J.R. Smith has literally been shanghaied away for the foreshortened 2011-2012 NBA season: rolling ILH sandbox thread

per was rated the most predictive of irl wins out of any stat in this study i read fwiw

average per of a team's players?

im sure it was weighted by minutes but yeah basically

any idea where I can read it?

i mean i dont really remember the specifics but it mustve been something like that - ill try to dig it up for you using deep googling - but just not right now

they concluded the best judgement of player quality was how many minutes they got, aka coaches knew better than any stat which players were good, but out of the stats per was the best

Paul Millsapp had the 5th best PER last year... There's cherry picking and then there's picking a cherry the size of a watermelon.

yeah its far from perfect, but i do think there are lots of players that it judges more accurately than the prevailing wisdom, obvs its p much useless as far as defense goes - and milsap is 29th on list im looking at fwiw

yeah I fucked up on basketball reference somehow

the worst part of PER is that it rewards taking more shots on a lower percentage (which to me means being LESS efficient), which is why it shits on a guy like Afflalo.

like kevin martin top 20 gtfo

well it considers usage which imo is necessary because there are guys who only take a few open threes a game, which are not looks a first scoring option on team will get, besides that tho its all abt efficiency, which is why martin is rated so high, he scores tons points on v few shots

but yeah it for sure over rates martin

I don't know how anyone could scroll through the top 200 PER list and not think "this stat is broken."

tyrus thomas is 51st and afflalo is 176th.

it's not an efficiency measure at all, it's an efficacy measure.

(and not a great one)

like why is tyrus 51st? Cuz when he's on the court a lot of shit happens. None of the shit that happens is efficient.

curry down

ugh his ankles are made of confetti.

His whole rep is based on stats! If you read Hollinger for actual game analysis ur REALLY doing it wrong.

― Matt Armstrong, Tuesday, December 20, 2011 11:08 PM (44 minutes ago)

the stats he uses/invents are an attempt to quantify what he sees with his own eyes - and his analysis of strategy/player skills is very good, you're fronting iud - in fact this is essentially what any statistic does!

as ice says PER is not a perfect stat but it is the best single stat that exists, which is the entire justification for its existence in the first place!

like when espn did that player rankings countdown he wrote a column saying why he thought rose had a better year than cp3 even though paul had a marginally better PER - and then followed it up with research into cp3's assist quality which he included in his player profile - the guy is not some robot deluded by his own genius, in fact he's totally self-aware and very funny which is another reason he is awesome

you have it backwards, his belief in PER leads him to have some really delusional player evaluations.

If he sees some guy who has a high PER and isn't getting playing time he absolutely rips into their GM/Coach (I think he's even done this with Tyrus!)

is not a perfect stat but it is the best single stat that exists
― k3vin k., Tuesday, December 20, 2011 10:57 PM (17 minutes ago) Bookmark Permalink

It is arguably the best single stat used by ESPN.com, sure. I'm sure the Mavs front office has 20 stats that blow it out of the fucking water.

not everyone is right all the time dude! he is the best basketball writer there is & a force for good in the universe!

It's not about a batting average, it's about a stat guru who designed a "player efficiency rating" that doesn't reward efficiency.

in what universe is Afflalo not one of the most efficient players? It's just bizarre that he kept that as the name. Bizarre and pretty dumb

he's not a bad writer per se and I actually agree with him a lot in terms of his opinions that aren't stat-driven. But his stats are junk.

btw apparently we're going to be seeing a "disgusting and graphic" photo of monta ellis soon.

http://www.ibabuzz.com/warriors/2011/12/20/monta-ellis-warriors-to-be-in-sexual-harrassment-suit/#respond

gee last year for example he played starters minutes but didn't score, rebound, or assist particularly well, when he decided to shoot he was pretty effective but when you're playing 34 minutes a game and averaging a 12.6/3.6/2.4 as a 6-5 guard you're pretty much an average player on offense when it comes to like, making your team score points which is what wins basketball games. he is a very 'efficient' player but given his low usage rate he's not an especially effective player - what you said upthread is otm in this case, lol semantics tho

I was going to say, he has a nice TS% but he's pretty average or even subpar in most ways.

i guess that first effective should be 'efficient'

PER probably slightly underrates afflalo, he's a good team player but doesn't have the ball in his hands enough to be anything more than a solid starter, you're prob overrating him

tho afflalo literally acknowledged his "low usage rate" in an interview a few months ago so if he can take some more shots without sacrificing too much efficiency he could be a pretty good player, which he is not rght now

what you said upthread is otm in this case, lol semantics tho

― k3vin k., Tuesday, December 20, 2011 11:38 PM (6 minutes ago) Bookmark Permalink

I think it's not too much to ask that when you build your rep around a stat you also know the meaning of the words in your stat.

especially if you're a writer.

are you done trolling now

you mean disagreeing with you?

Efficiency has a broader meaning in this context - if you had a guy who only took one shot a game but always made it, I don't think anyone would talk about his amazing efficiency.

Efficiency has a broader meaning in this context - if you had a guy who only took one shot a game but always made it, I don't think anyone would talk about his amazing efficiency.

― milo z, Tuesday, December 20, 2011 11:53 PM (25 seconds ago) Bookmark Permalink

sure they would!

they'd just beg him and his team to put him in that situation more.

efficiency is a lack of waste: turnovers, missed FGs, wasted possessions, poor fouls (arguably). PER rewards production far far more than this, which is why, for example, Dwight Howard is number 2. He's extremely productive, not extremely efficient.

Hollinger defines PER as a "player's per-minute productivity" - which isn't surprising, you can post amazing percentage stats but if you're not making use of them it's not really all that efficient. And players are certainly punished for being woefully inefficient.

You seem to equate efficiency pretty much entirely with TS%?

no, turnovers are just as important.

PER is absolutely a productivity stat, which is why its name is dumb.

not really, it's productivity per minute, which is certainly a suitable definition of efficiency

So TS% and turnovers. Which happen to be the only stats where Afflalo is outstanding (TS%) or good (Turnover Rate). He doesn't score a lot of points or create a lot of plays or excel (statistically) anywhere defensively. Which would, one might say, make him an efficient shooter but not particularly efficient anywhere else in his game.

but if the productivity includes a lot of waste, it's not efficient.

Monta you idiot.

What ever happened to the Monta for Iggy trade?

efficiency is a lack of waste: turnovers, missed FGs, wasted possessions, poor fouls (arguably). PER rewards production far far more than this, which is why, for example, Dwight Howard is number 2. He's extremely productive, not extremely efficient.

― Matt Armstrong, Wednesday, December 21, 2011 12:56 AM (3 minutes ago)

production is a key part of efficiency!

there is a balance, like no one would say that a corner shooter who plays 30 mins a game & shoots 50% from 3 day in and day out and does nothing else is a good player, he is literally a waste of space even though he scores very efficiently. similarly someone who takes 50 shots a game and averages 30 ppg on 24% shooting would not be a good use of playing time

for example to extrapolate reggie miller who posted insane TS% his whole career, even led the league a couple times, as a guard, was never in the top 15 or so i don't think in PER, which seems fair

if one player has a ratio of 20:1 in production to waste, he is more efficient than one who has a ratio of 200:50, even if the other player is much more productive.