Everything Norman Powell

[DanH]

Well, another way to look at it is you start (and close with) your 5 best players. Right now they don't look that far off:

https://www.basketball-reference.com...01&idx=players

Fred better on-ball defender and on steals, Norm better rebounder, able to switch onto bigger wing players, lengthier closing out on perimeter shooters. Drtg of both is the same at 106

Psst, don't look at bball-ref's DRTG. It's essentially valueless.

Norm is the worst team rebounder on the team no matter his individual rebounding numbers.

Fred is significantly better than Norm.
​​

[Tmac2darak]

https://www.thestar.com/amp/sports/r...-is-today.html

Norm credits kobe

[inthepaint]

Psst, don't look at bball-ref's DRTG. It's essentially valueless.

Norm is the worst team rebounder on the team no matter his individual rebounding numbers.

Fred is significantly better than Norm.
​​

oh I see. We should just focus on the stats that support our opinion and dismiss the other as valueless, got it. I've seen you post for years, but haven't learned how to do that yet. Like, Norm has a better PER, better rebounding per 100 possessions, better 3pt%, better 2p% (significantly better eFG% as a result), better TS%, higher TRB% more points/36, gets blocked at the rim way less than Fred, but those are all worthless facts.

[golden]

Psst, don't look at bball-ref's DRTG. It's essentially valueless.

Norm is the worst team rebounder on the team no matter his individual rebounding numbers.

Fred is significantly better than Norm.
​​

Pssst... why don't we use a better stat like RPM? You know... that advanced stat that always makes Lowry look great. Makes us appreciate all the hidden things that he does that don't show up in the boxscore. That stat is valueless too, right?

Guess what? Norm is significantly better than FVV in both ORPM and DRPM. TD is pretty awesome, too. I'm not complaining... we have 3 very good SGs. Norm is arguably more valuable than FVV, because he doesn't crater off the bench and scores in multiple ways (transition, draw fouls, from 3, etc...)

I'm not saying Norm is better than FVV, but there are certainly numbers that support that opinion.





http://www.espn.com/nba/statistics/rpm/_/position/2

[SkywalkerAC]

Stepping on up, got that gameface goin'.

[DanH]

Pssst... why don't we use a better stat like RPM? You know... that advanced stat that always makes Lowry look great. Makes us appreciate all the hidden things that he does that don't show up in the boxscore. That stat is valueless too, right?

Guess what? Norm is significantly better than FVV in both ORPM and DRPM. TD is pretty awesome, too. I'm not complaining... we have 3 very good SGs. Norm is arguably more valuable than FVV, because he doesn't crater off the bench and scores in multiple ways (transition, draw fouls, from 3, etc...)

I'm not saying Norm is better than FVV, but there are certainly numbers that support that opinion.





http://www.espn.com/nba/statistics/rpm/_/position/2

RPM is pretty much the worst of the attempts to replicate large sample APM/RAPM results in a more stable small sample. Besides BPM, which is barely an attempt. It's very tied into box score priors, and throws away a lot of the adjusted plus minus data that is supposed to be the entire point of an impact stat approach. They even overhauled it lately to try to fix its issues and its still a poor correlation to actual APM stats.

PIPM is the best number on the market right now. RAPM is noisy but can be of value, but is also hard to source up to date numbers.

It's also important to remember the crucial context of every impact stat, much like production stats - they measure how effective the player was exactly how he's been used. Powell has slid nicely into that ideal role for Fred (that is, primarily a spot up shooter and transition scorer, with a little bit of secondary/tertiary creation), while Fred has been asked to do way more than that, and Powell has spent very little time as the key point of attack defender in any lineup while Fred takes on a lot of that when he's in the lineup.

[DanH]

oh I see. We should just focus on the stats that support our opinion and dismiss the other as valueless, got it. I've seen you post for years, but haven't learned how to do that yet. Like, Norm has a better PER, better rebounding per 100 possessions, better 3pt%, better 2p% (significantly better eFG% as a result), better TS%, higher TRB% more points/36, gets blocked at the rim way less than Fred, but those are all worthless facts.

No, we should focus on good stats, and choose the appropriate ones for the topic we are discussing. Picking random box score stats to compare is a terrible way to compare any two players.

Players provide value one of two ways. They produce (generate points, rebounds, assists, raw numbers that are necessary to add up to a team's total production - if you can't score and can't rebound as a team, you can't win, so it has value), but that's also not all they do. Players can provide value that doesn't show up in individual production. That's why most smart advanced metric assessment of players includes looking at both things. But it requires different statistical approaches to evaluate the two different things.

I've made no argument against Norm being a better raw producer this year. His scoring has been undeniably great this year, and his efficiency is insane. None of what you bring up above is worthless. It's just all looking at one part of the game - individual production. That individual production, though, is at best half of what a player provides to his team (as it really mostly just measures offence), and realistically is far less than that, as players can provide significant value offensively without racking up box score stats. A pessimistic take would suggest it is 1/10th of the game, if we are thinking primarily about individual scoring, but that's too low IMO. Probably fair to say it's somewhere between 25-50% of most players' value.

But evaluating how a player impacts his team while generating his own production is the other part of the puzzle, and there's a reason why most of the effort in the advanced stats development over the last half decade has been poured into this part of the puzzle - it's the part that's been missing, we've had a handle on production for ages, but more importantly, it's probably the more important measure for the vast majority of players outside of the biggest star producers (the best of which you hope measure out well here as well).

By the way, the reason I call basketball-reference's DRTG valueless is this: do you know what it describes? I'll give you two hints. First hint: basketball-reference's ORTG measures a players' individual efficiency - it attempts to translates a player's individual box score production into a per-100-possession point efficiency. Just like the team level numbers we are used to. So if a player generally contributes 110 points of value (through their own scoring, assists, rebounding, etc) in 100 possessions they use up (through FGA's, TOV's, etc), they have a 110 ORTG. It's a little more complex than that but that's the idea. Second hint is: the logical assumption is that DRTG would therefore be the same thing but for the players they guard. The second hint is that it is not that at all. In fact, it is not anything logical at all. It's a fictional concept called a stop rate, based only on box score defensive numbers, used as an adjustment to the team's overall defensive rating. It's not related to the ORTG at all, and is the most derided "advanced" defensive number available. I'd bet you didn't know that when you were quoting it - which of course is part of the reason why it is a valueless stat, most people who use it don't know what it is, mostly because once you find out what it is, you stop using it. Sort of like when people discovered that the "contested rebound percentage" that NBA.com used to track was not the percentage of contested rebounds that a player wins, but the percentage of their rebounds that were contested, and people completely abandoned the stat once they knew what it really meant.

I certainly didn't intend to suggest that everything you were presenting was valueless. Just the bball-ref DRTG. It really is without real value in a basketball discussion.

Anyway, Norm's individual offensive production has certainly been far above Fred's this year. His 116 individual ORTG is pretty high up on the team, while Fred's 111 is still good but definitely a step down from Norm. But one of the things you hope for from a player who is putting up good offensive efficiency and production, would be to have a noticeable improvement in the offence because of them. And yet Norm in the impact stats (I use PIPM here, it's easier than digging up RAPM) has measured out a slightly lesser offensive contributor than Fred (though both are slightly above average) in terms of the team offence actually improving when he is on the court (adjusting, as the APM stats do, for who he plays with and against). While Fred grades out as solidly above average defensively, while Norm is a below average defender - again, within the context of the defensive roles they tend to be assigned, which are tougher for Fred in general IMO. Personally that matches my eye test, but eye tests can be exceedingly flawed. That role assignment also colours those individual offensive production numbers - Fred has been tasked with a level of creation that is simply not in his skill set, in one of Nurse's typical role-stretching gambits. Norm has for the most part been able to operate in his ideal role with a lot of catch and shoot and tertiary action, attacking moving defences or in transition. Providing production in that role has real value - but trying to compare production and efficiency without considering that context is where just throwing up box score comparisons kind of falls short.

[DanH]

There's actually a lot of depth we could dive into with NBA.com's tracking stats for shot types, etc, to compare Norm and Fred's exceedingly different offensive roles this season, but the quickest way is this simple demonstration.

% of 3PM's that are assisted:
Norm: 94%
Fred: 70%

% of 2PM's that are assisted:
Norm: 68%
Fred: 18%

[golden]

RPM is pretty much the worst of the attempts to replicate large sample APM/RAPM results in a more stable small sample. Besides BPM, which is barely an attempt. It's very tied into box score priors, and throws away a lot of the adjusted plus minus data that is supposed to be the entire point of an impact stat approach. They even overhauled it lately to try to fix its issues and its still a poor correlation to actual APM stats.

PIPM is the best number on the market right now. RAPM is noisy but can be of value, but is also hard to source up to date numbers.

It's also important to remember the crucial context of every impact stat, much like production stats - they measure how effective the player was exactly how he's been used. Powell has slid nicely into that ideal role for Fred (that is, primarily a spot up shooter and transition scorer, with a little bit of secondary/tertiary creation), while Fred has been asked to do way more than that, and Powell has spent very little time as the key point of attack defender in any lineup while Fred takes on a lot of that when he's in the lineup.

Context: you keep ignoring the fact that Fred craters without Pascal.

[DanH]

Context: you keep ignoring the fact that Fred craters without Pascal.

Except... I don't? I've been pretty clear that Fred is being asked to do too much creation and is not being used correctly while in that role. Siakam on the court is when he is least asked to do that. It's not a coincidence. I mean Gasol is our best defender by most measurements but I'm not asking him to guard the opposing point guard and then judging him on the results. If Fred could be our primary guard defender, run the offence as he has this year, and also score as efficiently as Powell in a much tougher shot creation role, all while being not reliant on playing beside a star, yeah, I guess that would be better. He'd also be a shoe in for a max contract of course and would be underpaid on that contract.

[golden]

There's actually a lot of depth we could dive into with NBA.com's tracking stats for shot types, etc, to compare Norm and Fred's exceedingly different offensive roles this season, but the quickest way is this simple demonstration.

% of 3PM's that are assisted:
Norm: 94%
Fred: 70%

% of 2PM's that are assisted:
Norm: 68%
Fred: 18%

Let’s just ignore FG% shall we?

2P%
Norm: .593
Fred: .430

That difference is insane.

But Fred tries hard. He really, really does. He also plays a lot with Pascal and Kyle. That kinda, sorta helps.

[golden]

Except... I don't? I've been pretty clear that Fred is being asked to do too much creation and is not being used correctly while in that role. Siakam on the court is when he is least asked to do that. It's not a coincidence. I mean Gasol is our best defender by most measurements but I'm not asking him to guard the opposing point guard and then judging him on the results. If Fred could be our primary guard defender, run the offence as he has this year, and also score as efficiently as Powell in a much tougher shot creation role, all while being not reliant on playing beside a star, yeah, I guess that would be better. He'd also be a shoe in for a max contract of course and would be underpaid on that contract.

Don’t you think that opposing defences have something to do with Fred’s role? They run him off the 3-pt line and challenge him to beat them at the rim all the time. It’s not like he can always pick and choose. He’s limited by size and athleticism and has to take what the defense gives him. Context..... remember?

[TheWaterboy]

Here's the real context in regards to Norm and Fred, in my opinion. Barring the effect of salary cap pressure and free agency competition, we are likely to retain both Fred and Norm. We are also likely to start transitioning Kyle into the Serge role (can start when needed, but feasts on backups) when his athleticism dictates, most likely somewhere between age 36 and 38, especially if Kyle is willing to take Serge dollars (20 mil) over a 3 year term after his current extension is done. When this happens, Fred is likely to slide into the starting point guard role, with Norm becoming the starting shooting guard. The real question, at least to me, is not "which player is a better player now?" but "can we develop both Norm and Fred in their current roles to play complementary roles in the future if they are our future guard tandem?"

[DanH]

Don’t you think that opposing defences have something to do with Fred’s role? They run him off the 3-pt line and challenge him to beat them at the rim all the time. It’s not like he can always pick and choose. He’s limited by size and athleticism and has to take what the defense gives him. Context..... remember?

I think that opposing defences have a heck of a lot less to do with it than the role the team has him in. Unless they are willing to let other guys eat them up without sending help, they can't run him off the line when he's off ball like he should be. His FGA's inside are not a result of being run off the line. They are a result of him initiating action, where the attack means he's driving inside of his own volition, which is just not putting him in a position to succeed. Obviously the intent is for him to improve there but that doesn't mean it's his ideal usage or that his current effectiveness is indicative of his peak effectiveness.

[DanH]

Let’s just ignore FG% shall we?

2P%
Norm: .593
Fred: .430

That difference is insane.

But Fred tries hard. He really, really does. He also plays a lot with Pascal and Kyle. That kinda, sorta helps.

Who is ignoring FG%? The entire point of my post was to point out one of the many reasons WHY their efficiencies are different.