Bio not provided
This is fantastic. Writing D3 directives for Angular has been on my todo list for a while now.
7 months, 2 weeks ago on Sky Blue Trades | AngularJS + D3.js = Radian
@ascreamingacrossthecourt Ooh, that's interesting. Another idea I had (for another day) is to do this in 3-D with %AST being the third axis. Perhaps, there would be a frontier plane to look for!
7 months, 4 weeks ago on Lists! The League's Best Scorers in 2013 according to Scoring Index
@plarson01 USG is definitely coming too!
11 months, 1 week ago on Introducing NBA WOWY!
@plarson01 I'll definitely be putting possessions up, but minutes will come later. For basketball reasons. ;)
@JakeNichols Right now, because you can do any arbitrary combination that doesn't make much sense. But if you want to do that, just run the query without any filter and do the comparison manually. Hey, it's better than nothing right?
I have an idea about letting the user chain together multiple queries for comparison, so eventually you'll be able to do this automagically. That may take a bit of time, though.
@ivanbe I would argue that you are actually implicitly following the rules already. We might could only settle the ceiling comparison (Granger vs. Pierce vs. whomever) by actually constructing such a bet in real life. ;)
1 year ago on On Ceilings and Floors and Betting
@ivanbe LeBron or Bird. Is Bird the ceiling for LeBron? Is LeBron already better?
@ivanbe "If you actually asked everyone to put money on this game, then of course nobody's going to say LeBron. And then you just have a boring game with safe bets and little imagination."
It actually does sound like we're playing the same game. I'm just giving it formal rules.
@creedofhubris Good suggestion. I haven't done that yet, but I should. I would assume the TP11 ratings will be similarly shrunk, but it would be interesting to see the ∆ values.
1 year, 1 month ago on Bayesian True Power Ratings for the NFL
@creedofhubris I don't know. I read somewhere that this has been one of most unpredictable seasons in recent memory. I've seen models that are producing crazy results like Denver being #1.
@creedofhubris They are and those won't come up until more games are played. It's the bias-variance tradeoff.
@creedofhubris Good idea!
1 year, 2 months ago on Starting work on Bayesian football power ratings
@creedofhubris The Vegas line on BAL-HOU was -7.5. If you picked Houston to beat the spread, you would have won. Bayes predicted the spread to be -4.3, meaning "Houston beats the spread". That is the sense in which "BEAT" should be understood here.
@bjlevy34 " The more games you play, the smaller the standard deviation will be even with the exact same variability in performance."
This is not true. The mean of the s.d. upon repeated samplings will be the same whether the sample size is small or large. What will change is the distribution of the s.d. and the C.I.'s. For large sample sizes, the s.d. will approach it's "true" value by LLN. For small sample sizes, the s.d. could be smaller or larger than the true value, or it could be the same. We don't know. Go try it out for yourself with a RNG. Calculate the s.d. for 100 values sampled from a U(0,1) distribution and do that 1000 times. Do the same exercise for 1000 values 1000 times. You'll see that in the second case, the s.d. is just about the same in each trial, but the for the first case the s.d. will vary substantially. Sometimes it will be higher, sometimes lower, but the mean of those trials will be the same as the mean of the larger trials.
This is not to say that I disagree with your main point, however. In general, we should not trust smaller sample sizes as much as larger ones. So if Aaron could take this a step further, calculate confidence intervals, yada, yada, maybe we'd see a real effect, or maybe not.
1 year, 2 months ago on Player Capsules (Plus): The Flickering Candle of Stephen John Nash
@mediasres "If you are telling me that your spread sheet was partial and your rankings were somewhat provisional"
That is what I've been trying to tell you. In this post I only showed players with > 25% USG, but I went back later and included all players. That's why on the other spreadsheet you'll see Artis Gilmore with a 70% TS 19% USG season on the list.
1 year, 3 months ago on Ranking the All-Time Great Scoring Seasons in the 3-PT Era by Their Distance from Greatness
@mediasres Kelly Tripucka's best season rates #325 on the full list.
@mediasres Maggette in '04 had the #452 best season. That's nowhere near as good as Barkley's '87 season according to my metric.
@mediasres I already explained in the other post that Barkley's '87 season is rated #60 out of all 10,000 data points. It's just not on the spreadsheet included in this post. Maggette's highest rated season is #106 (2010) when he shot 61.5% TS on 26.7% USG.
Now go cherry pick the next one you have a problem with.
@mediasres There would be a very high correlation between TS% and PER, that's the problem. By definition, PER is correlated to shooting efficiency. USG is not.
Maggette is a very good scorer. He can't do anything else. But he can score. I have no problem with that.
@mediasres If you know of a stat that will give "perfect" results, let me know.
It's not PER, though. Or PPG.