16 Draymond or 17 Westbrook

[Special_Puppy]

Does Jokic in the playoffs? Do these other players have playoffs series played mostly without the other best player leading to these numbers AND a regular season like Draymond's AND a Finals series like his including game 7? You're waiving away a lot.

No, I’m specifically waving away the idea of drawing any real conclusion from numbers in a 359-minute sample. And that’s true whether the data for someone in a sample like that looks good, bad, or somewhere in the middle. It’s just a sample we’d expect to be very noisy. It’s not completely useless and can be thrown on the pile of info we have, but on its own it’s not worth very much. I actually largely agree with your conclusion about Draymond though, because we do actually have other information that looks great for him too.

It's 359 minutes with playoff series played primarily without the other best player on the back of an incredible regular season. Do you know of comparable results that are also "noisy" like this?

Most playoff on/off analysis which is unfortunate since lineup data is so useful and we care about the playoffs more than the regular season

[Special_Puppy]

No, I’m specifically waving away the idea of drawing any real conclusion from numbers in a 359-minute sample. And that’s true whether the data for someone in a sample like that looks good, bad, or somewhere in the middle. It’s just a sample we’d expect to be very noisy. It’s not completely useless and can be thrown on the pile of info we have, but on its own it’s not worth very much. I actually largely agree with your conclusion about Draymond though, because we do actually have other information that looks great for him too.

It's 359 minutes with playol5ayerff series played primarily without the other best player on the back of an incredible regular season. Do you know of comparable results that are also "noisy" like this?

Yeah 359 minutes in playoffs is a useful data sample for the purpose of evaluating someone withour their co star

Cannot help but wonder if such a sample size would be so nitpicked with different players involved

If you don’t trust the small playoff sample in 2016…From 2015-2017, Draymond has a +9.6 net rating without Curry or Durant on the court in a 1400 minute sample. Curry has a +7.5 net rating without Draymond or Durant on the court in that same time period (1100 minute sample).

https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

[lessthanjake]

It's 359 minutes with playol5ayerff series played primarily without the other best player on the back of an incredible regular season. Do you know of comparable results that are also "noisy" like this?

Yeah 359 minutes in playoffs is a useful data sample for the purpose of evaluating someone withour their co star

Cannot help but wonder if such a sample size would be so nitpicked with different players involved

If you don’t trust the small playoff sample in 2016…From 2015-2017, Draymond has a +9.6 net rating without Curry or Durant on the court in a 1400. Curry has a +7.5 net rating without Draymond or Durant on the court in that same time period (1100 minute sample).

https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Yeah, see that gets to a sample size that starts to be potentially meaningful. Draymond’s impact numbers in that 2015-2017 timeframe are incredibly impressive even in larger samples. And his 2016 year was pretty clearly his best, so I think he was very likely significantly more impactful that year than he was overall in that three-year span. As for those specific numbers, I will note that the minutes these guys were left on the court without each other were a bit different—Draymond’s minutes without Steph/Durant were a decent bit more likely to be minutes where the Warriors had more starters in than the other team, and the Warriors with Draymond did do super well in those minutes (while the Warriors did better with Steph and no Durant/Draymond than with Draymond and no Durant/Steph in minutes with no starter advantage). That’s a bit nitpicky, but all that (and more) is the sort of thing that ends up inherently factored into RAPM and is how Steph ends up above Draymond in RAPM in that timeframe despite the numbers you provided. Regardless, Draymond’s impact numbers in that timeframe are super impressive, and I don’t even really disagree with HCL’s general conclusion about him—just the weight he was putting on specific small-sample data.

[homecourtloss]

It's 359 minutes with playol5ayerff series played primarily without the other best player on the back of an incredible regular season. Do you know of comparable results that are also "noisy" like this?

Yeah 359 minutes in playoffs is a useful data sample for the purpose of evaluating someone withour their co star

Cannot help but wonder if such a sample size would be so nitpicked with different players involved

If you don’t trust the small playoff sample in 2016…From 2015-2017, Draymond has a +9.6 net rating without Curry or Durant on the court in a 1400 minute sample. Curry has a +7.5 net rating without Draymond or Durant on the court in that same time period (1100 minute sample).

https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Draymond without Curry, 2015-2019 playoffs



Draymond without Curry or Durant

[Special_Puppy]

Yeah 359 minutes in playoffs is a useful data sample for the purpose of evaluating someone withour their co star

Cannot help but wonder if such a sample size would be so nitpicked with different players involved

If you don’t trust the small playoff sample in 2016…From 2015-2017, Draymond has a +9.6 net rating without Curry or Durant on the court in a 1400 minute sample. Curry has a +7.5 net rating without Draymond or Durant on the court in that same time period (1100 minute sample).

https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Draymond without Curry, 2015-2019 playoffs



Draymond without Curry or Durant

For what it’s worth:

Draymond without Curry or Durant from 2015 to 2019 (RS+PO included): +4.8 over a 2300 minute sample

Curry without Draymond or Durant from 2015 to 2019 (RS+PO included): +7.3 over a 1400 minute sample


https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2018-19,2017-18,2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

[homecourtloss]

If you don’t trust the small playoff sample in 2016…From 2015-2017, Draymond has a +9.6 net rating without Curry or Durant on the court in a 1400 minute sample. Curry has a +7.5 net rating without Draymond or Durant on the court in that same time period (1100 minute sample).

https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Draymond without Curry, 2015-2019 playoffs



Draymond without Curry or Durant

For what it’s worth:

Draymond without Curry or Durant from 2015 to 2019 (RS+PO included): +4.8 over a 2300 minute sample

Curry without Draymond or Durant from 2015 to 2019 (RS+PO included): +7.3 over a 1400 minute sample


https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2018-19,2017-18,2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Curry is a better player with a strong case for top 10 all time for career and peak, but the data, especially the playoffs data, show that Draymond is highly underrated with a case for top 40 peak (at least) all time.

[Special_Puppy]

Draymond without Curry, 2015-2019 playoffs



Draymond without Curry or Durant

For what it’s worth:

Draymond without Curry or Durant from 2015 to 2019 (RS+PO included): +4.8 over a 2300 minute sample

Curry without Draymond or Durant from 2015 to 2019 (RS+PO included): +7.3 over a 1400 minute sample


https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2018-19,2017-18,2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Curry is a better player with a strong case for top 10 all time for career and peak, but the data, especially the playoffs data, show that Draymond is highly underrated with a case for top 40 peak (at least) all time.

He was 40th all time in the last peaks project and yeah Draymond at his peak was a top 6-9 player in the league

[Djoker]

The Curry-Green lineup permutations are unbalanced. Because of Curry missing a bunch of 1st round games in 2016 and 2018, the Green ON Curry OFF sample comes largely from early round games. Green missed one playoff game in the 2016 Finals so a bigger chunk of Curry ON Green OFF data is from later rounds which are tougher. Not an apples to apples comparison.

[theonlyclutch]

Using 'lineup permutations' to critique Draymond's impact in this context when the data largely comes from the Dray-led warriors steamrolling playoff teams (to a roughly +8/9 level) is... odd.

[One_and_Done]

You can build a playoff team around Westbrook. You can't say the same about Draymond. I say that as one of the biggest Westbrook critics out there.

[lessthanjake]

Draymond without Curry, 2015-2019 playoffs



Draymond without Curry or Durant

For what it’s worth:

Draymond without Curry or Durant from 2015 to 2019 (RS+PO included): +4.8 over a 2300 minute sample

Curry without Draymond or Durant from 2015 to 2019 (RS+PO included): +7.3 over a 1400 minute sample


https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2018-19,2017-18,2016-17,2015-16,2014-15&SeasonType=All&PlayerIds=201939,203110,201142&Leverage=Medium,High,VeryHigh

Curry is a better player with a strong case for top 10 all time for career and peak, but the data, especially the playoffs data, show that Draymond is highly underrated with a case for top 40 peak (at least) all time.

I feel like this is a pretty unremarkable conclusion to be driving at. Draymond having a case for top-40 peak is, I think, a pretty mainstream view on the PC board (though I grant you that it probably isn’t mainstream with the general public, and I’m on board with saying he’s underrated by the general public). Indeed, he was ranked 40th in the last peaks project. I guess if his standing compared to all past players was held constant, that might put him at #42 now, since 2025 SGA would surely go above him, and 2024 Luka might leapfrog him. But that still squarely leaves him in the “a case for top 40 peak” territory. He’s being compared in this thread to someone who was #43 in the last peaks project, though, so he’s being compared to another guy with a case for top 40 peak, meaning your conclusion here doesn’t definitively answer the question of the thread. In other words, I think we’re mostly all on the same general page about where Draymond’s peak stands, but that doesn’t really answer the question of the thread.