I apologize if this has already been posted.
------------------
The difficult question facing most general managers is whether or not to rebuild or not to rebuild. The Knicks, in particular, are faced with choosing to make a run for an early playoff exit or dump the excessive contracts and hope for a high draft pick. There are many factors that go into such a decision (job security of the GM, money generated from a playoff appearance, draft prospects available) that we cannot always quantify. One part of the equation that is worth investigating more is how much each slot in the draft is worth. "Worth" is obviously a slippery term but there are some ways of figuring out how much of a drop-off each slot in the draft is versus the picks ahead of it.



With this in mind, we went through every NBA Draft from 1980 to the present to determine the average career length (in games) of each pick, the number of All Star years produced at each sloth, and the number Hall of Famers produced at each slot. There are some holes with this method, we took average career length from 1980 to 1997 because in the recent drafts most players are still active and games played is a wash. We threw out a couple of outliers to the data (Len Bias and Ricky Berry who both died very early in their careers). Whether a player is Hall of Famer can be a subjective judgment but we think we were pretty fair in predicting who is likely to be in the Hall. The biggest problem is that the talent procurement system has changed since 1980 and even 1990 so using such data from times were high picks tended to crap out more often may be misleading. Still, we do have general faith in that scouting for the top picks pretty similar so that this is not a glaring problem. With all the caveats stated, here is the data:


Pick # Career Length Hall of Fame All Stars
1 722.72 7 70
2 714.12 4 38
3 666.33 4 48
4 685.28 0 18
5 678.94 4 35
6 523.56 0 6
7 609.89 2 7
8 559.28 0 15
9 693.44 2 16
10 524.67 0 4
11 695.78 1 15
12 539.22 0 3
13 605.72 2 14
14 482.83 2 17
15 364.72 0 2
16 469.33 1 12
17 368.22 1 8
18 517.71 1 9
19 351.39 0 1
20 406.89 0 0
21 361.39 0 4
22 253.67 0 0
23 351.72 0 1
24 436.94 0 6
25 203.33 0 5
26 182 0 1
27 286 1 2
28 240.22 0 0
29 386.94 0 0



As might be expected, the number one pick is, by far, the best pick to have. In fact, the first pick has almost double the All Star appearances and Hall of Fame players as the second pick. The more interesting question is where the steep talent drop-offs occur in the draft. It appears that there is a big difference between picks three and four. This may be a fluke because pick five is about as strong in quality as picks two and three. On the other hand, this blip at pick four might also indicate that GMs tend to gamble more with pick four (ie the fourth of the consensus top four falls into the GMs lap and he feels compelled to take the consensus pick whereas the number five the GM might have a clean slate in terms of expectation with the consensus picks already gone). In either case, it is clear that after pick five, the quality of the draft, on average, decreases markedly.



Picks six through 13 also seem to have similar worth. Pick 13 is the last lottery pick, by the way. Once you make the playoffs, even as an eighth seed, your picks likelihood of contributing decreases by a tangible amount (career average of 482 games versus 605 for the 13 pick). Thus, teetering on the edge of the playoffs is not a great strategy because it really hampers your ability to find free talent in the draft with which to rebuild. Once you get in to playoff teams, there are random fluctuation in career lengths, pick 29 has a longer career length on average than picks 21-28 (with the exception of pick 24). There is also little chance of finding anything more than a solid contributor after pick 19. There is only one arguable Hall of Famer picked after pick 19 and that is Dennis Rodman who is about as weird a find as a team could have.



After looking at all the data, we must conclude that there are defined talent tiers in the draft in terms of probability of finding a good player. Pick one is a tier onto itself and much higher than any other pick in any of the categories that we investigated. Picks two through five offer similar levels of success in finding an All Star or long term starter. Picks six through 13 would be the third tier, where there is a chance to find a good player. There are then two mini-tiers at 14-19 and 20-29 where finding a decent player is a crap shoot. If you are Scott Layden and you are capped out, there is little chance of finding good young players perpetually picking at pick 15. The quicker way to rebuild is to dump the bad contracts (if possible) at go for the high lottery. Of course, there are other political reasons why he will never do this but his strategy will only prolong the malaise in New York until they finally have to rebuild anyway.



The other interesting thing to note is that there is not a steady decrease in likelihood of success of the pick. There is a general decrease but you will find points of decline, like pick four and picks six and seven, followed by increase in success for a few picks afterwards. This is likely because GMs have made some high risk-high return picks at the same points in the draft every year(big man projects like Rich King or attitude issues like Isaiah Rider). The conclusion one must draw is that the high risk strategy has not been fruitful and that certainty of contribution is something GMs should welcome over junk bond-type prospects.



As the season goes on we will look at the draft numbers a little deeper but I hope you enjoyed this first look.

Very very interesting. Kind of emphasizes the fact that if you don't have high picks, you're not getting a francise-level star.

Seems like the NBA has two ways to get a franchise-level star: Have a terrible team and get a high pick or get under the cap and get a big free agent for max $$$.

We need to pick one way or the other, maybe even find a way to try for both. I hope we are trying to get under the cap by the end of 2008 and that we get a few high lottery picks in the next few years.

one reason for a pick up in numbers as you make the way down is the quality of guard play picks up after the bigs are gone. Jordan went no. 3 after Hakeem and Sam Bowie. 2 years ago Dwane Wade went 5th after the bigs were gone.

and manu