APBRmetrics Forum Index APBRmetrics
The statistical revolution will not be televised.
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Basketball-Reference.com Reboot
Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next
 
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion
View previous topic :: View next topic  
Author Message
jkubatko



Joined: 05 Jan 2005
Posts: 603
Location: Columbus, OH

PostPosted: Fri Jan 25, 2008 11:40 pm    Post subject: Reply with quote

mateo82 wrote:
Sorry, I have to complain about the move to "per 36".


As I wrote in my blog, I can’t win here.

Quote:
All it does is ensure that every pundit and fan alike is going to keep debating players based on per game statistics.


I don't understand how that would change if I started displaying per 48 minute statistics.

Quote:
Why not just go with per 48? Per 48 makes since. It's the length of a basketball game.


How many players actually play 48 minutes per game?

Look, you prefer per 48 to per 36, someone else prefers it the other way around, someone else wants per 40, etc. This is just a way to display the player’s per minute production, so any multiplier is in some sense arbitrary. We can expect a solid starter to play about 36 minutes per game. That seems to me to be a reasonable standard. It's not a reasonable standard to you. C'est la vie.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Chronz1



Joined: 22 May 2006
Posts: 134

PostPosted: Sat Jan 26, 2008 8:58 am    Post subject: Reply with quote

Per 36 is fine, it actually lets you visualize what a player would be averaging in those minutes. No one could ever player 48 minutes,
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 446
Location: Phoenix

PostPosted: Sat Jan 26, 2008 12:49 pm    Post subject: Reply with quote

Chronz1 wrote:
Per 36 is fine, it actually lets you visualize what a player would be averaging in those minutes. No one could ever player 48 minutes,

"

It seems that there are two goals in expressing stats in a per-minute format: to put every player on a level field for comparisons, and to show the quality of a player's performance in and of itself. We can use anything for the first, but the second requires using a realistic standard.
Back to top
View user's profile Send private message
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 713
Location: Seattle

PostPosted: Sat Jan 26, 2008 1:48 pm    Post subject: Reply with quote

Given that per-X-minute statistics are arbitrary, my take is the important thing is to have consistency of use. It seemed like we had pretty well settled on using 40 minutes within the community. Is anyone else using 36?
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
DLew



Joined: 13 Nov 2006
Posts: 72

PostPosted: Sat Jan 26, 2008 3:27 pm    Post subject: Reply with quote

Justin,

How easy would it be to add a menu where the user can select per 36, 40 or 48? That way people could go with whatever they like.
Back to top
View user's profile Send private message
jkubatko



Joined: 05 Jan 2005
Posts: 603
Location: Columbus, OH

PostPosted: Sat Jan 26, 2008 3:41 pm    Post subject: Reply with quote

Kevin Pelton wrote:
Given that per-X-minute statistics are arbitrary, my take is the important thing is to have consistency of use. It seemed like we had pretty well settled on using 40 minutes within the community. Is anyone else using 36?


The consensus in this thread seemed to be that people preferred per 36 minute rates.

DLew wrote:
How easy would it be to add a menu where the user can select per 36, 40 or 48? That way people could go with whatever they like.


This is certainly do-able. It's just a question of whether or not I want to spend the time doing it. Changing from per-36 to per-whatever is a matter of changing one line of code. Giving people options is more complicated.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
mateo82



Joined: 06 Aug 2005
Posts: 209

PostPosted: Sun Jan 27, 2008 2:23 pm    Post subject: Reply with quote

jkubatko wrote:
mateo82 wrote:
Sorry, I have to complain about the move to "per 36".


As I wrote in my blog, I can’t win here.


I would put both per 40 and per 48.

jkubatko wrote:
Quote:
Why not just go with per 48? Per 48 makes since. It's the length of a basketball game.


How many players actually play 48 minutes per game?

Look, you prefer per 48 to per 36, someone else prefers it the other way around, someone else wants per 40, etc. This is just a way to display the player’s per minute production, so any multiplier is in some sense arbitrary. We can expect a solid starter to play about 36 minutes per game. That seems to me to be a reasonable standard. It's not a reasonable standard to you. C'est la vie.


Per 48 is not arbitrary. It is the length of a basketball game. It doesn't reflect what a player would ever perform in a game, and it's not supposed to. It's used because per-minute is practically unreadable. Per 40 or per 36 or any other number is arbitrary by the definition of the word arbitrary, no matter how good of a reason you think you have for using one. To me that's a deal-breaker, I'm not using arbitrary numbers in analytical analysis no matter how "reasonable" I or anyone else thinks it is.

Now, if people really insist on having an arbitrary per-X stat, fine, I can live with it though i'll never use it myself. I think giving per48 as well is a good compromise. But it's your site. It's just a shame that there aren't any good stat sites out there that give per48 stats.
Back to top
View user's profile Send private message
Eli W



Joined: 01 Feb 2005
Posts: 354

PostPosted: Sun Jan 27, 2008 3:44 pm    Post subject: Reply with quote

If you really want to avoid arbitrariness, why not flip things up?

Instead of points per time period, use time period per point.

For example, Kevin Garnett:
19.8 pts/36 min -> 1 minute, 49 seconds per point
10.2 reb/36 min -> 3:32 per rebound
3.9 ast/36 min -> 9:13 per assist
etc.

All the information is preserved, and there's no arbitrariness. Of course, now lower numbers are good instead of bad (except for things like turnovers, where higher numbers are now good instead of bad).

Be sure to send me a royalties check once you make the switch.
_________________
Eli W. (formerly John Quincy)
CountTheBasket.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
mateo82



Joined: 06 Aug 2005
Posts: 209

PostPosted: Sun Jan 27, 2008 4:38 pm    Post subject: Reply with quote

Not bad, but it's missing the "pop" factor. If I see a per 48 stat and it immediately pops out about how good it is, same for the per 40 stat (unfortunately, it's unavoidable). But I'd be scratching my head for the first couple of months figuring out whether a basket every 3:38 seconds was acceptable or not. Wink And no one would have a clue what I was talking about.

Still, I really like it and, no joke, I might actually start doing this.
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 972
Location: Durham, NC

PostPosted: Sun Jan 27, 2008 6:44 pm    Post subject: Reply with quote

mateo82 wrote:
Not bad, but it's missing the "pop" factor. If I see a per 48 stat and it immediately pops out about how good it is, same for the per 40 stat (unfortunately, it's unavoidable). But I'd be scratching my head for the first couple of months figuring out whether a basket every 3:38 seconds was acceptable or not. Wink And no one would have a clue what I was talking about.

Still, I really like it and, no joke, I might actually start doing this.

Your opinion of what "pops" and what doesn't seems like the very definition of arbitrary to me.
Back to top
View user's profile Send private message Send e-mail AIM Address
Mike G



Joined: 14 Jan 2005
Posts: 1794
Location: Delphi, Indiana

PostPosted: Sun Jan 27, 2008 7:07 pm    Post subject: Reply with quote

I'm just guessing, but I get the feeling the majority of players can play up to 36 minutes in a game without showing fatigue, and only a minority can do so for 40. The difference is just 4 minutes, but that's 50% more rest.

By 'show fatigue', of course I'm referring to how well per-minute rates hold up. Players who regularly go 40+ mpg have the advantage of playing more minutes vs reserves. So their per-minute stats may also scale (down) to 36 reliably.

Most games go 48 minutes; some go 53, or 58, etc. The average is about 48.5 .
Most players who register 48 in a game do so in overtime games. Per-48 doesn't mean anything more than to indicate 20% of a team's player-minutes (regulation game). 36 minutes is 15% of total. It's also 3 quarters of play.
_________________
40% of all statistics are wrong.
Back to top
View user's profile Send private message Send e-mail
jkubatko



Joined: 05 Jan 2005
Posts: 603
Location: Columbus, OH

PostPosted: Sun Jan 27, 2008 10:36 pm    Post subject: Reply with quote

mateo82 wrote:
I would put both per 40 and per 48.


And then what do I do when people say that they want per 36? Or per 32? Or per whatever?

Quote:
To me that's a deal-breaker, I'm not using arbitrary numbers in analytical analysis no matter how "reasonable" I or anyone else thinks it is.


This is the second time that you've written something like this and I still can't figure out what you mean. In general, any conclusions drawn from an analysis of per minute statistics can be extended to any multiplier. For example, if Player A scores more points per minute than Player B, then Player A scores more points per 36 minutes, more points per 40 minutes, more points per 48 minutes, etc.

Quote:
Now, if people really insist on having an arbitrary per-X stat, fine, I can live with it though i'll never use it myself.


Do you ever use Player Efficiency Rating (PER)? Based on what you wrote above you shouldn't, as an arbitrary multiplier (that is, 15) is used to scale PER. How about True Shooting Percentage (TS%)? TS% is arbitrarily divided by 2 (i.e., multiplied by 0.5) to make it look like a shooting percentage. If you are consistent with the statement above, then you should never use either of those stats.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Ben



Joined: 13 Jan 2005
Posts: 222
Location: Iowa City

PostPosted: Tue Jan 29, 2008 1:19 pm    Post subject: Reply with quote

My apologies if I missed this, but do you still have leaderboards for MVP award shares?
Back to top
View user's profile Send private message
Chronz1



Joined: 22 May 2006
Posts: 134

PostPosted: Wed Jan 30, 2008 1:03 am    Post subject: Reply with quote

jkubatko wrote:
mateo82 wrote:
I would put both per 40 and per 48.


And then what do I do when people say that they want per 36? Or per 32? Or per whatever?

Isnt it obvious? Create a toolbar where we can select any arbitrary number we want. Razz
Back to top
View user's profile Send private message
jkubatko



Joined: 05 Jan 2005
Posts: 603
Location: Columbus, OH

PostPosted: Wed Jan 30, 2008 7:06 am    Post subject: Reply with quote

Ben wrote:
My apologies if I missed this, but do you still have leaderboards for MVP award shares?


Oops, I need to get that back up there.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Display posts from previous:   
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion All times are GMT - 5 Hours
Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next
Page 5 of 7

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2005 phpBB Group