In our previous discussion and look into absolute and relative quality of draft picks, we ended up creating an approximate expected WSP (Win Share Percentage) for each draft pick slot that allows for comparisons across years and picks. Now we are going to try and take a look and see if we can put a number on what teams draft well and do not draft well. Something to keep in mind is that obviously a player might not spend the majority of their career let alone their entire career with the team who drafted them. However, if you draft a player who turns out great, even if you lose them later in free agency, that was a quality draft choice. The player leaving is independent of the good/bad of the pick itself.
What is a “Good” Draft?
First, we can acknowledge that there is purely an eye-test with regards to drafting well. We know that players like LeBron, KAT, and Ben Simmons are worthy number one overall picks. We also know players like Jimmy Butler and Draymond Green definitely worked out better than expected. However, determining single pick success with obvious All-NBA types is fairly easy. What about a team over ten years with differing number of picks and in different slots? What is better, drafting LeBron James and Anthony Bennett, or drafting James Harden and Russell Westbrook? What about Robert Covington, Dario Saric, and Joel Embiid? I am not entirely sure these have a definite answer, but it is an interesting question.
Who Drafted Where?
In order to begin this investigation, we need to understand the distribution of draft picks by team and draft slot from 2000-2015. Table 1 is stratified by lottery, non-lottery first, and second based on general conventional wisdom. Warning, do not look at the Brooklyn Nets row or you will go blind.
Again, I want to make clear that these are entirely based on if the team made the selection. Brooklyn obviously had picks at one point but turned them into the Old Celtics 2.0 and Danny Ainge’s current stockpile of 6’8” switchy guys.
Couple interesting points of zero analytical strength
- Dirk Nowitzki means you will not be in the lottery
- The zero lottery picks for the Spurs are a nice cut off that misses David Robinson and Tim Duncan, two pretty acceptable players
- I forgot how terrible Golden State was prior to We Believe and the current juggernaut
- RIP Vancouver and Seattle
What is Expected for Each Pick Slot?
Figure 1 uses a graphical presentation with a Loess curve showing the uptick in expected WSP in the lottery and high lottery picks. This is a nice visualization of why Picks 1, 2, 3, etc. are values higher than 12, 13, 14 in a non-linear manner. Also late first draft picks with a WSP greater than 7.5% were labeled, just so you can see who those players are.
Now, using our ability to predict what the WSP “should be” based on draft pick position, we can see what teams have historically done well or poorly. If we use Andre Iguodala as an example, the Sixers obviously drafted him, but did not get all of his productivity since he went to Golden State. However, that was still a quality player that was drafted. Looking at draft data in aggregate like this requires keeping situations like that in mind. Figure 3 took the sums of expected and actual WSP for each team’s first round picks (2000-2015) and subtracted them to get a residual, or the difference between the observed and predicted. Of note, these take into consideration draft day trades.
Based on this data, the Milwaukee Bucks are the “worst” drafting team from 2000-2015 and the San Antonio Spurs are the “best”. Table 4 below shows each first round pick from 2000-2015 and their WSP, predicted WSP, and residual for the Spurs and Bucks. The scaled residual ranking covers both teams, so you can see across both teams how the picks stack up. Warning though, the Bucks picks, Giannis aside, are terrifying.
The results from these two figures brings us back to the question we asked at the start - is it better to draft well all the time or just really nail one or two picks and the rest just does not matter. With the foundation now set, we will look into that exact question in the third installation in the series.
Coming soon...On the Clock Part 3: How Many Hits Does it Take?