D'Angelo Russell Discussion: Traded To The Nets

Great to see you back @trodgers We missed the data analysis.

 

All that is left... is for @D-Fish Man to spill the beans already.

 

This is an actual interesting problem. If I was the Phd Advisor, I would have suggested a very different approach.

I would look for players with a high variance then the normal distribution with some other filters (career length has to be long enough, some.minimal.minutes?, Injury adjustment?). Then I would look to see if those players the better years correlated with better performances. Essentially see if the dudes whose year to year performance changed a lot or had real outlier years did that match a contract year

 

yeah, and the reason given for the sampling approach was basically "it's too hard to do other stuff". in other words, the author knows this wasn't the optimal way to answer the question.

as i said before, i'd definitely look at already completed careers, starting probably in the 90s (semi-arbitrary there, might consult previous CBAs to look for major shifts in general contract trends). i don't think you can select based on within-player variance, as that might actually suppress the effect if there is one. i'm also not sure if i'd use advanced metrics as outcome variables until the last few years. we need measures that the players themselves would be aware of, imo. pre-2010-15, i'm guessing nobody was shooting to up their VORP. but they were looking for points and rebounds and assists and steals.

so limit the time frame, maybe sample only players who had at least three contracts that lasted at least 2 years each?

but again, someone could question all of my choices, too.

anyway, fun that someone got to do this as a thesis project in any case. i'd totally supervise something like this--i actually did have a student who was on our basketball team and did a project wherein she connected on-court measures with some personality data for players. that was pretty fun.

 

I never was a big DLo fan, but he does fit well with AD and LeBron because he can shoot. I love to find out stats like he’s the second best mid-range shooter in the League. When it comes to analytics, they are just dead wrong about the mid range game being a waste of time. Two of the most unstoppable, dominant scorers ever in Kobe and MJ were masters of the mid range game.

 

I don't think he changed much...

 

to quote the dude...you're not wrong, walter...

yeah, i think the issue is when you don't have the data points to begin with or can't agree on what are the relevant data points.

when dean oliver and morey and whomever showed up touting "analytics" as the true path, i was always dubious. not because i distrust science or because i prefer to construct narratives around anecdotal data, but because they were suggesting that they had ways to look at data that were superior and would give them a big advantage. i think that's proven false over time. "analytics" in basketball isn't even about the analysis, it's about gathering more and better objective data points that are actually summarized in a very simple and straightforward manner. as teams and outside parties have begun simply gathering more data (charting all shots and numerous contextual elements in addition to traditional stats), we have the opportunity to know who makes more shots with a defender near them from the left side of the floor and who misses boxouts and who flubs the help scheme on defense. interestingly, when someone said they want to hear what coaches or gms say, it was really this because at the basic level of turning qualitative to quantitative information (e.g., "that's a steal" or "that's a missed box out") there's often still subjectivity. so your analyses are only as good as the inputs, which aren't of high fidelity in many cases.

i'm reminded of a chapter i read on gathering animal behavior data years ago, wherein an attempt to objectively define "charging" in chimpanzees resulted in the systematic and inaccurate coding of playing chase as an aggressive behavior. the lesson isn't that we can't or shouldn't attempt to quantify some things, but just that some things are really hard to quantify and that the process of quantification cannot escape subjectivity (i.e., had they allowed coders to use their own subjective judgment of chimpanzee intent, they'd have gotten better "objective" data points!).

anyway, i'm rambling now. my lay hypothesis is that contract years do affect players, but not in a systematic way, and thus you won't find systematic evidence for it. in such cases, it's probably best to evaluate on a case-by-case basis using the individual's relevant history than to rely on popular trends.

 

I can't wrap my head around having a player on the roster who puts up 14 3s a game and you're not crying in a corner or turning up a bottle because of it.

 

In his defense, Gobert isn't good on offense and the Wolves shouldn't have traded for him. lol

 

So uncalled for how Minnesota media is throwing DLo under the bus now! He’s already off the team, they should be professional and let bygones be bygones! Has anyone stopped to consider how DLo is going to explain this character assassination to his kids?

:RW Barfbag:

 

lol, i do think it's funny that a week ago dlo was considered a suitable replacement for chris paul by some in the press, and now he's mike conley's b**** and sucks. i wonder what happened?

 

I disagree with automatically criticizing individual examples as not good data analysis: when you have a very heterogenous group and hard to quantify metric (player performance) looking at large datasets very easily will wash out an effect that occurs in a subgroup. Think of primary or secondary prevention in medicine: most things like daily ASA show no benefit when looking at a larger group or even harm but when you narrow it down to folks who have had heart attacks you see a significant benefit.

As noted in my next post: I would approach the data analysis very differently then what the paper does

 

Honestly this makes me feel better about DLo just like that video of him being annoyed not getting the ball; he actually seems to care about the team doing well now.
That was always the big fear with him was he just one of those dudes who played ball because he wanted the money and fame and not actually do the things to win

 

Same. I think if he is dialed in and wants to win then he has a chance here to really shine.

 

Not to mention all the Kardashian units are once again just a phone call and short drive away.

 

Was it me hallucinating last night or did Russ’s hair look like LA Rams horns?