APBRmetrics

I know it's a bit early, but I've released my top 100 NBA draft prospects based on a series of regressions I did. The methodology is somewhat similar to Hollinger's Draft Rater with the biggest difference being I used RAPM (for those who have played long enough) and blended SPM/RAPM (for those who haven't) rather than PER. My predictors could be different to (I don't think he ever released exactly what he used). Anyway, I thought everyone might find this interesting, so here it is. I plan on updating it once a week.

http://shutupandjam.net/draft-rankings/

James

jbrocato23 wrote:I know it's a bit early, but I've released my top 100 NBA draft prospects based on a series of regressions I did. The methodology is somewhat similar to Hollinger's Draft Rater with the biggest difference being I used RAPM (for those who have played long enough) and blended SPM/RAPM (for those who haven't) rather than PER. My predictors could be different to (I don't think he ever released exactly what he used). Anyway, I thought everyone might find this interesting, so here it is. I plan on updating it once a week.

http://shutupandjam.net/draft-rankings/

James

When I get deep into my draft stuff a month before the draft - it'll be interesting to see how our over & under rated match up compared to mocks.

James, are your projected NBA impact ratings per 100 possessions?

Thanks for the data. I will look through it later.

Crow,

Yes the projections are impact per 100 possessions.

Statman,

Absolutely. I'll be interested to see what you come up with.

Those numbers on the top guys seem way too high to be rookie year expected values.

You have about 15 guys with numbers higher than any rookie this year.

There have only been 3-4 rookie seasons in the past 10 years as good as the top 3-4 expected values in your list. Lebron, Durant, etc were nowhere near that good in their first year in the NBA.

talkingpractice,

I'm not trying to predict rookie year values, I'm trying to predict career values.

Edit: I also wanted to add that the numbers are high even for most years right now because it is still early in the season. I fully expect the top values to regress to the mean a little by the time the season ends (for example, I highly doubt four guys will be 2+).

James

jbrocato23 wrote:talkingpractice,

I'm not trying to predict rookie year values, I'm trying to predict career values.

Ahhhhhhhhhhhhhhhhhhhhhhh, I getcha now. What more precisely do you mean by career values? Is your number their expected max/prime value, or their average career value?

talkingpractice wrote: Ahhhhhhhhhhhhhhhhhhhhhhh, I getcha now. What more precisely do you mean by career values? Is your number their expected max/prime value, or their average career value?

Average career values. In my dataset, for example, only nine players in the past 11 years project as +3 or better: Kyrie, Durant, Wade, Anthony Davis, Kevin Love, Chris Paul, Derrick Rose, Jrue Holiday, and Demarcus Cousins. Kyrie might need an asterisk because he played all of 300 minutes in college. I know I have three people projected as +3 or better in this draft class right now, but again, I don't expect that to be the case at the end of the season. Edit: Added some guys I somehow left out...

As the draft nears, these numbers continue to solidify. I'm pretty confident at this point that Otto Porter, Marcus Smart, and Nerlens Noel (assuming he recovers from his injury reasonably well) are the three best prospects in this draft.

I see you have Archie Goodwin's projected NBA impact at a level that yields a 28th draft ranking. I guess that could happen. That projection seems much wiser than draftexpress's current listing at 18th (and I think he was higher earlier), though I would definitely be looking at many other players even at 28 as alternatives to him. I've seen a number of his games and I don't think he is anywhere close to being a NBA level contributor yet and your NCAA impact estimate agrees with that. Decision-making, shot selection and accuracy, defensive attention & effort and overall game to game consistency are all issues. His NBA impact projection is much above his NCAA impact and that might be appropriate all things considered; but my gut reaction based on limited information is that his NBA impact projection still be on the high side of my expectations for him.

jbrocato23 wrote:As the draft nears, these numbers continue to solidify. I'm pretty confident at this point that Otto Porter, Marcus Smart, and Nerlens Noel (assuming he recovers from his injury reasonably well) are the three best prospects in this draft.

Trey Burke?

Impact true PGs cannot be ignored.....

Statman wrote:
Trey Burke?

Impact true PGs cannot be ignored.....

Yes, and Trey Burke. Who I coincidentally just wrote a post about: http://shutupandjam.net/2013/03/13/unde ... rey-burke/

Crow wrote:I see you have Archie Goodwin's projected NBA impact at a level that yields a 28th draft ranking. I guess that could happen. That projection seems much wiser than draftexpress's current listing at 18th (and I think he was higher earlier), though I would definitely be looking at many other players even at 28 as alternatives to him. I've seen a number of his games and I don't think he is anywhere close to being a NBA level contributor yet and your NCAA impact estimate agrees with that. Decision-making, shot selection and accuracy, defensive attention & effort and overall game to game consistency are all issues. His NBA impact projection is much above his NCAA impact and that might be appropriate all things considered; but my gut reaction based on limited information is that his NBA impact projection still be on the high side of my expectations for him.

Yeah, my best advice to Goodwin would be to stay in college for another year. Unless of course he's actually not very good, then maybe he should come out now before his stock drops to much, haha. That's the tricky thing about being a one-and-done who doesn't do particularly well your only year in college. I, for one, wouldn't take this type in the draft - I'd rather they prove that they can be productive.

What are you using for NCAA impact? ASPM?

What did the top 20 for the 2012 class look like last year?