His WAR is a mere .9. OK, I think we all agree that W-L is a team one, not his doing. Nonetheless, how much better than 17-9 would a great pitcher be? If the pitcher is 20-6 with great peripherals, it actually has little difference in the standings. So qualitatively the latter pitcher is better, I want him more etc etc. But quantitatively... how many wins “above” Lewis can one predict?
|
|
|
Post by hongkyongnae on Aug 19, 2018 19:03:38 GMT -5
A closer blows a 4-3 lead in the top of the 9th and gives up 4 ER. His team rallies in the bottom of the 9th to score 5 runs and walk off with the win. The closer who gave up 4 ER in his one inning now gets credited with the "win". The win literally happened, and he gets the credit, but in reality he was the least valuable player to play within the game on either side. This records a real life event, but doesn't give you any idea of the player's value. Good thread and this post, in particular, brings up a good point on the value of using wins to judge pitchers. As another example, Hector Velázquez is currently 7 and 1 for the Sox. Chris Sale is 12 and 4. I think the difference between these two is greater than the win/loss record might indicate. In the spirit of learning from the thread, would some folks here please provide some explanation of what they believe to be better indicators a pitcher's worth. Thanks.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 19:36:26 GMT -5
A closer blows a 4-3 lead in the top of the 9th and gives up 4 ER. His team rallies in the bottom of the 9th to score 5 runs and walk off with the win. The closer who gave up 4 ER in his one inning now gets credited with the "win". The win literally happened, and he gets the credit, but in reality he was the least valuable player to play within the game on either side. This records a real life event, but doesn't give you any idea of the player's value. Good thread and this post, in particular, brings up a good point on the value of using wins to judge pitchers. As another example, Hector Velázquez is currently 7 and 1 for the Sox. Chris Sale is 12 and 4. I think the difference between these two is greater than the win/loss record might indicate. In the spirit of learning from the thread, would some folks here please provide some explanation of what they believe to be better indicators a pitcher's worth. Thanks. I think there are two things going on: one is quality and the other is value. That is, wins are not necessarily an expressuon of quality (nor are they always NOT). DeGrom is a great pitcher with few wins on a bad team. That said, I think one can get some sense of value in the overall record the team has in a pitcher’s starts — at least in so far as I would argue in an extreme case — take the Colby Lewis one — that a better quality pitcher wouldn’t likely have that much better ultimate results. Now, there are deeper factors (are you eatibg innings or sapping the bullpen, for example). But if a guy is 17-9, another guy can’t “replace” him with that many more wins! So in the most fundamenral value, he’s pretty valuable. Put differently... if a team wins 100 games, its pitchers must have been good enough. That is different from being “good” in the Chris Sale sense... but it makes it hard to say there is much room for wins “above” that staff. Or to your example: of course for many reasons Sale is much better than Velasquez. But it is also hard to see how many more wins “above” the 7-1 record another pitcher would be in Velasquez’s role. More doninant? Sure? But this is where I see a divide between replacement values and excellence.
|
|
|
Post by jimed14 on Aug 19, 2018 19:46:05 GMT -5
You're criticizing WAR because it doesn't do what you want it to do, measure value a player has to his team. So come up with your own stat. WAR was never designed for that. It's more designed for judging a player's value on a league-wide basis for a particular season.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 19:49:54 GMT -5
You're criticizing WAR because it doesn't do what you want it to do, measure value a player has to his team. So come up with your own stat. WAR was never designed for that. It's more designed for judging a player's value on a league-wide basis for a particular season. The link you sent says precisely that it measures value to the player’s team. I quoted it above. That’s in the definition. And that value seems to be what matters, really. If a guy has a 5.00 ERA but his team wins every game he pitches, as Bill James says it may be luck, but so what? What replacenent, no matter how good, would do better?
|
|
|
Post by jimed14 on Aug 19, 2018 19:57:22 GMT -5
You're criticizing WAR because it doesn't do what you want it to do, measure value a player has to his team. So come up with your own stat. WAR was never designed for that. It's more designed for judging a player's value on a league-wide basis for a particular season. The link you sent says precisely that it measures value to the player’s team. I quoted it above. That’s in the definition. And that value seems to be what matters, really. If a guy has a 5.00 ERA but his team wins every game he pitches, as Bill James says it may be luck, but so what? What replacenent, no matter how good, would do better? Anyone who pitched better would be better. If you want to argue that, we're done. This is the quote from the article that you seem to think is not in the article:
|
|
|
Post by umassgrad2005 on Aug 19, 2018 20:08:13 GMT -5
War is wins over replacement level. So you need to set replacement level, which a guy set at 52 wins per year using bwar. So a 0-162 team would actually be a team with negative 52 bwar, a hundred win team would have 48 bwar. It's not 100%, but its been shown to be really close. Its win above or below a replacement level team or a 52 win team.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 20:16:22 GMT -5
The link you sent says precisely that it measures value to the player’s team. I quoted it above. That’s in the definition. And that value seems to be what matters, really. If a guy has a 5.00 ERA but his team wins every game he pitches, as Bill James says it may be luck, but so what? What replacenent, no matter how good, would do better? Anyone who pitched better would be better. If you want to argue that, we're done. This is the quote from the article that you seem to think is not in the article: I saw that. But as I quoted above, its definition contradicts that directly. I’m not disagreeing about qualitative difference. I’m not even arguing that many of the elements that go into WAR don’t help measure that. I am, however, suggesting that WAR is flawed in the end as a kind of universal means of exchange.
|
|
|
Post by jimed14 on Aug 19, 2018 20:22:20 GMT -5
Yeah, so let's talk about pitcher wins instead. Progress.
|
|
|
Post by hongkyongnae on Aug 19, 2018 20:23:47 GMT -5
Anyone who pitched better would be better. If you want to argue that, we're done. This is the quote from the article that you seem to think is not in the article: I saw that. But as I quoted above, its definition contradicts that directly. I’m not disagreeing about qualitative difference. I’m not even arguing that many of the elements that go into WAR don’t help measure that. I am, however, suggesting that WAR is flawed in the end as a kind of universal means of exchange.I do not think anyone in this thread disagrees with you on the bolded point. I do think they are arguing that WAR has value above and beyond the traditional stats and that they open up the game of baseball to a better understanding. But then, speaking for others is always a risky venture.
|
|
|
Post by umassgrad2005 on Aug 19, 2018 20:28:01 GMT -5
Anyone who pitched better would be better. If you want to argue that, we're done. This is the quote from the article that you seem to think is not in the article: I saw that. But as I quoted above, its definition contradicts that directly. I’m not disagreeing about qualitative difference. I’m not even arguing that many of the elements that go into WAR don’t help measure that. I am, however, suggesting that WAR is flawed in the end as a kind of universal means of exchange. I don't think anyone thinks war is perfect, if players are close it might not tell you who is better. That being said that doesn't seem to be your issue, your issue is you want to look at crazy extremes. In a large sample size a guy with a 5.0 ERA will never win every game. If that happened do you really think he wouldn't be the luckiest guy in baseball? That seems to be your take on it, that if that happend it was skill and not luck. If you believe that, than advanced stats are not for you! You just don't believe in anything but what happens(results) and you should just except that.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 20:42:37 GMT -5
I saw that. But as I quoted above, its definition contradicts that directly. I’m not disagreeing about qualitative difference. I’m not even arguing that many of the elements that go into WAR don’t help measure that. I am, however, suggesting that WAR is flawed in the end as a kind of universal means of exchange. I don't think anyone thinks war is perfect, if players are close it might not tell you who is better. That being said that doesn't seem to be your issue, your issue is you want to look at crazy extremes. In a large sample size a guy with a 5.0 ERA will never win every game. If that happened do you really think he wouldn't be the luckiest guy in baseball? That seems to be your take on it, that if that happend it was skill and not luck. If you believe that, than advanced stats are not for you! You just don't believe in anything but what happens(results) and you should just except that. Of course I chose an extreme example for ease of math. But DeGrom works: Mets have won 11 of his starts, and he has almost 8 wins above replacement. This suggests his replacement would have a record well below their overall winning percentage. That to me is a significant issue. As for using it as a comparison.... let’s take the example of Jason Kendall in 1998. He was 20th in bWAR in the NL with 5.6 (incidentally, the Pirates won 69 games). This puts him 2.3 ahead of Todd Helton, who is better in every offensive stat except stolen bases. And Helton is, at that point, a future gold glover, so it ought not be defense. Is Kendall the better player that year?
|
|
|
Post by incandenza on Aug 19, 2018 22:05:14 GMT -5
I don't think anyone thinks war is perfect, if players are close it might not tell you who is better. That being said that doesn't seem to be your issue, your issue is you want to look at crazy extremes. In a large sample size a guy with a 5.0 ERA will never win every game. If that happened do you really think he wouldn't be the luckiest guy in baseball? That seems to be your take on it, that if that happend it was skill and not luck. If you believe that, than advanced stats are not for you! You just don't believe in anything but what happens(results) and you should just except that. Of course I chose an extreme example for ease of math. But DeGrom works: Mets have won 11 of his starts, and he has almost 8 wins above replacement. This suggests his replacement would have a record well below their overall winning percentage. That to me is a significant issue. As for using it as a comparison.... let’s take the example of Jason Kendall in 1998. He was 20th in bWAR in the NL with 5.6 (incidentally, the Pirates won 69 games). This puts him 2.3 ahead of Todd Helton, who is better in every offensive stat except stolen bases. And Helton is, at that point, a future gold glover, so it ought not be defense. Is Kendall the better player that year? Todd Helton had a 114 wRC+ and Kendall had a 136. Kendall's OBP was 30 points higher. So... yes?
|
|
|
Post by scottysmalls on Aug 19, 2018 22:47:18 GMT -5
I don't think anyone thinks war is perfect, if players are close it might not tell you who is better. That being said that doesn't seem to be your issue, your issue is you want to look at crazy extremes. In a large sample size a guy with a 5.0 ERA will never win every game. If that happened do you really think he wouldn't be the luckiest guy in baseball? That seems to be your take on it, that if that happend it was skill and not luck. If you believe that, than advanced stats are not for you! You just don't believe in anything but what happens(results) and you should just except that. Of course I chose an extreme example for ease of math. But DeGrom works: Mets have won 11 of his starts, and he has almost 8 wins above replacement. This suggests his replacement would have a record well below their overall winning percentage. That to me is a significant issue. As for using it as a comparison.... let’s take the example of Jason Kendall in 1998. He was 20th in bWAR in the NL with 5.6 (incidentally, the Pirates won 69 games). This puts him 2.3 ahead of Todd Helton, who is better in every offensive stat except stolen bases. And Helton is, at that point, a future gold glover, so it ought not be defense. Is Kendall the better player that year? I’m so confused by your argument here. Teams performances vary from game to game, so yeah of course if you looked at the game logs you aren’t going to find that a replacement level player would have won exactly 8 less games than DeGrom on the Mets this year. That’s not what the stat is trying to do, it’s attempting to take that context out of it and show you how valuable/talented the player is/was in a vacuum. No one is saying WAR is perfect or should be the end all be all, but you seem to have a fundamental misunderstanding of what it is trying to accomplish. If you want a statistic that shows you precisely how much a players performance (excluding defense) impacted the team’s odds of winning games, with the context included than look at WPA. Your argument against WAR seems to be that it doesn’t do something it doesn’t try to do. If a team happens to win every game started by an awful pitcher that doesn’t mean the pitcher is good it just means the team scored a lot of runs. I don’t understand why you want to value players less based on their teammates’ performances which they cannot control but that’s your prerogative, but you have to get that this is what’s WAR is trying to counteract.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 22:49:11 GMT -5
Of course I chose an extreme example for ease of math. But DeGrom works: Mets have won 11 of his starts, and he has almost 8 wins above replacement. This suggests his replacement would have a record well below their overall winning percentage. That to me is a significant issue. As for using it as a comparison.... let’s take the example of Jason Kendall in 1998. He was 20th in bWAR in the NL with 5.6 (incidentally, the Pirates won 69 games). This puts him 2.3 ahead of Todd Helton, who is better in every offensive stat except stolen bases. And Helton is, at that point, a future gold glover, so it ought not be defense. Is Kendall the better player that year? Todd Helton had a 114 wRC+ and Kendall had a 136. Kendall's OBP was 30 points higher. So... yes? Helton has the Colorado penalty. But fine... the singles hitter who walks at a fractionally higher rate is worth 2 more wins. But my other issue is: is Kendall’s WAR higher because he is a catcher? If so, it makes the comp apples and oranges (a home run by a catcher is worth more than one by a first baseman... even though they are both a run).
|
|
|
Post by scottysmalls on Aug 19, 2018 22:55:51 GMT -5
Todd Helton had a 114 wRC+ and Kendall had a 136. Kendall's OBP was 30 points higher. So... yes? Helton has the Colorado penalty. But fine... the singles hitter who walks at a fractionally higher rate is worth 2 more wins. But my other issue is: is Kendall’s WAR higher because he is a catcher? If so, it makes the comp apples and oranges (a home run by a catcher is worth more than one by a first baseman... even though they are both a run). A catcher that hits 20 home runs is more valuable than a first baseman that does, all else being equal, because it’s harder to find another catcher that can do that than it is a first baseman. The positional adjustments might not be perfect, but they exist so that you can compare value across positions.
|
|
|
Post by Oregon Norm on Aug 19, 2018 23:03:52 GMT -5
I'm not sure where you're getting your data but your making a case that doesn't hold water. A lot of folks thought Kendall was the next big thing at catcher back then and for a few years he was. He stole 26 bases, batted .327 with an on base pct of .411 while slugging .473 over 149 games. Did I mention he did that as a catcher?
It's not an extreme case at all. The value proposition of WAR holds up beautifully. That's before we talk about the real extreme, the distortion of offensive numbers in Denver.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 23:06:22 GMT -5
] I’m so confused by your argument here. Teams performances vary from game to game, so yeah of course if you looked at the game logs you aren’t going to find that a replacement level player would have won exactly 8 less games than DeGrom on the Mets this year. That’s not what the stat is trying to do, it’s attempting to take that context out of it and show you how valuable/talented the player is/was in a vacuum.
Again though: if the 8 is not literal, what does it represent? And what is the vacuum if it actually makes judgmental “adjustments”? How is it a measure of value if we don’t take seriously what it purports to represent?
No one is saying WAR is perfect or should be the end all be all, but you seem to have a fundamental misunderstanding of what it is trying to accomplish.
Perhaps, but it is often used as shorthand to represent overall quality. Earlier in the year, when Trout was accumulating WAR, there was talk he might have the best season ever, despite that in a number of areas he was not likely on pace to have his own best season.
If a team happens to win every game started by an awful pitcher that doesn’t mean the pitcher is good it just means the team scored a lot of runs. I don’t understand why you want to value players less based on their teammates’ performances which they cannot control but that’s your prerogative, but you have to get that this is what’s WAR is trying to counteract. [/quote]
I think you are misunderstanding me: my point is that talent and success are two different things. the mediocre pitcher who has great success is still mediocre in ability. I have low expectations for next year etc. We all know guys can just have those years. At the same time, a statistic that says someone else would have helped a team win more.... seems dubious if the team is already maxing out.
In other words DeGrom is great. The Mets wins and losses in his starts are about their average in all their pitchers’ starts. So it doesn’t seem like he is making any real difference on that team. If the idea is he’s win more elsewhere, I’d say duh. If the idea is he’s worth more wins to the Mets, I’d say — only in that vacuum you mention.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 23:19:21 GMT -5
Helton has the Colorado penalty. But fine... the singles hitter who walks at a fractionally higher rate is worth 2 more wins. But my other issue is: is Kendall’s WAR higher because he is a catcher? If so, it makes the comp apples and oranges (a home run by a catcher is worth more than one by a first baseman... even though they are both a run). A catcher that hits 20 home runs is more valuable than a first baseman that does, all else being equal, because it’s harder to find another catcher that can do that than it is a first baseman. The positional adjustments might not be perfect, but they exist so that you can compare value across positions. But that does mean that 5 WAR for a catcher is different than 5 WAR for a third baseman. So take 2003 with my pal Kendall. He has 6 HRs, 84 runs,58 RBIs, slash .325/.399/.416 OPS+ 112 WAR of 4.5 Player X, same year: 27 HRs, 103 runs, 106 RBIs, .305/.402/.517 OPS+ 137 WAR 4.4 My point is — is player X equal to Kendall? To me, it isn’t close. Kendall is very good for a catcher, but not nearly as good as the guy at a power position.
|
|
|
Post by umassgrad2005 on Aug 19, 2018 23:22:24 GMT -5
I don't think anyone thinks war is perfect, if players are close it might not tell you who is better. That being said that doesn't seem to be your issue, your issue is you want to look at crazy extremes. In a large sample size a guy with a 5.0 ERA will never win every game. If that happened do you really think he wouldn't be the luckiest guy in baseball? That seems to be your take on it, that if that happend it was skill and not luck. If you believe that, than advanced stats are not for you! You just don't believe in anything but what happens(results) and you should just except that. Of course I chose an extreme example for ease of math. But DeGrom works: Mets have won 11 of his starts, and he has almost 8 wins above replacement. This suggests his replacement would have a record well below their overall winning percentage. That to me is a significant issue. As for using it as a comparison.... let’s take the example of Jason Kendall in 1998. He was 20th in bWAR in the NL with 5.6 (incidentally, the Pirates won 69 games). This puts him 2.3 ahead of Todd Helton, who is better in every offensive stat except stolen bases. And Helton is, at that point, a future gold glover, so it ought not be defense. Is Kendall the better player that year? You are getting so hung up on Degrom and his win total, while completely overlooking the fact that wins are a horrible way of judging a pitcher. Instead of looking at just wins, look how he's pitched and things he can't control like run support. That is what war is doing. You sound like old school guys 20 years ago that only wins matter and those guys get the Cy Youngs, while totally overlooking things a pitcher can't control. Put Degrom on a good team he'd have a lot more wins, its rather simple. The whole concept is based on replacement level for a given position. The bar is higher at a corner OF position and much lower at catcher. Just go look at the corner OF players and the catchers as a group. Its rather easy to see. The top corner OF will almost always out hit the top catcher. Yet there are ton of very good corner OF and only a few good catchers. Making a top catcher that can hit and play D incredibly valuable. Do to the averages a replacement level corner OF will out hit a replacement level catcher by a ton. So who's more valuable the couple elite catchers or the large group of corner OFs?
|
|
|
Post by umassgrad2005 on Aug 19, 2018 23:28:50 GMT -5
Todd Helton had a 114 wRC+ and Kendall had a 136. Kendall's OBP was 30 points higher. So... yes? Helton has the Colorado penalty. But fine... the singles hitter who walks at a fractionally higher rate is worth 2 more wins. But my other issue is: is Kendall’s WAR higher because he is a catcher? If so, it makes the comp apples and oranges (a home run by a catcher is worth more than one by a first baseman... even though they are both a run). You are looking at that all wrong, based on one game. Catchers hit way less HRs than guys at 1B do. So 20 HRs from a catcher is worth more than 30 HRs from a guy at 1B. Mainly because the catcher is hitting a much higher amount above replacement level than the guy at 1B. Also position and defensive value matter, certain positions give you the chance to add more value.
|
|
|
Post by umassgrad2005 on Aug 19, 2018 23:32:35 GMT -5
A catcher that hits 20 home runs is more valuable than a first baseman that does, all else being equal, because it’s harder to find another catcher that can do that than it is a first baseman. The positional adjustments might not be perfect, but they exist so that you can compare value across positions. But that does mean that 5 WAR for a catcher is different than 5 WAR for a third baseman. So take 2003 with my pal Kendall. He has 6 HRs, 84 runs,58 RBIs, slash .325/.399/.416 OPS+ 112 WAR of 4.5 Player X, same year: 27 HRs, 103 runs, 106 RBIs, .305/.402/.517 OPS+ 137 WAR 4.4 My point is — is player X equal to Kendall? To me, it isn’t close. Kendall is very good for a catcher, but not nearly as good as the guy at a power position. Both players are equal in how much better they are from replacement level and both would add the same amount of wins to a team. Player X is playing at a position where the replacement level bar is very high.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 23:38:58 GMT -5
Of course I chose an extreme example for ease of math. But DeGrom works: Mets have won 11 of his starts, and he has almost 8 wins above replacement. This suggests his replacement would have a record well below their overall winning percentage. That to me is a significant issue. As for using it as a comparison.... let’s take the example of Jason Kendall in 1998. He was 20th in bWAR in the NL with 5.6 (incidentally, the Pirates won 69 games). This puts him 2.3 ahead of Todd Helton, who is better in every offensive stat except stolen bases. And Helton is, at that point, a future gold glover, so it ought not be defense. Is Kendall the better player that year? You are getting so hung up on Degrom and his win total, while completely overlooking the fact that wins are a horrible way of judging a pitcher. Instead of looking at just wins, look how he's pitched and things he can't control like run support. That is what war is doing. You sound like old school guys 20 years ago that only wins matter and those guys get the Cy Youngs, while totally overlooking things a pitcher can't control. Put Degrom on a good team he'd have a lot more wins, its rather simple. The whole concept is based on replacement level for a given position. The bar is higher at a corner OF position and much lower at catcher. Just go look at the corner OF players and the catchers as a group. Its rather easy to see. The top corner OF will almost always out hit the top catcher. Yet there are ton of very good corner OF and only a few good catchers. Making a top catcher that can hit and play D incredibly valuable. Do to the averages a replacement level corner OF will out hit a replacement level catcher by a ton. So who's more valuable the couple elite catchers or the large group of corner OFs? I’m not talking about his win total. I’m talking about the team’s record in games he pitches. It’s different. And... a replacement would have his team. So you can’t factor that out. I have no beef with “elite” catchers, except that what makes an elite catcher is far less than other positions. Hence, for example, Swihart being exciting as a catching prospect but not as an outfielder. Because catchers can give less and be good catchers. Again, it is apples and oranges. But it means that 5 WAR means different things by position.
|
|
manfred
Veteran
Posts: 11,405
Member is Online
|
Post by manfred on Aug 19, 2018 23:43:10 GMT -5
But that does mean that 5 WAR for a catcher is different than 5 WAR for a third baseman. So take 2003 with my pal Kendall. He has 6 HRs, 84 runs,58 RBIs, slash .325/.399/.416 OPS+ 112 WAR of 4.5 Player X, same year: 27 HRs, 103 runs, 106 RBIs, .305/.402/.517 OPS+ 137 WAR 4.4 My point is — is player X equal to Kendall? To me, it isn’t close. Kendall is very good for a catcher, but not nearly as good as the guy at a power position. Both players are equal in how much better they are from replacement level and both would add the same amount of wins to a team. Player X is playing at a position where the replacement level bar is very high. That is crazy to say Kendall brings more wins to his team than player X. He is significantly worse almost entirely across the board. That one is a left fielder might mean you could get a different slugger to fill his position, but so what? My point is: X may be worth 4ish wins more (who knows) than another LF, and Kendall may be worth more than another catcher... but if leftfielders are — on average — worth more than catchers, then X ought to be assigned a different value from Kendall. Is he docked for opportunity cost?
|
|
|
Post by incandenza on Aug 20, 2018 0:31:53 GMT -5
Both players are equal in how much better they are from replacement level and both would add the same amount of wins to a team. Player X is playing at a position where the replacement level bar is very high. That is crazy to say Kendall brings more wins to his team than player X. He is significantly worse almost entirely across the board. That one is a left fielder might mean you could get a different slugger to fill his position, but so what? My point is: X may be worth 4ish wins more (who knows) than another LF, and Kendall may be worth more than another catcher... but if leftfielders are — on average — worth more than catchers, then X ought to be assigned a different value from Kendall. Is he docked for opportunity cost? Well I don't know who this X is, but you haven't given us his baserunning and defensive stats, so there's that. Also one thread of your argument seems to be "WAR is not perfect at assessing a player's value," which every reasonable person agrees with. And while I am on record as being somewhat skeptical of the positional adjustments WAR makes, the basic intuition is reasonable isn't it? Not that many people can catch, and even fewer catchers can hit. If you get a catcher that can hit really well, you're getting a lot of value out of a position other teams don't get value from. Compare that to left fielders, which are a relative dime a dozen. If what you're trying to measure is the value that a player adds, then the WAR approach seems like a decent approximation. If what you're trying to measure is something else, then... use a different stat. But don't blame WAR for not measuring that other thing. EDIT: Well I went and figured it out - Player X is Chipper Jones. He had, per fangraphs, a defensive rating of -13.5. His UZR was -6.8, meaning he was just a lousy fielder (if you believe UZR). Which raises a point: I notice in your counterarguments you only look at offensive stats, but shouldn't defense and baserunning count too? WAR thinks so! How would you incorporate those other aspects of the game into your evaluation of a player?
|
|
|