On February 21 2012 04:04 cLutZ wrote: OKC just seems so incomplete. 1 Great player, 2 very good players, and a cadre of incomplete players. Could they win, yes, but it will be a struggle.
are we watching the same team?
1. westbrook is great b. ibaka, harden, perkins should all fulfill your very good (how many centers would you honestly take over perkins its a position in termoil) III. they are deep (but a little less without maynor TT)
On February 20 2012 02:08 MassHysteria wrote: @ x2fst:: The problem is you are trying to label TS% as some kind of absolute "statistic" to determine how good a player is, or trying to say that TS% is a much more complete formula (and it is)and that is not what was in context. Statistics are used in basketball to help understand occurrences/situations, not be the absolute truth.
So Ace isn't saying FG% is a better statistic than TS% for determining overall shooting impact in a statistical sense (b/c it isn't), he is saying it is more useful in comparing players we have in mind and want to compare them. If I say " ya West shot 45% from the field while taking 20 shots, shot 5-6 from the line, while taking no 3-point shots", that description would be much more useful in comparing him to player x than saying " ya West TS% was 52".
on the first bit, this is not what i said at all, i have literally no idea where how you came to that incredibly wrong conclusion (well i do actually, you assumed it cos it suits your viewpt to believe it). all i said was that ts%>fg% as a stat, which is true in almost every context
on the 2nd paragraph, it's wrong in that sense too. ts% is clearly better than fg% in comparing 2 similar players who get similar offensive opportunities. your example is precisely the reason we use stats, lol at going through gamelogs for an entire seasons worth of data (which is what you have to do to analyse bball in any useful way, and the boxscore is only the beginning), for that production saying that he had 52ts is far more useful than citing 45fg% anyway, and i've already provided an example of why this is the case
On February 20 2012 02:20 Jibba wrote: Yes, the idea of compounding everything into a single statistic is stupid, especially when it hides important nuances. Fucking Hollinger & co. They're trying to follow baseball's lead, except baseball did it with the help of actual statisticians. It's like NA players copying Korean build orders, without understanding why.
there's a line between what's actually happening and trying to boil it down to single digit stats, and obviously bball is too complex for that to be an ideal approach. but humans are incredibly flawed quanititative analysts unless they have the aid of computers are are trained, if i had to choose a guy to bet games for me, and i had a panel of armchair analysts vs a guy like hollinger, i'd take hollinger every time with absolute confidence. his approach may be flawed but it's 100x more accurate the random nonsense that untrained human brains produce when it comes down to actual performance, people need to realise that every time they criticise. and i'm no big fan of hollinger or his methods
tbh i find it really surprising that a forum full of videogame players (aka bunch of nerds basically, i'm one too clearly) isn't more interested in analysing this stuff from a quantitative standpoint. i mean the fact that fg% is inferior to ts% for analysis is so unbelievably first level it's shocking to me that anyone doesn't immediately recognise why
tbh i find it really surprising that a forum full of videogame players (aka bunch of nerds basically, i'm one too clearly) isn't more interested in analysing this stuff from a quantitative standpoint. i mean the fact that fg% is inferior to ts% for analysis is so unbelievably first level it's shocking to me that anyone doesn't immediately recognise why
Why should we? Quantitative analysis often ruins the experience of being a fan.
On February 22 2012 09:57 x2fst wrote: tbh i find it really surprising that a forum full of videogame players (aka bunch of nerds basically, i'm one too clearly) isn't more interested in analysing this stuff from a quantitative standpoint. i mean the fact that fg% is inferior to ts% for analysis is so unbelievably first level it's shocking to me that anyone doesn't immediately recognise why
It's not real quantitative analysis, and it's no more biased than what any of us come up with. There exists real quantitative analysis, such as what GMs and players have access to, and then there exists showy numbers that glare selection biases and show no understanding of statistical methods. That's what PER is. It's completely incomparable to things like OBP.
For a forum full of nerds, PER is even more insulting because it shows how little Hollinger understands, but it's fed to non-nerds as if it has merit because it's in number format.
In 2010, Charlie Villanueva was the Piston's PER leader. That's truly all the information you need to disbelieve its usefulness. If you just use (points + rebounds + assists)/playing time, you'll come up with a similarly useful rating (ie terrible.)
On February 20 2012 02:08 MassHysteria wrote: @ x2fst:: The problem is you are trying to label TS% as some kind of absolute "statistic" to determine how good a player is, or trying to say that TS% is a much more complete formula (and it is)and that is not what was in context. Statistics are used in basketball to help understand occurrences/situations, not be the absolute truth.
So Ace isn't saying FG% is a better statistic than TS% for determining overall shooting impact in a statistical sense (b/c it isn't), he is saying it is more useful in comparing players we have in mind and want to compare them. If I say " ya West shot 45% from the field while taking 20 shots, shot 5-6 from the line, while taking no 3-point shots", that description would be much more useful in comparing him to player x than saying " ya West TS% was 52".
on the first bit, this is not what i said at all, i have literally no idea where how you came to that incredibly wrong conclusion (well i do actually, you assumed it cos it suits your viewpt to believe it). all i said was that ts%>fg% as a stat, which is true in almost every context
on the 2nd paragraph, it's wrong in that sense too. ts% is clearly better than fg% in comparing 2 similar players who get similar offensive opportunities. your example is precisely the reason we use stats, lol at going through gamelogs for an entire seasons worth of data (which is what you have to do to analyse bball in any useful way, and the boxscore is only the beginning), for that production saying that he had 52ts is far more useful than citing 45fg% anyway, and i've already provided an example of why this is the case
Yes I actually agree with you that ts%>fg% as a shooting stat, which is what I meant to say. Apologies for the wrong conclusion, I take it back.
As for 2nd paragraph, I must say that using it as an entire seasons worth of data gives us a nice snapshot on the player and that saying he had 52ts is obviously more useful that saying 45fg% (for the season). Specially only if we were using that number as our single source of information. Knowing how basketball really works though, I would still want see the FG and FT and 3PT stats/numbers(FG and FT attempts, etc.) anyways, because it just provides a better sense of what their overall game is. Which kind of makes citing TS a little obsolete in certain cases when comparing the "game" of certain players.
On February 22 2012 09:57 x2fst wrote: tbh i find it really surprising that a forum full of videogame players (aka bunch of nerds basically, i'm one too clearly) isn't more interested in analysing this stuff from a quantitative standpoint. i mean the fact that fg% is inferior to ts% for analysis is so unbelievably first level it's shocking to me that anyone doesn't immediately recognise why
It's not real quantitative analysis, and it's no more biased than what any of us come up with. There exists real quantitative analysis, such as what GMs and players have access to, and then there exists showy numbers that glare selection biases and show no understanding of statistical methods. That's what PER is. It's completely incomparable to things like OBP.
For a forum full of nerds, PER is even more insulting because it shows how little Hollinger understands, but it's fed to non-nerds as if it has merit because it's in number format.
In 2010, Charlie Villanueva was the Piston's PER leader. That's truly all the information you need to disbelieve its usefulness. If you just use (points + rebounds + assists)/playing time, you'll come up with a similarly useful rating (ie terrible.)
I know Duncan, Parker, Ginobili and Splitter aren't playing but still disappointed by such a blow out... I cant watch the games but Portland scored 90 points with 4 minutes left in the 3rd, what are they doing???
Over 100 points in 3 quarters... wow. Pop is not going to be happy. They're supposed to be one of the deeper teams in the league, no way they should allow that.
On February 22 2012 13:47 MilesTeg wrote: Over 100 points in 3 quarters... wow. Pop is not going to be happy. They're supposed to be one of the deeper teams in the league, no way they should allow that.
Big interview on NBA.com with Commissioner David Stern by David Aldridge on topics like the franchises like Kings & Hornets, Linsanity, the CP3 trade, D12 trade talk (and FAs in general), possible Seattle franchise, the new CBA & how long he'll stay on in charge of the NBA.
EPT
