2012 NBA Draft - Part III

[DCZards]

I think I'm still hanging by a thread to reality, but shouldn't some small percentage, maybe half that of the general population, be Republican? If not, why not? After all, they are the 1% 'ers.

Given the demographics and background of most NBA players, I don't think they become Repubs simply because they are now 1%'ers.

[Zonkerbl]

We should dig up the political thread. African Americans are OVERWHELMINGLY Democrats, no matter where they live. Furthermore, urban residents tend to be democrats, no matter what their skin tone. Most NBA players are African American, so the ones that care about politics will most likely be Democrats. White players who grow up in rural areas are the most likely to be Republican, and that's a very small proportion of the current NBA population, since very few Hoosiers actually make it to the NBA. Charles Barkley at one point publicly considered running for governor of Alabama as a Republican, but he recanted later.

So besides the fact that Republicans are all evil and wrong and stupid, the demographics of NBA players make them extremely unlikely to be Republican.

I'm curious why it would matter? Does being evil and wrong and stupid make you a better bball player?

[Nivek]

M'sieur Nivek -- please explain what YODA entails and how one computes (divines?) it!! Thanks.

Other folks have done a reasonable job hypothesizing about the process. It started as a way to take Ed Weiland's benchmarks and turn it into a system where I could plug in a guy's numbers and voila -- have a rating. The first try sucked. So did the second and the third and the fourth. So, I put Ed's research aside and started from the ground up. Considering the amount of work I've put into it, I'm not willing to just give it away.

The basics are these: I take the player's stats, adjust for strength of schedule (25 & 12 at Lehigh isn't worth as much as 25 & 12 at Kentucky), and then account for other factors such as length, quickness/speed, strength, vertical, intangibles. I give a modest bonus to freshmen. And that's about it.

What's interesting is that by putting aside Weiland's findings, I think I may have largely replicated them -- or at least "discovered" much the same thing.

Back to those physical attributes/intangibles for a moment -- whenever possible I try to use stuff like draft camp measurements and times. It helps a lot to keep those assessments as objective as possible -- avoiding absurdities like the Kevin Love vs. Michael Beasely absurdity I posted about someplace in this thread. (To summarize: the two measured almost identical in length and posted almost identical scores in the physical tests. Yet the "eye test" scouting reports said Love wasn't long enough and lacked the athleticism necessary to overcome that "problem" whereas Beasley was an elite athlete with an NBA ready body.)

Intangibles can be LOTS of things including on-court stuff like motor, exceptional hoops IQ, outstanding defense; as well as off-court stuff like substance abuse, legal troubles, breaking team rules, etc.

[Chocolate City Jordanaire]

CCJ, Chad Ford likes your man Barton:

Andy (New Jersey)


Who do you see as the potential biggest steal of the 2nd round?
Chad Ford
(1:11 PM)


I'm intrigued with Memphis' Will Barton. Has a NBA game. Good athlete. Very long. Has a great mid-range game. Rebounded very well for a guard. A lot of talent there. You could argue he was better than Jeremy Lamb this year.

http://espn.go.com/sportsnation/chat/_/ ... -chad-ford

Maybe I should do my own social media thingy and start my own blog.

[Chocolate City Jordanaire]

CCJ, Chad Ford likes your man Barton:

Andy (New Jersey)


Who do you see as the potential biggest steal of the 2nd round?
Chad Ford
(1:11 PM)


I'm intrigued with Memphis' Will Barton. Has a NBA game. Good athlete. Very long. Has a great mid-range game. Rebounded very well for a guard. A lot of talent there. You could argue he was better than Jeremy Lamb this year.

http://espn.go.com/sportsnation/chat/_/ ... -chad-ford

Maybe I should do my own social media thingy and start my own blog.

In all seriousness, we all have opinions and many times people read what we say in this blog. Ford's a guy who knows who is going to be drafted and when a lot of the time. He got to where he is by being good. (IIRC Ford lives on the Big Island BTW) I bet he reads tons and tons of scouting reports, and anything else search engines bring up.

This time of year, I want the Wizards to draft the right guys to help the team. So, I put energy into posting about certain player who I feel are maybe not getting the publicity and the love from the scouts that they deserve. I post a lot. I wouldn't discount the possibility of Ford or anyone else reading what I and many others who blog are posting. Same way some guy really did know how good Lin would be, I knew about Faried.

gesa2, I don't know what led to Ford's appraisal of Will Barton, but obviously, I agree with him.

Will Barton's a top-15 to top-25 player in this draft, but he's the player I would like to see the Wizards draft as a SG of the future with Wall. Morris Almond is also a guy I want back with the Wizards. Will is a slasher, Morris, once he gets confident will be able to score more as a catch and shoot player.


(Just in case Chad is reading, gotta give Mo Almond some more good press, to follow.).

Morris showed himself to be a hustle player--5 steals in 26 minutes is impressive, even against Heat backups. Along with being a hustle player, Morris is a cerebral player who played 67 minutes with only 1 turnover. Morris Almond plays hard and he thinks the game well. He's grown and is still developing into a player that can help a team. Now he can bring something other than scoring.

Mo's Wizards totals: 14 points, 8 rebounds, 2 assists, 7 steals, 1 turnover in 67 minutes. Needs to shoot better and to correct the mechanics on his shot, but you KNOW the guy can score.

[Chocolate City Jordanaire]

The day Kevin Harlan, while covering a Wizards game said, "Michael and the Jordanaires," I darn near fainted. I thought I came up with the nickname, until doclinkin pointed out there was a country and western group called the Jordanaires.

Still, I think the guy probably saw the name "Jordanaires' and thought it would be a good way to associate all things involved with the Wizards. That's what I thought back when MJ was running the Wizards. It really worked out well the years when Eddie was the coach. The name still had relevance.

Now, I feel really good about the way the Wizards are going. I do feel like fans can have a voice.

[Nivek]

It's definitely a system that I dig, but like the man himself I'm a little concerned about the component of the system that deals with rating a players physical tools and all that mental makeup jazz in that it seems to be working pretty well retrospectively (which is how you do this stuff, try to develop a system that predicts the past really well) but it's hard to not be influenced by what you know of a player's career when you're trying to retroactively shrink his head as a collegiate. I mean if you're rating Drew Gooden or someone do you rate him as "sort of a basketball moron" based on what he did at Kansas, was that evident at Kansas? Anyway this isn't news to Kev, but intangibles drool anyway.

I kinda addressed this in my last post, but yeah -- this is the part that worries me most. First, on the level of -- I'm still not sure that athleticism is a key predictor of success in the NBA beyond a certain minimum threshold. A productive player who's also a great athlete? Sign him up. A productive player who's not a great athlete? He's doing something right. A great athlete who's not productive? Well, why not? That athleticism is supposed to give him an advantage. If it's not, there's some kind of disconnect that needs to be understood.

Second, rating players retrospectively can get dicey when there are no measurements available. I do my best not to let what a guy did in the NBA affect how I do that retro-assessment. When in doubt, I default to neutral. In the case of a guy like Drew Gooden (who is not in YODA yet), my standard would be to rate him only on the information that was available before the draft. So, no -- he wouldn't get dinged for becoming a "basketball moron" in the NBA.

The impossible thing to predict is a guy's work ethic once he reaches the NBA. Everyone who gets into the league still needs to get better. I think there are LOTS of guys who can be quality players if they work hard enough. Karl Malone, for example, had grade consistent with a mid-2nd round pick (in a 60-player draft), but worked his ass off (and maybe used steroids) to become an all-time great. Michael Redd rated as a borderline 2nd rounder, and improved when he got to the league.

Conversely, Mike Sweetney had a top 10 grade and ate his way out of the league. Marcus Fizer was productive against good competition as a junior -- YODA liked him as a mid-first round pick. He had a couple decent years, but basically a meh career. (Although, at least YODA wouldn't have picked him 4th overall.)

[Chocolate City Jordanaire]

I recall posting that Micheal Sweetney would be ROY. I missed worse on Sweetney than with any other player. That season I watched a lot of NCAA ball, and when Georgetown played Carmelo's Syracuse team, Sweetney had a huge game. IIRC Sweetney had a monster NIT. I thought he would be great in the NBA, better than Carmelo. Man was I wrong! I still think Sweetney can play if he can ever get his weight down. The talent is there for him to be a solid NBA role player. (Another player who it looked like I missed on can be at least a role player, Morris Almond).

Nivek, I missed on Fizer, too. Another similar player destroyed D-League and did get a cameo with the Nets this season, Andre Emmett. I felt like my projection of how Emmett would do in the NBA missed.

The numbers from NCAA ball don't tell everything.

Sometimes players get caught in a numbers game with the teams they go to. Sometimes the system is wrong for their game or they have to play a role they are not comfortable with. Fizer is not much of a defender IIRC, for example. Players who bust or who do not meet expectations usually have a lack of athleticism necessary to compete at a high level in the NBA.

[montestewart]

The impossible thing to predict is a guy's work ethic once he reaches the NBA. Everyone who gets into the league still needs to get better. I think there are LOTS of guys who can be quality players if they work hard enough. Karl Malone, for example, had grade consistent with a mid-2nd round pick (in a 60-player draft), but worked his ass off (and maybe used steroids) to become an all-time great. Michael Redd rated as a borderline 2nd rounder, and improved when he got to the league.

Would be interesting to look at a broad range of players who, like Redd, Malone, Drexler, Stockton, etc. were chosen appropriately or too low (or even too high, as was perhaps the case with Malone) based on your system, but in hindsight turned out to be bargains at their draft position, to see if there are recurring commonalities that might enable better predicting of work ethic and/or other attributes that contribute to long term success.

After the Celtics victory, I read something (can't remember where) on Paul Pierce that mentioned him being drafted too low based on his physical specimen, apparently overlooking intangibles that might have predicted his later success. No idea whether there was anything to support that position.

[payitforward]

Nivek, if you get a chance can you see what Chris Jackson / Mahmoud Abdul-Rauf looked like in YODA?

...YODA rates him ...

Is YODA published? Or, rather, I suppose I just mean is it "public?" I.e. will you share it? And if so, does it require software to compute (i.e. regression analysis, etc.)?

If not public, have you at least described it somewhere and to some degree? Love to learn more.

Thanks!

[payitforward]

oooops! Sorry to miss your post above, Nivek.

[Nivek]

YODA had Emmett with a late first rating, but that's without benefit of physical measurements beyond standing reach (which was about average for a SG).

[Nivek]

oooops! Sorry to miss your post above, Nivek.

I thought you were almost always correct.

First the Fat Boys break up, now this.

[Chocolate City Jordanaire]

The impossible thing to predict is a guy's work ethic once he reaches the NBA. Everyone who gets into the league still needs to get better. I think there are LOTS of guys who can be quality players if they work hard enough. Karl Malone, for example, had grade consistent with a mid-2nd round pick (in a 60-player draft), but worked his ass off (and maybe used steroids) to become an all-time great. Michael Redd rated as a borderline 2nd rounder, and improved when he got to the league.

Would be interesting to look at a broad range of players who, like Redd, Malone, Drexler, Stockton, etc. were chosen appropriately or too low (or even too high, as was perhaps the case with Malone) based on your system, but in hindsight turned out to be bargains at their draft position, to see if there are recurring commonalities that might enable better predicting of work ethic and/or other attributes that contribute to long term success.

After the Celtics victory, I read something (can't remember where) on Paul Pierce that mentioned him being drafted too low based on his physical specimen, apparently overlooking intangibles that might have predicted his later success. No idea whether there was anything to support that position.

Jeremy Lin didn't look like an NBA player. He could have thoroughly outplayed Wall three straight times and nobody would have even said he's a first rounder. Also, Jeremy Lin went to Harvard. Who gets drafted from there? I think the average mind looks for familiar patterns of appearance. Most go along with the crowd. Players from the big schools that are on television always get the benefit of a doubt and they get drafted too high. I'm surprised Damian Lillard is going in the lottery, coming from Weber State. I am NOT surprised by the Austin Rivers hype. Most basketball fans are not all that well informed. The players who get drafted too low usually are just relative unknowns, or they don't look the part to scouts.

Other guys who get drafted too low are the ones who have some negatives. Four-year seniors always have their share of critics. I doubt Jared Sullinger is worse as a sophomore than he was as a freshman, but right now his draft stock is in free fall. Perry Jones could be the same way. Other negatives could be lack of team success. Dion Waiters would have gone mid lottery IMO if Syracuse didn't fall apart after Melo was ruled ineligible.

Here is what might make certain players in this draft go too low. Not my words, but what their critics are saying:

-Jae Crowder is not an elite athlete at SF. He doesn't have a good handle.
- Draymond Green is too short and will have to make a position.
- Marcus Denmon is not a pure PG
- Will Barton is too thin, not a good three point shooter
- Kevin Jones only did it one year at WVU.
--John Henson is way too skinny

monte, hopefully, the players above turn out to be pretty good players. I bet several will be drafted too low.

[payitforward]

Catching up on a few things:

1. The Jordannaires backed up Elvis (in case that hadn't been mentioned).

2. YODA -- Alas, I'm going to be skeptical of any prognostication system that tries to factor in the significance of "height, length..." etc. There is no data suggesting they have *any* significance at all. Do a bell-shaped curve of SG productivity, in other words, and it will show a *random* distribution of height, wingspan, etc. Random. Ditto for other positions. Do the league's tallest, longest teams have the best win-loss record?

3. Ben Gordon -- this guy is *way* overpaid. You cannot create a good team by accumulating overpaid players because they have "skills" you need. In fact, in a future post, I'm going to try to debunk the whole notion that you build a good team *at all* by trying to get all the "skills" you think are necessary. Coming soon (but probably not today).

4. If we get Davis, can we trade Booker? Yeah man!!

[DCZards]

Sometimes players get caught in a numbers game with the teams they go to. Sometimes the system is wrong for their game or they have to play a role they are not comfortable with.

I'm believe this is the case with Austin Rivers. Yes, he's immature, selfish and lacking some basic bball skills. He should have stayed in college for at least another year to grow up and work on those skills. But with his competitiveness, quickness, bball IQ, reported work ethic and his pedigree I'm betting that Rivers is going to be a very, very good NBA player, maybe one of the best 6-7 players to come out of this draft. Doc's son will at least be worthy of the top 10-15 pick that most mock drafts that I've seen have him slotted at.

[Zonkerbl]

A few comments:

The statistically rigorous thing to do is choose a metric that is highly correlated with success as an NBA player (reg season minutes played + playoff minutes played) and regress that against all the historical (pre-NBA career) data you can muster, including interactions (points*rebounds, for example), on as many NBA players as you can get data for. The indicators that contribute significantly to NBA success will rise to the top, and the regression coefficients will tell you what weights to use.

I would include anything measurable -- SAT scores, GPA in college, whether they grew up in an urban or rural environment, single parent household, parents' income, anything you can get. Body mass index? Anything.

Another thing: There are lots of failures. Lots of Darko Milicic's and Kwame Brown's and Sweetneys. What predicts failure?

For Darko the big red flag was he couldn't get playing time on his European team. Kwame Brown was obviously a numbskull and that's probably reflected in his SAT score. Sweetney had weight issues.

By the way -- running the regression is easy, collecting the data is not. If someone is willing to collect the data I'd be willing to do the regression for you.

[Ed Wood]

Thanks Zonker, that's what I was trying to convey with the post about analysis of variance, that you can basically plug whatever you want in so long as you can convert it into some sort of discrete variable and that actually running the test to determine whether or not a dude's hair style significantly co-varies with his performance ain't not thing, the real challenge is finding a reliable source for all of the biographical stuff and having the time and inclination to record it all.

Plus there is the issue of what exactly you'd want to use as your measurable for NBA quality. I can see the appeal of keeping things simple with something like playing time but that's not going to weed out guys like Al Thornton or Yi who've managed to accrue a whole bunch of playing time by being young and ostensibly promising on really terrible teams while really being absolute balls as NBA players and actively (unless Yi's adjusted plus minus wizardry isn't a statistical mirage) making those bad teams worse. Maybe using a large enough sample would marginalize that sort of stuff enough that you'd get what you wanted but maybe some sort of weighted minutes played value with an adjustment using something akin to ws48 or PER or adjusted +/- or whatever your preferred rate-based poison.

[pancakes3]

don't be too quick to say "any" significance. it's not a random (read: uniform) distribution. there is a bell curve.

[nate33]

2. YODA -- Alas, I'm going to be skeptical of any prognostication system that tries to factor in the significance of "height, length..." etc. There is no data suggesting they have *any* significance at all. Do a bell-shaped curve of SG productivity, in other words, and it will show a *random* distribution of height, wingspan, etc. Random. Ditto for other positions. Do the league's tallest, longest teams have the best win-loss record?

Height and length certainly have a significance. Burly, floorbound 6-7 big men routinely dominate the college ranks (think Lonny Baxter) but rarely make an impact in the pros. The college ranks are littered with 6-2 shooting guards who can't crack an NBA rotation unless they learn PG skills.

I agree that height and length can be overweighted when evaluating a prospect. You don't pick Player A over Player B solely because Player B's wingspan is 2 inches longer. However, you do need to verify that a player has the minimum size and athleticism to compete in the league. You don't want a 6-1 SG or a 6-8 C, even if their numbers were a bit better than their taller brethren.