Search found 876 matches

by DSMok1
Wed Apr 17, 2019 5:17 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

Did you just try to minimize squared error the first time around? If you're looking to reign in outliers you could look at raising your p in your L^p norm minimization? Not sure how much that'd change your average error or how well it'd work, but worth looking into. I did originally, yes. Because I...
by DSMok1
Wed Apr 17, 2019 5:15 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

RyanRiot wrote:
Wed Apr 17, 2019 4:20 pm
If you're comfortable with including a position variable, you can try including a position and 3PA rate interaction like ATC used in Dredge.
That's a good thought, I'll look into that.
by DSMok1
Mon Apr 15, 2019 1:48 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

This sheet shows the existing BPM vs. this new RAPM basis I will be using. The major issues I see: Point Guards are systematically undervalued Post players are not estimated as accurately (this could be because more of their value is from defense). https://public.tableau.com/views/PeriodRAPM/Dashboa...
by DSMok1
Sun Apr 14, 2019 12:25 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

"Yes, standard box score. " Does this mean you are retreating from the use of "advanced" stats like Reb%, Usg%, Ast% ? These are pretty simple ratios of raw player totals, and team / opponent totals, minutes adjusted. ORtg and DRtg are slightly more complex, just including possession estimates. Eit...
by DSMok1
Sat Apr 13, 2019 11:22 am
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

I think with boxscore is fine and very helpful to those who hasn’t have play by play or another tracking system, BPM still rolling the models to make predictions so I believe is pretty good. I totally agree. There are many applications for Box Plus/Minus beyond the NBA that do not have comprehensiv...
by DSMok1
Sat Apr 13, 2019 12:41 am
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

Mike G wrote:
Fri Apr 12, 2019 10:31 pm
I think if your center gets five 3-pointers while your PG gets 5 offensive put-backs, you have the same number of points as you would have if the roles reverse.
Right. I believe positions are only useful in assessing defensive credit. On offense I don't think positions add anything.
by DSMok1
Fri Apr 12, 2019 10:25 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

It is relatively easy to determine position strictly from the box score. No, height/weight would not be included. It seems kinda circular at that point. Not to say it wouldn't work. Would you lean towards a small or large number of positions/roles considered? (eg guard/wing/big vs Whitehead's 23 of...
by DSMok1
Fri Apr 12, 2019 9:07 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

eminence wrote:
Fri Apr 12, 2019 8:59 pm
I'm not sure how position would be implemented reliably from the box score, height/weight or what?
It is relatively easy to determine position strictly from the box score. No, height/weight would not be included.
by DSMok1
Fri Apr 12, 2019 8:33 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

To do this project, I have worked with an NBA team (special thanks to them!) to develop an improved RAPM basis. This basis provides average RAPM over 6 year eras, and handles aging/role changes via a Bayesian prior. These shorter eras allow a far better coverage of outliers (like LeBron). Can you g...
by DSMok1
Fri Apr 12, 2019 7:17 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

I added one way of doing it above. Your quick responses are encouraging. My weights try to account for general substitution patterns and who guys switch onto the most. Reverse the substitution related proportions for subs. I will be working with season-level data, so something based on individual m...
by DSMok1
Fri Apr 12, 2019 6:45 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

No play by play stats is a big choice. Assuming you stick with that, it appears shot defense is either same for everybody regardless of who you guard or if you are even on court or not via a team level adjustment, or not include at all... or estimated in a semi-complicated but doable and maybe help...
by DSMok1
Fri Apr 12, 2019 6:29 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

Look at how much assist rate over role / position (PG, wing, big of primary, secondary, tertiary) contributes to team assist rate differential vs. league average. If you dominate the ball but don't make team above average, you probably aren't as good as your assist rate alone suggests. This is an i...
by DSMok1
Fri Apr 12, 2019 6:27 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Re: Reconstructing Box Plus/Minus

When we say 'box-score' that means min/pts/oreb/dreb/ast/stl/blk/tov/pf along with shooting %'s? What size sample are you going for maximum 'accuracy' on? Ranging from single game to full season(more?)? Should the other still be weighted? Also, the 5yr RAPM looks really nice! I love Korver at near ...
by DSMok1
Fri Apr 12, 2019 4:55 pm
Forum: APBRmetrics
Topic: Reconstructing Box Plus/Minus
Replies: 32
Views: 1612

Reconstructing Box Plus/Minus

Hello: I have been preparing for some time now to reconstruct Box Plus/Minus (BPM), with a goal of addressing the major existing issues. Here are some issues that have been identified (I will add more to this post as more are brought forward): Poor handling of outliers on offense Mishandling of inte...
by DSMok1
Fri Apr 12, 2019 1:16 pm
Forum: APBRmetrics
Topic: 2018-19 lineup analysis
Replies: 31
Views: 10400

Re: 2018-19 lineup analysis

Here is the final lineup regressed net ratings for the 100 most played lineups from @bbstats (Nathan Walker): https://twitter.com/bbstats/status/1116003714928730114 https://lh3.googleusercontent.com/VrJHHVDbz5hTbweg7hJCk63bL0xftJMJ-1Tted02qBWj6M4mKIVM_FIw8nUNq1OZt6FoDNtEUlapP6DyqmrIV47iO6Ypx7FwGw6Cf...