Search found 294 matches

by xkonk
Sat Mar 04, 2017 4:07 pm
Forum: APBRmetrics
Topic: Centers: "dying breed" but heavily represented in RPM top 40
Replies: 4
Views: 3492

Re: Centers: "dying breed" but heavily represented in RPM to

Guessing based on the high impact and low minutes, maybe coaches are only playing centers when they know they'll be successful. And/or, after a couple year trend towards "small ball", perhaps you have to be a pretty good center to get on the floor at all these days. If it's the latter, you might exp...
by xkonk
Sat Feb 18, 2017 2:26 am
Forum: APBRmetrics
Topic: p-hacking
Replies: 14
Views: 10090

Re: p-hacking

So how would having independent data sets reduce the chance of an accidental p<0.05 result on an individual trial? I don't think it would on an individual test per se, but if your results differed across sets or you used an independent set for out-of-sample testing and noticed a big drop in accurac...
by xkonk
Fri Feb 17, 2017 1:30 am
Forum: APBRmetrics
Topic: p-hacking
Replies: 14
Views: 10090

Re: p-hacking

... If we have a bunch of researchers independently checking similar hypotheses, then we expect 1 in 20 of them to get a p value of 0.05 or less by accident, right? Under some circumstances I could envision this being true, but certainly not if the researchers were using different data sets or if t...
by xkonk
Thu Feb 16, 2017 12:15 am
Forum: APBRmetrics
Topic: p-hacking
Replies: 14
Views: 10090

Re: p-hacking

Perhaps the more important thing to take away is that if you start by p-hacking, you don't necessarily have a great chance at finding the pattern again in out-of-sample testing. I'd make an even stronger statement: the researcher will almost always find much less statistical significance in out-of-...
by xkonk
Wed Feb 15, 2017 1:40 am
Forum: APBRmetrics
Topic: p-hacking
Replies: 14
Views: 10090

Re: p-hacking

The article is right in that doing what he described is poor statistical practice. It basically invalidates the p value, so if that's your basis for deciding what's "real" or important, you're in bad shape. But Nate and Crow are right that it could be the start of a more rigorous analysis that could...
by xkonk
Tue Aug 09, 2016 11:52 pm
Forum: APBRmetrics
Topic: Setting Up Excel File for RAPM
Replies: 33
Views: 28578

Re: Setting Up Excel File for RAPM

That's why it's crucial to share the source PBP dataset, the regression-ready data and coding steps to reproduce the results. Unfortunately none shares that. They just publish the results and update them time to time and that's it. I just want to second this thought along the lines that one of my m...
by xkonk
Sat Jul 09, 2016 3:54 pm
Forum: APBRmetrics
Topic: Help needed with Master Thesis
Replies: 3
Views: 3171

Re: Help needed with Master Thesis

There's an R blog network where I've seen occasional NBA posts. You could try digging around there. http://www.r-bloggers.com/
by xkonk
Thu Jun 02, 2016 2:26 am
Forum: APBRmetrics
Topic: Including team lead in APM/RAPM calculations?
Replies: 2
Views: 3178

Re: Including team lead in APM/RAPM calculations?

You could start by looking through this thread viewtopic.php?f=2&t=8501&hilit=team+lead
by xkonk
Wed Apr 06, 2016 10:29 am
Forum: APBRmetrics
Topic: (Adjusted) Impact on Win Probability
Replies: 43
Views: 26297

Re: (Adjusted) Impact on Win Probability

Now, at what point could the math lead to weird results? 108% doesn't exists at all, but comes from a misinterpretation of the results. Taking the results from J.E. for Green for example, simply means that with Green playing all 200 possessions (100 offensive and 100 defensive possessions) of a the...
by xkonk
Mon Apr 04, 2016 9:45 am
Forum: APBRmetrics
Topic: (Adjusted) Impact on Win Probability
Replies: 43
Views: 26297

Re: (Adjusted) Impact on Win Probability

You simply replace the scoring margin with changes of the win probability in the response vector (usually named y). And then you run the regression normally. In that way you get win probabilities instead of APM values for each player. The interpretation would actually be that a player would raise/l...
by xkonk
Sun Apr 03, 2016 6:34 pm
Forum: APBRmetrics
Topic: (Adjusted) Impact on Win Probability
Replies: 43
Views: 26297

Re: (Adjusted) Impact on Win Probability

This suggests that having Draymond Green on the floor for the entire game raises your probability of winning by 13.2% (assuming 200 possessions, which is actually a little high) So you took Draymond's coefficient and... what? Averaged/integrated through all the values of time left in a game? Assume...
by xkonk
Wed Mar 02, 2016 1:16 am
Forum: APBRmetrics
Topic: Who are the optimal opponents for scratching starters?
Replies: 7
Views: 6267

Re: Who are the optimal opponents for scratching starters?

I think it depends on your team's quality/situation. If you're a top playoff team that needs wins to get or hold a certain position, I think you would rest against crummy opponents because you're still likely to get the win. But if you're a top team that's running away with things, I think you would...
by xkonk
Wed Feb 17, 2016 12:11 am
Forum: APBRmetrics
Topic: Kaggle March madness competition for 2016 posted
Replies: 7
Views: 6157

Re: Kaggle March madness competition for 2016 posted

I believe that in general, Kaggle asks winners and other high-ranking teams to write a blog post describing their approach to all contests. But I don't think they require that to be anything super-specific.
by xkonk
Sat Jan 23, 2016 2:25 am
Forum: APBRmetrics
Topic: RAPM request thread
Replies: 155
Views: 184354

Re: RAPM request thread

If the code you posted there is what you actually ran, then problem #1 is that the 1415prior in 'offset=1415prior' doesn't exist (you named it prior1415) and can't exist because variable names can't start with a numeral.
by xkonk
Fri Jan 08, 2016 12:30 pm
Forum: APBRmetrics
Topic: Cumulative lineup height RAPM
Replies: 13
Views: 20889

Re: Cumulative lineup height RAPM

JE, how does this fit in with your post on small-ball vs twin towers lineups? My memory is that the effects there were more symmetric, whereas here there would be a net benefit for playing tall lineups (i.e., the spread you mention here of ~6 points for defense vs. ~3 for offense).