Moderators Popular Post AHF Posted May 27, 2014 Moderators Popular Post Report Share Posted May 27, 2014 (edited) I just took a look at this since it has been a frequent subject of discussion on here. For practical purposes, I used Win Shares and Win Shares/48 as the two measuring sticks. To control for draft pick, I am using a net rank number to review so the calculation works like this: Kawhi Leonard - Drafted #15 - Career Win Share Rank In Draft Class #1 = +14 Lebron James - Drafted #1 - Career Win Share Rank In Draft Class #1 = 0 Danillo Gallinari - Drafted #6 - Career Win Share Rank In Draft Class #10 = -6 So for each draft class, I ranked them by WS and WS/48 and did a net of their draft spot against their performance ranking. I then added the numbers together for each class and totaled them for a decade worth of draft picks for the 2002 through 2011 drafts. Here are the net numbers: 2002 - 2011 Drafts College Players Net WS Rank: +18 College Players Net WS/48 Rank: -6 Note: College players underperformed their draft slot on a per minute basis but totaled enough production to outperform their draft slot in total win shares. International Players Net WS Rank: -39 International Players Net WS/48 Rank: +2 Note: International players outperformed their draft slot on a per minute basis but the cost for missed years overseas was underperforming on total win share production over their careers. High School Players Net WS Rank: +21 High School Players Net WS/48 Rank: +4 Note: High school players were the only group to both outperform their draft slot on a per minute and career total basis. 2002 - 2011 Draft Volatility: College Players Positive WS Rank Drafts: 4 positive, 6 negative College Players Positive WS/48 Rank Drafts: 6 positive, 4 negative Note: This was interesting as it was the opposite of what you would expect. While college players were net positive over 10 total drafts in career production, they were net positive in only 4 of those 10 drafts. While they were net negative over 10 years in WS/48 rank, they were positive in 6 of 10 drafts. International Players Positive WS Rank Drafts: 5 positive, 1 negative International Players Positive WS/48 Rank Drafts: 4 positive, 5 negative, 1 neutral Note: Pretty much a wash. High School Players Positive WS Rank Drafts: 2 positive, 2 negative High School Players Positive WS/48 Rank Drafts: 2 positive, 2 negative Note: A wash. Conclusion: I really don't see much difference between the college and international players here as far as draft value for time on the floor. The difference comes into play with the international players playing fewer career minutes due to additional time spent overseas which results in couple of draft slots value gap for career win share numbers. Edited May 27, 2014 by AHF 6 Link to comment Share on other sites More sharing options...
Moderators AHF Posted May 27, 2014 Author Moderators Report Share Posted May 27, 2014 (edited) The focus here was not so much on whether a player from a given category was more likely to be a star or not based on them being drafted in round 1 from college, high school or the international pools. It was really more on how picks from those pools performed against their draft slot. The reason I did that is because expectations should differ significantly based on draft slot and draft class. More should be expected from a top 10 pick than a pick in the 20s. What might be good value in a terrible draft class might be terrible value in a great one. Here is are a few players from the 2003 draft class to give an example for methodology: Draft Slot -- Player -- Career WS --Class WS Rank -- Net Draft Slot/Class Rank 3 -- Carmelo Anthony -- 83.1 WS -- 4 -- (-1) 18 -- David West -- 70.1 WS -- 5 -- +13 7 -- Kirk Hinrich -- 50.9 WS -- 6 -- +1 21 -- Boris Diaw -- 43.1 WS -- 7 -- +14 12 -- Nick Collison -- 41.0 WS -- 8 -- +4 29 -- Josh Howard -- 38 WS -- 9 -- +20 9 -- Mike Sweetney -- 7.8 WS - 21 -- (-12) 2 -- Darko Milicic -- 7.1 WS -- 22 -- (-20) 19 -- Sasha Pavlovic -- 7.0 WS -- 23 -- (-4) 17 -- Zarko Planinic -- 1.1 WS -- 26 -- (-9) 26 -- Ndubi Ebi -- 0.1 WS -- 27 -- (-1) 15 -- Reece Gaines -- -0.6 WS -- 29 -- (-14) Edited May 27, 2014 by AHF 1 Link to comment Share on other sites More sharing options...
Admin Posted May 27, 2014 Report Share Posted May 27, 2014 Uh oh using Win Shares to prove something. Certain people aren't going to like that. Clever way of ranking these guys though and good work on putting it all together. Link to comment Share on other sites More sharing options...
Moderators AHF Posted May 27, 2014 Author Moderators Report Share Posted May 27, 2014 Uh oh using Win Shares to prove something. Certain people aren't going to like that. Clever way of ranking these guys though and good work on putting it all together. I wasn't going to manually enter a whole lot of data into a spreadsheet. That would take more time than I am willing to devote. In contrast, I could pull draft slot, WS and WS/48 off of basketball reference without too much work. I could drop that info into excel and then sort and generate rank data without too much manual typing. That also let me do it for an entire decade to get a better sample size. Someone who prefers another metric can do that one if they want. PER and WS are not perfect numbers by any stretch of the mind but they have the advantage of being intended for use by themselves. Other numbers are context dependent. If you looked at points, for example, you need to look at TS% for efficiency before that means much (see Kyle Korver vs. Evan Turner) and you will probably want to also compare that against points/minute. If you look at assists, you need to start looking at turnovers, assist %, usage rate, etc. to get more of a picture of whether this is someone making creative passes and protecting the ball or someone who simply hogs the ball and therefore ends up with a good number of assists by virtue of sheer volume of touches (i.e., the Stephon Marbury / Allen Iverson type of non-playmaking assist totals). This isn't a definitive study or anything, but intended to be food for thought. 1 Link to comment Share on other sites More sharing options...
Admin Posted May 27, 2014 Report Share Posted May 27, 2014 Can you do it for PPG? That's the only thing that matters :saythat: Link to comment Share on other sites More sharing options...
TheNorthCydeRises Posted May 27, 2014 Report Share Posted May 27, 2014 (edited) I just took a look at this since it has been a frequent subject of discussion on here. For practical purposes, I used Win Shares and Win Shares/48 as the two measuring sticks. To control for draft pick, I am using a net rank number to review so the calculation works like this: Kawhi Leonard - Drafted #15 - Career Win Share Rank In Draft Class #1 = +14 Lebron James - Drafted #1 - Career Win Share Rank In Draft Class #1 = 0 Danillo Gallinari - Drafted #6 - Career Win Share Rank In Draft Class #10 = -6 So for each draft class, I ranked them by WS and WS/48 and did a net of their draft spot against their performance ranking. I then added the numbers together for each class and totaled them for a decade worth of draft picks for the 2002 through 2011 drafts. Here are the net numbers: 2002 - 2011 Drafts College Players Net WS Rank: +18 College Players Net WS/48 Rank: -6 Note: College players underperformed their draft slot on a per minute basis but totaled enough production to outperform their draft slot in total win shares. International Players Net WS Rank: -39 International Players Net WS/48 Rank: +2 Note: International players outperformed their draft slot on a per minute basis but the cost for missed years overseas was underperforming on total win share production over their careers. High School Players Net WS Rank: +21 High School Players Net WS/48 Rank: +4 Note: High school players were the only group to both outperform their draft slot on a per minute and career total basis. 2002 - 2011 Draft Volatility: College Players Positive WS Rank Drafts: 4 positive, 6 negative College Players Positive WS/48 Rank Drafts: 6 positive, 4 negative Note: This was interesting as it was the opposite of what you would expect. While college players were net positive over 10 total drafts in career production, they were net positive in only 4 of those 10 drafts. While they were net negative over 10 years in WS/48 rank, they were positive in 6 of 10 drafts. International Players Positive WS Rank Drafts: 5 positive, 1 negative International Players Positive WS/48 Rank Drafts: 4 positive, 5 negative, 1 neutral Note: Pretty much a wash. High School Players Positive WS Rank Drafts: 2 positive, 2 negative High School Players Positive WS/48 Rank Drafts: 2 positive, 2 negative Note: A wash. Conclusion: I really don't see much difference between the college and international players here as far as draft value for time on the floor. The difference comes into play with the international players playing fewer career minutes due to additional time spent overseas which results in couple of draft slots value gap for career win share numbers. This is the kind of stuff I used to do and got blasted for. Glad to see you come up with your own "unscientific" statistical method to come up with a decent analysis. Edited May 27, 2014 by TheNorthCydeRises Link to comment Share on other sites More sharing options...
Admin Posted May 27, 2014 Report Share Posted May 27, 2014 This is the kind of stuff I used to do and got blasted for. Glad to see you come up with your own "unscientific" statistical method to come up with a decent analysis. NCJSI anyone? Link to comment Share on other sites More sharing options...
DrReality Posted May 27, 2014 Report Share Posted May 27, 2014 Think I slept through this in stats class? Link to comment Share on other sites More sharing options...
Moderators AHF Posted May 27, 2014 Author Moderators Report Share Posted May 27, 2014 This is the kind of stuff I used to do and got blasted for. Glad to see you come up with your own "unscientific" statistical method to come up with a decent analysis. It is a direct comparison of draft slot versus expected performance in two different statistical categories designed to measure total performance. You would expect that the #1 pick would produce the most win shares. You would expect the last pick to produce the fewest. The metrics at issue are established by the statistical experts - not by me. All I am doing is comparing production within a draft class to where someone was drafted within the draft class. Link to comment Share on other sites More sharing options...
Guest Posted May 28, 2014 Report Share Posted May 28, 2014 2002 - 2011 Draft Volatility: College Players Positive WS Rank Drafts: 4 positive, 6 negative College Players Positive WS/48 Rank Drafts: 6 positive, 4 negative Note: This was interesting as it was the opposite of what you would expect. While college players were net positive over 10 total drafts in career production, they were net positive in only 4 of those 10 drafts. While they were net negative over 10 years in WS/48 rank, they were positive in 6 of 10 drafts. International Players Positive WS Rank Drafts: 5 positive, 1 negative International Players Positive WS/48 Rank Drafts: 4 positive, 5 negative, 1 neutral Note: Pretty much a wash. High School Players Positive WS Rank Drafts: 2 positive, 2 negative High School Players Positive WS/48 Rank Drafts: 2 positive, 2 negative Note: A wash. Internationals only show up for 6 drafts, or is it 10? I get that the High Schoolers only have 4 drafts in your sample. I also do not have any idea what it means to be negative/positive on net (interpretation, not definition). Every year, your sum of AHFI must be equal to 0 so I'm not too surprised your conclusions are all a wash. But I am also not sure what this tells you. Link to comment Share on other sites More sharing options...
Moderators AHF Posted May 28, 2014 Author Moderators Report Share Posted May 28, 2014 Internationals and college are all 10. If college players were significantly more likely to outplay their draft slot then you would expect to see a significant positive. High school players did outplay their draft slot and that number likely gets better with more years (the latter part of that I haven't run but is my guess). These 10 years suggest to me that Internationals are very comparable to college players for who hits their draft slot if they actually come over but there is a higher risk they don't play as much as a college draft pick. On other threads it has been suggested that international players are a waste of a pick but I don't see that it these numbers. 1 Link to comment Share on other sites More sharing options...
DS17Fanboy Posted May 29, 2014 Report Share Posted May 29, 2014 (edited) On other threads it has been suggested that international players are a waste of a pick but I don't see that it these numbers. Often with the argument that they dont become "Allstars". You know that proffesional basketball qualification that gets its stamp of approval from 12-21 year olds with enough internet savy and fanboy effort to actually go and spend some time electing an Allstar(no offense). Often though Yao Ming (or was it Ming Yao) got this badge of honor. Since there are lots of Chinese and him being paraded around there on TV etc allot. Dirk on the other hand got most his Allstar appearances trough the coach-pick. Not enough Germans. To many NBA fans to caught up with dunks, superior athleticism and the herd mentality ESPN etc creates. All in all Dirk, Mutombo, Olajuwon, Deng, Divac, Ibaka, Gasol Bro's, Gortat, Ilgauaskas, Sabonis, Rik Smits, Kukoc, Schrempf, Ginobili, Parker, Bogut, Nene prove an international pick is always worth considering. Thx for the OP showing some statistical evidence(since many int players are more team oriented skillset wise). Edited May 29, 2014 by DS17Fanboy 2 Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now