Nylon Calculus: Measuring the value of future draft picks

The Thunder and Pelicans received numerous future first-round draft picks in trades this offseason. What can we say about the value of picks far in the future?

In the few years before the 2019 offseason, front offices seemed reticent to part with future unprotected first-round draft picks. No one wanted to repeat the mistakes of the Brooklyn Nets, who infamously traded three unprotected first-round draft picks and a pick swap to the Boston Celtics in 2013 as part of a trade that set the franchise back for years.  But that all changed this offseason.

In this rendition of free agency, multiple teams dug into their asset pools to complete blockbuster trades to land superstars. The Los Angeles Lakers got their dream big man in Anthony Davis, but they had to send the New Orleans Pelicans three first-round draft picks (including the No. 4 overall pick in this year’s draft), an additional future first-round pick swap, and talented youngsters Brandon Ingram, Lonzo Ball, and Josh Hart.

Their cross-town rivals, the L.A. Clippers, acquired Paul George from the Oklahoma City Thunder but had to part with an astounding five first-round draft picks (including two Miami picks they previously owned), swap rights in two additional years, and Shai Gilgeous-Alexander and Danilo Gallinari. The Houston Rockets added superstar point guard Russell Westbrook, but at the cost of sending the Thunder two lightly protected first-rounders, two pick swaps, and Chris Paul.

Professional basketball is a sport in which one star player can have a tremendous impact on a team’s fortunes, so it is hard to criticize any franchise too heavily for making an aggressive move to acquire such a player.  Still, it is interesting that the teams getting the stars were willing to part with so much of their future draft equity. Moreover, the years when some of these picks and swaps will convey are quite far into the future. Remarkably, the Pelicans and Thunder now have a vested interest in the performance of some of their rivals in the 2024, 2025, and 2026 seasons.  This made me wonder: what can we say about the value of a draft pick so far out in time?

To answer this question, I collected the regular season win-loss records of all teams from the 2000 to 2019 regular seasons and ranked the teams each year from worst record to best record. Thus I basically computed the pre-lottery draft position of each team for each of these seasons, where the team is ranked in the draft pecking order before any of the lottery balls are drawn. For the sake of simplicity, I ignored how the two-conference structure messes with the draft order by ensuring that the teams that miss the playoffs, rather than the 14 worst records, end up entering the draft lottery.

Next, I matched where a team finished in a given season with where it finished in the following season, two seasons after the current season, three seasons after, and so on.  As an example to illustrate this, let’s take the 2009-10 Orlando Magic. That team, led by Dwight Howard at the peak of his powers, went 59-23 and won the second-most games in the regular season, behind only the Cleveland Cavaliers. They thus had a pre-draft lottery position of 29 (i.e. they had the 29th-worst first-round draft pick before the lottery drawing). But three short years later, the 2012-2013 Orlando Magic won only 20 regular-season games and finished with the worst regular-season record, and thus the No. 1 pre-draft lottery position. A team with the rights to this pick would have been quite pleased. This example becomes a data point in the analysis where the No. 29 pick became the No. 1 pick three seasons later.

This is at the heart of the analysis. We want to determine the relationship, if one exists, between current draft position and future season draft position. How likely are strong teams now to be strong in the future?

One way to do this is to take the correlation between a team’s pre-lottery draft position in a given season and its pre-lottery draft position in subsequent future seasons.  I ended up taking this correlation for each year up to five years in the future.  The chart below shows the results.

This chart makes intuitive sense. A team’s record next year is highly related to its record this year.  Good teams don’t usually become really bad in one season and bad teams don’t usually become really good (unless they acquire LeBron). The left-most point depicting a 0.62 correlation represents this. But teams can, of course, change more over a longer time period. The correlation between pre-lottery draft position for the current year and pre-lottery draft position two years in the future is only 0.39. We have a sense of how good teams will be two years in the future but things can certainly change.

If we go further into the future, the picture gets hazier. Three and four years into the future, the correlation with this year’s pre-draft lottery position falls to just 0.18 and 0.08 respectively.  By the time we are five years in the future, the correlation is close to 0.

I think the plots below are enough good way of visualizing the uncertainty of far away drafts.  Each point represents one season of a team from 2000 to 2019. The x-axis is the team’s pre-lottery draft position in that particular year, binned into five-spot increments. The y-axis is that same team’s pre-lottery draft position in the following year (left graph) or in five year’s time (right graph).

We can clearly see in the left graph that teams at the top and bottom of the standings in a given year tend to stay relatively close to that position in the following year, with some noted exceptions.  But in the right graph, we see no clear pattern.  Translated: five years into the future, a current contender could easily disintegrate and a team at the bottom of the standings could climb to the top.  If past history if any guide, once we get four or more years into the future we really have little idea where any of the teams will be picking.

All this research is interesting, but what does it mean for the Oklahoma City Thunder and the New Orleans Pelicans, our two beneficiaries of large draft pick hauls?  Well, my surface-level analysis tells us that the picks eventually conveyed could really end up anywhere in the first round.  Just because the Lakers, Clippers, and Rockets have strong teams now does not necessarily mean they will be good (or bad) performers in the future, particularly in the distant future of four or more years away.

In other words, we should widen our view of the possible returns.

Another thing we need to keep in mind is that controlling multiple picks from the same franchise is a high-risk, high-reward proposition. If the team whose draft picks you own has a bad spell, you can reap a huge reward (think Nets/Celtics). But, if the team giving up the picks remains in the playoffs, then the upside of a potential high lottery pick vanishes. Essentially, future picks of the same team are more correlated with one another than a mix of future picks from different teams. This is actually one of the major advantages of the Thunders’ return over the Pelicans’. The Thunder acquired a diversified portfolio of future Clipper, Rocket, and Heat picks while the Pelicans are heavily invested in just the Lakers’ future.

It also should be noted that there are key considerations for each of the teams giving up the picks, namely the length of the contracts and the ages of the key players. The Thunder have to be happy that Kawhi Leonard signed only a three-year contract in which the third year is a player option, rather than the maximum allowable four years. He and star teammate Paul George are only committed to the Clippers through the 2020-21 season, leaving open the tantalizing possibility of them both opting out the year before the first unprotected Clipper pick conveys in 2022.

The Pelicans should be comforted by the fact that LeBron James is turning 35 this season, though Anthony Davis is only entering his prime at age 26. The Houston Rockets star duo of James Harden and Russell Westbrook are both 30, meaning they cannot be counted on to produce in the 2024 and 2026 seasons when the Rockets’ top four protected picks are owed to the Thunder. Moreover, age shows up as an important variable in this type of analysis. In a basic linear regression, I found that teams with a higher average age tend to do poorer in future seasons than would be predicted by their current record (and vice-versa for younger teams).

When it comes to a first-round draft pick owed many seasons in the future, we really have little idea where the pick will fall.