APBRmetrics

eminence wrote: ↑Sun Dec 20, 2020 5:05 pm I've never bothered with a simulated season myself, but more power to y'all that do.

How do you do it?

If you expect your team to have avg MOV of +5, they should win 65% of their games, or 53 of 82.
A simulated game-by-game season might give you 48 or 58, and a bunch of such 'seasons' ought to add up to an avg of 53.

Maybe a schedule can be biased by somehow getting more back-to-backs and longer layoffs. But some players do well in those situations, while others do not. Have whole team-seasons been shown to benefit or suffer by the arrangement of 82 opponent-dates?

Was that an answer to mine?

Mike G wrote: ↑Mon Dec 21, 2020 1:00 am If you expect your team to have avg MOV of +5, they should win 65% of their games, or 53 of 82.

Please define any abbreviations you use. There are so many I have no idea which one you mean. What is MOV? I had to google it. So that we can be on the same page.

Anyway....if you have an expectation that means you somehow calculated an average. Either from simulation or from past data or some empirical knowledge or from some magic ball.

Mike G wrote: ↑Mon Dec 21, 2020 1:00 am A simulated game-by-game season might give you 48 or 58, and a bunch of such 'seasons' ought to add up to an avg of 53.

In theory. The advantage of random simulations (lets call them Monte Carlo) is that they can give you many things beyond a single value (mean). They can give you the whole distribution (variance, kurtosis, skewness etc).


Obviously in situations that you have a fixed schedule you only simulate the outcomes.
In the case that you do not have a partial fixed schedule or none at all, then given some rudimentary rules you can also simulate the schedule in your Monte Carlo step.

vzografos wrote: ↑Sat Dec 19, 2020 9:12 am

vzografos wrote: ↑Tue Dec 15, 2020 11:43 pm

DSMok1 wrote: ↑Tue Dec 15, 2020 10:23 pm
I used Kevin Pelton's minutes.

I saw those too.
Not sure how accurate they are but if you want some additional projections I am going to calculate all the (per-season average) boxscore stats for 2020-21. I expect to have them in the next couple of days. Dont know if they will be any good but I would be interested to see how your team win projections change (if at all).

If anyone is interested you can find my minutes projections here:
https://drive.google.com/file/d/1juLaPO ... sp=sharing
as an alternative to Kevin Pelton's numbers.

These projections do not incorporate any player injuries before the season, or any changes to player status (i.e. left the team, league, retired etc). Any players that were given 0 minutes from Pelton's method have been highlighted because maybe there is additional information that I dont have so you need to pay attention to those.

You have given many minutes to players who are not on teams anymore.

vzografos wrote: ↑Sat Dec 19, 2020 9:12 am

vzografos wrote: ↑Tue Dec 15, 2020 11:43 pm

DSMok1 wrote: ↑Tue Dec 15, 2020 10:23 pm
I used Kevin Pelton's minutes.

I saw those too.
Not sure how accurate they are but if you want some additional projections I am going to calculate all the (per-season average) boxscore stats for 2020-21. I expect to have them in the next couple of days. Dont know if they will be any good but I would be interested to see how your team win projections change (if at all).

If anyone is interested you can find my minutes projections here:
https://drive.google.com/file/d/1juLaPO ... sp=sharing
as an alternative to Kevin Pelton's numbers.

These projections do not incorporate any player injuries before the season, or any changes to player status (i.e. left the team, league, retired etc). Any players that were given 0 minutes from Pelton's method have been highlighted because maybe there is additional information that I dont have so you need to pay attention to those.

These projections do not incorporate any player injuries before the season, or any changes to player status (i.e. left the team, league, retired etc).

So you didn't incorporate any of the information that actually matters in making a good win projection? what is the point of fancy math if your model is telling me Axel Toupane (waived) will play 12MPG for the Warriors this year?

Guess Hayward is healthy enough, switching back to original predictions (changed my predictions post).

I have updated my projections. The minutes projections are updated to Kevin Pelton's latest https://drive.google.com/file/d/1ZcQABH ... v1-Ad/view

I removed RAPTOR from my blend. It now only includes PIPM, RPM, and BPM.

I have also added Strength of Schedule to the win totals. This generally hurts teams in the West, particularly near the bottom of the pack.

Here are my updated projections:

Eastern Conference:
MIL 48.9
BKN 45.9
MIA 44.5
TOR 43.2
PHI 41.9
BOS 39.3
IND 37.2
ATL 36.2
WAS 34.3
ORL 28.0
CHI 27.5
CHA 26.3
NYK 23.0
DET 21.9
CLE 21.6

Western conference:
LAL 48.2
UTA 45.8
LAC 44.6
DAL 43.6
DEN 41.2
HOU 41.0
NOP 39.6
POR 39.5
MEM 34.4
PHX 34.3
MIN 34.0
SAC 31.5
GSW 30.0
SAS 29.4
OKC 23.5

liminal_space wrote: ↑Mon Dec 21, 2020 4:02 pm
So you didn't incorporate any of the information that actually matters in making a good win projection? what is the point of fancy math if your model is telling me Axel Toupane (waived) will play 12MPG for the Warriors this year?

You can remove those manually. I am sure you are able to do that

Hoping we get some more entries, don't be shy everyone!

By 7 pm Eastern tomorrow preferred. About 9:30 pm Eastern would be ok to accommodate different schedules and encourage last minute participation.

A day or two late is usually allowed with excuse offered / asked and without multiple objections, as long as it doesn't look fishy related to any early injuries or surprise developments.

I used a random number generator:

Good luck to everyone this year!

just a bunch of prawns