RealGM

Anyone have a link to one ?

It's going to be hard finding one.

you can make one yourself prety easily in excel

ChicagoStrong wrote:It's going to be hard finding one.

I did some googling before making this thread and I was surprised I couldn't find one.

you probably can't find one because it depends on the team stats and league average stats. There are sites that have the up to date numbers updated nightly, but only the full season numbers.

Surprised there is nothing out for this, even a small web app. Is PER patented?

http://www.crezbasketball.com/

It's not free but it's available.

I made an attempt at one in Excel a few years back, but I'm off to work now. I'll see if I can find it later.

Probably no one bothers making one since Hollinger jiggers the system every once in awhile if they don't match up with what he thinks is right.

i've got the formula entered etc. but the only thing i'm not sure about is how to calculate stats for the league average player

dunno whether to just divide season totals of a given stat by the number of games played times 13 (players on the roster) or to use a subtler methodology

https://docs.google.com/spreadsheet/ccc ... jV1E#gid=0

it works reasonably well, lebron's per is 35, durant's is 28, wade's is 29, westbrook's is 24 etc. just enter numbers from bbref in the second column

you guys can edit it so lol let's see how long it lasts

You don't really need to make it that complicated. I ran a regression of the data from 2004/05 to 2011/12 on the basketball-reference.com PER numbers, that's what it boils down to:

PER = 3301.2*(3.498*FGM-1.097*FGA+1.412*3PM+1.861*FTM-0.538*FTA+1.062*ORB+0.422*DRB+0.968*AST+1.559*STL+1.062*BLK-1.529*TO-0.479*PF)/(Pace*Min)

That gives a RMSE of 0.3. So, the league average isn't that big of a deal. For this season for example, multiplying the result with 1.044 gives a RMSE of 0.1.

Pace can be easily calculate using team data: Pace=FGA+0.4*FTA+TO-ORB, that is a good enough estimate.

If you want to calculate PER on a game-to-game basis, you can devide 106.8 by the average ORtg for both teams and use that as a multiplier. That exercise for example, gives us for the Utah vs. Dallas games:

I uploaded the spread sheet I did back in 2007 here:

Link

The sheet was made using basketball-reference.com´s formulas back in 2007. I don´t know if there have been significant revisions during the past 5 years, but Kobe's PER from 2005-06 of 28.0 works out pretty well with the spread sheet.

tkb wrote:I uploaded the spread sheet I did back in 2007 here:

Link

The sheet was made using basketball-reference.com´s formulas back in 2007. I don´t know if there have been significant revisions during the past 5 years, but Kobe's PER from 2005-06 of 28.0 works out pretty well with the spread sheet.

can you re-up this?

rockmanslim wrote:

tkb wrote:I uploaded the spread sheet I did back in 2007 here:

Link

The sheet was made using basketball-reference.com´s formulas back in 2007. I don´t know if there have been significant revisions during the past 5 years, but Kobe's PER from 2005-06 of 28.0 works out pretty well with the spread sheet.

can you re-up this?

Yeah just clicked on it and says it is unavailable. I think filefactory only keeps stuff for 6 months.

Here's an iOS app for one, pretty accurate based on what I've done so far:
https://itunes.apple.com/us/app/per-calculator-basketball/id1115329917?mt=8

JetDarlingNg wrote:Here's an iOS app for one, pretty accurate based on what I've done so far:
https://itunes.apple.com/us/app/per-calculator-basketball/id1115329917?mt=8

I use this as well. Pretty simple to use