@craig44 said:
Do you not think that playing a more demanding defensive position can have a negative impact on a players offense?
If you think it can, then you must agree that in order to level the playing field for player comparisons we need to adjust for position.
I do agree, but that's not what the positional adjustment for WAR does. The adjustment isn't made to account for how physically demanding the position is, but for how important the position is defensively. There's some correlation between the two, but not much. Also, the WAR position adjustment is based on average players at each position while all the rest of WAR is based on comparisons to replacement level players at that position, so it introduces an apples/oranges problem.
But I think you're still missing the most important point - defensive WAR is crap, with or without the positional adjustment. The defensive adjustment just makes it crappier. If you must use it, use it to compare two players at the same position in the same era. That removes the positional adjustment from the equation and reduces the impact of the other deficiencies in the system.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
Yount played 9 full seasons at short and averaged 27 errors per season. That's a lot. I saw him play is lterallly hundreds of games during that time. Good fielding shortstop never entered my mind. Great hitter at short always did.
mark
Walker Proof Digital Album Fellas, leave the tight pants to the ladies. If I can count the coins in your pockets you better use them to call a tailor. Stay thirsty my friends......
Yes It was brutal. He MAYBE played his way to average.
mark
Walker Proof Digital Album Fellas, leave the tight pants to the ladies. If I can count the coins in your pockets you better use them to call a tailor. Stay thirsty my friends......
@craig44 said:
Ok, so if war is crap, how should we adjust for position? It is clear that we need to in order for comparisons to be valid.
There are two ways of approaching this, one purely subjective and the other mostly objective.
The subjective way is to approach it like Bill James did in Win Shares. Under James' system, position is completely ignored when compiling Win Shares. A Win Share reflects a player's contribution to winning a game, and it makes no difference to the team whether a HR is hit by the CF, the SS or the pitcher. Each offensive event is assigned exactly the same value regardless of the player who contributes it. And James then publishes each player's Win Shares for each season, orders them, compares them, etc. But James also published a list of the players who he believed were the 100 best of all-time. He used Win Shares as the foundation for that list, but he also included a positional adjustment to make the list look like he thought it ought to look. Obviously, that was subjective, and he says as much. Without it, I don't know if any catchers would have made the top 100; maybe Bench and Berra would have cracked the bottom 50. James felt that was wrong so he "fixed" it. I like the way James handled it; he publishes the unadjusted numbers, so anyone is free to adjust them as they see fit.
The objective way starts by addressing the fundamental question of how the position someone plays effects their offense. There is a mountain of evidence that playing catcher, especially the further back in time you go, has a tremendous negative effect. Catchers don't play nearly as many games as players at other positions, and playing catcher is just plain exhausting. But beyond that, what evidence is there that any other player's offense is suppressed because of how physically demanding his position is? Certainly, it requires great skill to play SS well, but I don't believe that Ozzie Smith came to the plate unprepared to hit because of the physical strain of playing his position. And there's no evidence that any SS ever did, at least that they were any more exhausted than the OFs who had run down a fly ball or two, or the third baseman who had dived to snare a ball and thrown it 100 feet as hard as he could. Anyway, the objective way to approach the question is to search for evidence in the statistics that players at certain positions have shorter careers than players at other positions; that they hit worse when they come up early in an inning when they are presumably still shaking off the effects of playing the field shortly before that; that they are more prone to injuries than players at other positions; etc. Again, I don't think you'd have much trouble finding evidence for all of this, and maybe more, for catchers. I am skeptical that you would find evidence for anything at all at any other position. As far as know, nobody has ever tried to do this, but I think the "right" adjustments could be found this way. I don't know what the right adjustments are, but I know that the adjustments in WAR are wrong.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
Dallas, I think you may be on to something with your objective way of positional adjusting. I have often thought of doing something similar, but can't come close to having the time to devote to such a project. In lue of doing that level of research, why is it not valid to rank contemporary players against same position players as a baseline?
With the catching position, fatigue is clearly a factor. With skill positions like 3b, ss, 2b, I think the amount of preperation/practice necessary to play the position well can cut into time that player could have used for batting instruction/practice that an outfielder/1b/dh type wouldn't have to do.
We know there are factors involved because some positions just can't produce offensively like others. The trick is adjusting for those factors so we can fairly compare players across positions. For example: of the 500 club members, only 2 played predominantly at a "skill" position. Only 5 members of the 3000 hit club played predominantly at "skill" positions. I excluded Aroid from both lists for obvious reasons. The other members of both clubs were either outfielders/1b or split time between positions.
@craig44 said:
Dallas, I think you may be on to something with your objective way of positional adjusting. I have often thought of doing something similar, but can't come close to having the time to devote to such a project. In lue of doing that level of research, why is it not valid to rank contemporary players against same position players as a baseline?
Nothing wrong with that at all; it's when you try to compare players at two different positions that WAR really falls apart.
With the catching position, fatigue is clearly a factor. With skill positions like 3b, ss, 2b, I think the amount of preperation/practice necessary to play the position well can cut into time that player could have used for batting instruction/practice that an outfielder/1b/dh type wouldn't have to do.
The issue with catcher, as I see it, is that Johnny Bench was a great hitter. Had his manager played him at any other position than catcher his offensive stats would have been greatly improved. That's not true of Ozzie Smith. Had his managers moved him to first or second I see no reason to think that his hitting would have improved. You're saying that Ozzie might have been a better hitter had he played first base for his entire life - going back to high school or earlier, and spending more time in the batting cage - but there's no way to know if that's true and certainly no way to quantify it.
We know there are factors involved because some positions just can't produce offensively like others. The trick is adjusting for those factors so we can fairly compare players across positions. For example: of the 500 club members, only 2 played predominantly at a "skill" position. Only 5 members of the 3000 hit club played predominantly at "skill" positions. I excluded Aroid from both lists for obvious reasons. The other members of both clubs were either outfielders/1b or split time between positions.
The primary reason that skill position players don't hit as many homers, or hit as well generally, is that they aren't as big. I don't think it's reasonable to adjust for that. If we're trying to decide how great a player Johnny Bench was, I do think it's reasonable to consider what he would likely have done had his managers played him at another position. I think it's reasonable to do that for any player at any position, I just don't think the answer changes for any position but catcher.
But note how often I'm saying "I think", or words to that effect. Ultimately, comparing offense and defense and comparing across positions is going to involve at least some subjectivity. Mickey Mantle was better than Nellie Fox. I'm positive about that despite their playing two different positions. I think Robin Yount was better than Dave Winfield, but those two were just too close to ever say that with complete confidence and I don't think there is any such thing as a stat that will change that. The best approach, I think, is to focus on the stats that matter, with no adjustments or other subjective tinkering, recognize what they tell us about how much each player contributed, and then argue about the close calls.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
Yes, often, shortstops are physically smaller. But not so much for third baseman. Why have only two 3b hit 500 hrs? Many 3b are big guys. Size is not always a factor. Look at some great sluggers, mantle/mays/aaron/banks were not huge guys. They show that bat speed/hand quickness is as important as bulk.
The issue with catcher, as I see it, is that Johnny Bench was a great hitter. Had his manager played him at any other position than catcher his offensive stats would have been greatly improved. That's not true of Ozzie Smith. Had his managers moved him to first or second I see no reason to think that his hitting would have improved. You're saying that Ozzie might have been a better hitter had he played first base for his entire life - going back to high school or earlier, and spending more time in the batting cage - but there's no way to know if that's true and certainly no way to quantify it.
Ozzie is a bit of an extreme example. Don't you think ripken, Larkin, Sandberg, Morgan, could have been more productive hitters had they been playing first or dhing? Same idea as bench.
@craig44 said:
Ozzie is a bit of an extreme example. Don't you think ripken, Larkin, Sandberg, Morgan, could have been more productive hitters had they been playing first or dhing? Same idea as bench.
It's possible, although I'm skeptical that the degree to which it's true is worth measuring. In any event, I think it's exceedingly unlikely that the "right" positional adjustment for OF, 2B, SS and 3B would be materially different. So maybe the "right" thing to do is to give catchers a positive adjustment, 1B and DH a negative adjustment and leave everyone else alone. That gets us most of the way to "leave everyone alone", which is what I still think the best approach is.
This is for you @thisistheshow - Jim Rice was actually a pretty good player.
Comments
I do agree, but that's not what the positional adjustment for WAR does. The adjustment isn't made to account for how physically demanding the position is, but for how important the position is defensively. There's some correlation between the two, but not much. Also, the WAR position adjustment is based on average players at each position while all the rest of WAR is based on comparisons to replacement level players at that position, so it introduces an apples/oranges problem.
But I think you're still missing the most important point - defensive WAR is crap, with or without the positional adjustment. The defensive adjustment just makes it crappier. If you must use it, use it to compare two players at the same position in the same era. That removes the positional adjustment from the equation and reduces the impact of the other deficiencies in the system.
Ok, so if war is crap, how should we adjust for position? It is clear that we need to in order for comparisons to be valid.
George Brett, Roger Clemens and Tommy Brady.
Yount played 9 full seasons at short and averaged 27 errors per season. That's a lot. I saw him play is lterallly hundreds of games during that time. Good fielding shortstop never entered my mind. Great hitter at short always did.
mark
Fellas, leave the tight pants to the ladies. If I can count the coins in your pockets you better use them to call a tailor. Stay thirsty my friends......
You have to remember that yount learned ss essentially at the major league level. He had accumulated 94 of those errors by the time he was 20.
George Brett, Roger Clemens and Tommy Brady.
Yes It was brutal. He MAYBE played his way to average.
mark
Fellas, leave the tight pants to the ladies. If I can count the coins in your pockets you better use them to call a tailor. Stay thirsty my friends......
There are two ways of approaching this, one purely subjective and the other mostly objective.
The subjective way is to approach it like Bill James did in Win Shares. Under James' system, position is completely ignored when compiling Win Shares. A Win Share reflects a player's contribution to winning a game, and it makes no difference to the team whether a HR is hit by the CF, the SS or the pitcher. Each offensive event is assigned exactly the same value regardless of the player who contributes it. And James then publishes each player's Win Shares for each season, orders them, compares them, etc. But James also published a list of the players who he believed were the 100 best of all-time. He used Win Shares as the foundation for that list, but he also included a positional adjustment to make the list look like he thought it ought to look. Obviously, that was subjective, and he says as much. Without it, I don't know if any catchers would have made the top 100; maybe Bench and Berra would have cracked the bottom 50. James felt that was wrong so he "fixed" it. I like the way James handled it; he publishes the unadjusted numbers, so anyone is free to adjust them as they see fit.
The objective way starts by addressing the fundamental question of how the position someone plays effects their offense. There is a mountain of evidence that playing catcher, especially the further back in time you go, has a tremendous negative effect. Catchers don't play nearly as many games as players at other positions, and playing catcher is just plain exhausting. But beyond that, what evidence is there that any other player's offense is suppressed because of how physically demanding his position is? Certainly, it requires great skill to play SS well, but I don't believe that Ozzie Smith came to the plate unprepared to hit because of the physical strain of playing his position. And there's no evidence that any SS ever did, at least that they were any more exhausted than the OFs who had run down a fly ball or two, or the third baseman who had dived to snare a ball and thrown it 100 feet as hard as he could. Anyway, the objective way to approach the question is to search for evidence in the statistics that players at certain positions have shorter careers than players at other positions; that they hit worse when they come up early in an inning when they are presumably still shaking off the effects of playing the field shortly before that; that they are more prone to injuries than players at other positions; etc. Again, I don't think you'd have much trouble finding evidence for all of this, and maybe more, for catchers. I am skeptical that you would find evidence for anything at all at any other position. As far as know, nobody has ever tried to do this, but I think the "right" adjustments could be found this way. I don't know what the right adjustments are, but I know that the adjustments in WAR are wrong.
Dallas, I think you may be on to something with your objective way of positional adjusting. I have often thought of doing something similar, but can't come close to having the time to devote to such a project. In lue of doing that level of research, why is it not valid to rank contemporary players against same position players as a baseline?
With the catching position, fatigue is clearly a factor. With skill positions like 3b, ss, 2b, I think the amount of preperation/practice necessary to play the position well can cut into time that player could have used for batting instruction/practice that an outfielder/1b/dh type wouldn't have to do.
We know there are factors involved because some positions just can't produce offensively like others. The trick is adjusting for those factors so we can fairly compare players across positions. For example: of the 500 club members, only 2 played predominantly at a "skill" position. Only 5 members of the 3000 hit club played predominantly at "skill" positions. I excluded Aroid from both lists for obvious reasons. The other members of both clubs were either outfielders/1b or split time between positions.
George Brett, Roger Clemens and Tommy Brady.
Nothing wrong with that at all; it's when you try to compare players at two different positions that WAR really falls apart.
The issue with catcher, as I see it, is that Johnny Bench was a great hitter. Had his manager played him at any other position than catcher his offensive stats would have been greatly improved. That's not true of Ozzie Smith. Had his managers moved him to first or second I see no reason to think that his hitting would have improved. You're saying that Ozzie might have been a better hitter had he played first base for his entire life - going back to high school or earlier, and spending more time in the batting cage - but there's no way to know if that's true and certainly no way to quantify it.
The primary reason that skill position players don't hit as many homers, or hit as well generally, is that they aren't as big. I don't think it's reasonable to adjust for that. If we're trying to decide how great a player Johnny Bench was, I do think it's reasonable to consider what he would likely have done had his managers played him at another position. I think it's reasonable to do that for any player at any position, I just don't think the answer changes for any position but catcher.
But note how often I'm saying "I think", or words to that effect. Ultimately, comparing offense and defense and comparing across positions is going to involve at least some subjectivity. Mickey Mantle was better than Nellie Fox. I'm positive about that despite their playing two different positions. I think Robin Yount was better than Dave Winfield, but those two were just too close to ever say that with complete confidence and I don't think there is any such thing as a stat that will change that. The best approach, I think, is to focus on the stats that matter, with no adjustments or other subjective tinkering, recognize what they tell us about how much each player contributed, and then argue about the close calls.
Yes, often, shortstops are physically smaller. But not so much for third baseman. Why have only two 3b hit 500 hrs? Many 3b are big guys. Size is not always a factor. Look at some great sluggers, mantle/mays/aaron/banks were not huge guys. They show that bat speed/hand quickness is as important as bulk.
The issue with catcher, as I see it, is that Johnny Bench was a great hitter. Had his manager played him at any other position than catcher his offensive stats would have been greatly improved. That's not true of Ozzie Smith. Had his managers moved him to first or second I see no reason to think that his hitting would have improved. You're saying that Ozzie might have been a better hitter had he played first base for his entire life - going back to high school or earlier, and spending more time in the batting cage - but there's no way to know if that's true and certainly no way to quantify it.
Ozzie is a bit of an extreme example. Don't you think ripken, Larkin, Sandberg, Morgan, could have been more productive hitters had they been playing first or dhing? Same idea as bench.
George Brett, Roger Clemens and Tommy Brady.
It's possible, although I'm skeptical that the degree to which it's true is worth measuring. In any event, I think it's exceedingly unlikely that the "right" positional adjustment for OF, 2B, SS and 3B would be materially different. So maybe the "right" thing to do is to give catchers a positive adjustment, 1B and DH a negative adjustment and leave everyone else alone. That gets us most of the way to "leave everyone alone", which is what I still think the best approach is.