2014 Draft Projection Models

[VJL]

Re: Wiggins

Just to help think about it. If I bump Wiggins' STL rate from 1.1 to 2.0 his score jumps up to 7.7. Still behind quite a few guys, but in the range where he may still be a star. That said, once we get to that level of assessment we also need to talk about how dependent Wiggins is on transition offense for whatever semblance of scoring efficiency he has shown. This is a bad thing (http://www.hickory-high.com/?p=11629). So while I think we can cut Wiggins some slack based on system issues, I am equally ready to dock him for his terrible performance in the half-court which is hidden by transition scoring.

[Dr Positivity]

Good call. I was wrong to paint Lillard as a case of the "surprise shooter". Nearly 40% from range on 7 attempts and 88% FT is elite.

re: Parsons and Lowry
Parsons is an excellent example. He fooled the model completely (1.4 EWP). Terrible FT shooter, mediocre 3pnt shooter.. not to mention his block and steal rates. The model liked Lowry quite a bit (6.4), but mostly because of the steals. I agree that the shooter was difficult to predict and largely account for his surprising level of impact.

On the general point... I do think one of the better drafting heuristics is to look for guys who are a good shot away from being great players. Andre Roberson was my guy who fit that mold in the last draft. All he needs is a corner three, which there is reason to believe he can do, and you have a starting 3. Dipo is another one who makes sense. Noah Vonleh is the guy I would highlight for that idea in the draft.

I agree but I feel virtually all players have a "shooting upside" improvement in them anyways. For example take 3 shooters in 2014 draft - Smart (poor shooter), Wiggins (decent shooter), Stauskas (terrific shooter). It is true that Smart can turn it around and become a pretty decent 3pt shooter. However the probability of Smart doing this, may be the same as Wiggins making the leap from decent shooter to a Paul George like threat where he's a legitimately awesome shooter. And may be similar to the probability of Stauskas making the leap from great shooter, to being an insane shooter like Klay Thompson and Damian Lillard. So I don't really use it as a reason to like Smart more than the other two. Because a great shooter can make a very similar "out-doing of expectations" (and Klay is a great example of this actually. He would have gone higher if rated this elite of a shooter). However, I guess you could make a case for the chance Smart jumps all the way up to the George to Klay group, or Wiggins jumps up to the Klay group, but the probability of both is just too low to count on. Also admittedly when a shooter looks as good as a Steph Curry or Kevin Durant in college it's also fair to say they don't apply to the "they could be --> in the category" since they were already expected to just about max out

Or to put it another way. In my system I am going to rate Smart a 4 Wiggins a 6 and Stauskas an 8 as shooters (on a scale to 11, don't ask why). So because Wiggins has the longshot chance to get to 8 or 9 and Stauskas to 10 or 11 it means as much as Smart's puncher's chance at getting to 6 or 7

[deanondraft]

I think VJL's point makes a lot of sense and it's something I had been considering ever since THJ has been so much better than expected.

On a scale of 1-11 Stauskas has to be at least a 10. He can improve his FT% and likely will, but he's shooting 44% from 3 in college. It's not like he's going to shoot 50% from 3 as a pro, whereas a 34% college shooter could end up as a 40% NBA shooter.

[deanondraft]

so i just posted on my blog about how frequent technical fouls correlate positively with advanceds stats in the NBA, especially on defense: http://deanondraft.com/2014/02/10/check ... us-smarts/

i don't know how much you can do with this, but if you want to try to play around with prospect technicals statsheet keeps track of em going back to 2000-2001:

http://statsheet.com/mcb/players/stats/ ... -2014&min=

[VJL]

Dean:

That is pretty interesting, and at least a little surprising. I don't think I will work Ts into the model, but I will keep my eye on that pattern. I share your interest in getting at the mental aspect of prospects, I really think there is a lot there. Many of the extreme cases of a player not living up to his projections could be explained by either off-court issues or "big and athletic guy who never really loved basketball". I imagine you don't get Ts if you don't care.

[deanondraft]

I agree that is the underlying logic. The correlations I got could be that stars get more technicals because they have the sense of entitlement of being an NBA star, so those correlations may have some selection bias. I'm trying to do a prospect analysis here, and just looked at 2001-2007 I separated technical receiving players into value and busts arbitrarily based on my perception.

Value:
Yr Pick Player WS WS/48 Techs
2005 4 Chris Paul 110 0.246 3
2007 2 Kevin Durant 82.8 0.203 2
2003 3 Carmelo Anthony 79.1 0.137 3
2001 13 Richard Jefferson 74.9 0.124 3
2003 18 David West 67.1 0.14 4
2005 30 David Lee 62.1 0.146 2
2001 25 Gerald Wallace 59.4 0.118 2
2004 26 Kevin Martin 57.1 0.148 3
2001 14 Troy Murphy 50.3 0.121 2
2007 9 Joakim Noah 45.3 0.168 2
2003 29 Josh Howard 38 0.119 7
2005 1 Andrew Bogut 36.9 0.115 2
2002 45 Matt Barnes 31.5 0.102 3
2007 31 Carl Landry 30.8 0.153 2
2007 22 Jared Dudley 28.5 0.111 3
2006 11 J.J. Redick 28 0.13 2
2005 21 Nate Robinson 27.7 0.1 2
2003 38 Steve Blake 25.1 0.068 4
2005 37 Ronny Turiaf 18.6 0.115 2
2003 8 T.J. Ford 16.9 0.068 2
2002 52 Rasual Butler 16.4 0.052 2
2002 49 Darius Songaila 15.9 0.083 3

Bust
Yr Pick Player WS WS/48 Techs
2004 3 Ben Gordon 33 0.081 2
2006 3 Adam Morrison -1.4 -0.021 4
2006 4 Tyrus Thomas 12.9 0.078 2
2007 7 Corey Brewer 11.3 0.049 1
2001 7 Eddie Griffin 10.9 0.078 2
2006 8 Rudy Gay 33.9 0.081 2
2004 8 Rafael Araujo -0.4 -0.013 5
2003 9 Mike Sweetney 7.8 0.103 5
2006 12 Hilton Armstrong 3.6 0.053 3
2003 15 Reece Gaines -0.6 -0.049 2
2004 16 Kirk Snyder 4.5 0.061 5
2006 17 Shawne Williams 4.9 0.056 2
2002 19 Ryan Humphrey -0.4 -0.027 4
2006 20 Renaldo Balkman 7.2 0.113 2
2005 20 Julius Hodge 0 0.003 4
2001 22 Jeryl Sasser 0.7 0.032 2
2001 23 Brandon Armstrong -0.6 -0.039 2
2007 25 Morris Almond 0 0.005 1
2004 29 David Harrison 3.2 0.057 4

There were some early 2nd rounders I left off the bust list (Daniel Ewing, Omar Cook, Vincent Yarbrough, Robert Archibald, Steve Logan, James White all picked in the 29-32 range and got multiple T's in college). But to my naked eye, the overachievers in the value section appeared to overachieve more than the busts busted. Not sure how to quantify this though- I wanted to measure vs player stats compared to average at each draft slot, but it will be tainted by guys like Araujo being overdrafted and Bogut being a weak #1 overall pick even though he actually panned out. Hard to do much with it given such a SSS though

[deanondraft]

Is there a way to squeeze Euro guys into the model?

I'd be really interested to see how Nurkic, Saric, and Capela measure up with various SOS's (say +10, +15, and +20 if we are measuring in terms of SRS). I know it's not the best form of analysis since it's a diff game in Europe, but I suspect Nurkic will break your model and I'd like to know by how much

[Statman]

Is there a way to squeeze Euro guys into the model?

I'd be really interested to see how Nurkic, Saric, and Capela measure up with various SOS's (say +10, +15, and +20 if we are measuring in terms of SRS). I know it's not the best form of analysis since it's a diff game in Europe, but I suspect Nurkic will break your model and I'd like to know by how much

I plan on eventually getting foreign players into my model - but I'd probably need some backing by then to get all the data, there are more than a few leagues worldwide. It'd be a huge but interesting (to me) project.

[VJL]

Would love to add in the Euros, but it would be a big task. Pelton did Euros last season and I like his methodology. Gets roughly similar results for college guys compared to my model, so for now I am just happy trusting him. Unfortunately, that won't be coming until much closer to the draft.

[Statman]

Would love to add in the Euros, but it would be a big task. Pelton did Euros last season and I like his methodology. Gets roughly similar results for college guys compared to my model, so for now I am just happy trusting him. Unfortunately, that won't be coming until much closer to the draft.

I don't think there is any chance I'll tackle international players this draft - I'd need to compile league totals of every league I have a player I'd want to consider (need to adjust for that leagues pace and statkeeping - like assists being credited much less).

I kinda want to look at Exum, if I can properly get all his league's player stats - I could at least see where he ranks in his league rating wise at his young age - and see how his ratings beak down after all the adjustments. If there are a multitude of under 22 year old players in the league whose ratings are about as good or better than his - OR if he has a bad statistical outlier (say very low rebounding rate and/or defensive stop rate, after adjustments) - then I could say I wouldn't draft him high, but that's probably the best I could do this season.

Now, if Exum is better than every guy in that league under 23 or even 22, and no red flags on his statistical skillset breakdown - then I guess he very well may be worth a top pick.

[Dr Positivity]

The sample size for these international leagues is just too small. Only a select number of players have gone Euroleague or ACB to NBA and of course, usually at ages different than the prospect. So even studies guys like Hollinger have done on Euro to NBA translation was built on too tiny a SS

[Statman]

The sample size for these international leagues is just too small. Only a select number of players have gone Euroleague or ACB to NBA and of course, usually at ages different than the prospect. So even studies guys like Hollinger have done on Euro to NBA translation was built on too tiny a SS

Agreed.

But, you can see where that player ranks within the league he played and especially where he ranked in his age group. You can also create a similarity score (after all adjustments) to actual NBA players - not as much in terms of quality but in terms of style. The biggest thing is - if there are many players in his own league at a similar age that rate out as well as he does - he shouldn't be seen as a high level prospect. So, some foreign guys can maybe be weeded out by the analytics, while others may still look good enough to remain at least as draftable prospects.

Like I said before - if I (or probably any number of us here) worked for the Timberwolves a couple years back - they wouldn't have ever considered drafting an undraftable 27 year old guy from Africa - because the moment the scout offered him up as a prospect, the ratings would have been done for his league, and the first thing that we would have found is his real age (since the two sites I went to to look up his stats had his real age).

Of course, that scout could have done 5 minutes of extra online research - but hey, he gets paid to travel the world and look at players, not play on the internet.

It seems so backwards to me - why not have a guy spend a month or so crunching the #'s and rate EVERY player in EVERY pro league that MIGHT have a prospect - confirm the ages of the ones that "stand out", and THEN send scouts to scout them - going first to the areas that have the most possible quality prospects. Many early prospects can be found and ratings done w/ all the U/16, U/18, etc international competitions also, although the sample sizes are much smaller. As years go by, prospect trends can be tracked. If a scout finds a guy not on his list at some event - that guy/league can be added to be tracked in terms of his ratings. There wouldn't be the not knowing who to draft in the 2nd round - so go with that guy Tom saw for one half in a game in Africa.

[Dr Positivity]

Well TWolves scouting is something else, lol. The only thing that could've made David Kahn's draft history more depressing is when he had an interview saying the Jonny Flynn pick wasn't his fault, because all his scouts said he was the top PG talent in the draft and Kahn said ok, I just got here so I'll go with what my experienced scouts think (Notably, as a Raptors fan that's exactly what happened with Rob Babcock and the Araujo pick too, he got pushed into it having done almost no work personally on the draft. Hiring a first time GM a few weeks before a draft is a bad idea!)

[Bobbofitos]

It seems so backwards to me - why not have a guy spend a month or so crunching the #'s and rate EVERY player in EVERY pro league that MIGHT have a prospect - confirm the ages of the ones that "stand out", and THEN send scouts to scout them - going first to the areas that have the most possible quality prospects. Many early prospects can be found and ratings done w/ all the U/16, U/18, etc international competitions also, although the sample sizes are much smaller. As years go by, prospect trends can be tracked. If a scout finds a guy not on his list at some event - that guy/league can be added to be tracked in terms of his ratings. There wouldn't be the not knowing who to draft in the 2nd round - so go with that guy Tom saw for one half in a game in Africa.

this seems like the intelligent way of going about things, of course.

[Dr Positivity]

Anyways if you're interested I posted a big board - tried to include all the relevant players I could right now, ended up with 36 rated players http://asubstituteforwar.wordpress.com/ ... ch-update/ My methodology rates talent level and not if a player is going to reach it, albeit I believe if a player is in the NBA long enough (I said 6,000 minutes+ in the article) the likelihood of him reaching his talent level or near it becomes very high. (in reality there's probably less than 15-20 players in the NBA who aren't killing machines in terms of work ethic/conditioning/on court effort/etc. - once you get past Beasley, Charlie, Blatche, Josh and JR Smith, Bynum, Bargs, Gay, Green, Hedo, Javale, Cousins, the list of enigmatic character players almost entirely dries up). From there it's a matter of actually rating talent correctly, which is its own debate - my opinion is skill level and instincts make up 2/3s of talent, which some people may disagree with - while within physical tools, I tend to rate length as a smaller portion than others (Interestingly, if one considers physical tools 1/3 of talent level, then length maybe 1/3 of physical tools as sharing importance with athleticism and strength, you get a number like 1/9 of the total package for length. A very small number considering the draft sites that treat length as singularly important to make or break a prospect on its own. But there is a reason why Kevin Love is a superstar and Hasheem Thabeet can't play)

Top 5

1. C Joel Embiid
2. SG Nik Stauskas
3. PF Noah Vonleh
4. PF Julius Randle
5. SG Jordan Adams