By Alex Koenig
The NBA Draft Lottery was instituted by the league in 1985 after accusations that teams were deliberately tanking to secure the first pick in the draft. The logic went that it was worth a team’s while to give up on a season if it meant the chance at getting a premier, franchise-altering, talent.
After a bit of tweaking the league decided in 1994 upon the current weighting of the lottery based on in-season finish:
But has this really had the desired effect of discouraging teams from tanking? Last year there was only a seven-win difference between the last place Minnesota Timberwolves and the 5^{th}/6^{th} Sacramento Kings and New Jersey Nets. Over an 82-game season, it’s not all that hard to lose seven games you could have won, especially when you know it could be the difference between a 25% chance at the first pick and a 6.3% chance.In other words, as a #1 seed (i.e. the worst team in the league) you have a one in four chance of landing the top pick and a guarantee that you will at worst have the 4^{th} overall pick. Similarly, the #2 seed is guaranteed at least the 5^{th} pick and so on.
However, this assumes that there is a sizable difference between the first pick and the sixth pick. Is that the case?
Looking at the career Win-Shares of all players drafted from 1985 through 2006 (2006 was the cut-off because it allowed for enough time to pass for the “average” NBA career to run its course) the answer is emphatically yes.
Below are the expected Win-Share values for each of the 14 lottery picks.
Pick |
Expected Win-Shares |
1 |
78.08 |
2 |
54.17 |
3 |
48.13 |
4 |
52.90 |
5 |
56.75 |
6 |
25.11 |
7 |
34.11 |
8 |
31.56 |
9 |
42.88 |
10 |
43.00 |
11 |
32.03 |
12 |
18.77 |
13 |
40.00 |
14 |
21.02 |
That’s an expected difference of 52.97 career Win-Shares between the 1^{st} and 6^{th} picks. As a barometer, that’s roughly the difference between getting Tony Parker (79.1) and Josh Childress (25.2).
But being the number one seed doesn’t guarantee getting the number one pick, and getting the number one pick doesn’t mean you’ll deliver, after all Michael Olowokandi (with a whopping career WS of 2.5) was a number one pick and Jeff Hornacek (108.9) was the 46^{th} overall pick. That being said, the guarantee of being in the top of the lottery is significant.
Below is the expected career WS value of each seed before the lottery – i.e. only with the probabilities.
Seed |
Expected Win-Shares |
1 |
58.62 |
2 |
57.80 |
3 |
56.19 |
4 |
52.10 |
5 |
44.49 |
6 |
35.39 |
7 |
37.33 |
8 |
36.25 |
9 |
43.80 |
10 |
42.63 |
11 |
31.96 |
12 |
20.57 |
13 |
40.06 |
14 |
21.69 |
The draft lottery and the scarcity, evaluation and necessity of talent in the NBA definitely favor the worst four teams in the league. Of the last 21 MVP awards handed out, 13 were given to former top-four picks. Of those 13 only Shaquille O’Neal, who was drafted 1^{st} overall by the Magic and won an MVP in 2000 for the Lakers, won the award for a team other than the one that drafted him.
This gives credence to the notion that a top pick can have a transformative effect on a franchise. Going from one of the four worst teams in the league to a team sporting the league’s Most Valuable Player is rather transformative.
Furthermore, of the last 21 NBA champions, only two teams had the top-contributor (i.e. the players with the highest WS) not be a top-four pick. Those teams were the 2011 Mavericks (Dirk Nowitzki, 9^{th} overall pick) and the 2008 Celtics (Kevin Garnett, 5^{th} overall pick).
If the ultimate goal of an NBA franchise is winning a title, then it appears a necessary step along the way is bottoming out for a top-four pick. The Houston Rockets have had a winning record every season since 2002, but in that time have never made it passed the second round of playoffs.
It’s hard to accept, but maybe it’s about time teams like the Rockets, and others, took notice of the virtual necessity to tank.
EDIT:
If the median WS values are used (rather than the mean) the discrepancy is even greater:
Expected Value by pick:
1 – 65.9
2 – 54.1
3 – 50.9
4 – 47
5 – 36.5
6 – 19.1
7 – 30.5
8 – 24.5
9 – 23
10 – 32.9
11 – 23.2
12 – 14.9
13 – 14.1
14 – 9.6
Expected Value by Seed:
1 – 53.94
2 – 51.47
3 – 47.77
4 – 42.31
5 – 35.55
6 – 30.80
7 – 32.80
8 – 27.48
9 – 26.22
10 – 32.96
11 – 23.62
12 – 15.88
13 – 14.94
14 – 10.43
Out of curiosity, could you quickly calculate the median returns on both pick number and seed number? I’m thinking perhaps some outliers are playing around with the means…
The median returns on each pick are as follows:
1 – 65.9
2 – 54.1
3 – 50.9
4 – 47
5 – 36.5
6 – 19.1
7 – 30.5
8 – 24.5
9 – 23
10 – 32.9
11 – 23.2
12 – 14.9
13 – 14.1
14 – 9.6
Thanks for pointing this out, the results are even more drastic
Yeah, good article! Really shows how losing badly can quickly translate into winning! Now it makes me wonder about the tendencies of the teams that consistently lose in the first and second rounds. It seems like there’s an obvious gravitation from mediocrity to mediocrity. Maybe a nice Markov chain…
Are the win shares you list the expected win shares based off of a model, or are they the mean of the historical data you used? From a theoretical standpoint, it doesn’t make sense for the 13th pick to have a higher expected return than the 6th overall pick, but I guess that’s been the case over the past 21 years?
Yea these are based of the mean of the historical data. I’m going to check with the median soon…
Kobe is probably the dude messing up the mean for the 13th pick. That guy is such a jerk. Speaking of, he is so pissed that Pau lead the Lakers in win-shares in 2010, so that he didn’t get mentioned here.
Hey alex this is super cool I think you agree that the Columbia Spectator sucks and they do attack articles on you. Why do you not come join the #1 most prestigous campus publication at Columbia University? specsucks!!! You can email me at thecloakedmask@gmail.com if u want more info.
Pingback: Lottery Not Always Random « Double Guns