APBRmetrics

There is a lot more Adjusted +/- projects available now than a year or two ago.

Other research options are still out there, including versions of multi-season APM just against playoff level teams. Many other splits are of course possible but I won't enumerate a big list of them.

The last Factor of RAPM not yet duplicated and updated from Joe Sill's set would be Adjusted Impact on Team Turnovers (Own and Forced). I would have interest in that in itself and to complete the set.

Another option might multi-season Coaching only APM against playoff level teams at anytime or just in the playoffs. The Coaching APM is currently at the full Offensive and Defensive Efficiency level but potentially it could be taken down to the Factor level. That could be helpful with coaching assessments and coaching work plans and ultimately various player management decisions.

Of course any action to further improve the predictive power of the metric by whatever combination of means would be welcome.

Crow wrote:Other research options are still out there, including versions of multi-season APM just against playoff level teams.

I don't think that would be a very useful endeavor. Less data is almost always a bad thing. And I think "record against top teams" has been shown to not improve forecasting on a team level.

Coaching [..] the Factor level.

Agreed that this can be useful

"I think "record against top teams" has been shown to not improve forecasting on a team level."

That was the finding on average by a fairly simple analysis and is not necessarily the end of efforts to find a basis to predict actual playoff performance (perhaps especially late rounds of it) better.

And some players, some stars and some teams might be notably better or worse than average against top teams (or some split of it) and it might mean something in some cases and it might be better to have that information about their placement on the distribution than not.



Team level Adjusted Efficiencies and lineup efficiencies have been calculated but Adjusted Factors could be calculated for these levels.

A team that tracked play calls and / or what actually happened play by play could potentially, with a lot of effort, take the overall and factor Adjusted efficiencies down to the Synergy play type level for the whole team and the most used lineups. Evan has given the raw data but the Adjusted data might be some degree of enhancement, if one were pursuing success in this business as far as one could. Ultimately it would be important to know what plays were run; that choice influences the results and not all data segments for players or sub-units or full lineups will be equally advantaged or disadvantaged by play-calling. Data analysis that is blind to play-calling is handicapped to a degree. The addition of the coach to the Adjusted +/- analysis may help some for year to year player change but it doesn't address the potentially variable coaching play-calling performance (offensive and defensive) within seasons.

One other thing I want to do is handling possessions differently in which the defending team is over team foul limit. I'm sure offensive efficiency skyrockets once that happens. Defenders get unfairly punished at that point, they might not even be those that committed the fouls

Yes, the foul-committing player may have fouled out, or he may be sat down due to perceived foul trouble.
If you've been inserted between the first and second FTA, there's a good chance you're at minus-1 before a second has ticked off the clock.