Ty Lawson traded to Houston

[moofs]

Sorry I didn't respond earlier. Family issues and whatnot.

And boxscoregeeks... just to the whole site. They're really not a terribly good analytics site, they've shown in the past an unwillingness to look at anything differently than they do and have said quite a few outlandish things.

I actually agree with you to a degree on this. The guys who run it are mind-numbingly annoying and can indeed be very arrogant (in fairness, I've been told I come off that way a lot of times myself, so... whatever that says about anything).
As I mentioned, though, I went through proofs on the math behind their WP numbers from Berri myself, so I trust the metric. I haven't done any re-evaluations since they implemented positional adjustments to the equation, but their site claims to have the data calculated, largely from the original formula, so I run with it.

They absolutely think Deandre Jordan is an MVP candidate and the metric they used, when called out for flaws, is left unaddresssed. Most of the statistical community seems to have them sort of ignored due to their misgivings.

I actually agree with them on that. As far as the flaws, you know any good references?
I mean, the statistical community as a whole cite PER and PM/RAPM/RPM/APM/etc a lot, so ... While I consider Alvarez a bit of a douche, he's also a statistics-based software engineer at Google, so I know his pedigree probably doesn't suck, at the least. Add that to my own work (which I did take quite a while on, if only once), and you can at least understand why it's been my semi-lazy go-to move for a long time now.

http://bkref.com/tiny/Ucs4I

For a minute, I thought something was wrong here, but then noticed your URL only has 2014-2015. Yeah, Lawson's TS% was middling last year (literally - it was .01% less than average for PGs), but for their careers, each is better at:
Lawson: TS (+3.6%), 3P (+6.5%), Fouls (-0.8/48), TO (-1.6/48),
Westbrook: Stl (+0.5/48), Reb (+2.9/48), Blk (+0.3/48)

Westbrook may be an above-average rebounder, but he's a (slightly) below-average shooter, and he shoots a LOT. I love rebounding, but I don't think it overcomes the TS%.
(going by the Dean Oliver rating, which may or may not still be accepted, TS% is 40% of winning, TOs 25%, Rebs 20%, and FTs 15% - which would have Lawson winning out on the top two metrics, and Westbrook on the bottom two)

Also, impact stats have him as a huge plus/minus boost who helps teammates, much moreso than Lawson (check RPM from last year, its literally the difference from Harden to Evans).
http://espn.go.com/nba/statistics/rpm/_/sort/RPM/position/1

I'm into RPM about as much as I'm into macro-ec, which is to say: not at all (you probably guessed that already). When there are unquantifiable factors involved, I regard the idea that they can be controlled for as essentially being an assertion of magic. Especially when a lot of the control involves really low sample sizes. At macro-ec levels, I tend to go by best guess deduction (with confirmational tests if possible) and hope I'm right. I don't believe there's another option.

Maybe that explains some of the ridiculousness?

[bondom34]

Sorry I didn't respond earlier. Family issues and whatnot.

And boxscoregeeks... just to the whole site. They're really not a terribly good analytics site, they've shown in the past an unwillingness to look at anything differently than they do and have said quite a few outlandish things.

I actually agree with you to a degree on this. The guys who run it are mind-numbingly annoying and can indeed be very arrogant (in fairness, I've been told I come off that way a lot of times myself, so... whatever that says about anything).
As I mentioned, though, I went through proofs on the math behind their WP numbers from Berri myself, so I trust the metric. I haven't done any re-evaluations since they implemented positional adjustments to the equation, but their site claims to have the data calculated, largely from the original formula, so I run with it.

They absolutely think Deandre Jordan is an MVP candidate and the metric they used, when called out for flaws, is left unaddresssed. Most of the statistical community seems to have them sort of ignored due to their misgivings.

I actually agree with them on that. As far as the flaws, you know any good references?
I mean, the statistical community as a whole cite PER and PM/RAPM/RPM/APM/etc a lot, so ... While I consider Alvarez a bit of a douche, he's also a statistics-based software engineer at Google, so I know his pedigree probably doesn't suck, at the least. Add that to my own work (which I did take quite a while on, if only once), and you can at least understand why it's been my semi-lazy go-to move for a long time now.

http://bkref.com/tiny/Ucs4I

For a minute, I thought something was wrong here, but then noticed your URL only has 2014-2015. Yeah, Lawson's TS% was middling last year (literally - it was .01% less than average for PGs), but for their careers, each is better at:
Lawson: TS (+3.6%), 3P (+6.5%), Fouls (-0.8/48), TO (-1.6/48),
Westbrook: Stl (+0.5/48), Reb (+2.9/48), Blk (+0.3/48)

Westbrook may be an above-average rebounder, but he's a (slightly) below-average shooter, and he shoots a LOT. I love rebounding, but I don't think it overcomes the TS%.

Also, impact stats have him as a huge plus/minus boost who helps teammates, much moreso than Lawson (check RPM from last year, its literally the difference from Harden to Evans).
http://espn.go.com/nba/statistics/rpm/_/sort/RPM/position/1

I'm into RPM about as much as I'm into macro-ec, which is to say: not at all (you probably guessed that already). When there are unquantifiable factors involved, I regard the idea that they can be controlled for as essentially being an assertion of magic. Especially when a lot of the control involves really low sample sizes. At macro-ec levels, I tend to go by best guess deduction (with confirmational tests if possible) and hope I'm right. I don't believe there's another option.

Maybe that explains some of the ridiculousness?

No prob, and hope all is well w the family!


As for the issues w/ BSG and wins produced, a few articles:

https://ascreamingcomesacrossthecourt.wordpress.com/2012/01/09/how-wins-produced-fails-in-being-the-magical-assessor-of-player-value/

The main issue with Wins Produced is how the authors and supporters discuss the model like it’s flawless. It’s an interesting method that can compete with Win Shares or PER, but to claim that it can explain wins with a 95% accuracy is ridiculous. All they’re doing is dressing up a team’s efficiency numbers and allotting each piece to a player, and then summing those results to show how close they’re correlated to wins. Well, of course they are; everyone knows a team’s points scored and allowed per possession can explain wins. If you have the audacity to ask them how defense can be explained by steals, blocks and rebounds, you’ll get a response about how your little brain can’t comprehend a counterintuitive result. My advice is to view Wins Produced as an interesting summation of box score stats, and not to sway your complete view of a player. Using Wins Produced as your only method in evaluation a player is like zooming in on one feature of an animal. Maybe that one part can lead you to a conclusion about the rest of the animal, but you could likely become the blind man holding the elephant’s trunk and believe it’s a snake instead of something much more powerful.

And an APBR thread:

http://www.apbr.org/metrics/viewtopic.php?f=2&t=25

One of the main points of Wages of Wins is that according to the evidence presented in the book, it appears that NBA decision-makers are irrational. They write: "It is not that people in the NBA are lazy or stupid. It is just that the tools at their disposal do not allow them to see the value of the various actions players take on the court."

The argument that supports this conclusion is they show that Wins Produced explains wins much better than NBA Efficiency, but NBA Efficiency is much more closely tied to salaries and All-Rookie Team voting (done by coaches). Thus, according to their evidence, it is irrational for teams to not be using something like Wins Produced to make their decisions.

OK, but the difficulty with this argument is that it hinges on Wins Produced better explaining wins than NBA Efficiency. Wins Produced does a great job explaining wins because of their team defense adjustment, but the authors admit that this adjustment has very little effect on their relative rankings of players. So if it doesn't matter much for the relative rankings of players, I just don't see how it can be used as a justification for the methodology. To me, that whole exercise raises a big red flag about the validity of using the prediction of team wins as a barometer for a metric for individual players.

But then if we move to another barometer - adjusted plus/minus ratings - we see that Wins Produced only performs better than NBA Efficiency if position adjustments are used for Wins Produced but not for NBA Efficiency. That significantly changes the story of much of their book. Instead of a story about NBA teams overvaluing scorers, their story becomes one that NBA decision-makers are irrational because they don't properly position adjust.

Moreover, the authors provide little justification for their position adjusting, especially in relation to how important it is to their metric. They argue that big players would have difficulty filling the roles of guards; i.e. a team could not play all centers. But if centers truly are worth more than guards as their unadjusted Wins Produced suggests, this would not be the only reaction of NBA decision-makers. Instead of playing centers at guard, what would happen would be that they would pay centers more than guards - which is precisely what does happen. So rather than proving conventional wisdom wrong, maybe the authors have provided justification for conventional wisdom.

Also, re Westbrook/Lawson (which is still completely crazy), there isn't a reasonable debate to say Lawson is better, it literally would be like saying Evans is better than Harden. Westbrook out rebounds, out assists, and is a better defender. He scores much more easily and given his usage is remarkably efficient. Lawson posts a mildly better TS on way lower usage, and for his usage Westbrook was crazy good. Unless you really think Lawson is a top 2-3 PG in the league I have a hard time with that. As well, every player he plays with gets better when he's around, I did this on the PC board to show him making teammates better moreso than John Wall, and I'd take Wall well over Lawson too.

Who has Westbrook made look better? It's not about Durant taking every shot.

Everyone? The offense fell off a cliff when he was off court, it was a 10 point difference.

Westbrook On:

http://nbawowy.com/kk66qchqp9#/kk66qchqp9

112 O Rating

Points per shot:
Morrow: 1.22
Ibaka: 1.13
Kanter: 1.23
Durant: 1.26
Roberson: 1.11
Adams: 1.11

Off:

http://nbawowy.com/druga9vlrhm#/druga9vlrhm

101 O Rating

Points per shot:
Morrow: 1.20
Ibaka: 1.04
Kanter: 1.17
Durant: 1.24
Roberson: 0.77
Adams: 1.06

So to recap that, Westbrook is better on offense, better on defense, and makes his teammates better. I mean I can't really say much else because that's an incredible POV to take. If you'd say you'd take Lawson over Westbrook maybe 5 years ago, OK, but Westbrook got better, Lawson didn't. BPM by year:

Lawson:
0, 1.9, 3.3, 1.4, 1.7, 0.5

Westbrook:
-.1, 2.1, 5, 3.2, 5, 6.4, 11

So one guy was pretty good young, and stayed the same or got worse. The other improved massively to a point where he's nearing MVP candidacy.

[bondom34]

Oh, and to address some other points, having a statistical degree doesn't mean you know how a stat applies to basketball, that's part of the issue with Berri/Alvarez. They're unwilling to move on advice from others who have really good input over on APBR and elsewhere. And I'm hoping the Deandre Jordan MVP thing was sarcasm....

[Mr. E]

You guys are reminding me of the law firm from Parks & Recreation

[moofs]

Oh, and to address some other points, having a statistical degree doesn't mean you know how a stat applies to basketball, that's part of the issue with Berri/Alvarez. They're unwilling to move on advice from others who have really good input over on APBR and elsewhere.

No doubt. I run into this problem with people in general, even (perhaps "especially", even) within math and science fields, a LOT.

I loved this quote, "The macroscopic model can show you the entire galaxy but blurred and hazy; the microscopic will focus on a few spiral arms of the galaxy with great detail and clarity. You can’t assume you know exactly what the rest of the galaxy looks like from only a small section."

The shortcomings you listed are actually issues I have always had with the stat myself (there are probably quotes to be dug up of me saying as much).

I think assists are ... weird, and defense isn't quantified very well (or "hardly at all").
The reason I stick with it in spite of those two issues is that there is not yet any perfect catch-all stat that I'm aware of (given the natures of economic trade-offs and human behavior, there likely won't ever be).

Your points regarding on/off pps (particularly if it involves the same opponents) is one I believe I've also made before.

So per the "lazy" bit. At this point, I don't know how I'd significantly monetize the adjustments I'd want to make (which would involve a LOT of work), so I go with the simpler option on the closest known, and typically work in other variables as I see fit / am bored

And I'm hoping the Deandre Jordan MVP thing was sarcasm....

I plea the 5th (for now).

[bondom34]

Oh, and to address some other points, having a statistical degree doesn't mean you know how a stat applies to basketball, that's part of the issue with Berri/Alvarez. They're unwilling to move on advice from others who have really good input over on APBR and elsewhere.

No doubt. I run into this problem with people in general, even (perhaps "especially", even) within math and science fields, a LOT.

I loved this quote, "The macroscopic model can show you the entire galaxy but blurred and hazy; the microscopic will focus on a few spiral arms of the galaxy with great detail and clarity. You can’t assume you know exactly what the rest of the galaxy looks like from only a small section."

The shortcomings you listed are actually issues I have always had with the stat myself (there are probably quotes to be dug up of me saying as much).

I think assists are ... weird, and defense isn't quantified very well (or "hardly at all").
The reason I stick with it in spite of those two issues is that there is not yet any perfect catch-all stat that I'm aware of (given the natures of economic trade-offs and human behavior, there likely won't ever be).

Your points regarding on/off pps (particularly if it involves the same opponents) is one I believe I've also made before.

So per the "lazy" bit. At this point, I don't know how I'd significantly monetize the adjustments I'd want to make (which would involve a LOT of work), so I go with the simpler option on the closest known, and typically work in other variables as I see fit / am bored

And I'm hoping the Deandre Jordan MVP thing was sarcasm....

I plea the 5th (for now).

Yeah, I was turned of WP a while back and never could really follow the logic of it, but on/off to me makes some inherent sense as I can see what effect a guy has on the rest of the lineup. Assists and D are tough as well (though I think PM is the one thing to best quantify defense). And on/off is another question entirely. All stats have flaws, but some are bigger than other and when nearly every metric is saying the exact same thing (one player being better than another by a pretty good margin) and when I've seen guys play and one is still clearly better, its just off to say the opposite. Either way for the Westbrook thing he's clearly a better defender by any standard, Lawson is really limited there. That said, there's still really not a comparison offensively as well when Westbrook just had one of the better offensive PG seasons. That said there seems to be a fascinating non basketball point to the posts here as well which is kind of nice too .

Edit:

As for catchalls, I really like a mix of maybe 2/3 things:

RPM, BPM, and WS

Each has its own flaws, but when taken together they cover relatively well. RPM is my main defensive metric, WS and BPM have issues there but I try to not gauge them for D. They do however show offense reasonably well.

Those 3 plus some on/off data are pretty solid. And FWIW, all of it says the same for the PGs in question.

Second edit: So I finally looked at WP/48 and the guy ranked after Lawson sort of shows why I'm not a fan. Prigs is not in any way, even on a per minute basis, nearly as good as Lawson or anyone above him (or frankly nearly anywhere below him). If that's not some proof, I don't know what is .

[moofs]

Second edit: So I finally looked at WP/48 and the guy ranked after Lawson sort of shows why I'm not a fan. Prigs is not in any way, even on a per minute basis, nearly as good as Lawson or anyone above him (or frankly nearly anywhere below him). If that's not some proof, I don't know what is .

GAME 7, BABY!!!
There's the pudding.

Also the only time we were able to pull ahead of the Clippers was when that accursed oversized manchild sat on his rump.

[bondom34]

Second edit: So I finally looked at WP/48 and the guy ranked after Lawson sort of shows why I'm not a fan. Prigs is not in any way, even on a per minute basis, nearly as good as Lawson or anyone above him (or frankly nearly anywhere below him). If that's not some proof, I don't know what is .

GAME 7, BABY!!!
There's the pudding.

[moofs]

As for catchalls, I really like a mix of maybe 2/3 things:
RPM, BPM, and WS
Each has its own flaws, but when taken together they cover relatively well. RPM is my main defensive metric, WS and BPM have issues there but I try to not gauge them for D. They do however show offense reasonably well.

My ideal stat would be:
1. Net shooting impact%
2. Net shot qty (necessary to gauge impact of 1)
3. Net posessions

The obvious question, of course, is HOW THE HELL?
It'd be totally awesome, though.

[bondom34]

As for catchalls, I really like a mix of maybe 2/3 things:
RPM, BPM, and WS
Each has its own flaws, but when taken together they cover relatively well. RPM is my main defensive metric, WS and BPM have issues there but I try to not gauge them for D. They do however show offense reasonably well.

My ideal stat would be:
1. Net shooting impact%
2. Net shot qty (necessary to gauge impact of 1)
3. Net posessions

The obvious question, of course, is HOW THE HELL?
It'd be totally awesome, though.

I think the shooting impact would be covered by the wowy stuff though? And quantity and possessions are kept already, unless I'm misunderstanding.


The issue is there's still much more to offense than shooting (rebounding, passing, getting guys open), and it doesn't take D into account at all, which is why the plus/minus stuff makes some sense to be used in conjunction to me. As for shooting impact, I'd add this:

http://nyloncalculus.com/2015/07/25/scoring-efficiency-by-touch-frequency/

What you can see is that, true shooting percentages aside, Hassan Whiteside and Russell Westbrook are both insanely efficient given how often they are involved in their teams’ offenses. Each player leads their grouping in points per half court touch differential, and by a huge margin. On the other end of the spectrum you can see players like Andre Roberson and Joakim Noah, in the low usage group, and Dante Exum and Matthew Dellavedova, in the high usage group, who scored far less frequently than we would expect.

That and some nbawowy helps which is where I got the stuff I mentioned above. The stuff they do w/ Synergy now is pretty sweet though.

[moofs]

My ideal stat would be:
1. Net shooting impact%
2. Net shot qty (necessary to gauge impact of 1)
3. Net posessions

The obvious question, of course, is HOW THE HELL?
It'd be totally awesome, though.

I think the shooting impact would be covered by the wowy stuff though? And quantity and possessions are kept already, unless I'm misunderstanding.

w.r.t. assists being sketchy, pps impact, opposing player ts%, assisted defense, etc.

The issue is there's still much more to offense than shooting (rebounding, passing, getting guys open), and it doesn't take D into account at all, which is why the plus/minus stuff makes some sense to be used in conjunction to me. As for shooting impact, I'd add this:

See above
rebounding (net possessions - possessions retained, possessions acquired, possessions lost, etc)
passing (net shooting %/net shot qty - how often and to what degree are the shooting % of others improved, and how unique is it to the player [if possible, i.e. vo{team}rp/vo{average|median}rp])
getting guys open (net shooting %/net shot qty - improving pps, shot qty, reducing possessions lost)
and it doesn't take D into account at all (net shooting % / net possessions - reducing TS% of opposing player/team, removing possessions from opposing team, scaring the f**ing dogs**t out of opposing players trying to attain a better TS% probability shot, etc.)

I'm just grouping it a bit differently.

The "HOW THE HELL?" came from accounting for goofy things like, e.g., a prime KG flying all over the court to help out just about everyone on defense, accounting for it being basically impossible to score inside 6 feet with Mark Eaton in the middle, Ben Wallace/Deandre Jordan (71%!!) basically only attempting to score on dunks, etc.
Among other "little" technical difficulties...

http://nyloncalculus.com/2015/07/25/scoring-efficiency-by-touch-frequency/

[bondom34]

My ideal stat would be:
1. Net shooting impact%
2. Net shot qty (necessary to gauge impact of 1)
3. Net posessions

The obvious question, of course, is HOW THE HELL?
It'd be totally awesome, though.

I think the shooting impact would be covered by the wowy stuff though? And quantity and possessions are kept already, unless I'm misunderstanding.

w.r.t. assists being sketchy, pps impact, opposing player ts%, assisted defense, etc.

The issue is there's still much more to offense than shooting (rebounding, passing, getting guys open), and it doesn't take D into account at all, which is why the plus/minus stuff makes some sense to be used in conjunction to me. As for shooting impact, I'd add this:

See above
rebounding (possessions)
passing (net shooting %/net shot qty)
getting guys open (net shooting %/net shot qty)
and it doesn't take D into account at all (net shooting % / net possessions)

I'm just grouping it a bit differently.

The "HOW THE HELL?" came from accounting for goofy things like, e.g., a prime KG flying all over the court to help out just about everyone on defense, accounting for it being basically impossible to score inside 6 feet with Mark Eaton in the middle, Ben Wallace/Deandre Jordan (71%!!) basically only attempting to score on dunks, etc.
Among other "little" technical difficulties...

http://nyloncalculus.com/2015/07/25/scoring-efficiency-by-touch-frequency/

Ah, sort of got it now, but that's partially where plus/minus is attempting to account for things, just indirectly. Its not perfect, and I'm not its biggest believer, but if you use it with the box score and some context you get a pretty good idea. In general its served as the best baseline metric for me to go with and then use the rest from there.

[moofs]

(Added details - you respond too quickly for my typical edit window allowance )

[bondom34]

Adjusted plus minus is attempting to....at least to show how a team performs when a guy is on/off, which ultimately is the goal. Outscore the opponent. Doesn't really matter whos got more assists or boards or a better TS, you score more and win. Gotta be quicker to edit I'd recommend the PC board here, some great stats discussion.

[Nebula1]

Anybody know when he gets to Houston?

[MaxRider]

Anybody know when he gets to Houston?

no idea
he is in the middle of his DUI rehab in California
i think he has to attend it daily
so i don't think he can leave anytime soon

[inquisitive]

Anybody know when he gets to Houston?

mid August or so...then he will check in with John Lucas i think.

[inquisitive]

Lawson tearing it up....can't wait to see how he gets into the paint

[youtube]https://www.youtube.com/watch?v=ODPgg3dopg0[/youtube]

[inquisitive]

lawson out of rehab?

[Nebula1]

lawson out of rehab?

This is quietly going down as the best offseason deal sans LMA and Monroe.

1. Aldridge to Spurs
2. Monroe to Bucks
3. Lawson to Rockets
4. Jordan to Clippers


I can't wait.