Breaking -- The human race has waived Josh Smith: 2014-15 regular season thread part 2

right, i guess a better way to put it is, it sorta baffles me (although maybe it shouldn't) that various nba writers spend a lot of time talking about him beyond "he is bad" and try to pitch this "he is bad but also compelling" argument, because all things being equal we should not be interested in a player who exemplifies constant mediocrity and frequent shittiness coupled with a complete and utter denial of said shittiness

like remember when a not insignificant number of people were saying that JR would prob be more damaging to cleveland than waiters would be to OKC, and that JR was the shit sandwich cleve needed to eat to get iman shumpert? i didn't understand that and it ended up being markedly not true

it's not that waiters is "bad but compelling" -- he's bad but important. he has been a pivotal role player on two of the most important teams in the nba this season!

think about why no one ever talks about derrick williams

i talk abt derrick williams nearly constantly

While working out this week, I joined an interesting conversation about analytics with some other applied mathematicians and statisticians. It mostly concerned Kirk Goldsberry’s presentation during the 2014 Joint Statistical Meetings about advances in basketball analytics. Everyone agreed it was embarrassing. He didn’t show recorded observations (“proprietary”) or predictive models (“proprietary”). Moreover, he couldn’t identify one accurate prediction. Why do basketball bloggers give people shit about being skeptical about the recent analytics trend? It’s a completely reasonable view. Ergh.

proprietary work is a problem across sports analytics i think

e.g. this tweet doesn’t even make sense:

https://twitter.com/billoram/status/570764951578488832

(However, he does set himself up for a joke about imaginary numbers.)

proprietary work is a problem across sports analytics i think

Fine. But not showing predictions? Finance is secretive too, but it’s easy to know when a model works and when it doesn’t.

Here’s some related trivia that I love that few people know: Donald Knuth started modeling Case Institute of Technology basketball games in the fifties (thirty years before Bill James’ Abstract).

i'm not familiar w/ the presentation you're talking about (the rebounding one?) nor do i know remotely nearly enough about statistics as a practice to argue about it, so i will cede the floor

`Eye in the Sky: The Player Tracking Revolution in Sports Analytics`

i'm sure you could reasonably argue that some of the methods being used in NBA statistical analysis are in need of improvement - admittedly idk enough about them to argue on a micro scale about specific concepts - but the macro idea that analytics are bad in the NBA full stop and should be ignored or given minimal credence, which is what someone like byron scott is saying, is problematic imo

like scott isn't skeptical because he thinks the methods being used don't work and could if improved, he seems to literally not believe that statistical analysis beyond the most basic numbers - PPG, FG percentage, etc. - has any place in the sport whatsoever

Why is that problematic? A model that’s too faulty or too complicated confuses otherwise reasonable judgement e.g. http://archive.wired.com/techbiz/it/magazine/17-03/wp_quant?currentPage=all.

like scott isn't skeptical because he thinks the methods being used don't work and could if improved, he seems to literally not believe that statistical analysis beyond the most basic numbers - PPG, FG percentage, etc. - has any place in the sport whatsoever

Yeah. I understand that. But is there any evidence that any non-traditional model has motivated wins (or any measure of success)? Honestly, I don’t know and would like to see some (i.e. I like basketball and statistics).

most of the skepticism analytics does seem to be pretty ignorant and old school, some of that is prob ppl who know a lot about basketball being dismissive rather than bothering to make an actual argument which is fine and understandable, tho i think the argument is basically over at this point in that most teams use fancy analytics to some extent

i do suspect tho that a lot of the new analytics are bad to varying extents for various reasons, one only has to think of sacramentos stats dudes making that presentation in favor of drafting nick stauskas to understand in how many ways this sort of thing can go horribly wrong

the most interesting thing to me is to what extent its very hard to figure out how good an individual player is because he exists within the context of a team which is comprised of not just other players but also a includes a coaching scheme, feel like this is super fundamental and very underrated by numbers ppl who are maybe too influenced by the example of baseball which is essentially a series of one on one matchups, im not sure that analytics will ever fully be able to control for team which is p cool imo

lagoon otm, like analytics aren't THE tool, they're A tool, they can be incorrectly deployed (sacramento, potentially philly). the constantly used case study for hoops analytics success is the houston rockets but that model obviously doesn't work without its centerpiece player, harden, and morey has said as much.

i think it's useful to remind ourselves that "analytics" is not this one monolithic thing and no one to my knowledge is presenting or pursuing a unified theory of ball. it strikes me as very ignorant to deny that analytics has any purpose esp. in the realm of suggesting answers to more focused questions (to take a very simple example "where are the most efficient places to shoot from on the floor?")

I hate this game. /hopsoffbandwagon

imo, analytics should be used as a corrective to judgment, not as the basis for judgment.

iow, if you think that player X has a certain skill set, then analytics can give you hard numbers to put against that skill set. if the numbers don't agree with your basic impression, you need to go back and rethink your judgment. but if you base your judgment solely on the numbers, then there's nothing you can use to correct whatever the numbers tell you.

pacers uniform league pass alert

https://pbs.twimg.com/media/B-4QhUVVIAAyPML.jpg:large

http://www.nba.com/bulls/90s-night

James recorded 40 points, five rebounds and five assists in a game for the 41st time in his career, breaking a tie with Elgin Baylor for fourth most such games in NBA history. According to the Elias Sports Bureau, the three players ahead of James are Oscar Robertson (71), Michael Jordan (62) and Wilt Chamberlain (61).

lol nice xp

keep seeing headlines abt rose on 4-6 week rehab schedule but he's having the mensicus removed so what are the chances that he's going to be able to play productively again? my dad had his meniscus removed in the 80s and if he pivots wrong his knee just kind of pops open, so my only pt of ref is informing a much more difficult road back to balling

Depends how bad it is. Fully removed or just the damaged portion removed. I've had damaged meniscus removal and was playing within 6 weeks and have felt nothing since.

wade had his meniscus removed when he was like in college and had a hof pro career but says he regrets it now. rose could *theoretically* be better than he's been for awhile w/o it but the drop-off as he ages is steeper

but this prob comes down to an ever body is different situation and it's probably a lock that rose with have the yips for quite awhile after coming back

rose just had a partial removal so its prob not that bad, chris paul had the same thing and didnt miss a step

yips r another story tho

hope he doesn't end up with a case of that rondo and shy from any contact (tho that's already sort of been the haps this season)

ya for sure

feel like he was starting to get back in the zone tho

Man pat Riley just can't quit that Beasley

Gsw taking out their frustrations from last night on the raps

Bogut nice touch with the jump hook.

Is Iguodala an injury scratch?

lebron not playing v pacers tonight must've needed a booster shot

Warriors ginna start strategically sitting guys Kerr said after the all star break

xp Bron got nothing to show the pacers, best he sits. Their bench shld be able to take this one

i love isaiah thomas

Man pat Riley just can't quit that Beasley

otm

i love goran's game so much. so many `pretty good plays`. 7 layups.

celtics update from my celtics-fan girl who is at the game:

https://www.evernote.com/shard/s3/sh/1a9ee6d2-a7ff-42cc-986d-51cb7605a1c8/4d4907cb7b4efd2f340786f89f7dcbc3/res/e06a073f-1a9e-400b-9c01-7e6b7b997064/skitch.png

try that again:

They're not playing so well!
Even the jr celtics are lousy

killing the hornets in the 4th rn

ya. flipping back and forth with mia/no. isaiah thomas is something. forgot celtics picked up jerebko too.

hearing tommy attempt to say jonas jerebko is a blessing

i bet super cool beas and mr. riely occasionally listen to some really neat records together

Brian Robb ‏@CelticsHub 5m5 minutes ago
Isaiah Thomas leads the NBA with 6.2 points per game in the fourth quarter. He has 12 points already tonight in the frame.

worst shot of all time?

https://vine.co/v/O2jxmtiJqhU

Somehow the worst shot and pass simultaneously