Going to WAR...for Classic0

[Ozziesfan41]

Its absurd people think Jason Heywards 2015 7 war 13 homerun 60 RBI 79 Run season was not just better but much better than Bichettes 1995 1.2 War 102 run 40 homerun 128 RBI season its just ridiculous and if both of them were coming off those seasons you can bet Bichette would get the much bigger contract than the elite war heyward. Its just an absurd stat

Do you not think Heyward would have had much better offensive stats if he was transported back to 1995 and got to play half his games at Coors Field?

In addition to Bichette, Walker had a phenomenal year in 1995 (36/101) while two other non-hall of famers Galarraga (31/106) and Castilla (32/90) put up MVP "middle of the order bat" type of traditional stats. Four players with 30+ HRs, four players with 90+ RBI, three players with 100+ RBI.

Same exact thing in 1999 (four 30+ HR hitters, four 100+ RBI hitters). You can't just look at HR and RBI and say the mid to late 90s Rockies were the best collection of hitters ever assembled, that's blatantly ignoring massive amounts of context.

The Rockies as a team in 1995 had the following splits . . .
Home - .316/.383/.556/.939, 134 HR, 485 R
Away - .247/.315/.384/.700, 66 HR, 300 R

. . . weird. Their road OBP is on par with their home BA. Their road SLG is on par with their home OBP. They hit more than twice as many HRs at home as they did on the road. What happened?

Okay we can use players from the same year Bautista 4.8 war 108 runs 40 home runs 114 RBI Heyward 7 war 79 runs 13 home runs 60 rbi and you’re telling me Heyward was better then Bautista because of war that’s ridiculous no one in their right mind would have taken Heywards production over Bautistas except maybe the war champions

I've already outlined the differences between Heyward's bWAR of 7 and his fWAR of 5.6 in 2015, driven mostly by different defensive inputs. Both DRS (used in bWAR) and UZR (used in fWAR) think he was great defensively (which I think we all agree with), but DRS thinks he stood out from the pack by another 11 runs saved, or 1.1 WAR. I'm not sure which one is more correct and which one is less correct, but I've repeatedly said I use fWAR and that DRS typically has a much wider range of defensive outcomes.

By fWAR in 2015 . . .
Heyward - 5.6 fWAR
Bautista - 5.2 fWAR

That's well within the range of saying there's not much between them, when looking at fWAR, which I think is a conclusion most could get behind even if they prefer how one player creates their value (more certainty in offensive metrics, etc.). On rare occasions bWAR and fWAR have big differences in value. For position players that almost always is for defensive outliers (on either end of the spectrum). I'd encourage you in these instances to check fWAR and see if those results align more with your expectation.

So check and use whatever war stat I want to use that aligns with my beliefs lol yea war is super reliable then. Oh one war says Heyward is way better the other says he’s closer. Both are wrong in my option I would take Bautista any day over Heyward and so would just about every major league team probably. So I’ll just agree to disagree because neither of us will change our minds

[rbirules]

Its absurd people think Jason Heywards 2015 7 war 13 homerun 60 RBI 79 Run season was not just better but much better than Bichettes 1995 1.2 War 102 run 40 homerun 128 RBI season its just ridiculous and if both of them were coming off those seasons you can bet Bichette would get the much bigger contract than the elite war heyward. Its just an absurd stat

Do you not think Heyward would have had much better offensive stats if he was transported back to 1995 and got to play half his games at Coors Field?

In addition to Bichette, Walker had a phenomenal year in 1995 (36/101) while two other non-hall of famers Galarraga (31/106) and Castilla (32/90) put up MVP "middle of the order bat" type of traditional stats. Four players with 30+ HRs, four players with 90+ RBI, three players with 100+ RBI.

Same exact thing in 1999 (four 30+ HR hitters, four 100+ RBI hitters). You can't just look at HR and RBI and say the mid to late 90s Rockies were the best collection of hitters ever assembled, that's blatantly ignoring massive amounts of context.

The Rockies as a team in 1995 had the following splits . . .
Home - .316/.383/.556/.939, 134 HR, 485 R
Away - .247/.315/.384/.700, 66 HR, 300 R

. . . weird. Their road OBP is on par with their home BA. Their road SLG is on par with their home OBP. They hit more than twice as many HRs at home as they did on the road. What happened?

Okay we can use players from the same year Bautista 4.8 war 108 runs 40 home runs 114 RBI Heyward 7 war 79 runs 13 home runs 60 rbi and you’re telling me Heyward was better then Bautista because of war that’s ridiculous no one in their right mind would have taken Heywards production over Bautistas except maybe the war champions

I've already outlined the differences between Heyward's bWAR of 7 and his fWAR of 5.6 in 2015, driven mostly by different defensive inputs. Both DRS (used in bWAR) and UZR (used in fWAR) think he was great defensively (which I think we all agree with), but DRS thinks he stood out from the pack by another 11 runs saved, or 1.1 WAR. I'm not sure which one is more correct and which one is less correct, but I've repeatedly said I use fWAR and that DRS typically has a much wider range of defensive outcomes.

By fWAR in 2015 . . .
Heyward - 5.6 fWAR
Bautista - 5.2 fWAR

That's well within the range of saying there's not much between them, when looking at fWAR, which I think is a conclusion most could get behind even if they prefer how one player creates their value (more certainty in offensive metrics, etc.). On rare occasions bWAR and fWAR have big differences in value. For position players that almost always is for defensive outliers (on either end of the spectrum). I'd encourage you in these instances to check fWAR and see if those results align more with your expectation.

So check and use whatever war stat I want to use that aligns with my beliefs lol yea war is super reliable then. Oh one war says Heyward is way better the other says he’s closer. Both are wrong in my option I would take Bautista any day over Heyward and so would just about every major league team probably. So I’ll just agree to disagree because neither of us will change our minds

I look at fWAR almost exclusively, mostly for the way their handle the run suppression side of baseball (plus I like fangraphs' site much better than BBR), so I rarely look at bWAR. But these conversations come up from time to time about a poster freaking out about a handpicked comparison of players and it's almost always looking at bWAR. The reason is almost always driven by the defensive side of things. Consider this a PSA that these comparisons are usually so far apart because of DRS. I'm not going to tell people which model to use, though I would hope you have reasons for picking the one you do besides "that's the website I go to for baseball stats", but if an outcome looks odd check the other model to see if it agrees. If it doesn't (for position players) it's likely due to defense and you can easily dig into that if needed. (For pitchers the two models evaluate them differently so you'll see differences there pretty often.)

Nobody is stopping you from taking Baustista's 2015 over Heyward's. They are close enough that either player could be better. Defensive stats have "error bars" in their results (larger ones than offensive stats), and take much longer to stabilize so we sometimes get weird outliers for a single season.

I'm not sure what I'm not changing my mind about, or unwilling to change my mind about (that defense matters?), but whatever.

[Ozziesfan41]

Its absurd people think Jason Heywards 2015 7 war 13 homerun 60 RBI 79 Run season was not just better but much better than Bichettes 1995 1.2 War 102 run 40 homerun 128 RBI season its just ridiculous and if both of them were coming off those seasons you can bet Bichette would get the much bigger contract than the elite war heyward. Its just an absurd stat

Do you not think Heyward would have had much better offensive stats if he was transported back to 1995 and got to play half his games at Coors Field?

In addition to Bichette, Walker had a phenomenal year in 1995 (36/101) while two other non-hall of famers Galarraga (31/106) and Castilla (32/90) put up MVP "middle of the order bat" type of traditional stats. Four players with 30+ HRs, four players with 90+ RBI, three players with 100+ RBI.

Same exact thing in 1999 (four 30+ HR hitters, four 100+ RBI hitters). You can't just look at HR and RBI and say the mid to late 90s Rockies were the best collection of hitters ever assembled, that's blatantly ignoring massive amounts of context.

The Rockies as a team in 1995 had the following splits . . .
Home - .316/.383/.556/.939, 134 HR, 485 R
Away - .247/.315/.384/.700, 66 HR, 300 R

. . . weird. Their road OBP is on par with their home BA. Their road SLG is on par with their home OBP. They hit more than twice as many HRs at home as they did on the road. What happened?

Okay we can use players from the same year Bautista 4.8 war 108 runs 40 home runs 114 RBI Heyward 7 war 79 runs 13 home runs 60 rbi and you’re telling me Heyward was better then Bautista because of war that’s ridiculous no one in their right mind would have taken Heywards production over Bautistas except maybe the war champions

I've already outlined the differences between Heyward's bWAR of 7 and his fWAR of 5.6 in 2015, driven mostly by different defensive inputs. Both DRS (used in bWAR) and UZR (used in fWAR) think he was great defensively (which I think we all agree with), but DRS thinks he stood out from the pack by another 11 runs saved, or 1.1 WAR. I'm not sure which one is more correct and which one is less correct, but I've repeatedly said I use fWAR and that DRS typically has a much wider range of defensive outcomes.

By fWAR in 2015 . . .
Heyward - 5.6 fWAR
Bautista - 5.2 fWAR

That's well within the range of saying there's not much between them, when looking at fWAR, which I think is a conclusion most could get behind even if they prefer how one player creates their value (more certainty in offensive metrics, etc.). On rare occasions bWAR and fWAR have big differences in value. For position players that almost always is for defensive outliers (on either end of the spectrum). I'd encourage you in these instances to check fWAR and see if those results align more with your expectation.

So check and use whatever war stat I want to use that aligns with my beliefs lol yea war is super reliable then. Oh one war says Heyward is way better the other says he’s closer. Both are wrong in my option I would take Bautista any day over Heyward and so would just about every major league team probably. So I’ll just agree to disagree because neither of us will change our minds

I look at fWAR almost exclusively, mostly for the way their handle the run suppression side of baseball (plus I like fangraphs' site much better than BBR), so I rarely look at bWAR. But these conversations come up from time to time about a poster freaking out about a handpicked comparison of players and it's almost always looking at bWAR. The reason is almost always driven by the defensive side of things. Consider this a PSA that these comparisons are usually so far apart because of DRS. I'm not going to tell people which model to use, though I would hope you have reasons for picking the one you do besides "that's the website I go to for baseball stats", but if an outcome looks odd check the other model to see if it agrees. If it doesn't (for position players) it's likely due to defense and you can easily dig into that if needed. (For pitchers the two models evaluate them differently so you'll see differences there pretty often.)

Nobody is stopping you from taking Baustista's 2015 over Heyward's. They are close enough that either player could be better. Defensive stats have "error bars" in their results (larger ones than offensive stats), and take much longer to stabilize so we sometimes get weird outliers for a single season.

I'm not sure what I'm not changing my mind about, or unwilling to change my mind about (that defense matters?), but whatever.

I don’t expect you to change your mind and I’m not changing my mind about not overvaluing a corner outfielders defense who only makes one to two catches per game and most of those are routine that just about any outfielder would make. I’m also not going to start loving war when I can say I agree with this war so I’ll use it but I don’t agree with that war so I won’t use that it proves how unreliable it is

[ClassicO]

I'm making assumptions because I can't see behind the curtain of advanced metrics. And the more recent, more accurate ones didn't even exist in the 1990s. What would you suggest I do? I'm showing on a rate basis he caught less balls in 1999 than he did in 1998 or 1997, he also made more errors. Those have consequences do they not? It's not ruled an error if the other team didn't gain a base (dropped fly ball leading to a hit or bad throw leading to an extra base or two).

". . . that probably had no bad outcomes as far as the actual game." Now who's making assumptions? How do you know none of those plays impacted a game? For that matter, why should we count hits, HRs, RBIs, if they don't affect a game? If a team wins 15-3 and is already up 7-3 when Bichette hits a 3 run HR and gets a double later on, should we count that? This is an absurd argument. We still count hits that don't drive in runs in BA, OBP, SLG, etc.

You don't know when a hit will lead to the start of, or sustaining of, a rally, just like you don't know if a caught fly ball might stop one (or a missed fly ball might start or sustain one).

There's certainly times when using context dependent stats (RBI, runs, RE24, WPA) makes sense, but I would say in the vast majority of situations context neutral stats (BA, OBP, SLG, OPS, wOBA, wRC+, WAR, etc.) are preferred.

He had 1.86CH per GAME…Watching games how many times are there balls hit that require a GG speedster to field to keep that runner from advancing to the “extra” base. Very RARE…..therefore it makes very little sense that top of league offensive output is Negated to the point of turning negative in ANY rating system because of LF defense. Huge overcompensation. You even admit that bWAR is weighs much more heavily than fWAR.,…given that how do you know they both don’t put too much weight on D??

How many times do you see a ball caught at the edge of a fielder's range? Fairly often. How often do you see a slow OF just barely not get to a ball that you think should have been caught, or maybe should have been caught? A fair amount over 162 games. One of those a week is 26 XBHs a season that a defender is giving back to the other team, to use an extreme example. Bichette caught (on a rate basis) 15-30 fewer balls than he himself did in his age 33 and 34 seasons. Is it crazy to think at age 35 he's slowing down and not getting to as many balls anymore? And that most of those balls are going for at least a single if not a double in Coors?

1.86 CH/game is 301 over the course of a season. Giving away 10-20 could easily be 15-30 "total bases" given back.

bWAR does not weight defense much more heavily than fWAR, I believe they both account for 7% of total WAR (14% of run suppression, the other 86% going to pitchers). UZR and DRS are both zero sum stats. If one player is well above average at a position another player, or several players, have to be (well) below average to balance the scales (kind of like a perfectly balanced see saw). I'm saying DRS has more outliers (on both ends of the spectrum) whereas UZR tends to have more players closer to the middle of the seesaw to balance things out. I'm not even saying DRS (and hence bWAR) is wrong, I'm just saying that's what I see in the metrics and thus we see wider ranges of DRS and bWAR than we do in UZR and fWAR, typically.

Excellent stuff. You're WAY over the capacity of the complainers. Melville still uses rudimentary stats -- or mostly none, just the same blowhard template.

[ScotchMIrish]

Designated hitters don’t lose runs. My gawd.

And they do nothing to prevent runs (another important aspect of the game). And the positional +/- is all about comparisons.
Is a LF as valuable overall as a SS - or DH?

This. It's the relativity to other positions that matters. They could have designed it with DH having a 0 positional adjustment and C having a +30 adjustment, but then they'd have to lower the baseline later, like they do when they go from RAA (runs above average) to RAR (runs above replacement), so instead they just centered all positions around 0 (some above, some below).

Numbers accumulated while being a DH should not count in calculating WAR. I would say the same regarding career totals. DH has a much lower risk of injury because he sits on the bench for the entire game except for the 4 or 5 times he walks to the batters box. He doesn't need to own a glove. He isn't a baseball player. He is a batter.

Yes, your disdain for the DH has long been established. Saying numbers from that position should not count is just an absurd position to take without any real argument for doing so other than "I don't like the DH". The DH gets a massive positional adjustment for the very reason you described, they contribute nothing in the field.

The DH isn't a baseball player just like a man at a driving range isn't a golfer.

[Goldfan]

I'm making assumptions because I can't see behind the curtain of advanced metrics. And the more recent, more accurate ones didn't even exist in the 1990s. What would you suggest I do? I'm showing on a rate basis he caught less balls in 1999 than he did in 1998 or 1997, he also made more errors. Those have consequences do they not? It's not ruled an error if the other team didn't gain a base (dropped fly ball leading to a hit or bad throw leading to an extra base or two).

". . . that probably had no bad outcomes as far as the actual game." Now who's making assumptions? How do you know none of those plays impacted a game? For that matter, why should we count hits, HRs, RBIs, if they don't affect a game? If a team wins 15-3 and is already up 7-3 when Bichette hits a 3 run HR and gets a double later on, should we count that? This is an absurd argument. We still count hits that don't drive in runs in BA, OBP, SLG, etc.

You don't know when a hit will lead to the start of, or sustaining of, a rally, just like you don't know if a caught fly ball might stop one (or a missed fly ball might start or sustain one).

There's certainly times when using context dependent stats (RBI, runs, RE24, WPA) makes sense, but I would say in the vast majority of situations context neutral stats (BA, OBP, SLG, OPS, wOBA, wRC+, WAR, etc.) are preferred.

He had 1.86CH per GAME…Watching games how many times are there balls hit that require a GG speedster to field to keep that runner from advancing to the “extra” base. Very RARE…..therefore it makes very little sense that top of league offensive output is Negated to the point of turning negative in ANY rating system because of LF defense. Huge overcompensation. You even admit that bWAR is weighs much more heavily than fWAR.,…given that how do you know they both don’t put too much weight on D??

How many times do you see a ball caught at the edge of a fielder's range? Fairly often. How often do you see a slow OF just barely not get to a ball that you think should have been caught, or maybe should have been caught? A fair amount over 162 games. One of those a week is 26 XBHs a season that a defender is giving back to the other team, to use an extreme example. Bichette caught (on a rate basis) 15-30 fewer balls than he himself did in his age 33 and 34 seasons. Is it crazy to think at age 35 he's slowing down and not getting to as many balls anymore? And that most of those balls are going for at least a single if not a double in Coors?

1.86 CH/game is 301 over the course of a season. Giving away 10-20 could easily be 15-30 "total bases" given back.

bWAR does not weight defense much more heavily than fWAR, I believe they both account for 7% of total WAR (14% of run suppression, the other 86% going to pitchers). UZR and DRS are both zero sum stats. If one player is well above average at a position another player, or several players, have to be (well) below average to balance the scales (kind of like a perfectly balanced see saw). I'm saying DRS has more outliers (on both ends of the spectrum) whereas UZR tends to have more players closer to the middle of the seesaw to balance things out. I'm not even saying DRS (and hence bWAR) is wrong, I'm just saying that's what I see in the metrics and thus we see wider ranges of DRS and bWAR than we do in UZR and fWAR, typically.

Excellent stuff. You're WAY over the capacity of the complainers. Melville still uses rudimentary stats -- or mostly none, just the same blowhard template.

“Defensive stats have "error bars" in their results (larger ones than offensive stats), and take much longer to stabilize so we sometimes get weird outliers for a single season.”
I believe this is what I’ve been debating for 16 pg…………….
And Remember RBI says bWAR skews more to D in their WAR ratings and he only uses fWAR…….
So acting like there isn’t a bias, which is exactly what we’ve been debating, is comical…..

[ScotchMIrish]

Designated hitters don’t lose runs. My gawd.

And they do nothing to prevent runs (another important aspect of the game). And the positional +/- is all about comparisons.
Is a LF as valuable overall as a SS - or DH?

This. It's the relativity to other positions that matters. They could have designed it with DH having a 0 positional adjustment and C having a +30 adjustment, but then they'd have to lower the baseline later, like they do when they go from RAA (runs above average) to RAR (runs above replacement), so instead they just centered all positions around 0 (some above, some below).

Numbers accumulated while being a DH should not count in calculating WAR. I would say the same regarding career totals. DH has a much lower risk of injury because he sits on the bench for the entire game except for the 4 or 5 times he walks to the batters box. He doesn't need to own a glove. He isn't a baseball player. He is a batter.

You have to score runs to win baseball games. Why would you NOT count an offensive player's offensive contributions?

That's absurd.

The Cardinals catcher is barely above the Mendoza line. He plays because Herrera can't catch. WAR doesn't factor the drag on the offense from Herrera's inability to play a position.

[rbirules]

I'm making assumptions because I can't see behind the curtain of advanced metrics. And the more recent, more accurate ones didn't even exist in the 1990s. What would you suggest I do? I'm showing on a rate basis he caught less balls in 1999 than he did in 1998 or 1997, he also made more errors. Those have consequences do they not? It's not ruled an error if the other team didn't gain a base (dropped fly ball leading to a hit or bad throw leading to an extra base or two).

". . . that probably had no bad outcomes as far as the actual game." Now who's making assumptions? How do you know none of those plays impacted a game? For that matter, why should we count hits, HRs, RBIs, if they don't affect a game? If a team wins 15-3 and is already up 7-3 when Bichette hits a 3 run HR and gets a double later on, should we count that? This is an absurd argument. We still count hits that don't drive in runs in BA, OBP, SLG, etc.

You don't know when a hit will lead to the start of, or sustaining of, a rally, just like you don't know if a caught fly ball might stop one (or a missed fly ball might start or sustain one).

There's certainly times when using context dependent stats (RBI, runs, RE24, WPA) makes sense, but I would say in the vast majority of situations context neutral stats (BA, OBP, SLG, OPS, wOBA, wRC+, WAR, etc.) are preferred.

He had 1.86CH per GAME…Watching games how many times are there balls hit that require a GG speedster to field to keep that runner from advancing to the “extra” base. Very RARE…..therefore it makes very little sense that top of league offensive output is Negated to the point of turning negative in ANY rating system because of LF defense. Huge overcompensation. You even admit that bWAR is weighs much more heavily than fWAR.,…given that how do you know they both don’t put too much weight on D??

How many times do you see a ball caught at the edge of a fielder's range? Fairly often. How often do you see a slow OF just barely not get to a ball that you think should have been caught, or maybe should have been caught? A fair amount over 162 games. One of those a week is 26 XBHs a season that a defender is giving back to the other team, to use an extreme example. Bichette caught (on a rate basis) 15-30 fewer balls than he himself did in his age 33 and 34 seasons. Is it crazy to think at age 35 he's slowing down and not getting to as many balls anymore? And that most of those balls are going for at least a single if not a double in Coors?

1.86 CH/game is 301 over the course of a season. Giving away 10-20 could easily be 15-30 "total bases" given back.

bWAR does not weight defense much more heavily than fWAR, I believe they both account for 7% of total WAR (14% of run suppression, the other 86% going to pitchers). UZR and DRS are both zero sum stats. If one player is well above average at a position another player, or several players, have to be (well) below average to balance the scales (kind of like a perfectly balanced see saw). I'm saying DRS has more outliers (on both ends of the spectrum) whereas UZR tends to have more players closer to the middle of the seesaw to balance things out. I'm not even saying DRS (and hence bWAR) is wrong, I'm just saying that's what I see in the metrics and thus we see wider ranges of DRS and bWAR than we do in UZR and fWAR, typically.

Excellent stuff. You're WAY over the capacity of the complainers. Melville still uses rudimentary stats -- or mostly none, just the same blowhard template.

“Defensive stats have "error bars" in their results (larger ones than offensive stats), and take much longer to stabilize so we sometimes get weird outliers for a single season.”
I believe this is what I’ve been debating for 16 pg…………….
And Remember RBI says bWAR skews more to D in their WAR ratings and he only uses fWAR…….
So acting like there isn’t a bias, which is exactly what we’ve been debating, is comical…..

Yeah, rating defense of every player in the league in a systematic way is difficult, but it's better than not accounting for defense or just subjectively doing it in our heads based on the three games we've seen that player play against the Cardinals during an entire season.

Again, bWAR does not skew more favorably to defense in it's WAR model. I said the opposite, they give the same total credit to defense, IIRC. bWAR defensive input tends to have a wider range of individual player values, total defensive WAR league wide is the same.

I use fWAR because of how it treats pitchers. I have no issue with how bWAR treats position players.

[Goldfan]

I'm making assumptions because I can't see behind the curtain of advanced metrics. And the more recent, more accurate ones didn't even exist in the 1990s. What would you suggest I do? I'm showing on a rate basis he caught less balls in 1999 than he did in 1998 or 1997, he also made more errors. Those have consequences do they not? It's not ruled an error if the other team didn't gain a base (dropped fly ball leading to a hit or bad throw leading to an extra base or two).

". . . that probably had no bad outcomes as far as the actual game." Now who's making assumptions? How do you know none of those plays impacted a game? For that matter, why should we count hits, HRs, RBIs, if they don't affect a game? If a team wins 15-3 and is already up 7-3 when Bichette hits a 3 run HR and gets a double later on, should we count that? This is an absurd argument. We still count hits that don't drive in runs in BA, OBP, SLG, etc.

You don't know when a hit will lead to the start of, or sustaining of, a rally, just like you don't know if a caught fly ball might stop one (or a missed fly ball might start or sustain one).

There's certainly times when using context dependent stats (RBI, runs, RE24, WPA) makes sense, but I would say in the vast majority of situations context neutral stats (BA, OBP, SLG, OPS, wOBA, wRC+, WAR, etc.) are preferred.

He had 1.86CH per GAME…Watching games how many times are there balls hit that require a GG speedster to field to keep that runner from advancing to the “extra” base. Very RARE…..therefore it makes very little sense that top of league offensive output is Negated to the point of turning negative in ANY rating system because of LF defense. Huge overcompensation. You even admit that bWAR is weighs much more heavily than fWAR.,…given that how do you know they both don’t put too much weight on D??

How many times do you see a ball caught at the edge of a fielder's range? Fairly often. How often do you see a slow OF just barely not get to a ball that you think should have been caught, or maybe should have been caught? A fair amount over 162 games. One of those a week is 26 XBHs a season that a defender is giving back to the other team, to use an extreme example. Bichette caught (on a rate basis) 15-30 fewer balls than he himself did in his age 33 and 34 seasons. Is it crazy to think at age 35 he's slowing down and not getting to as many balls anymore? And that most of those balls are going for at least a single if not a double in Coors?

1.86 CH/game is 301 over the course of a season. Giving away 10-20 could easily be 15-30 "total bases" given back.

bWAR does not weight defense much more heavily than fWAR, I believe they both account for 7% of total WAR (14% of run suppression, the other 86% going to pitchers). UZR and DRS are both zero sum stats. If one player is well above average at a position another player, or several players, have to be (well) below average to balance the scales (kind of like a perfectly balanced see saw). I'm saying DRS has more outliers (on both ends of the spectrum) whereas UZR tends to have more players closer to the middle of the seesaw to balance things out. I'm not even saying DRS (and hence bWAR) is wrong, I'm just saying that's what I see in the metrics and thus we see wider ranges of DRS and bWAR than we do in UZR and fWAR, typically.

Excellent stuff. You're WAY over the capacity of the complainers. Melville still uses rudimentary stats -- or mostly none, just the same blowhard template.

“Defensive stats have "error bars" in their results (larger ones than offensive stats), and take much longer to stabilize so we sometimes get weird outliers for a single season.”
I believe this is what I’ve been debating for 16 pg…………….
And Remember RBI says bWAR skews more to D in their WAR ratings and he only uses fWAR…….
So acting like there isn’t a bias, which is exactly what we’ve been debating, is comical…..

Yeah, rating defense of every player in the league in a systematic way is difficult, but it's better than not accounting for defense or just subjectively doing it in our heads based on the three games we've seen that player play against the Cardinals during an entire season.

Again, bWAR does not skew more favorably to defense in it's WAR model. I said the opposite, they give the same total credit to defense, IIRC. bWAR defensive input tends to have a wider range of individual player values, total defensive WAR league wide is the same.

I use fWAR because of how it treats pitchers. I have no issue with how bWAR treats position players.

I thought you wrote that bWAR weighs more heavily to D thus the 2015 Heyward 7WAR/Carp 4.8…..and the much smaller gap with fWAR

[Melville]

Its absurd people think Jason Heywards 2015 7 war 13 homerun 60 RBI 79 Run season was not just better but much better than Bichettes 1995 1.2 War 102 run 40 homerun 128 RBI season its just ridiculous and if both of them were coming off those seasons you can bet Bichette would get the much bigger contract than the elite war heyward. Its just an absurd stat

Fact.

[Melville]

Designated hitters don’t lose runs. My gawd.

And they do nothing to prevent runs (another important aspect of the game). And the positional +/- is all about comparisons.
Is a LF as valuable overall as a SS - or DH?

This. It's the relativity to other positions that matters. They could have designed it with DH having a 0 positional adjustment and C having a +30 adjustment, but then they'd have to lower the baseline later, like they do when they go from RAA (runs above average) to RAR (runs above replacement), so instead they just centered all positions around 0 (some above, some below).

Numbers accumulated while being a DH should not count in calculating WAR. I would say the same regarding career totals. DH has a much lower risk of injury because he sits on the bench for the entire game except for the 4 or 5 times he walks to the batters box. He doesn't need to own a glove. He isn't a baseball player. He is a batter.

You have to score runs to win baseball games. Why would you NOT count an offensive player's offensive contributions?

That's absurd.

The Cardinals catcher is barely above the Mendoza line. He plays because Herrera can't catch. WAR doesn't factor the drag on the offense from Herrera's inability to play a position.

Correct.
WAR and its adherents can't understand the baseball symphony.
It sees only isolated instruments.
Displays a fundamental lack of understanding of the game.

[Melville]

It was my pleasure to introduce this topic during the break.
It was every bit as engaging and revealing as I designed it to be.
My thanks to all those who participated - and my thanks extend equally to both those who understand and appreciate the symphony of baseball, and those who settle for believing that reading the individual notes is a worthy substitute for the real thing.
Now, the break is over.
Tine to PLAY BALL.
I am only and always about the game.

[Goldfan]

I'm making assumptions because I can't see behind the curtain of advanced metrics. And the more recent, more accurate ones didn't even exist in the 1990s. What would you suggest I do? I'm showing on a rate basis he caught less balls in 1999 than he did in 1998 or 1997, he also made more errors. Those have consequences do they not? It's not ruled an error if the other team didn't gain a base (dropped fly ball leading to a hit or bad throw leading to an extra base or two).

". . . that probably had no bad outcomes as far as the actual game." Now who's making assumptions? How do you know none of those plays impacted a game? For that matter, why should we count hits, HRs, RBIs, if they don't affect a game? If a team wins 15-3 and is already up 7-3 when Bichette hits a 3 run HR and gets a double later on, should we count that? This is an absurd argument. We still count hits that don't drive in runs in BA, OBP, SLG, etc.

You don't know when a hit will lead to the start of, or sustaining of, a rally, just like you don't know if a caught fly ball might stop one (or a missed fly ball might start or sustain one).

There's certainly times when using context dependent stats (RBI, runs, RE24, WPA) makes sense, but I would say in the vast majority of situations context neutral stats (BA, OBP, SLG, OPS, wOBA, wRC+, WAR, etc.) are preferred.

He had 1.86CH per GAME…Watching games how many times are there balls hit that require a GG speedster to field to keep that runner from advancing to the “extra” base. Very RARE…..therefore it makes very little sense that top of league offensive output is Negated to the point of turning negative in ANY rating system because of LF defense. Huge overcompensation. You even admit that bWAR is weighs much more heavily than fWAR.,…given that how do you know they both don’t put too much weight on D??

How many times do you see a ball caught at the edge of a fielder's range? Fairly often. How often do you see a slow OF just barely not get to a ball that you think should have been caught, or maybe should have been caught? A fair amount over 162 games. One of those a week is 26 XBHs a season that a defender is giving back to the other team, to use an extreme example. Bichette caught (on a rate basis) 15-30 fewer balls than he himself did in his age 33 and 34 seasons. Is it crazy to think at age 35 he's slowing down and not getting to as many balls anymore? And that most of those balls are going for at least a single if not a double in Coors?

1.86 CH/game is 301 over the course of a season. Giving away 10-20 could easily be 15-30 "total bases" given back.

bWAR does not weight defense much more heavily than fWAR, I believe they both account for 7% of total WAR (14% of run suppression, the other 86% going to pitchers). UZR and DRS are both zero sum stats. If one player is well above average at a position another player, or several players, have to be (well) below average to balance the scales (kind of like a perfectly balanced see saw). I'm saying DRS has more outliers (on both ends of the spectrum) whereas UZR tends to have more players closer to the middle of the seesaw to balance things out. I'm not even saying DRS (and hence bWAR) is wrong, I'm just saying that's what I see in the metrics and thus we see wider ranges of DRS and bWAR than we do in UZR and fWAR, typically.

Excellent stuff. You're WAY over the capacity of the complainers. Melville still uses rudimentary stats -- or mostly none, just the same blowhard template.

“Defensive stats have "error bars" in their results (larger ones than offensive stats), and take much longer to stabilize so we sometimes get weird outliers for a single season.”
I believe this is what I’ve been debating for 16 pg…………….
And Remember RBI says bWAR skews more to D in their WAR ratings and he only uses fWAR…….
So acting like there isn’t a bias, which is exactly what we’ve been debating, is comical…..

As I’ve been writing for probably half of the 17pgs theres a bias in the defensive equations that then cascade through these manipulated analytics
For AOF, RBI, Quinc, Classic
https://www.purplerow.com/2020/1/23/209 ... oors-field

Not speaking to RBI when I say fellas……sometimes stop with arrogance and gotcha and use your common sense. Enjoy

[An Old Friend]

Designated hitters don’t lose runs. My gawd.

And they do nothing to prevent runs (another important aspect of the game). And the positional +/- is all about comparisons.
Is a LF as valuable overall as a SS - or DH?

This. It's the relativity to other positions that matters. They could have designed it with DH having a 0 positional adjustment and C having a +30 adjustment, but then they'd have to lower the baseline later, like they do when they go from RAA (runs above average) to RAR (runs above replacement), so instead they just centered all positions around 0 (some above, some below).

Numbers accumulated while being a DH should not count in calculating WAR. I would say the same regarding career totals. DH has a much lower risk of injury because he sits on the bench for the entire game except for the 4 or 5 times he walks to the batters box. He doesn't need to own a glove. He isn't a baseball player. He is a batter.

You have to score runs to win baseball games. Why would you NOT count an offensive player's offensive contributions?

That's absurd.

The Cardinals catcher is barely above the Mendoza line. He plays because Herrera can't catch. WAR doesn't factor the drag on the offense from Herrera's inability to play a position.

That’s the fault of the front office, not Herrera.

Misdirected Hostility

[ScotchMIrish]

Designated hitters don’t lose runs. My gawd.

And they do nothing to prevent runs (another important aspect of the game). And the positional +/- is all about comparisons.
Is a LF as valuable overall as a SS - or DH?

This. It's the relativity to other positions that matters. They could have designed it with DH having a 0 positional adjustment and C having a +30 adjustment, but then they'd have to lower the baseline later, like they do when they go from RAA (runs above average) to RAR (runs above replacement), so instead they just centered all positions around 0 (some above, some below).

Numbers accumulated while being a DH should not count in calculating WAR. I would say the same regarding career totals. DH has a much lower risk of injury because he sits on the bench for the entire game except for the 4 or 5 times he walks to the batters box. He doesn't need to own a glove. He isn't a baseball player. He is a batter.

You have to score runs to win baseball games. Why would you NOT count an offensive player's offensive contributions?

That's absurd.

The Cardinals catcher is barely above the Mendoza line. He plays because Herrera can't catch. WAR doesn't factor the drag on the offense from Herrera's inability to play a position.

That’s the fault of the front office, not Herrera.

Misdirected Hostility

No hostility. I don't care enough to get emotional about it. Just stating the obvious flaw with WAR.

[ScotchMIrish]

Designated hitters don’t lose runs. My gawd.

And they do nothing to prevent runs (another important aspect of the game). And the positional +/- is all about comparisons.
Is a LF as valuable overall as a SS - or DH?

This. It's the relativity to other positions that matters. They could have designed it with DH having a 0 positional adjustment and C having a +30 adjustment, but then they'd have to lower the baseline later, like they do when they go from RAA (runs above average) to RAR (runs above replacement), so instead they just centered all positions around 0 (some above, some below).

Numbers accumulated while being a DH should not count in calculating WAR. I would say the same regarding career totals. DH has a much lower risk of injury because he sits on the bench for the entire game except for the 4 or 5 times he walks to the batters box. He doesn't need to own a glove. He isn't a baseball player. He is a batter.

You have to score runs to win baseball games. Why would you NOT count an offensive player's offensive contributions?

That's absurd.

The Cardinals catcher is barely above the Mendoza line. He plays because Herrera can't catch. WAR doesn't factor the drag on the offense from Herrera's inability to play a position.

Correct.
WAR and its adherents can't understand the baseball symphony.
It sees only isolated instruments.
Displays a fundamental lack of understanding of the game.

I understand the desire to come up with a player value but as you correctly said it's not that simple.