Visualizing The NBA Draft History

The NBA draft is nothing new. It doesn’t surprise us nowadays. It is just another event (albeit exciting as no other) in the year-round basketball calendar. Some teams win, some others lose, a bunch of names are put inside ping pong balls, and a big-and-hard-to-understand machine calls the fate of different franchises. Easy enough.

Or not.

Since 1947, when the first NBA (or should we say BAA?) draft took place, a grand total of 7961 names have entered the history books as part of the process and what it entails. Sadly, not all of them have been ultimately stitched onto basketball jerseys, as as much as 4635 players of those once drafted never ever got to play even in a single NBA game.

It is ridiculous, and by today’s standards impossible, but that means that 58 percent of the draft picks were wasted in some way by the franchises that made them. In other words, that means that currently all of the second round picks would vanish from the process and even a couple of picks from the first round would never make it to the Association either. Kind of crazy.

---

With the 2019 NBA Draft right around the corner, I thought it’d be interesting to try and analyze what has happened through the history of the event from day one to this day. Obviously, things were much different back in 1947, so instead of looking at those drafts in combination with those more recent I’ll start looking at data from 1974 on, which marks the start of 10-round long drafts.

Barring the 1977 event, which went for eight rounds, and up until 1985 the draft saw no changes in length. By that year it was shortened to seven rounds, and finally in 1989 it adopted the current two-round format.

Keep in mind though, and as we have already learnt, that not every pick turns into an NBA player even if their name will remain forever attached to at least one of the league’s franchises.

Historically since 1947, 58 percent of the draftees have never debuted. Here is the side-by-side comparison of drafted prospects and players that ultimately got to play at least one game in the NBA:

This chart is certainly telling. Since 1974, there has been a wide variation of players being selected by teams in the draft. Long gone are the days in which drafts lasted for eight or ten rounds. But even in those years, no draft class ever produced more than 68 players that ultimately played an NBA game!

Think about it. Today we have events spitted in two rounds of 30 picks each. Since 1974 we have had drafts in which up to 228 players were selected, yet only in six occasions more than 60 prospects from the same class became NBA players eventually.

The 1986 draft produced the most pro players to play in the Association with 68, although nine of them never reached 20 games played. Of the 68 selected players, 14 only lasted one season in the NBA and 20 were out of the league by their second year.

In the “modern” era, the case of the 1998 draft is the most appealing, given that 58 of the 60 selected players went on to play at least a game. The back-to-back picks to miss out were Andrew Betts (No. 50) and Corey Brewer (No. 51).

---

Is there a relation between draft slot and production?

If the draft follows an order, and that order is based on win-loss percentage (with a lottery factored in to try and prevent in-season tanking), we would suppose there must be a relation between drafting higher and getting better players, right?

While it ultimately comes down to each franchise’s ability at the time of drafting, the relation should be there. Number one picks are expected to be the best of the class, while late picks normally have little to no expectations attached to them.

Thanks to Basketball-Rerefence.com data we can see how the relation between pick and Win Shares have evolved through the years. For that, I plotted each draft slot and the average WS/48 produced by players picked in that spot (actually I multiplied the WS/48 of each spot for the number of players picked historically at that slot, then added all of the values together). Keep in mind though, that I only included players with at least 40 games played in the data set to avoid outliers and values that could potentially skew the output. Here are the results:

No wonder how the chart looks like. The highest-drafted players have historically produced the most once in the NBA, while those drafted later and later have seen their production diminish in comparison.

Of course, this is not set in stone and although working with 45 years of data makes things regress to the mean and erases outliers from the equation generating a reliable average, there have been cases and cases. Everybody will always remember Michael Olowokandi as one of the worst No. 1 picks ever, the same as we will forever praise the selection of Manu Ginobili with the 57th pick in the 1999 event by the San Antonio Spurs.

---

Speaking of Manu, can we objectively measure draft steals?

That of Manu Ginobili is not a unique pick in the history of NBA drafts. Every year, although it may take long periods of time to acknowledge and actually prove it, some steal happens and a lucky (or should we say hard-working?) franchise sees someone turn into a much better player than anyone predicted back on draft day and the weeks prior to it.

We can pinpoint steals just by eye-analysis. The same as we did with Ginobili we can say of Isaiah Thomas (No. 60), Kyle Korver (No. 51) or Marcin Gortat (No. 57) since the turn of the millennium. But can we objectively measure draft steals?

There is a simple way we can actually do it. In order to it, I created a ridiculously simple metric I called “Scout Points”, or SP. It is calculated by multiplying the draft position a player was taken at by his WS/48. That way, players with high WS/48 numbers rank high by default, and they see their SP boosted by being picked late in the draft.

Here are the leaders in SP since 1974 (min. 3,250 SP):

As you can see, we were not far off deeming Ginobili a steal, but there have been even better robberies in the history of the draft. Namely, three of them:

• Mario Elie, No. 160 pick by Milwaukee in 1985: Although this is objectively the best steal ever, it wasn’t for the Bucks. Although the team drafted him in 1985, Elie didn’t play an NBA game until 1990 and he did so for the 76ers on a 10-day contract after he had spent his early professional days playing in Ireland, the USBL, Argentina, Portugal, the CBA and finally the WBL. By no means is Elie a superstar, but the value he provided during his career after being selected at the 160th spot was more than incredible, getting him three rings.
• Bill Laimbeer, No. 65 pick by Cleveland in 1979: After getting what would become a two-time champ and four-time All-Star, the Cavaliers spent only two seasons in deciding to trade Bill away to Detroit. It is no surprise, really, given that once drafted he decided to play a year in Italy before putting on the Cavs jersey for the first time in 1980. Cleveland thought they’d be better without him and we all know what happened then and the reasons Detroit became a much happier team with the transaction in place.
• Sedale Threatt, No. 139 pick by Philadelphia in 1983: In similar fashion to Elie, we probably would never pick Threatt over many more players in any draft, even knowing the outcome, but the numbers tell us that for a No. 139 pick the performance was definitely over the expectations and no team that ever put him on court regretted that decision.

One way to avoid players that really didn’t amount to anything in terms of becoming superstars and can’t-miss players on a daily basis is to multiply the PS value of each player by their total career WS. That way we bump down low WS-producers such as Elie and Threatt and can focus on the very best steals ever, although they were gotten with higher picks:

The pair of Karl Malone and John Stockton have now joined Lambeer at the top of the order. Normal, considering both produced more than 200 WS over their careers and although never got to win the chip they were as close as it gets to it.

Other notable names to make the cut are those of Kobe Bryant, Dirk Nowitzki, Tony Parker, and Reggie Miller. You probably didn’t consider them “steals” by the letter of the law but they indeed can be grouped there.

The four of them were first round picks, yes, but they shouldn’t have dropped as many spots as they did. Even Nowitzki, a No. 9 draftee, had to hear the names of Olowokandi, Robert Traylor, and Raef LaFrentz (among others) before his. Not only that, but Milwaukee, who ultimately made the pick and will always have their name attached to the German, traded him (and Steve Nash) to Dallas in exchange of Robert Traylor (No. 6 that year). Ouch.

---

There you have it.

The draft has been part of the NBA cycle for quite some time now, established even before the league became what we know today. There have been busts, there have been found gems, steals, and incredible drops, and we still remain fascinated about it each and every year.

Good for us, the latest edition of it is only mere days away, so get ready for another night full of surprises on the stage and stories to develop in the most unexpected ways during the following years.

Not going to lie here: I can’t wait for Adam Silver to step to the podium and get things running again.