'15-16 RealGM Player of the Year Discussion Thread

[SideshowBob]

What about Duncan 2000 or Kobe 2013?

Same. And 15 Durant and 09 Garnett and on and on.

[RSCD3_]

What about Duncan 2000 or Kobe 2013?

Same. And 15 Durant and 09 Garnett and on and on.

I share the same opinion, If the goal is theoretically how much you can help your team win in the playoffs and injuries reduce that number to zero then it's not really that much in added career value for that season.

[Quotatious]

I share the same opinion, If the goal is theoretically how much you can help your team win in the playoffs and injuries reduce that number to zero then it's not really that much in added career value for that season.

I would vehemently disagree. It's just extremely unfair to punish a player who misses the playoffs because of a serious injury, one that is too serious to play through. Duncan and Bryant were IMO top 5 players in '00 (TD) and '13 (KB), regardless if they could play in the postseason or not. 74 (TD) and 78 games (KB) of a top 5 caliber player is A LOT of added career value, for me.

I don't really share the opinion that the main goal for every player is to help his team win in the playoffs as much as possible. I think it can easily be said that the main goal for every player is to help his team win every game he plays in, as much as possible, be it RS or PS, and TD/KB really did an awful lot for their teams in those regular seasons.

If we applied that playoff availability rule, we would get some very counter-intuitive results - for instance, it means that '86 or '95 Jordan was more valuable than '00 Duncan or '13 Bryant, because he could play in the postseason, and TD/KB could not. It completely ignores the fact that MJ played only 17-18 RS games, and TD/KB played 70-80 games, not to mention they were both superior on a per-game basis.

Vast majority of every player's career consists of regular season games, and for me, every game matters.

2009 Garnett and especially 2015 Durant, is a different story. They didn't play nearly as many games as TD and KB - the latter two guys played almost an entire season.

[Texas Chuck]

yeah I'm with Q on this. If we rule out great seasons by guys unavailable for the playoffs due to injury do we rule out seasons where their team fails to qualify for the playoffs? Because if you lift your team a lot(say 05 KG for example) but its not enough to get the team into the playoffs why is that more valuable that Duncan or Kobe playing great and getting their team into the dance--even if they are no longer there to help?

I mean if the playoffs are all that matter, then Kobe and Duncan should rate as more valuable because your team being in teh playoffs even without its best player has more of a chance to win than a team not in.

Feels like a slippery slope to start down imo.

[SideshowBob]

But that doesn't seem to be the actual case, at least based on ElGee's research on title odds:

Link

(3) Regular Season Player Health matters less than you think.

In the RS, for a normal portability 5 SIO player, playing the whole year results in a 21.4% chance to win the title. Playing half the year? An 20.2% chance. Playing even 10% of the year still results in an 18.0% chance to win the title, assuming the player is playing at a +5 SIO level in the RS and in the PS.

Why? Because the SRS differential the player created in the playoffs is more important than the HCA advantage lost. The majority of below average teams will never see the PS with such a player missing most of the year, but almost every time a player is on an above average team (51% of teams since 1986) his teammates will have qualified for the playoffs. Think Wilt Chamberlain in 1970 or Michael Jordan in 1986 and 1995.

The better the player, the more missing time will hurt him (because of the likelihood of losing HCA in the later rounds against better teams). An 8 SRS player added to a random team gives them a 45% chance of winning title if he's healthy all year. If he plays 10% of the RS and then the playoffs, a 32% chance of winning a title.

It would seem that simply being able to be available to improve a team during postseason time is in most cases so valuable that it would gives any random team a better chance at a title than the same player playing the RS, getting the team into the postseason and then missing it.

Key here is that I'm holding a probabilistic perspective though (which I understand is a philosophical difference that I have with a lot of folks - fair enough). I'm not only trying to concern myself with the player's particular situation in a given year but rather how they'd perform across situations.

@Q: Same point I suppose, its a fundamental difference in criteria. The thought process for focusing on how much a player helps the team win a title/in the playoffs is this: high level teams win more often -> I want to consider players who contribute to high value teams a bit more valuable -> such players tend to have more additive/non-redundant skills -> I want to value players with additive/non-redundant skills a bit more than those without them, all else held equal -> I end up aligning my criteria to value how much a player contributes to lifting a team's title odds (and by definition level of play in the playoffs, adjusted for context of course).

[Texas Chuck]

But that doesn't seem to be the actual case, at least based on ElGee's research on title odds:

Link

(3) Regular Season Player Health matters less than you think.

In the RS, for a normal portability 5 SIO player, playing the whole year results in a 21.4% chance to win the title. Playing half the year? An 20.2% chance. Playing even 10% of the year still results in an 18.0% chance to win the title, assuming the player is playing at a +5 SIO level in the RS and in the PS.

Why? Because the SRS differential the player created in the playoffs is more important than the HCA advantage lost. The majority of below average teams will never see the PS with such a player missing most of the year, but almost every time a player is on an above average team (51% of teams since 1986) his teammates will have qualified for the playoffs. Think Wilt Chamberlain in 1970 or Michael Jordan in 1986 and 1995.

The better the player, the more missing time will hurt him (because of the likelihood of losing HCA in the later rounds against better teams). An 8 SRS player added to a random team gives them a 45% chance of winning title if he's healthy all year. If he plays 10% of the RS and then the playoffs, a 32% chance of winning a title.

It would seem that simply being able to be available to improve a team during postseason time is in most cases so valuable that it would gives any random team a better chance at a title than the same player playing the RS, getting the team into the postseason and then missing it.

Key here is that I'm holding a probabilistic perspective though (which I understand is a philosophical difference that I have with a lot of folks - fair enough). I'm not only trying to concern myself with the player's particular situation in a given year but rather how they'd perform across situations.

I agree with the bolded obviously. Assuming the team can still get in without you, I'd rather have my best player in teh playoffs even if my seeding is far worse. If the only goal is championships it doesn't matter if I have to play the best team in round 1 or round 3, I still have to get through them(discounting the times someone else upsets them).

But this doesn't answer my main question--what are you doing with great seasons on bad teams? Is 05-07 KG worth a 0 each year? And if this is the case for those who share your take on this, can you still consider KG as a top 15(or higher) player all-time with 3 very strong prime years removed?

If feels to me like yes, 05-07 KG has to be a zero whereas the Kobe and Duncan seasons referenced should be a small positive number, right? Because while the odds are very long for the Lakers and Spurs those years, they aren't impossible. So shouldn't Kobe and Duncan get like a 0.25*or something?

*totally made up number as I don't know the methodology being used to calculate these scores, but just putting something in greater than zero.

[SideshowBob]

I agree with the bolded obviously. Assuming the team can still get in without you, I'd rather have my best player in teh playoffs even if my seeding is far worse. If the only goal is championships it doesn't matter if I have to play the best team in round 1 or round 3, I still have to get through them(discounting the times someone else upsets them).

But this doesn't answer my main question--what are you doing with great seasons on bad teams? Is 05-07 KG worth a 0 each year? And if this is the case for those who share your take on this, can you still consider KG as a top 15(or higher) player all-time with 3 very strong prime years removed?

No, the only time I'd end up giving a 0 is in the injured during the PS time-frame scenario. Good season on a bad team is counted normally unless they're physically unable to play during playoff time, regardless of whether their specific team makes the playoffs or not, because again, if they were playing on an average team (0 SRS or 41-41 without them) then they would be improving the title odds of that team (so 05-07 Garnett, 15 Westbrook, etc. are not counted as 0 years, just slightly discounted for missed RS time).

If feels to me like yes, 05-07 KG has to be a zero whereas the Kobe and Duncan seasons referenced should be a small positive number, right? Because while the odds are very long for the Lakers and Spurs those years, they aren't impossible. So shouldn't Kobe and Duncan get like a 0.25*or something?

*totally made up number as I don't know the methodology being used to calculate these scores, but just putting something in greater than zero.

Sure, I'd have to work out the math but that makes sense. You're correct the improvement in odds would be small but would be non-zero. I'd guess the reason ElGee uses 0 though is that the odds in that case are so small that it's negligible. It seems like missing time/playing sub-par during postseason time is just really highly detrimental. I've seen him penalize >50% for simply playing half the postseason injured at a lower than established level.

[Texas Chuck]

No, the only time I'd end up giving a 0 is in the injured during the PS time-frame scenario. Good season on a bad team is counted normally unless they're physically unable to play during playoff time, regardless of whether their specific team makes the playoffs or not, because again, if they were playing on an average team (0 SRS or 41-41 without them) then they would be improving the title odds of that team (so 05-07 Garnett, 15 Westbrook, etc. are not counted as 0 years, just slightly discounted for missed RS time).

Thanks for the back and forth on this btw. Appreciate you putting up with my questions on this.

Okay I get what you are saying about this--but this is one place where I struggle a bit with that approach. Again using KG just for ease, but his seasons there aren't taking place in a vacuum, right? They are taking place in a very specific context. We can see from 07 to 08 that his statistical footprint is significantly different in a variety of measurements--including importantly I would think in the on/off +/- stats.

So how do we "know" what KG(or any player) "impact" would be on this theoretical 0 SRS team? It can't simply be what it is on the weak Wolves teams? Nor should we assume in 08 that his "impact" would be the same as it is on a very strong Boston team. Where do we find the "truth"?

My other small issue is that this somewhat arbitrarily punishes players who are typically pretty healthy(Kobe for instance) who get unlucky in the timing of an injury as opposed to a player like Shaq who doesn't take great care of himself yet managed to always be available for the playoffs. If the idea of looking at this from a probability standpoint is to remove as many outside elements from the equation--which I again am assuming is part of the goal here--then IDK I'm just rambling it feels like at this point.

[SideshowBob]

Thanks for the back and forth on this btw. Appreciate you putting up with my questions on this.

No problem Way I see it you're keeping me in check.

Okay I get what you are saying about this--but this is one place where I struggle a bit with that approach. Again using KG just for ease, but his seasons there aren't taking place in a vacuum, right? They are taking place in a very specific context. We can see from 07 to 08 that his statistical footprint is significantly different in a variety of measurements--including importantly I would think in the on/off +/- stats.

So how do we "know" what KG(or any player) "impact" would be on this theoretical 0 SRS team? It can't simply be what it is on the weak Wolves teams? Nor should we assume in 08 that his "impact" would be the same as it is on a very strong Boston team. Where do we find the "truth"?

Yeah on a <0 SRS team impact would typically be inflated the lower below the baseline is and same vice-versa. Guy who shows up as +5 on a 20 win team probably isn't as good in a vacuum as a +5 on a 40 win team, who probably isn't as good as the +5 on a 60 win team.

How do we know? Well that's what we all do here isn't it . I guess putting it simply would be that I know in the back of my mind I've got a scale for everyone else, so the task then becomes coming up with how player X rates on offense and defense relative to everyone else I've already evaluated, and trying to look for similar skills/abilities/mentalities and see if I feel there's a need for nuanced differences among players (and their particular seasons) given the same or similar traits.

+/- and the like serve more to give me a general ballparks of impact to validate and/or cross-reference my evaluations after the fact, as well as give at least somewhat of a baseline for scale.

My other small issue is that this somewhat arbitrarily punishes players who are typically pretty healthy(Kobe for instance) who get unlucky in the timing of an injury as opposed to a player like Shaq who doesn't take great care of himself yet managed to always be available for the playoffs. If the idea of looking at this from a probability standpoint is to remove as many outside elements from the equation--which I again am assuming is part of the goal here--then IDK I'm just rambling it feels like at this point.

True, and I've thought about and discussed looking at health from a probabilistic standpoint as well - give players a health grade depending on how likely they are to go down at the crucial point of the season. Problem is that's going well beyond my scope of knowledge (mathematically I could tackle it, but I do not have the medical knowledge/resources, nor the time/reason to do so), so I just ultimately settled for what was already done by others thus far.

[bondom34]

As usual, great stuff. This is why I look forward to this thread every year.

[Doctor MJ]

yeah I'm with Q on this. If we rule out great seasons by guys unavailable for the playoffs due to injury do we rule out seasons where their team fails to qualify for the playoffs? Because if you lift your team a lot(say 05 KG for example) but its not enough to get the team into the playoffs why is that more valuable that Duncan or Kobe playing great and getting their team into the dance--even if they are no longer there to help?

I mean if the playoffs are all that matter, then Kobe and Duncan should rate as more valuable because your team being in teh playoffs even without its best player has more of a chance to win than a team not in.

Feels like a slippery slope to start down imo.

I would say first and foremost that you do what you think is right. I insist that people focus on how effective a guy actually was rather than what he could have been if he wasn't pissed off at the coach/owner/his dog/etc, but beyond that, your call.

For me personally it's not really a hard call. I'm not saying I outright ignore guys just because they got injured for the playoffs, but it is clearly time that that player is not contributing value when he was supposed to, and I think it's pretty much a given that we all penalize a guy for missed time.

How much should those penalties be? Up to you. Do I weigh some missed time more than others based on context? Yup, but you don't have to.

[NinjaSheppard]

Hypothetical for people who penalize players for getting hurt?

Say there is Player A and Player B and you think both guys are having similar level seasons. Player A loses in the first round but stays healthy. Player B wins in the first round and then gets hurt in the second round and can't finish the series.

Is it fair to penalize player B for getting hurt in games that player A didn't even play? If your answer is Yes does that mean you are rewarding player A for losing in the first round?

[Texas Chuck]

yeah I'm with Q on this. If we rule out great seasons by guys unavailable for the playoffs due to injury do we rule out seasons where their team fails to qualify for the playoffs? Because if you lift your team a lot(say 05 KG for example) but its not enough to get the team into the playoffs why is that more valuable that Duncan or Kobe playing great and getting their team into the dance--even if they are no longer there to help?

I mean if the playoffs are all that matter, then Kobe and Duncan should rate as more valuable because your team being in teh playoffs even without its best player has more of a chance to win than a team not in.

Feels like a slippery slope to start down imo.

I would say first and foremost that you do what you think is right. I insist that people focus on how effective a guy actually was rather than what he could have been if he wasn't pissed off at the coach/owner/his dog/etc, but beyond that, your call.

For me personally it's not really a hard call. I'm not saying I outright ignore guys just because they got injured for the playoffs, but it is clearly time that that player is not contributing value when he was supposed to, and I think it's pretty much a given that we all penalize a guy for missed time.

How much should those penalties be? Up to you. Do I weigh some missed time more than others based on context? Yup, but you don't have to.

Doc,

You do realize that isn't my actual view on the matter, right? I was simply trying to understand the position of those who were weighting seasons as a 0 when guys were injured in the playoffs. I was trying to understand if the playoffs were all that mattered how they evaluated seasons where a player's team didn't qualify for the playoffs.

I'm assuming you read the further dialogue Bob and I had on this where he helped me better understand his position making my post here mostly irrelevant.

I don't use the same approach, but I can now better appreciate where Bob and others are coming from.

[ElGee]

I'm surprised that 2015 Stephen Curry was only #38 on his list last year. Someone like Cowens vs Curry is a pretty hard sell in the favor of Cowens, even for those skeptical of how good Curry was last year.

Thoughts on 15 Curry's offense? Given you missed out on the POY thread. Seems like you regard him quite a bit lower than a lot of us (given he's getting traction in this project already), but I recall you saying you had a tough time gauging just how valuable he could be on offense.

EDIT: And 15 Harden and Westbrook while we're at it

Are 15 Harden and Westbrook that much better than 14? To clarify on Curry, I mean he would supplant Thompson from the list (who I'd probably have below Baylor). As to where, I'm not sure. Probably top 30, but not 20. I'm still not clear on Curry.

So sounds like he had him mid-20s which is lower than he went (#17) and much lower than where some of us had him (I had him top 15, I think Dr. Spaceman was at least making his case within the top 10).

I've had a hard time judging Curry because he's basically sui generis. I'm a little more comfortable now, but it's still a tricky thing for me to gauge and I'm waiting for the postseason (and perhaps next season) to really refine my opinion.

I'll quantify where I am to help make it clearer: Last year I thought he was around a +5.5 offensive player, obviously super portable. This would place him squarely in my top-10 offensive peaks of all-time. I could see an argument for slightly lower or slightly higher. This year, his increase in scoring efficiency alone is worth about +2 points mathematically. I think he's also warping the floor slightly more than last year because team are so hyper-sensitive to him at this point and he's better than ever at deep shooting (drawing more attention) and penetrating and finishing with those little floaters. I think he's at his peak in passing too. It's a very strong argument for GOAT offensive season, and depending on how far one takes it, unless you consider him a horrible defender (hard to see) that leaves you with a GOAT overall season.

However, the football fan in me is acting up. In the NFL, there are often briefly successful new tactics that are then dampened as teams develop a counter-strategy. Curry has tailed off in efficiency in the second half of the season. Variance? Fatigue? Or defenses trying to adapt a new strategy against him? (Is the book still out on this until next year even?) This isn't like the balancing act a defense faces against peak Shaq or Jordan -- their buckets are worth 2 points only. Curry's efficiency on 3-points shots is so extreme that it seems better for defenses to warp the floor (classically a defensive no-no) to get him off of these shots that are worth like 1.5 points (or more) per attempt. (The equivalent of Jordan or Shaq shooting 75%!)

As of now, there's still some malleability in how I judge his last 2 seasons as I'm synthesizing all this stuff.

[bondom34]

^^^
The GOAT.

[SideshowBob]

I've had a hard time judging Curry because he's basically sui generis. I'm a little more comfortable now, but it's still a tricky thing for me to gauge and I'm waiting for the postseason (and perhaps next season) to really refine my opinion.

I'll quantify where I am to help make it clearer: Last year I thought he was around a +5.5 offensive player, obviously super portable. This would place him squarely in my top-10 offensive peaks of all-time. I could see an argument for slightly lower or slightly higher. This year, his increase in scoring efficiency alone is worth about +2 points mathematically. I think he's also warping the floor slightly more than last year because team are so hyper-sensitive to him at this point and he's better than ever at deep shooting (drawing more attention) and penetrating and finishing with those little floaters. I think he's at his peak in passing too. It's a very strong argument for GOAT offensive season, and depending on how far one takes it, unless you consider him a horrible defender (hard to see) that leaves you with a GOAT overall season.

Agreed on the bolded. I've seen arguments that he was basically this good (or almost as good) last year already and that this season is simply a result of allowing him more freedom with how far he take his shooting. While I appreciate that train of thought, I can't ignore that I'm also observing improvements in his game - handle in single coverage is smoother and overall he's better at shaking/blowing by defenders and has been more creative at finishing at the rim and in the paint (better body control too I think, he used to convert acrobatic layups before but this season he's been a marvel at times) and (thanks ronnymac) he's made conscious/physical changes that have allowed him to be more accurate from distant range. IMO its clear that he's just better in a vacuum this year.

However, the football fan in me is acting up. In the NFL, there are often briefly successful new tactics that are then dampened as teams develop a counter-strategy. Curry has tailed off in efficiency in the second half of the season. Variance? Fatigue? Or defenses trying to adapt a new strategy against him?

If its the latter two (fatigue or defenses adjusting), its something to note, no? Just turned 28 with ~19000 MP (RS+PS), playing 34.2 minutes a game with a minute distribution that is fairly staggered across a single game - fatigue would be a little concerning IMO, it shouldn't be a problem at this point in his career, but I don't really think that's it.

Defenses figuring it out makes more sense, as does it taking considerable time for that to happen - he's pretty unprecedented. Plus we didn't get to observe Pop experimenting in a series last season, I want to wait and see what he comes up with and then how that could shape defenses next season (we've seen playoff schemes affect strategy in following seasons; think prototyping).

Also (depending on how you want to split it) GSW On Court w/Curry:

Pre-All Star: 120.0 ORTG
Post-All Star: 117.0 ORTG

None of that is to say that he necessarily needs to be heavily penalized ITO evaluation - more like it could be the difference between GOAT offense by a margin vs. GOAT offense by a smaller margin.

(Is the book still out on this until next year even?) This isn't like the balancing act a defense faces against peak Shaq or Jordan -- their buckets are worth 2 points only. Curry's efficiency on 3-points shots is so extreme that it seems better for defenses to warp the floor (classically a defensive no-no) to get him off of these shots that are worth like 1.5 points (or more) per attempt. (The equivalent of Jordan or Shaq shooting 75%!)

Absolutely. See above.

[kayess]

^^^
The GOAT.

Amen. I thought he'd stopped posting!

And of course he hits it out of the park with that post. Acknowledging the very real possibility of the GOAT RS by some distance, but still leaving room for adjustment.

@ElGee - because of the balancing act you refer to in your post, if this is a trend that continues will you still lean towards Shaq as the GOAT peak due to how much more difficult it was to stop him?

[lorak]

Defenses figuring it out makes more sense, as does it taking considerable time for that to happen - he's pretty unprecedented. Plus we didn't get to observe Pop experimenting in a series last season, I want to wait and see what he comes up with and then how that could shape defenses next season (we've seen playoff schemes affect strategy in following seasons; think prototyping).

I strongly disagree. It's not like Curry started to playing that way this year. He is taking almost 8 3PA per game since '13. Defenses had plenty of time to figure out how to defend him and in worst case they would have done it at the beginning of current season after Curry's MVP campaign. But they didn't, so it's very unlikely some drastic adjustments were done in mid season across all teams in the NBA. I mean, what exactly teams like Lakers, Jazz, Mavs, Spurs, Wolves and 76ers (all Curry's games below 60 TS% since March) did defensively differently than in first half of the season? Regression to the mean looks like more likely explanation.

[SideshowBob]

Defenses figuring it out makes more sense, as does it taking considerable time for that to happen - he's pretty unprecedented. Plus we didn't get to observe Pop experimenting in a series last season, I want to wait and see what he comes up with and then how that could shape defenses next season (we've seen playoff schemes affect strategy in following seasons; think prototyping).

I strongly disagree. It's not like Curry started to playing that way this year. He is taking almost 8 3PA per game since '13. Defenses had plenty of time to figure out how to defend him and in worst case they would have done it at the beginning of current season after Curry's MVP campaign. But they didn't, so it's very unlikely some drastic adjustments were done in mid season across all teams in the NBA. I mean, what exactly teams like Lakers, Jazz, Mavs, Spurs, Wolves and 76ers (all Curry's games below 60 TS% since March) did defensively differently than in first half of the season? Regression to the mean looks like more likely explanation.

Well in SA they were blitzing him really high, like a few feet off half court and switching everything w/him involved and then forcing him to drive into the bigs. Seemed to work.

But general point taken. So you think he was just on a hot streak earlier in the season and he's cooled off a bit?

[Dr Spaceman]

Defenses figuring it out makes more sense, as does it taking considerable time for that to happen - he's pretty unprecedented. Plus we didn't get to observe Pop experimenting in a series last season, I want to wait and see what he comes up with and then how that could shape defenses next season (we've seen playoff schemes affect strategy in following seasons; think prototyping).

I strongly disagree. It's not like Curry started to playing that way this year. He is taking almost 8 3PA per game since '13. Defenses had plenty of time to figure out how to defend him and in worst case they would have done it at the beginning of current season after Curry's MVP campaign. But they didn't, so it's very unlikely some drastic adjustments were done in mid season across all teams in the NBA. I mean, what exactly teams like Lakers, Jazz, Mavs, Spurs, Wolves and 76ers (all Curry's games below 60 TS% since March) did defensively differently than in first half of the season? Regression to the mean looks like more likely explanation.

Well in SA they were blitzing him really high, like a few feet off half court and switching everything w/him involved and then forcing him to drive into the bigs. Seemed to work.

But general point taken. So you think he was just on a hot streak earlier in the season and he's cooled off a bit?

Tangentially, I LOVED what I saw from Curry on Thursday night. He murdered them from floater range, which is a shot he can hit very consistently, and his adaptation to a pretty classic drive and kick style attack meant he tore up the interior with his passing as well. Just a brilliant ha,e all-around.

His below the arc game is criminally underrated, and may have been his biggest improvement year over year.

Quick aside: I've always loved Curry's resiliency, but this ability to seemingly shift his attack to counter the defensive scheme is becoming scary if it's a real thing. It's the closest thing I've seen to peak Jordan, and he's a more dangerous scorer in genera1l than Jordan was. I commented on this during the finals last year, but it seems like he can now do this game-to-game.