Team OFF DEF NET
GSW 6.83 4.16 10.99
SAS 2.40 4.68 7.08
LAC 3.99 2.92 6.91
HOU 6.50 -1.27 5.24
UTA 0.83 3.15 3.97
CLE 5.33 -1.47 3.86
TOR 2.84 0.59 3.43
MEM 1.17 1.99 3.16
BOS 2.95 -0.39 2.56
MIL 1.77 0.08 1.84
WAS 2.65 -1.04 1.61
OKC 0.17 1.26 1.43
POR 0.19 -0.23 -0.04
ATL -3.39 1.98 -1.41
The list below shows the difference between a team's projected playoff rating and their current Adjusted Net Rating on BBR. Positive values indicate their playoff rating is better than their BBR rating.
What was the basis for the projected minutes? Purely subjective, application of certain rules, based on late season availability & use or a combination? Can you share the projections for a few of the most important teams or the ones with the biggest variances?
The offensive and defensive ratings are based on the whole regular season. At this point based on play against playoff teams or specific opponents would be more relevant, imo. RPM would probably vary by quality of opponent I'd split out that way. (Tons of kinds of RPM splits would be interesting in spite of the greater noise. Use multi-yr to offset the in season splits.)
Interesting that the team which held the Warriors under 100 ppg in the Finals now has the worst defense of any playoff team.
Why would teams have worse absolute RPM in playoffs? Doesn't everyone use their better players more minutes then?
The Spurs are historic in allowing regular season rest/downtime for their stars, for the specific purpose of playoff readiness. How do they come in as most likely to fall short of their RS rate? They seem to have zero injuries.
There wasn't really a set formula for the minutes. It was slightly subjective in that I started with MPG for the season for players whose role hasn't changed significantly and used more recent MPG numbers for players whose role has changed recently, but the latter were relatively rare for these teams. I generally gave each team a 10 man rotation, except for a couple teams where I was reasonably confident that an additional player may receive rotation minutes. MPG were then adjusted so the total MPG for the rotation I selected was 240, with starters getting a bump up in minutes typically if the total was less than 240 and role players getting a slight drop if the total was greater than 240.
@MikeG
I revisited the usage adjustment I made to each team and realized it was too aggressive for the Spurs. The original post has been edited to reflect that change. Now all but 3 teams are within 1 point of their Adjusted Net Rating on BBR. Not that I was trying to force the projections into looking that way or anything. I didn't make the initial comparison until after I had finished rating all of the teams.
I think part of the issue for so many teams appearing to fall short of their Adjusted Net Rating on BBR is because the BBR value is calculated using least squares whereas the RPM values used in the projections are obviously found via ridge regression, which is likely to compress the ratings a bit more. Since these are primarily above average teams, they fall on the side of looking worse by minutes weighted RPM than they do by BBR's Net Rating. But since it's affecting pretty much every team equally (except teams like LAC who get CP3 back full time) it wouldn't really impact any projections you would make in regards to point spreads for most matchups.