RealGM Top 100 List 2017 -- #8

[Joao Saraiva]

He played fine offensively, it was the defense that didn't hold up, which if you want to hold against KG, fine.

But then you need to do the same against Duncan in 06.

Kevin Garnett vs Dallas 2001:
24 PPG 5 APG 4 TOPG 51.4ts% 12.4 ORB% 21.5 GmSC

KG playing fine on offense doesn't put up that much of a great volume in scoring, his efficiency is low, AST/TO ratio is bad.

Tim Duncan vs Dallas 06: (actually a great case for Dirk's peak year)
32.3 PPG 3.7 APG 3 TOPG 61.5ts% 10.8 ORB% 26.8 GmSC

Well, I think even you can see a big diference on their offense... I think the PPGxts% is pretty much self suficient.

Also Duncan had a positive ORTG/DRTG relation, with +11. Garnett was -3.

I also doubt KG ever had a series at 26.8 GmSC... for what it's worth. Do you really want to make a comparison here and say it's the same?!

LOL, I'm not saying they did the same, that's actually the best offensive series of Duncan's career (and possibly his worst defensive series as well).

But that's one series against Dallas...what if Garnett got to face Dallas multiple times, instead of the Spurs and Blazers for 3 straight years? How would his numbers look then?

Again, when we look at them against similar opponents, Duncan does not outshine Garnett (Garnett facing the 99 and 01 Spurs is like Duncan facing the 05 Pistons, and they both played the Lakers in 03 and 04).

And I'm not even touching +/- yet, where KG in the playoffs was amazing. Against Dallas in 02, KG had a near +50 on/off. It's like criticizing LeBron in 2015 against the Warriors, because his TS% looked bad...yeah, and he was still exerting monster impact. What more could KG do? I don't think I've ever seen anyone have a near +50 on/off in a playoff series other than KG. Maybe Westbrook this year?

EDIT: yup, Westbrook was like +63.

Well that shows you what you need to pull that off. Can you guess what it is? Make a comparison between the teams, the bench and the opponents faced. Also, if Cleveland's team without LBJ in 15 did even worse, does that say LeBron was having a bigger impact? Hell no. It just says that the guys without LBJ are playing even worse.

Also the load LeBron carried in the 15 finals is something KG never replicated, so it's actually a bad example.

Not saying it's not impressive, but I'm just going to say this again... ONE piece of evidence to analyze players ignoring everything else is a big big big mistake.

[70sFan]

Duncan actually faced some Blazers teams. He also faced 2003 Nets - 1st defense in the league.

KG didn't play against good defensive teams in 2004 and his scoring efficiency still dropped much compared to RS.

[therealbig3]

Say all you want about Nets, they were really good defensively (therealbig3 missed this series).

LOL, I'm a Nets fan. I remember watching that series, hoping the Nets could pull it off.

Their defensive ratings were somewhat overrated because of the conference they played in, and even though they were really good defensively, matchups matter. Duncan was primarily being guarded by Kenyon Martin in that series, who was extremely undersized. I recently rewatched that series, and 2 of his primary defenders were K-Mart and Aaron Williams, both of whom were 6'9". Jason Collins was a 2nd year player, and Dikembe Mutombo barely played.

Context matters there. I don't see KG having a hard time against that front line either.

I'm talking about legitimate twin tower defensive lineups, that could make life hell for opposing big men in the paint, like Malone/Shaq in 04, or Sheed/Sabonis, or Wallace/Wallace. That's usually the caliber of defense KG was facing in the playoffs during his Minnesota years, so it seems a little unfair to simply point to TS% and say he's not Duncan, when Duncan struggled to the same degree against those kinds of teams.

[mischievous]

Duncan actually faced some Blazers teams. He also faced 2003 Nets - 1st defense in the league.

KG didn't play against good defensive teams in 2004 and his scoring efficiency still dropped much compared to RS.

But +/-!!!!.

People act like Kg's massively better at non-scoring apects, so therefore Duncan's scoring efficiency means nothing.

Duncan averaged 5+ apg in those 03 playoffs. He may not have been Kg's equal at passing/playmaking but he was still damn good at it.

[therealbig3]

Duncan actually faced some Blazers teams. He also faced 2003 Nets - 1st defense in the league.

KG didn't play against good defensive teams in 2004 and his scoring efficiency still dropped much compared to RS.

He faced the Blazers once, in 99, and he was held to under 17 ppg on less than 55% TS. He wasn't even the leading scorer on his own team in that series IIRC, it was Robinson.

[rebirthoftheM]

Well that shows you what you need to pull that off. Can you guess what it is? Make a comparison between the teams, the bench and the opponents faced. Also, if Cleveland's team without LBJ in 15 did even worse, does that say LeBron was having a bigger impact? Hell no. It just says that the guys without LBJ are playing even worse.

Also the load LeBron carried in the 15 finals is something KG never replicated, so it's actually a bad example.

Not saying it's not impressive, but I'm just going to say this again... ONE piece of evidence to analyze players ignoring everything else is a big big big mistake.

There's also another issue with these indicators that assess on and off court impact n the playoffs. Because of the sample size issues, it ignores the critical factor of momentum.

Take for example, KG subs out at the 9 minute mark in the 2nd quarter. In the first 3 minutes of the 2nd quarter, the Wolves play really uninspiring basketball and see their lead cut. The momentum is against them... they are clearly trending downwards. KG then subs out, and then the trend continues, this time accelerating as the opponent is clearly surging off the back of their resurgent play to the start the 2nd. KG then returns at a stage, where although his team is still down, over the past 2-3 possessions, his team is trending upwards. He then runs with this, and the team wins as a result.

On/off will read this as KG having monster impact. His team collapsed with him off the floor (although it is clear they were trending downwards anyways in the last few 2-3 minutes with him on the floor prior to being subbed out) and then got better with him on the floor (although they were trending upwards in the last few possessions prior to being subbed in).

So there are line-up/match up problems with playoff on/off, but also momentum/trending issues. IMO, any reference to playoff on/off has to be contextualized very very well and be justified as accurately reflecting a players impact. Anything else is bad analysis.

[THKNKG]

.

Hmm... good thoughts. First thoughts of mine are that 1) during a "carry job," defense tends to suffer and 2) it's easier to maintain offensive consistency with a turnstile than with defense, as if you have the ball in your hands often, offense won't shift much. That's what makes WB 17/Kobe 06+07/etc. so impressive. KG was the anchor of their offense and defense, so beyond his peak, while having to anchor both, there would likely be some sort of regressing on one side of the ball (however small).

I'm not completely satisfied with that answer of mine, so let me think more about it, and I'll try to respond more later. I welcome thoughts in the meantime.

[Joao Saraiva]

Say all you want about Nets, they were really good defensively (therealbig3 missed this series).

LOL, I'm a Nets fan. I remember watching that series, hoping the Nets could pull it off.

Their defensive ratings were somewhat overrated because of the conference they played in, and even though they were really good defensively, matchups matter. Duncan was primarily being guarded by Kenyon Martin in that series, who was extremely undersized. I recently rewatched that series, and 2 of his primary defenders were K-Mart and Aaron Williams, both of whom were 6'9". Jason Collins was a 2nd year player, and Dikembe Mutombo barely played.

Context matters there. I don't see KG having a hard time against that front line either.

I'm talking about legitimate twin tower defensive lineups, that could make life hell for opposing big men in the paint, like Malone/Shaq in 04, or Sheed/Sabonis, or Wallace/Wallace. That's usually the caliber of defense KG was facing in the playoffs during his Minnesota years, so it seems a little unfair to simply point to TS% and say he's not Duncan, when Duncan struggled to the same degree against those kinds of teams.

We already saw KG's stats against Dallas. TS% didn't rise. And then you said... oh but if he faced them more times...

Against Denver in 04 KG was at 25.8 PPG on 50.8ts%. His ts% actually didn't rise... neither did his volume.

He also faced the Kings in 04.He was at 23.9 PPG and 51.2 ts%... so again the same scenario.

We saw him in Boston 08 and his ts% didn't rise a lot in a much better team situation and not carrying that much of a big load on offense. He still ended below 55ts% on that playoff run.

So basically, nothing suggests Garnett could actually be a better scorer than he was. He did what he did, and that's who he was. But here with KG it's all about the could have done... not what he actually did. If anything Karl Malone deserves more of a pass than KG, since he proved several regular seasons he could score at elite ts% and even in some playoff deep runs. But when it's Karl Malone I don't see the same type of approach.

[Joao Saraiva]

Well that shows you what you need to pull that off. Can you guess what it is? Make a comparison between the teams, the bench and the opponents faced. Also, if Cleveland's team without LBJ in 15 did even worse, does that say LeBron was having a bigger impact? Hell no. It just says that the guys without LBJ are playing even worse.

Also the load LeBron carried in the 15 finals is something KG never replicated, so it's actually a bad example.

Not saying it's not impressive, but I'm just going to say this again... ONE piece of evidence to analyze players ignoring everything else is a big big big mistake.

There's also another issue with these indicators that assess on and off court impact n the playoffs. Because of the sample size issues, it ignores the critical factor of momentum.

Take for example, KG subs out at the 9 minute mark in the 2nd quarter. In the first 3 minutes of the 2nd quarter, the Wolves play really uninspiring basketball and see their lead cut. The momentum is against them... they are clearly trending downwards. KG then subs out, and then the trend continues, this time accelerating as the opponent is clearly surging off the back of their resurgent play to the start the 2nd. KG then returns at a stage, where although his team is still down, over the past 2-3 possessions, his team is trending upwards. He then runs with this, and the team wins as a result.

On/off will read this as KG having monster impact. His team collapsed with him off the floor (although it is clear they were trending downwards anyways in the last few 2-3 minutes with him on the floor prior to being subbed out) and then got better with him on the floor (although they were trending upwards in the last few possessions prior to being subbed in).

So there are line-up/match up problems with playoff on/off, but also momentum/trending issues. IMO, any reference to playoff on/off has to be contextualized very very well and be justified as accurately reflecting a players impact. Anything else is bad analysis.

While I agree with what you said, that just shows that the stat by itself has a ton of flaws. Useful yes, context is needed like in any other stat (yes), but it's a mistake to assume the player is better than the other based on one stat alone.

[therealbig3]

Say all you want about Nets, they were really good defensively (therealbig3 missed this series).

LOL, I'm a Nets fan. I remember watching that series, hoping the Nets could pull it off.

Their defensive ratings were somewhat overrated because of the conference they played in, and even though they were really good defensively, matchups matter. Duncan was primarily being guarded by Kenyon Martin in that series, who was extremely undersized. I recently rewatched that series, and 2 of his primary defenders were K-Mart and Aaron Williams, both of whom were 6'9". Jason Collins was a 2nd year player, and Dikembe Mutombo barely played.

Context matters there. I don't see KG having a hard time against that front line either.

I'm talking about legitimate twin tower defensive lineups, that could make life hell for opposing big men in the paint, like Malone/Shaq in 04, or Sheed/Sabonis, or Wallace/Wallace. That's usually the caliber of defense KG was facing in the playoffs during his Minnesota years, so it seems a little unfair to simply point to TS% and say he's not Duncan, when Duncan struggled to the same degree against those kinds of teams.

We already saw KG's stats against Dallas. TS% didn't rise. And then you said... oh but if he faced them more times...

Against Denver in 04 KG was at 25.8 PPG on 50.8ts%. His ts% actually didn't rise... neither did his volume.

He also faced the Kings in 04.He was at 23.9 PPG and 51.2 ts%... so again the same scenario.

We saw him in Boston 08 and his ts% didn't rise a lot in a much better team situation and not carrying that much of a big load on offense. He still ended below 55ts% on that playoff run.

So basically, nothing suggests Garnett could actually be a better scorer than he was. He did what he did, and that's who he was. But here with KG it's all about the could have done... not what he actually did. If anything Karl Malone deserves more of a pass than KG, since he proved several regular seasons he could score at elite ts% and even in some playoff deep runs. But when it's Karl Malone I don't see the same type of approach.

Yeah, so essentially what we're seeing here is that Duncan tends to score more efficiently against weaker defenses than KG does, but against similar (aka better) opponents, they're about the same.

How important is Duncan scoring more efficiently against lesser defenses compared to KG's advantages everywhere else (and I do mean everywhere else, I don't think Duncan does anything on or better than KG's level other than ISO scoring)?

I think once we get to the stronger defensive teams, and both guys struggle with their shot, KG becomes the more valuable offensive weapon...and that's more important to me at the end of the day than being able to score 30 on 60% TS against a 6th-8th seed.

[therealbig3]

Well that shows you what you need to pull that off. Can you guess what it is? Make a comparison between the teams, the bench and the opponents faced. Also, if Cleveland's team without LBJ in 15 did even worse, does that say LeBron was having a bigger impact? Hell no. It just says that the guys without LBJ are playing even worse.

Also the load LeBron carried in the 15 finals is something KG never replicated, so it's actually a bad example.

Not saying it's not impressive, but I'm just going to say this again... ONE piece of evidence to analyze players ignoring everything else is a big big big mistake.

There's also another issue with these indicators that assess on and off court impact n the playoffs. Because of the sample size issues, it ignores the critical factor of momentum.

Take for example, KG subs out at the 9 minute mark in the 2nd quarter. In the first 3 minutes of the 2nd quarter, the Wolves play really uninspiring basketball and see their lead cut. The momentum is against them... they are clearly trending downwards. KG then subs out, and then the trend continues, this time accelerating as the opponent is clearly surging off the back of their resurgent play to the start the 2nd. KG then returns at a stage, where although his team is still down, over the past 2-3 possessions, his team is trending upwards. He then runs with this, and the team wins as a result.

On/off will read this as KG having monster impact. His team collapsed with him off the floor (although it is clear they were trending downwards anyways in the last few 2-3 minutes with him on the floor prior to being subbed out) and then got better with him on the floor (although they were trending upwards in the last few possessions prior to being subbed in).

So there are line-up/match up problems with playoff on/off, but also momentum/trending issues. IMO, any reference to playoff on/off has to be contextualized very very well and be justified as accurately reflecting a players impact. Anything else is bad analysis.

While I agree with what you said, that just shows that the stat by itself has a ton of flaws. Useful yes, context is needed like in any other stat (yes), but it's a mistake to assume the player is better than the other based on one stat alone.

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

[Joao Saraiva]

LOL, I'm a Nets fan. I remember watching that series, hoping the Nets could pull it off.

Their defensive ratings were somewhat overrated because of the conference they played in, and even though they were really good defensively, matchups matter. Duncan was primarily being guarded by Kenyon Martin in that series, who was extremely undersized. I recently rewatched that series, and 2 of his primary defenders were K-Mart and Aaron Williams, both of whom were 6'9". Jason Collins was a 2nd year player, and Dikembe Mutombo barely played.

Context matters there. I don't see KG having a hard time against that front line either.

I'm talking about legitimate twin tower defensive lineups, that could make life hell for opposing big men in the paint, like Malone/Shaq in 04, or Sheed/Sabonis, or Wallace/Wallace. That's usually the caliber of defense KG was facing in the playoffs during his Minnesota years, so it seems a little unfair to simply point to TS% and say he's not Duncan, when Duncan struggled to the same degree against those kinds of teams.

We already saw KG's stats against Dallas. TS% didn't rise. And then you said... oh but if he faced them more times...

Against Denver in 04 KG was at 25.8 PPG on 50.8ts%. His ts% actually didn't rise... neither did his volume.

He also faced the Kings in 04.He was at 23.9 PPG and 51.2 ts%... so again the same scenario.

We saw him in Boston 08 and his ts% didn't rise a lot in a much better team situation and not carrying that much of a big load on offense. He still ended below 55ts% on that playoff run.

So basically, nothing suggests Garnett could actually be a better scorer than he was. He did what he did, and that's who he was. But here with KG it's all about the could have done... not what he actually did. If anything Karl Malone deserves more of a pass than KG, since he proved several regular seasons he could score at elite ts% and even in some playoff deep runs. But when it's Karl Malone I don't see the same type of approach.

Yeah, so essentially what we're seeing here is that Duncan tends to score more efficiently against weaker defenses than KG does, but against similar (aka better) opponents, they're about the same.

How important is Duncan scoring more efficiently against lesser defenses compared to KG's advantages everywhere else (and I do mean everywhere else, I don't think Duncan does anything on or better than KG's level other than ISO scoring)?

I think once we get to the stronger defensive teams, and both guys struggle with their shot, KG becomes the more valuable offensive weapon...and that's more important to me at the end of the day than being able to score 30 on 60% TS against a 6th-8th seed.

Well, for a start Duncan didn't only score 30 on 60ts% against low seed teams.

Then about KG doing everything else better... well, I think calling his rim protection better than Duncan's is extremely biased to say the least.

I also don't think Duncan is a worse rebounder than KG.

Playmaking is where KG has an edge, but it's not like it's a major one.

Oh, and scoring is valuable. And scoring well against worse defenses is definitely important. That will cause more gravity effect - if a bad defensive team can make unefficient, the amount of adjustment to lower your production an elite team needs is definitely smaller, more upper seeds and maybe even some playoff appearences even with bad casts. Of course it is important.

EDIT: On that last part, just see Kobe. He took a bad cast to the playoffs in 06 and his scoring was probably his biggest asset to do it. Bringing up the volume KG never did, and at an efficiency KG couldn't usually achieve even in lesser volume.

Of course scoring matters and it's a big part of the game.

[rebirthoftheM]

.

Hmm... good thoughts. First thoughts of mine are that 1) during a "carry job," defense tends to suffer and 2) it's easier to maintain offensive consistency with a turnstile than with defense, as if you have the ball in your hands often, offense won't shift much. That's what makes WB 17/Kobe 06+07/etc. so impressive. KG was the anchor of their offense and defense, so beyond his peak, while having to anchor both, there would likely be some sort of regressing on one side of the ball (however small).

I'm not completely satisfied with that answer of mine, so let me think more about it, and I'll try to respond more later. I welcome thoughts in the meantime.

Defs think there would be some regressing, though again, it appears we are both in tentative agreement that his defensive indicators seem to be out of sync with KG's actual defensive play. I trust his offensive indicators far more.

I agree with you in general re offense and offensive players, although I think a player can have huge impact on offense without holding the ball for long/dribbling around. I'm thinking late 90s MJ, Dirk, and even in Kobe in 06 played a ton off the ball. Elite offensive players tend to control offenses even without touching the ball. Steph Curry recently has shown us the extent of this in the 3 ball era.

I'm still not 100% sure what to make of this, pending someone chiming in and telling us what occured with the Wolves in 05/06, and why KG's defensive impact didn't register as high.

[rebirthoftheM]

There's also another issue with these indicators that assess on and off court impact n the playoffs. Because of the sample size issues, it ignores the critical factor of momentum.

Take for example, KG subs out at the 9 minute mark in the 2nd quarter. In the first 3 minutes of the 2nd quarter, the Wolves play really uninspiring basketball and see their lead cut. The momentum is against them... they are clearly trending downwards. KG then subs out, and then the trend continues, this time accelerating as the opponent is clearly surging off the back of their resurgent play to the start the 2nd. KG then returns at a stage, where although his team is still down, over the past 2-3 possessions, his team is trending upwards. He then runs with this, and the team wins as a result.

On/off will read this as KG having monster impact. His team collapsed with him off the floor (although it is clear they were trending downwards anyways in the last few 2-3 minutes with him on the floor prior to being subbed out) and then got better with him on the floor (although they were trending upwards in the last few possessions prior to being subbed in).

So there are line-up/match up problems with playoff on/off, but also momentum/trending issues. IMO, any reference to playoff on/off has to be contextualized very very well and be justified as accurately reflecting a players impact. Anything else is bad analysis.

While I agree with what you said, that just shows that the stat by itself has a ton of flaws. Useful yes, context is needed like in any other stat (yes), but it's a mistake to assume the player is better than the other based on one stat alone.

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

It is huge? The sample size in the 04 playoffs where he was off was 86 minutes. 86 minutes!. That's minuscule and is subject to a whole lotta noise. What about match-up/line-up issues? Momentum/trending issues? End of game score running by an opponent?

Or are you using cumulative on/offs from multiple playoff runs (a lot of them being 1st round exits) to interpret a single years on/off? If so, IMO this would be a fallacy. What KG did in the 2002 playoffs has zero bearing in the 2004 Playoffs. They are two separate years, against separate teams, with their own unique circumstances. We need proper context to these on/off splits.

This isn't to say that KG was not having big impact in the playoffs. I mean, we know he was a high high impact player during his prime/peak. But the monster on/offs we see in the PS, even they occur across several years, again are not at all definitive. Sample sizes way way too small.

[Joao Saraiva]

There's also another issue with these indicators that assess on and off court impact n the playoffs. Because of the sample size issues, it ignores the critical factor of momentum.

Take for example, KG subs out at the 9 minute mark in the 2nd quarter. In the first 3 minutes of the 2nd quarter, the Wolves play really uninspiring basketball and see their lead cut. The momentum is against them... they are clearly trending downwards. KG then subs out, and then the trend continues, this time accelerating as the opponent is clearly surging off the back of their resurgent play to the start the 2nd. KG then returns at a stage, where although his team is still down, over the past 2-3 possessions, his team is trending upwards. He then runs with this, and the team wins as a result.

On/off will read this as KG having monster impact. His team collapsed with him off the floor (although it is clear they were trending downwards anyways in the last few 2-3 minutes with him on the floor prior to being subbed out) and then got better with him on the floor (although they were trending upwards in the last few possessions prior to being subbed in).

So there are line-up/match up problems with playoff on/off, but also momentum/trending issues. IMO, any reference to playoff on/off has to be contextualized very very well and be justified as accurately reflecting a players impact. Anything else is bad analysis.

While I agree with what you said, that just shows that the stat by itself has a ton of flaws. Useful yes, context is needed like in any other stat (yes), but it's a mistake to assume the player is better than the other based on one stat alone.

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

So Durant might never achieve that type of impact with GSW. Since when he's out, they might rely on lineups with Curry, Klay and Green. So the off numbers will usually look pretty good without him...

So basically he's doing nothing out there, even if it made GSW have probably the most dominant run ever in the playoffs with him scoring momentum baskets in the finals all series long.

Also Curry is much more useful than KD. Let's ignore GSW has much better forward solutions to replace KD than they have PG solutions for Curry...

And it goes on and on.

[therealbig3]

While I agree with what you said, that just shows that the stat by itself has a ton of flaws. Useful yes, context is needed like in any other stat (yes), but it's a mistake to assume the player is better than the other based on one stat alone.

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

It is huge? The sample size in the 04 playoffs where he was off was 86 minutes. 86 minutes!. That's minuscule and is subject to a whole lotta noise. What about match-up/line-up issues? Momentum/trending issues? End of game score running by an opponent?

Or are you using cumulative on/offs from multiple playoff runs (a lot of them being 1st round exits) to interpret a single years on/off? If so, IMO this would be a fallacy. What KG did in the 2002 playoffs has zero bearing in the 2004 Playoffs. They are two separate years, against separate teams, with their own unique circumstances. We need proper context to these on/off splits.

This isn't to say that KG was not having big impact in the playoffs. I mean, we know he was a high high impact player during his prime/peak. But the monster on/offs we see in the PS, even they occur across several years, again are not at all definitive. Sample sizes way way too small.

I'm not drawing conclusions off the PS +/-, I think they just tend to show the same thing that the RS +/- consistently showed us.

I brought up PS +/- in one series just to counter the point that KG didn't play well in that series. What +/- in a small sample size can tell us is if a player was helping his team or not. If KG posts a +47 on/off in a playoff series, I think it's safe to say that the team pretty much lived and died with him in that particular series. Drawing any conclusions other than that is tricky, but it does run counter to the idea that he didn't play well.

[therealbig3]

While I agree with what you said, that just shows that the stat by itself has a ton of flaws. Useful yes, context is needed like in any other stat (yes), but it's a mistake to assume the player is better than the other based on one stat alone.

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

So Durant might never achieve that type of impact with GSW. Since when he's out, they might rely on lineups with Curry, Klay and Green. So the off numbers will usually look pretty good without him...

So basically he's doing nothing out there, even if it made GSW have probably the most dominant run ever in the playoffs with him scoring momentum baskets in the finals all series long.

Also Curry is much more useful than KD. Let's ignore GSW has much better forward solutions to replace KD than they have PG solutions for Curry...

And it goes on and on.

LOL, and if you know anything about my opinions about KD and Curry, you would know that I actually feel pretty similarly to this.

But it's not because of ONE version of +/-. On/off is different than RAPM. They're both different than WOWY. And RPM has a little box score thrown in for good measure.

They're all calculated in different ways, and they ALL point to the same conclusion sometimes. Which makes it very convincing.

[Joao Saraiva]

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

It is huge? The sample size in the 04 playoffs where he was off was 86 minutes. 86 minutes!. That's minuscule and is subject to a whole lotta noise. What about match-up/line-up issues? Momentum/trending issues? End of game score running by an opponent?

Or are you using cumulative on/offs from multiple playoff runs (a lot of them being 1st round exits) to interpret a single years on/off? If so, IMO this would be a fallacy. What KG did in the 2002 playoffs has zero bearing in the 2004 Playoffs. They are two separate years, against separate teams, with their own unique circumstances. We need proper context to these on/off splits.

This isn't to say that KG was not having big impact in the playoffs. I mean, we know he was a high high impact player during his prime/peak. But the monster on/offs we see in the PS, even they occur across several years, again are not at all definitive. Sample sizes way way too small.

I'm not drawing conclusions off the PS +/-, I think they just tend to show the same thing that the RS +/- consistently showed us.

I brought up PS +/- in one series just to counter the point that KG didn't play well in that series. What +/- in a small sample size can tell us is if a player was helping his team or not. If KG posts a +47 on/off in a playoff series, I think it's safe to say that the team pretty much lived and died with him in that particular series. Drawing any conclusions other than that is tricky, but it does run counter to the idea that he didn't play well.

Well, I might have to say again we're talking about playing well against the standards of Hakeem, Shaq or Duncan. Particularly against Shaq and Hakeem, since Duncan is already in.

Nobody is saying he didn't have impact, or that he didn't help his team, or that he was a complete disaster out there.

[Joao Saraiva]

Who's doing this btw?

And it's disingenuous to keep referring to it as just one stat, as if it's basically the same as PPG or PER. It's a stat that encompasses everything that happens on the basketball court, and with enough sample size, it's very valuable and trumps the value of the box score, because it's showing us what those box score stats are actually translating into in terms of the final score.

And the sample size is huge with KG. In a variety of situations. That's why it's so convincing to so many of us.

So Durant might never achieve that type of impact with GSW. Since when he's out, they might rely on lineups with Curry, Klay and Green. So the off numbers will usually look pretty good without him...

So basically he's doing nothing out there, even if it made GSW have probably the most dominant run ever in the playoffs with him scoring momentum baskets in the finals all series long.

Also Curry is much more useful than KD. Let's ignore GSW has much better forward solutions to replace KD than they have PG solutions for Curry...

And it goes on and on.

LOL, and if you know anything about my opinions about KD and Curry, you would know that I actually feel pretty similarly to this.

But it's not because of ONE version of +/-. On/off is different than RAPM. They're both different than WOWY. And RPM has a little box score thrown in for good measure.

They're all calculated in different ways, and they ALL point to the same conclusion sometimes. Which makes it very convincing.

Expected, since that is the king stat for you. Just shows you actually evaluate players at least with A TON of emphasis on that stat. Big mistake.

Btw care to say anything about the scoring against worse teams or something? Or does it make KG look bad?

[therealbig3]

So Durant might never achieve that type of impact with GSW. Since when he's out, they might rely on lineups with Curry, Klay and Green. So the off numbers will usually look pretty good without him...

So basically he's doing nothing out there, even if it made GSW have probably the most dominant run ever in the playoffs with him scoring momentum baskets in the finals all series long.

Also Curry is much more useful than KD. Let's ignore GSW has much better forward solutions to replace KD than they have PG solutions for Curry...

And it goes on and on.

LOL, and if you know anything about my opinions about KD and Curry, you would know that I actually feel pretty similarly to this.

But it's not because of ONE version of +/-. On/off is different than RAPM. They're both different than WOWY. And RPM has a little box score thrown in for good measure.

They're all calculated in different ways, and they ALL point to the same conclusion sometimes. Which makes it very convincing.

Expected, since that is the king stat for you. Just shows you actually evaluate players at least with A TON of emphasis on that stat. Big mistake.

Btw care to say anything about the scoring against worse teams or something? Or does it make KG look bad?

I think it's a big mistake to rely so much on the box score like you do. I wonder where Adrian Dantley ranks for you.

I don't even know what you're talking about in the last sentence.