APBRmetrics

I know a few people have mentioned the need for minutes projections to run their systems. I uploaded mine as a Google Doc in case anyone is interested in using them as a starting point:
https://drive.google.com/file/d/0By9ByT ... sp=sharing

Thanks for sharing.

Is there something past or present explaining the methodology or will there be forthcoming? Are coaches assumed to be mostly following the past and conventions or actually pushing to maximize individual talent on the court, based on past data or performance projections?

They perhaps should be pushing to maximize lineup performance but so little discipline is used in what lineups are rolled out to the court that the data is far from as helpful as it might be with this.

Jared Sullinger @ 12 mpg ?

the celtics have a crowded frontcourt, having added david lee ($15 mil) and amir johnson ($12 mil) to go along with sullinger ($2.3), tyler zeller ($2.6), and kelly olynyk ($2.2), plus rookie jordan mickey ($1.2). someone is going to miss out on some playing time...

Kevin Pelton wrote:I know a few people have mentioned the need for minutes projections to run their systems. I uploaded mine as a Google Doc in case anyone is interested in using them as a starting point:
https://drive.google.com/file/d/0By9ByT ... sp=sharing

Thanks Kevin, I very well may use your minute projections in the next few weeks with my work (looks like I'll be starting back up) - I really hate trying to approximate what a coach or team may do with their minutes, I'm sure you've put a ton more thought into it than I can allow myself. I'm much more into what a team SHOULD do than what I think they will do, but the latter is more important for projections. Obviously I will give you credit for the minute projections any time I use them.

Hey, if at all possible - if you change up some minute projections (injuries, cuts, etc) - could you do a quick post in this thread? Thanks again for sharing, very thoughtful. If you ever need anything you think I may be able to help with, let me know and I'll do what I can.

Kevin Pelton wrote:I know a few people have mentioned the need for minutes projections to run their systems. I uploaded mine as a Google Doc in case anyone is interested in using them as a starting point:
https://drive.google.com/file/d/0By9ByT ... sp=sharing

Thanks so much Kevin!

Despite having no insight on the accuracy of these projections, I really hope everyone adapts these numbers (or some other fixed standard) when they are doing their 2016 win projections.

ampersand5 wrote:

Kevin Pelton wrote:I know a few people have mentioned the need for minutes projections to run their systems. I uploaded mine as a Google Doc in case anyone is interested in using them as a starting point:
https://drive.google.com/file/d/0By9ByT ... sp=sharing

Thanks so much Kevin!

Despite having no insight on the accuracy of these projections, I really hope everyone adapts these numbers (or some other fixed standard) when they are doing their 2016 win projections.

Agreed. It may be interesting (educational?) to see how team projection systems compare with everyone using the same exact player minute projections.

Thanks, Kevin, for doing this. If some trades are made before the season starts, and you want to re-issue your minutes projections -- would it be much trouble to put players' ages in the list?

It seems that in predicting, we end up "testing" various elements:
- Metric (or blend),
- aging curve, and
- minutes allocation estimate.

Uncertainty in these last 2 disallow for a proper assessment of the metric itself. But if we know a team's roster (to start the season), and we've agreed to adhere to a standard minutes allotment, we can project a team's average age. This would be pretty handy.

If someone wants to offer an accepted aging adjustment, that would be great. If anyone has estimates for rookies, based on their draft position, that would also be worth sharing.

These are some quick subjective takes on guys who I think will play notably more or less (at least 5 minutes per game variance) than Kevin's projections.

Schroeder up
Thabo way down
Muscala up to way up
T Robinson up
Mickey up
P Jones up to way up
Larkin down
D Sloan up
Boatright up
Olynyk up
Jerebko way up
B Jennings way up
Capela up
M Harrell up
Jordan Hill down
Stephenson down
K Bryant down
r Upshaw up
C Anthony down
Porzingis way up
N Collison up to way up
Mcgary up to way up
C Payne way up
Singler down
Staukas down to way down
H Thompson down to way down
C Landry up to way up
C Wood up to way up
Burke down
Burks down
DA Russell down
C Andersen up to way up
J Parker down
Garnett down
Ginobili down to way down
N Powell up
G Jerrett up

Any thoughts on Shump here?

Seemed relatively ok to me.

This should be updated for MKG and Shumpert injuries, along with a few others:
https://docs.google.com/spreadsheets/d/ ... sp=sharing

Team minutes add up to 19680, or 240 player-minutes per game.
With overtimes, there are typically around 242; so we should have another 2 mpg from someone on each team.

One could disregard this discrepancy, perhaps figuring ~2 mpg from 'replacement' types; but people have differences interpreting that.
One could multiply all these minutes by 1.0083

File now updated with a few injury tweaks and birthdays for most players (a few late additions to camp rosters do not have birthdays).

Kevin Pelton wrote:File now updated with a few injury tweaks and birthdays for most players (a few late additions to camp rosters do not have birthdays).

Thanks again Kevin - I think this is really going to save me some time.