EPTHe said he was gonna post his 300-line explanation. He just copied his post #250! Post #250 and #259 are the same. What the fuck is going on?
NBA Offseason 2017 - Page 14
| Forum Index > Sports |
|
Twinkle Toes
United States3605 Posts
He said he was gonna post his 300-line explanation. He just copied his post #250! Post #250 and #259 are the same. What the fuck is going on? | ||
|
JimmyJRaynor
Canada16711 Posts
On June 27 2017 02:45 JimmiC wrote: I will make bold projections and speak down to anyone who watches the sports or has participated in them for a lack of advanced understanding) i don't think this moves the discussion forward. quantity and quality of observations matter. + Show Spoiler + i've already stated that my Raptors projections over the long term turn out to be better than others in this thread partly because i watch the team more than any one in the thread. that is not a brag. it is difficult for 1 person to follow all 30 teams as closely as i can follow 1 cherry picked team that plays their games close to where i live. if someone's favourite team is phoenix and they watch all their games and live in Phoenix their projections about the Suns will be better than mine and far more detailed. one of my co-workers was born and raised in chicago and went to northwestern. he goes back to chicago 10 weeks a year. when i want to know WTF is going on with the bulls. i ask him.1 does any one have time to follow all 30 teams? On June 27 2017 02:53 Twinkle Toes wrote: Is this guy fucking serious? He said he was gonna post his 300-line explanation. He just copied his post #250! Post #250 and #259 are the same. What the fuck is going on? i accommodated JimmiC's concern that i was removing only 1 line of his entire post. now his entire post is there. its coherent logical rebuttals to the issues discussed. please note in the previous post i included a "Jargon Table" covering what "R" , "COM Objects" and "Visual Foxpro" do. if you want me to go into greater depth and detail i can do that after i'm done working today. following the development and improvement of the Adjusted Plus/Minus stat is a bit of a hobby of mine. | ||
|
xwoGworwaTsx
United States984 Posts
In one post he is championing APM and in another he is saying it is useless. Boasts about his use of stats and then say that he is sick of it. He is all over the place and I think he is being deliberately obscure in order to hide his ignorance. I don't want to waste anyone's time, so I'll be direct - JimmyJRaynor, I challenge you to a debate. Pls write in a clear, organized and specific language a statement that characterizes your understanding of sports analytics as it is used in the NBA so I can respond properly. | ||
|
JimmyJRaynor
Canada16711 Posts
i do not agree with (1) or (2). However, (1) +/-, APM , and xRAPM are all useless (2) advanced stats are discussed too much and i'm sick of them. if you combine (1) and (2) then there is no point in moving forward. | ||
|
JimmiC
Canada22817 Posts
| ||
|
nVme
952 Posts
| ||
|
Twinkle Toes
United States3605 Posts
On June 27 2017 03:34 JimmiC wrote: I admit it was amusing to watch him crash and burn. But no one is going to get the satisfaction we desire because he will just keep being him. Which is never admitting a mistake, moving the goal posts and keeping his comments so broad and varied that at some later date he can copy paste what ever worked out and claim he was right all along. I had tons of fun the last few says when he kept posting and everyone just ignored his shit. That is probably the best solution to as he says "keep the discussion moving". When we ignore him and post there is often some good discussion and you learn some shit and have a good time. When we engage him it becomes a shit storm that he seems to thrive/love. Yes, I'm done giving him attention. Although I'd be intently watching him getting beaten up by another person. ION. Draymond sure DPOY, RW sure MVP. | ||
|
JimmiC
Canada22817 Posts
| ||
|
JimmyJRaynor
Canada16711 Posts
Dan Rosenbaum and Engelmann do solid analytics work for basketball and i've included why that is the case in my previous posts. back to my stuff about the Raptors which got buried. http://fansided.com/2017/05/10/nylon-calculus-explaining-raptors-playoff-disasters/ Why do the Raptors play so badly in the playoffs. " the Raptors don’t fail when Lowry’s on the court in the playoffs; they fail when someone else is." "Using Jonas Valanciunas as a constant, the lineups are best without DeRozan entirely, and the ones with just him are worse offensively and are a disaster on defense. " | ||
|
ZenithM
France15952 Posts
This is explained here: http://fansided.com/2014/09/25/glossary-plus-minus-adjusted-plus-minus/ (original work from some math guys, you can find their OP/paper on the internet). What you should have done, JimmyJ, is to summarize this shit for us all to follow your anulytics ramblings. In fact, let me do it for you (the article still goes into more detail than necessary). We all know what basic +/- is (what JJ calls Vanilla +/-). You count the point differential with and without the player on the court. Problem is, in the NBA, teams tend to play certain sets of players consistently together, such that if a scrub is exclusively paired with awesome players, he will still have a great +/-. This is mostly why 5 of GSW's guys comfortably sit in the top 10 of this metric (or something absurd like that). Adjusted +/-, or APM, basically assesses the plus/minus of a player if you removed the personal impact of the 4 other dudes playing with him at all times. The above article has a nice low-dimension example to start with. You play 2v2 street ball, and you have 3 guys on your team (1 on the bench). You compute the collective plus/minus of each possible lineup as they perform together on the court, to get something like this: P1 + P2 = +10 P1 + P3 = +3 P2 + P3 = -1 The vanilla +/- of these 3 guys are all positive (13, 9, 2), so if you only looked at this number the 3 players would look pretty good. But what you'd prefer is the personal contribution from each player to each of the lineups they take part in. That's basically APM. If you look above, this is actually a system of 3 linear equations with 3 variables P1, P2, P3 which are this APM we're looking for. What's left for you is just to solve the system. With this many variables it's still high-school level math, so something like Gaussian elimination will make short work of it: P1 = 7, P2 = 3, P3 = -4 (P3 is not looking so good after all) But when you have 5 players on the floor, and all the potential lineups they form (imagine how many lineups there are over the course of a game, or even a season), AND if you actually take into account their different defenders as well (so 10 players for one game, but all the other players from the other teams over a season), the system quickly becomes unmanageable to solve using your usual "optimal solution" algorithms, and we enter the realm of "linear regression" (which in theory only approximates the solution). We happen to know several good methods to solve linear regression, capable of handling all of these player combinations, but the detail of those doesn't really matter. Overall this is indeed an interesting metric but I love how JimmyJ makes it sound fucking complicated, likely on purpose to hide his real understanding of these statistics. APM is in fact pretty intuitive. | ||
|
JimmyJRaynor
Canada16711 Posts
On June 27 2017 04:15 ZenithM wrote: Overall this is indeed an interesting metric but I love how JimmyJ makes it sound fucking complicated, likely on purpose to hide his real understanding of these statistics. APM is in fact pretty intuitive. i posted your same explanation previously from Nylon Calculus. its a good explanation. would you like the link of that previous TL.Net post? i also stated the only real "hocus pocus" in APM was the regression step. then someone disagreed with that comment. i can provide either a short synopsis or a thorough analysis. any how, i think its a useful stat .. i basically agree with ur perspective on how adjusted plus/minus works. | ||
|
andrewlt
United States7702 Posts
On June 27 2017 03:53 JimmiC wrote: Did you see some gave Isiah Thomas a vote for DPOY, clearly a troll but that guy should lose his vote Rolf. Yes watching him get roasted is fun but maybe we could just vote him off the island like in the hockey thread. Who votes for these awards this year? Still the media? They changed all star voting this year to add a players' vote but they didn't take it seriously. At least the media takes it seriously. Many players and ex-players complain frequently that they know more than the media does but the amount of joke votes that players submitted this year is ridiculous. | ||
|
JimmiC
Canada22817 Posts
| ||
|
JimmyJRaynor
Canada16711 Posts
or the all NBA defensive teams... o ya, Milwaukee is in the east. | ||
|
Scarecrow
Korea (South)9172 Posts
On June 22 2017 03:18 cLutZ wrote: its not like Danny Green, Pau Gasol, and Patty Mills are some sick defenders . | ||
|
ZenithM
France15952 Posts
| ||
|
xwoGworwaTsx
United States984 Posts
All you did here was "do you want me to post my explanation here? its long" or "i posted it earlier" or "yeah o said that". All bullshit. Thanks ZenithM. That JimmyJRaynor fails to even discuss elementary-level stat concepts is proof that he knows nothing about this at all. Then add all the dodging, misquoting, gross generalizations, and hiding in jargon. then it becomes clear that he is the worst kind of ignorant person - one who pretends that he is an expert and talks down others. I was reluctant to really push the issue after the way Twinkle Toes destroyed JimmyJRaynor after he bragged about watching games live. You could see it on page 13 here, post 245. I mean seriously who brags about this thing? And he retreats to his hole like a filthy sewer rat after he is called out on it. No reply to that until now. But I have to call him out too since it is clear he is bluffing. So, my challenge still stands until the award shows start - JimmyJRaynor debate me. Since you have trouble providing a summary of your knowledge of this subject matter, let me help you by making you answer this relevant, specific, on-point questions IN YOUR OWN WORDS - 1. What is the absolute measurement to a players value? 2. How do you account for secondary factors in the computation of stints? 3. Why is your version of APM better than all the other variations? 4. In your version of APM, what regression ridges are prioritized, and why are they being prioritized over others? 5. Could your version of APM be wrong or insufficient to the point of being inaccurate? There. If you do not answer these questions, everyone here will be happy if you admit you are wrong, or at least never bring up sports analytics ever again/ | ||
|
JimmyJRaynor
Canada16711 Posts
On June 27 2017 08:57 xwoGworwaTsx wrote: 3. Why is your version of APM better than all the other variations? its not. never said it was. i cleared an apparent misunderstanding you had about 3 stats being useless. I'll say it again +/-, APM and xRAPM are valuable. Some view them as useless. I think they have value and i've used them many times. guaranteed Engelmann's xRAPM ( ESPN Real Plus Minus) is light years ahead of the hobby level stuff i did in 2007 using a 2004 article published by Rosenbaum. here is the 2004 article i used as the basis for the APM analysis i did of NHL players 10 years ago. http://www.82games.com/comm30.htm some of the code is in Visual Foxpro, some of it is in "R". i'll dig through it when i have time. I'm not even sure if its Visual Foxpro 8 or 9. On June 27 2017 08:57 xwoGworwaTsx wrote: 5. Could your version of APM be wrong or insufficient to the point of being inaccurate? i do not think it is "wrong". however, its no where near as good as the stuff Engelmann, Rosenbaum and other Adjusted Plus/Minus analytics people are doing today. Its been 10 years. Guaranteed Rosenbaum had already improved on his methods before he published his stuff for the general public to examine in 2004. | ||
|
xwoGworwaTsx
United States984 Posts
TROLLOLOLLOLOL. | ||
|
xwoGworwaTsx
United States984 Posts
| ||
| ||
