PDA

View Full Version : Interesting stats



Julius Sour
03-13-2006, 02:23 AM
http://www.nba.com/pacers/news/plus_minus_0506.html

have a good day

Js

sweabs
03-13-2006, 02:30 AM
:laugh:

Plus-Minus are funny.

SoupIsGood
03-13-2006, 03:36 AM
The man that probably helps our team the most, on a per-night basis, is IMO granger, and he's something like last on that list.

Yup, +/- is funny. And screwy.

SoupIsGood
03-13-2006, 03:37 AM
(Larry would note that the top three are all white guys :eek:)

Lithfan
03-13-2006, 05:02 AM
I think this is most important stat that we have. Points and other things just don't matter. Most important thing is:
How did the team play when this particular player was on the floor?

And I think if you take the whole season stats it true.

Especially when its good for Saras ;)

No joking though.

PostArtestEra
03-13-2006, 05:10 AM
(Larry would note that the top three are all white guys :eek:)

Damnit, I was going to say that any statistic that puts 3 white guys as your best 3 players can't be accurate.

Rytas_Jega
03-13-2006, 06:09 AM
Exactly plus-minus statistics is worth nothing.

Wayne Gretzky, Robert Orr, Mario Lemieux, Chris Pronger, Peter Forsberg all worth nothing too.

Kestas
03-13-2006, 07:54 AM
http://www.nba.com/pacers/news/plus_minus_0506.html

have a good day

Js

:hmm:
I swear I did not read this before posting my reply to Able's comment in the other thread!

strange though..

Major Cold
03-13-2006, 10:34 AM
Okay plus minus means nothing...
Points for and against are the most important stats.

3Ball
03-13-2006, 11:37 AM
It certainly doesn't sound good when #1 and #3 are Ron Artest and Eddie Gill. What I'm still having trouble figuring out is how every single player on the team is below the team average. I mean, is that possible?

Raskolnikov
03-13-2006, 01:26 PM
It certainly doesn't sound good when #1 and #3 are Ron Artest and Eddie Gill. What I'm still having trouble figuring out is how every single player on the team is below the team average. I mean, is that possible?
Yes it is, but the term 'average' may be a little misleading here. Team average (row)/ Plus Minus (column) just gives the Pacers' total score difference (actually the team's plus-minus) from all games. Same goes for the categories minutes, games and last game. If you take that into account, then it's certainly possible, in fact it almost has to be like that.

The only two categories that really deserve the term 'average' are per 48 mins and per game because they are, actually, averaged. I also guess that, if you really want to drag some information out of the Plus Minus chart, you have to focus on those two categories.

As you can all see, Ron Artest leads them both.

:leaving:

Peck
03-13-2006, 02:15 PM
http://www.nba.com/pacers/news/plus_minus_0506.html

have a good day

Js


dude I don't care anything about the +/- stat.

All I care about is who is that in your avatar????:-o :-o :-o

Lithfan
03-13-2006, 02:20 PM
dude I don't care anything about the +/- stat.

All I care about is who is that in your avatar????:-o :-o :-o

I see what you mean. Thats one healthy lithuanian girl. Behind you Julius?

DeS
03-13-2006, 02:24 PM
Well, Plus/Minus is quite a usefull thing. Between us, Saras fans: when following it, You can easily track Saras playing PG / Saras playing SG intervals ;)

Since86
03-13-2006, 03:27 PM
Exactly plus-minus statistics is worth nothing.

Wayne Gretzky, Robert Orr, Mario Lemieux, Chris Pronger, Peter Forsberg all worth nothing too.

I don't know whether to laugh, or feel sorry for you.

You just put hockey's stats on par with basketball's......

Two very, very, very, very, very, very different sports.

Dr. Goldfoot
03-13-2006, 04:12 PM
The problem with +/- stats... the other four guys affect your number. Runi isn't the best player on the Pacers and Mel Mel isn't the worst. Ron Artest still has the highest +/- on the team. Eddie Gill has a postive number.


What if Scot Pollard comes into the game for 3 minutes...Jack hits 4 3's in a row... One off the inbound, one off an Foster offensive rebound on a Pollard airball, one off a steal by Granger and one off a Jack rebound outlet to Mel Mel who hits the trailing Jackson for a 3.


Explain to me how Pollard should recieve +12 on that sequence.

Kestas
03-13-2006, 04:15 PM
Explain to me how Pollard should recieve +12 on that sequence.

those stats are for 60 games, not for some sequence.

Moses
03-13-2006, 04:19 PM
Well, There you have it ladies and gentleman. I'll call Bird first thing in the morning and have Runi moved up to first string as soon as possible.

ChicagoJ
03-13-2006, 04:20 PM
Five-man plus/ minus: good

Individual plus/ minus: meaningless

A few seasons ago, I tried to keep a 5-man plus/ minus database, you can find the discussion in our archives (I think).

82games.com also keeps it; I believe they publish the data based on each team's top-ten 5-man lineups (in terms of minutes played) so its useful but still hard to make comparisons meaningful comparisons.

ChicagoJ
03-13-2006, 04:21 PM
those stats are for 60 games, not for some sequence.

His point is, why is that +12 for Pollard meaningful? (hint: it isn't). He only got the +12 because SJax was red hot and he wasn't on the bench.

Dr. Goldfoot
03-13-2006, 04:23 PM
right but assume Pollard gets hurt on the next play and he recieves a + 12 for the game. and that happens every game.... and after 60 games he would have a +720 doesn't seem right does it.

Anthem
03-13-2006, 05:06 PM
Try this: http://www.82games.com/0506/0506IND2.HTM

ChicagoJ
03-13-2006, 05:12 PM
We've had so much turmoil (injuries, idiots) that its hard to get anything out of those.

Our unit that has played the most minutes together includes a starter that wasn't even on the team at the beginning of calendar 2006 and has logged 213 minutess together.

The Nets, as just one example - their usual starting five has played together for over 700 minutes, and the Uncle Cliffy version also has more minutes than every Pacers combination.

SoupIsGood
03-13-2006, 05:25 PM
I hate those 5 man listings, because every time I look at Detroit's.

Nearly 1400 damn minutes logged under their best lineup. I hate them.

Lithfan
03-13-2006, 05:35 PM
Five-man plus/ minus: good

Individual plus/ minus: meaningless

A few seasons ago, I tried to keep a 5-man plus/ minus database, you can find the discussion in our archives (I think).

82games.com also keeps it; I believe they publish the data based on each team's top-ten 5-man lineups (in terms of minutes played) so its useful but still hard to make comparisons meaningful comparisons.

I have some knowledge about statistical analysis. Without too much thinking I don't see why the individual +- stats would benefit Saras. He is playing with the second squad and you are taking 60 games stats. So the randomal as well as absolute error margins here are insignificant. If there is an error, it would be absolute negative for Saras due to a second squad.

To the Pollard question - first he usually doesn't start, so he doesn't play with Jax. Moreover, if Jax does hit 4 frees it probably means Pollard made some decent screens. And finally if he didn't, this +12 thing will be the only plus he gets in 60 games.

So please give me a link, (I didn't find it by myself on PD) or some argument why individual +- stats are useless if you take a season results.

SoupIsGood
03-13-2006, 05:37 PM
So please give me a link, (I didn't find it by myself on PD) or some argument why individual +- stats are useless if you take a season results.


Here ya go


http://www.watchthefriggengames.com

Fool
03-13-2006, 05:39 PM
I hate those 5 man listings, because every time I look at Detroit's.

Nearly 1400 damn minutes logged under their best lineup. I hate them.

That fact isn't all good.

SoupIsGood
03-13-2006, 05:40 PM
That fact isn't all good.


Maybe not, but I'd gladly have that problem (the starters playing too much). :cool:

Lithfan
03-13-2006, 05:41 PM
Here ya go


http://www.watchthefriggengames.com

Thank you

www.dontbelieveIactuallyclickedit.ac.il (http://www.dontbelieveIactuallyclickedit.ac.il)

www.hatesoup.lt (http://www.hatethesoup.lt)
:laugh:

SoupIsGood
03-13-2006, 05:43 PM
Thank you

www.dontbelieveIactuallyclickedit.ac.il (http://www.dontbelieveIactuallyclickedit.ac.il)

www.hatesoup.lt (http://www.hatethesoup.lt)
:laugh:

:laugh:;)

Kestas
03-13-2006, 05:44 PM
His point is, why is that +12 for Pollard meaningful? (hint: it isn't). He only got the +12 because SJax was red hot and he wasn't on the bench.

thanks, but I know what he meant. however, i hate demagogy ;)

ChicagoJ
03-13-2006, 05:51 PM
:laugh: @ Lithfan and Soup.

Anyway, we had a great discussion on here during the early portion of the 2003-04 season, when I tried to keep my database up-to-date.

I failed; something about a day (and night, ugh) job.

That's the portion of the database that we lost in the original "fire".

As for individual plus/minus, its a flawed stat from the beginning as it is heavily influenced by the other four players on the court.

A lineup of (gratuitous props to the all-time best five-man lineup IMO) Andrew Toney, me, Julius, Bobby Jones and Mo Malone is not going to look much worse than a lineup of Andrew Toney, Maurice Cheeks, Julius, Jones and Malone. So I would rack up a big positive on the plus/minus chart. But that doesn't make me "almost as good" as Mo Cheeks.

The flaw is in the data collection. The data on the player you are trying to extrapolate is heavily influenced by something other than said player. Just summing up the flawed data for an entire season certainly doesn't make it meaningful.

ChicagoJ
03-13-2006, 05:55 PM
thanks, but I know what he meant. however, i hate demagogy ;)

Fine, but who's practicing demagogy here?

Lithfan
03-13-2006, 06:10 PM
:laugh: @ Lithfan and Soup.

Anyway, we had a great discussion on here during the early portion of the 2003-04 season, when I tried to keep my database up-to-date.

I failed; something about a day (and night, ugh) job.

That's the portion of the database that we lost in the original "fire".

As for individual plus/minus, its a flawed stat from the beginning as it is heavily influenced by the other four players on the court.

A lineup of (gratuitous props to the all-time best five-man lineup IMO) Andrew Toney, me, Julius, Bobby Jones and Mo Malone is not going to look much worse than a lineup of Andrew Toney, Maurice Cheeks, Julius, Jones and Malone. So I would rack up a big positive on the plus/minus chart. But that doesn't make me "almost as good" as Mo Cheeks.

The flaw is in the data collection. The data on the player you are trying to extrapolate is heavily influenced by something other than said player. Just summing up the flawed data for an entire season certainly doesn't make it meaningful.

Thats not exactly correct.
If you take into account that the other 4 players are changing randomly - which is true if you take enough stats (the whole season), then you get an ultimate influence of the particular player.

As we all know there are many things that cannot be measured by points or assists or rebounds....

For example, aRTEST best +- on Pacers was true as he was best defender and second best scorer.

ChicagoJ
03-13-2006, 06:14 PM
You're making the argument that Rick's substitution patterns are random?

Wow, I'd hate to be on that side of the argument.

:D

Kestas
03-13-2006, 06:17 PM
Fine, but who's practicing demagogy here?

if we go by your line that +- is flawed and distorts the true image, then it's Pacers home page that are practicing demagogy.
if we go by the line that the +- stats have some sort of meaning and are usefull, then the person trying to deny them by using some blured and rather stupid example would be the demagog.

Anthem
03-13-2006, 06:21 PM
You're making the argument that Rick's substitution patterns are random?

Wow, I'd hate to be on that side of the argument.

:D
Man, that deserves one of these: :brilliant:

I find myself consistently agreeing with Jay, Bball, and Peck. Have I become a Darksider?

Anthem
03-13-2006, 06:22 PM
if we go by your line that +- is flawed and distorts the true image, then it's Pacers home page that are practicing demagogy.
if we go by the line that the +- stats have some sort of meaning and are usefull, then the person trying to deny them by using some blured and rather stupid example would be the demagog.
Well, that's easy. The Pacers homepage is a PR piece. It's the obvious choice.

Lithfan
03-13-2006, 06:22 PM
You're making the argument that Rick's substitution patterns are random?

Wow, I'd hate to be on that side of the argument.

:D

His substitution pattern is not random in some sense.

But considering the large number of lineups Pacers had this year, every player is on court with other random 4 players.

For example Saras played in different 8 !!! lineups with pretty even distribution of minutes 80 max - 30 min :
http://www.82games.com/0506/0506IND2.HTM

Anthem
03-13-2006, 06:40 PM
For example Saras played in different 8 !!! lineups
Five of which have a negative plus-minus...

Lithfan
03-13-2006, 06:48 PM
Five of which have a negative plus-minus...
You see, thats exactly the thing about the stats - look at the total, don't look at small parts. His overall is best on the team isn't it?

ChicagoJ
03-13-2006, 07:07 PM
if we go by your line that +- is flawed and distorts the true image, then it's Pacers home page that are practicing demagogy.

Conrad and I have traded emails on this very topic. Hell, it may even be a very old QofD.

http://www.nba.com/pacers/news/question_040225.html

(not my question but useful nonetheless).

I don't really want to speak for him and I can't find the exact quotes in my email archive, but (1) five-man is much harder to track; and (2) the distinction that I find very important (which lineups are our best lineups) would be more important to the coaches than individual plus/ minus but that's not necessarily what they are trying to present on the website.

In other words, individual plus/ minus is okay for a retrospective look into how each player influenced a past game, just like PPG. But like PPG, it is not sufficient, and certainly not in isolation, for a prospective "who should get more minutes?" question.

However, the five-man plus/minus is the only stat that is sufficient to answer the "who should get more minutes?" question.

Jermaniac
03-13-2006, 07:15 PM
Here ya go


http://www.watchthefriggengames.comBest stat site in the world

ChicagoJ
03-13-2006, 07:18 PM
if we go by the line that the +- stats have some sort of meaning and are usefull, then the person trying to deny them by using some blured and rather stupid example would be the demagog.

Wait, are you calling my example "blurred and rather stupid?"

Lithfan
03-13-2006, 07:24 PM
Conrad and I have traded emails on this very topic. Hell, it may even be a very old QofD.

http://www.nba.com/pacers/news/question_040225.html

(not my question but useful nonetheless).

I don't really want to speak for him and I can't find the exact quotes in my email archive, but (1) five-man is much harder to track; and (2) the distinction that I find very important (which lineups are our best lineups) would be more important to the coaches than individual plus/ minus but that's not necessarily what they are trying to present on the website.

In other words, individual plus/ minus is okay for a retrospective look into how each player influenced a past game, just like PPG. But like PPG, it is not sufficient, and certainly not in isolation, for a prospective "who should get more minutes?" question.

However, the five-man plus/minus is the only stat that is sufficient to answer the "who should get more minutes?" question.

Thanks for the link.

So we agree upon its not meaningless?

1:25 ....
Good night

edit: Kestas was probably talking about 12 Pollard's + points example, but I'm not speaking for him...

ChicagoJ
03-13-2006, 07:29 PM
Not meaningless, just not as good as five-man.

Keep in mind, though, I'm the anti-stats guy.

Dr. Goldfoot
03-13-2006, 07:45 PM
My post was a hypothetical, illustrating how +/- stats in and of themselves show little to nothing on how a players has performed. If my scenario were to play out and the Pacers would have ended up losing by 40 points, if one were to only refer to the +/- of said game it would look like Pollard was the only player to contribute. Where as, those who actually watched the game and/or refered to all of the stats available would know otherwise. Assuming, in a 40 point loss, the rest of the players would most likely have negative stats for the game. I was just trying to shed a little light on how some players may or may not actually influence their own +/- stats. I'm sure Luc Longley, Randy Brown, Bill Wennington and any other role playing starter from the Jordan era Bulls would concur.

SoupIsGood
03-13-2006, 07:51 PM
I feel left out.

I have no been agreeing with much of anybody as of late.


I've been agreeing with Grace lately.... since she mainly seems to take a "I don't care" approach to these debates :D :p

Anthem
03-13-2006, 08:38 PM
I've been agreeing with Grace lately.... since she mainly seems to take a "I don't care" approach to these debates :D :p
Agreeing with someone because they can hurt you is no way to choose sides.