2021-22 RealGM All-Season Awards - Discussion Thread

[jalengreen]

funny thing about steph this postseason is how bad his FT shooting has been compared to his standard. we expect him to at least be a 90% ft shooter but hes only been at 80.5% in the playoffs which is shocking for him

(90% ft shooting would have his approximate TS% at 61.75% instead of 60.53%)

[Doctor MJ]

Thinking about 6th Man of the Year, finding it odd that neither Otto Porter nor Grant Williams really got any kind of consideration for the regular season award when at this point they seem like they should be pretty strong candidates.

[falcolombardi]

are we gonna have an all-playoffs team as category? cause i almost feel tempted at al horford msking at least a second team if there was one

[Doctor MJ]

are we gonna have an all-playoffs team as category? cause i almost feel tempted at al horford msking at least a second team if there was one

Not a part of this particular project, but it's something that could be its own project.

[Dutchball97]

How much importance do you guys place on being the best player in a series? I'm mainly looking at Steph with this tbh. He's been great this post-season and right now he seems to have the best cards among players still in it to make a push for the #1 spot for POY but I do think it is a bit concerning he arguably wasn't the best in any of the series so far. I don't think too many people would disagree with the notion Jokic and Luka were the best individual players in their series against GSW. Curry was the best in the Grizzlies series but only because Ja went down.

Now I know there are always people who rate Curry higher than I do due to his infamous gravity but I do have to say it's gonna be hard for me to elevate him over players who I think outplayed him head to head despite the Warriors still winning due to being much deeper. To a lesser degree I also feel like this about Tatum vs Giannis, unless Tatum leads the Celtics to a ring with an amazing performance it'll be hard to elevate him over Giannis after not really outperforming him head to head.

[Doctor MJ]

How much importance do you guys place on being the best player in a series? I'm mainly looking at Steph with this tbh. He's been great this post-season and right now he seems to have the best cards among players still in it to make a push for the #1 spot for POY but I do think it is a bit concerning he arguably wasn't the best in any of the series so far. I don't think too many people would disagree with the notion Jokic and Luka were the best individual players in their series against GSW. Curry was the best in the Grizzlies series but only because Ja went down.

Now I know there are always people who rate Curry higher than I do due to his infamous gravity but I do have to say it's gonna be hard for me to elevate him over players who I think outplayed him head to head despite the Warriors still winning due to being much deeper. To a lesser degree I also feel like this about Tatum vs Giannis, unless Tatum leads the Celtics to a ring with an amazing performance it'll be hard to elevate him over Giannis after not really outperforming him head to head.

So, I think if you think a player was the best player in a given series, that should matter.

I'll tell you though, when considering a losing player in that role, I tend to look to see evidence that the other team was struggling with said player's presence.

Here's the Per 100 +/- numbers for the Mavs in this series among guys playing 30 MPG or more:

Brunson -9.6
Finney -13.6
Bullock -19.3
Doncic -28.4

All standard caveats about standard size apply, but I do think it's important to understand that this isn't remotely a situation where Doncic is unstoppable but the Warriors are winning because Doncic can't play 48 MPG.

Rather, to this point, the Doncic minutes have been where the Warriors have been thriving the most.

To me this is part of the more general point that you cannot compare the volume of production between players to gauge how effective their play actually is. Doncic isn't putting up bigger box score numbers than any Warrior because he's playing "better", but rather because that's just how Dallas plays with Doncic on the court.

It's of course very impressive that Doncic can do this and as I try to keep making clear I'm not betting against Doncic's career greatness, but to me the notion that you could be the MVP of a series while your team gets destroyed this badly when you're on the court AND it isn't as bad when you're off, just is a non-starter of an argument.

[sp6r=underrated]

I rarely quote myself but I'm going too now.

My hesitancy with plus/minus is that we need an enormous sample size and I think even full 4 rounds of PS data isn't sufficient. If a player's box score stats look very similar in the RS & PS but I see a large difference in +/- I think there is a good chance we're looking at noise.

Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

[Dutchball97]

How much importance do you guys place on being the best player in a series? I'm mainly looking at Steph with this tbh. He's been great this post-season and right now he seems to have the best cards among players still in it to make a push for the #1 spot for POY but I do think it is a bit concerning he arguably wasn't the best in any of the series so far. I don't think too many people would disagree with the notion Jokic and Luka were the best individual players in their series against GSW. Curry was the best in the Grizzlies series but only because Ja went down.

Now I know there are always people who rate Curry higher than I do due to his infamous gravity but I do have to say it's gonna be hard for me to elevate him over players who I think outplayed him head to head despite the Warriors still winning due to being much deeper. To a lesser degree I also feel like this about Tatum vs Giannis, unless Tatum leads the Celtics to a ring with an amazing performance it'll be hard to elevate him over Giannis after not really outperforming him head to head.

So, I think if you think a player was the best player in a given series, that should matter.

I'll tell you though, when considering a losing player in that role, I tend to look to see evidence that the other team was struggling with said player's presence.

Here's the Per 100 +/- numbers for the Mavs in this series among guys playing 30 MPG or more:

Brunson -9.6
Finney -13.6
Bullock -19.3
Doncic -28.4

All standard caveats about standard size apply, but I do think it's important to understand that this isn't remotely a situation where Doncic is unstoppable but the Warriors are winning because Doncic can't play 48 MPG.

Rather, to this point, the Doncic minutes have been where the Warriors have been thriving the most.

To me this is part of the more general point that you cannot compare the volume of production between players to gauge how effective their play actually is. Doncic isn't putting up bigger box score numbers than any Warrior because he's playing "better", but rather because that's just how Dallas plays with Doncic on the court.

It's of course very impressive that Doncic can do this and as I try to keep making clear I'm not betting against Doncic's career greatness, but to me the notion that you could be the MVP of a series while your team gets destroyed this badly when you're on the court AND it isn't as bad when you're off, just is a non-starter of an argument.

Even though production and boxscore stats rely on a small sample size in a play-off series, I find +- to be even less reliable in such a small sample size. Especially considering how +- is influenced to a huge degree by what teammates and opponents are on the floor. I respect everyone's own judgement in valuing +- here but for me personally it doesn't move the needle nearly as much as someone's individual production, which while also influenced by outside factors, is still not nearly as reliant on other players as +-. I also didn't call Luka the MVP of the series, I agree that would be strange when his team is down 0-3. I called him the best individual player, to me that's a pretty significant difference.

In any case I didn't mean to make a definitive statement about Steph vs Luka, everyone is free to make their own judgement on that. I'm just asking IF you think a player was outplayed in a series but advanced due to either luck or just having a better team around them, how do you look at that.

To use another example let's look at 2002 with Duncan and Shaq. In the regular season they were clearly the best two players in the league imo with Duncan winning out on MVP due to Shaq missing 15 games, while Duncan (and Kidd who also managed to finish ahead of Shaq in MVP voting) played all 82. In any case they're the two main guys you look at going into the post-season and I'd argue they both delivered. In the retro POY project for 02 (held in 2010) the discussion was clearly between the two as well with Shaq just winning out due to more people putting extra value on winning the title than just individual performance (some people even had Kobe ahead of Duncan as well because of this). However when you look at the Lakers - Spurs series Duncan looks like the clear best player there so I'm inclined to go with Duncan over Shaq that year since they played at a similar level all year but I'd rate the head to head outperformance higher than continued team success. In a similar vein I'm leaning towards keeping Jokic ahead of Curry and Giannis ahead of Tatum and I'm unsure if either winning the title would be able to change my mind on that.

On that note, I wonder if there'd be interest in revisiting the retro POY some time. 2010 is a long time ago and the landscape of active posters has also changed, although I'm not sure how many outcomes would be different.

[Doctor MJ]

I rarely quote myself but I'm going too now.

My hesitancy with plus/minus is that we need an enormous sample size and I think even full 4 rounds of PS data isn't sufficient. If a player's box score stats look very similar in the RS & PS but I see a large difference in +/- I think there is a good chance we're looking at noise.

Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

Certainly fine to talk about small sample size issues, but let's be real here:

When we talk about how guys on losing teams should still get Finals MVP consideration, we're not talking about guys on teams getting swept.

[Outside]

I rarely quote myself but I'm going too now.

My hesitancy with plus/minus is that we need an enormous sample size and I think even full 4 rounds of PS data isn't sufficient. If a player's box score stats look very similar in the RS & PS but I see a large difference in +/- I think there is a good chance we're looking at noise.

Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

Good point in your first paragraph, but you lose me with the second. Raw plus/minus is not the One True Stat to assess a player's performance, and neither are counting stats. Both are blunt force tools considering sample size and the lack of context. Both should be included in the assessment, along with other factors. Which means this falls into an area that is anathema to the nature of this board -- subjectivity. Even if you rely on a variety of stats and metrics, it's subjective how much emphasis you put on one piece of data over another, and it's of course subjective as soon as you introduce "context." But I don't see how you can avoid it.

[sp6r=underrated]

I rarely quote myself but I'm going too now.

My hesitancy with plus/minus is that we need an enormous sample size and I think even full 4 rounds of PS data isn't sufficient. If a player's box score stats look very similar in the RS & PS but I see a large difference in +/- I think there is a good chance we're looking at noise.

Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

Certainly fine to talk about small sample size issues, but let's be real here:

When we talk about how guys on losing teams should still get Finals MVP consideration, we're not talking about guys on teams getting swept.

I wasn't talking about Finals MVP. It is an award I argued 13 years ago should be eliminated. I've changed my mind on a lot of NBA related things over those years but not that award.

viewtopic.php?f=6&t=893156

My point was +/- doesn't work in small sample sizes and can only be a very small part of player analysis when you are doing such analysis.

[eminence]

I would say Steph has pretty convincingly outplayed Luka in the WCF so far.

[Doctor MJ]

I rarely quote myself but I'm going too now.



Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

Certainly fine to talk about small sample size issues, but let's be real here:

When we talk about how guys on losing teams should still get Finals MVP consideration, we're not talking about guys on teams getting swept.

I wasn't talking about Finals MVP. It is an award I argued 13 years ago should be eliminated. I've changed my mind on a lot of NBA related things over those years but not that award.

viewtopic.php?f=6&t=893156

My point was +/- doesn't work in small sample sizes and can only be a very small part of player analysis when you are doing such analysis.

Okay to be clear:

There is a difference between using the +/- over a random 4-7 game stand to make a definitive statement about the player in general, and using +/- to hammer home the fact that a guy on a losing team was indeed losing when he was on the court.

If you personally believe that it's a total fluke that GS is up 3-0 right now, you go right ahead and make that argument.

If you personally believe that it's a total fluke that Dallas is losing the Doncic minutes, you go right ahead and make that argument.

But this is not me trying to shoehorn +/- into small sample size theater out of nowhere. When people think that the best guy in the series also happens to be on the team that is losing every game because his production is so high, it makes perfect sense to ask whether there's +/- data that suggests this is a "winning because of when Player X is on the bench". And that's not what the data is telling us.

[MyUniBroDavis]

I rarely quote myself but I'm going too now.



Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

Certainly fine to talk about small sample size issues, but let's be real here:

When we talk about how guys on losing teams should still get Finals MVP consideration, we're not talking about guys on teams getting swept.

I wasn't talking about Finals MVP. It is an award I argued 13 years ago should be eliminated. I've changed my mind on a lot of NBA related things over those years but not that award.

viewtopic.php?f=6&t=893156

My point was +/- doesn't work in small sample sizes and can only be a very small part of player analysis when you are doing such analysis.

What do you have against finals mvp?

[iggymcfrack]

This might be the first year where my top 5 for POY are exactly the same as they were at the end of the regular season. More and more, it’s looking like:

1. Jokic
2. Giannis
3. Embiid
4. Tatum
5. Curry

is just the right answer. Yeah, Butler’s been incredible in the playoffs and could easily move himself into the top 5 if he has some big games leading Miami to the Finals and Curry could move up to 4 with a big Finals and Tatum could even go up to 3 if he goes supernova while winning Finals MVP, but I think this is probably going to be my final order..

[sp6r=underrated]

Certainly fine to talk about small sample size issues, but let's be real here:

When we talk about how guys on losing teams should still get Finals MVP consideration, we're not talking about guys on teams getting swept.

I wasn't talking about Finals MVP. It is an award I argued 13 years ago should be eliminated. I've changed my mind on a lot of NBA related things over those years but not that award.

viewtopic.php?f=6&t=893156

My point was +/- doesn't work in small sample sizes and can only be a very small part of player analysis when you are doing such analysis.

Most ye

What do you have against finals mvp?

This is what I wrote 13 years ago. It still more or less captures my issue with the award.

First, what you did in the first three rounds matters just as much as what you did in the finals. Second, some years the final series is not the real challenge, look at 2007 when the Spurs met their greatest test in Phoenix or 1990 when Detroit's greatest test was Chicago.

Third, sample size sometimes a player just gets hot or cold from the field for a 2-3 game stretch, which in one series is enough to change who gets the award. This is not really a problem when your talking about in 18-26 stretch of games.

There is precedent for this: the ABA awarded a playoff MVP.

[eminence]

This might be the first year where my top 5 for POY are exactly the same as they were at the end of the regular season. More and more, it’s looking like:

1. Jokic
2. Giannis
3. Embiid
4. Tatum
5. Curry

is just the right answer. Yeah, Butler’s been incredible in the playoffs and could easily move himself into the top 5 if he has some big games leading Miami to the Finals and Curry could move up to 4 with a big Finals and Tatum could even go up to 3 if he goes supernova while winning Finals MVP, but I think this is probably going to be my final order..

Those could be my top 5, but don't see Embiid at clear #3. His playoffs was mediocre at best (POY standards), I don't see a clear separation from 4/5 during the RS.

To me he's the one in clearest danger of falling off the ballot for Butler or Luka.

[Doctor MJ]

I rarely quote myself but I'm going too now.

My hesitancy with plus/minus is that we need an enormous sample size and I think even full 4 rounds of PS data isn't sufficient. If a player's box score stats look very similar in the RS & PS but I see a large difference in +/- I think there is a good chance we're looking at noise.

Through the first two rounds if you heavily weighed plus/minus Luka was playing at an above GOAT level. And in the WCF he's played at a below replacement level. This level of variance just doesn't occur in basketball. In baseball elite hitters will have 2 week stretches were they are useless and random guys will play at ATG levels. That just doesn't occur much in hoops. If your metric is showing it occurring it probably isn't working.

Long/Short: counting stats give you far better idea of how players are playing than plus/minus stats in small sample sizes like the post-season

Good point in your first paragraph, but you lose me with the second. Raw plus/minus is not the One True Stat to assess a player's performance, and neither are counting stats. Both are blunt force tools considering sample size and the lack of context. Both should be included in the assessment, along with other factors. Which means this falls into an area that is anathema to the nature of this board -- subjectivity. Even if you rely on a variety of stats and metrics, it's subjective how much emphasis you put on one piece of data over another, and it's of course subjective as soon as you introduce "context." But I don't see how you can avoid it.

Great points.

I really want people to understand that my tendency to use +/-, and specifically raw +/-, in my posts, doesn't mean I'm using it as the end-all, be-all of my analysis. It's something objective I can point to that generally people aren't really thinking through.

I could just focus in my Curry vs Doncic comparison on the fact that Curry plays completely differently, it making sharp decisions right now, is attacking the paint better than he's done in the past, and is defending better than he's done in the past all while once again leading a more effective basketball team than we've ever seen Doncic do, but it's typically pretty hard to beat back a statistically-driven argument with qualitative arguments, and those arguing for Doncic basically by definition are extremely influenced by his box score production.

[falcolombardi]

Certainly fine to talk about small sample size issues, but let's be real here:

When we talk about how guys on losing teams should still get Finals MVP consideration, we're not talking about guys on teams getting swept.

I wasn't talking about Finals MVP. It is an award I argued 13 years ago should be eliminated. I've changed my mind on a lot of NBA related things over those years but not that award.

viewtopic.php?f=6&t=893156

My point was +/- doesn't work in small sample sizes and can only be a very small part of player analysis when you are doing such analysis.

Okay to be clear:

There is a difference between using the +/- over a random 4-7 game stand to make a definitive statement about the player in general, and using +/- to hammer home the fact that a guy on a losing team was indeed losing when he was on the court.

If you personally believe that it's a total fluke that GS is up 3-0 right now, you go right ahead and make that argument.

If you personally believe that it's a total fluke that Dallas is losing the Doncic minutes, you go right ahead and make that argument.

But this is not me trying to shoehorn +/- into small sample size theater out of nowhere. When people think that the best guy in the series also happens to be on the team that is losing every game because his production is so high, it makes perfect sense to ask whether there's +/- data that suggests this is a "winning because of when Player X is on the bench". And that's not what the data is telling us.

that sounds like an odd argument to use

not because of curry being or not better than luka, but because it is saying "if team A beats team B then best player in team A must have outplayed best player in team B" which is a weird reasoning to use

i dont like using who won a series to argue who was the better player in the series, even if in a majority of cases it may be the case (your star outplaying the other team star surely helps with winning)

[sp6r=underrated]

I wasn't talking about Finals MVP. It is an award I argued 13 years ago should be eliminated. I've changed my mind on a lot of NBA related things over those years but not that award.

viewtopic.php?f=6&t=893156

My point was +/- doesn't work in small sample sizes and can only be a very small part of player analysis when you are doing such analysis.

Okay to be clear:

There is a difference between using the +/- over a random 4-7 game stand to make a definitive statement about the player in general, and using +/- to hammer home the fact that a guy on a losing team was indeed losing when he was on the court.

If you personally believe that it's a total fluke that GS is up 3-0 right now, you go right ahead and make that argument.

If you personally believe that it's a total fluke that Dallas is losing the Doncic minutes, you go right ahead and make that argument.

But this is not me trying to shoehorn +/- into small sample size theater out of nowhere. When people think that the best guy in the series also happens to be on the team that is losing every game because his production is so high, it makes perfect sense to ask whether there's +/- data that suggests this is a "winning because of when Player X is on the bench". And that's not what the data is telling us.

that sounds like an odd argument to use

not because of curry being or not better than luka, but because it is saying "if team A beats team B then best player in team A must have outplayed best player in team B" which is a weird reasoning to use

i dont like using who won a series to argue who was the better player in the series, even if in a majority of cases it may be the case (your star outplaying the other team star surely helps with winning)

to put it another way, I don't have any confidence in a 4 game sample size to tell me that if Player A/Team A is on the court with Player B/Team B, and his team is outscoring them over that time period, that player A is outplaying Player B.

I don't even have confidence it gets me 10% of the way to establishing the thesis. All it tells me is Team A is outscoring Team B during these selected minutes. It just isn't large enough of a sample size. I have much more confidence in individual box score stats over a 4 game sample size to establish who is outplaying who.

And if I felt that a 4 game plus minus stat really was sufficient, I would just take the plus/minus data dump it into excel rank em and use that as my list. because you wouldn't need any other stat nor would you need any analysis. A ninth grader could figure it out.