NBA - What the hell is going on?

[Maithanet]

32 minutes ago, Tywin et al. said:

I think that's a perfect range. The question is how do we want to measure it. 6-9 seems like a good starting point for middle of the lottery, but I'm not sure how large of a range we should use for the middle of the draft, and if that range should include or exclude someone like Booker, who went 13th, and Rondo, who went 21st. I just did a quick check, only looking at if a player made at least one all-star game, and while there were more guys in the mid lottery than middle of the draft, it wasn't a huge difference, but, the size of the range matters. If you include the 10th pick the gap goes up, and it goes up even more if you can't include guys in the early 20s. I think Habs original point though is that you can reasonable predict expected all-star appearances for the first few picks, but after that the average between mid lottery picks and mid round picks is fairly small (I'm not sure if he was including each all-star appearance or just whether or not the player made one AS game).  

I am using picks 6 to 9 for mid lottery and picks 13 through 18 for mid-first.  In the 2006 to 2015 drafts, the mid lottery picks had 8 all-nba guys and 9 all stars of 40 players.  That's 20% and 22% respectively.  

The mid first picks had 2 all NBA guys and 6 all stars.  So that's 3% and 10% respectively.  

To me that's a big drop.  The #8 pick is worth more than double the #15 pick.  

[Mr. Chatywin et al.]

4 hours ago, Maithanet said:

I am using picks 6 to 9 for mid lottery and picks 13 through 18 for mid-first.  In the 2006 to 2015 drafts, the mid lottery picks had 8 all-nba guys and 9 all stars of 40 players.  That's 20% and 22% respectively.  

The mid first picks had 2 all NBA guys and 6 all stars.  So that's 3% and 10% respectively.  

To me that's a big drop.  The #8 pick is worth more than double the #15 pick.  

I used a larger window, which resulted in basically the same number of all-stars, but with an even worse percentage. I didn't focus on the all-NBA guys though because some of the names on there were pretty sketchy (though 2 of the 3 best players in these clusters were the mid round guys).

Interestingly, if you just look at the 2011-15 drafts, more all-stars came from the mid rounds than the middle of the lottery. If you expand it to include the 2016 and 2017 drafts, that takes it to 7 all-stars in the mid rounds compared to just 4, and it's 9 to 4 if you include picks 11-20. And lastly, if you take the last five years in that sample, it's 6-1. So perhaps Habs' point was to highlight the shift in value in more recent drafts.

[Proudfeet]

9 hours ago, Maithanet said:

I would say if anything NBA talent scouts are more likely to fall in love with high ceiling guys and ignore guys who are small or unexceptional athletes who already have a strong game.  That is IMO the only explanation for how Curry goes as late as he did after how dominating he was in college. 

I think that there's a balance there somewhere, and of course different GMs have different standards. There is certainly a ceiling discount generally applied to older players and then to supposed physical shortcomings, but teams don't just go for ceiling alone. I think that the worst example is probably limited to Thabeet? Most other lottery players I feel are expected to be able to start immediately and not just play some minutes off the bench.