/cdn.vox-cdn.com/uploads/chorus_image/image/60171067/usa_today_10904162.0.jpg)
The Sixers gave up a pick, #10 in the 2018 draft, and received in return not only Zhaire Smith but also Miami’s unprotected 2021 first-round selection. What is the difference in value between the pick the team gave up and the one it received? I perform some calculations below, and my finding is that on net the draft value the team lost was very modest; equivalent to a mid-to-high second-round pick. I then argue that Smith’s own value is much higher than many realize, probably equal to that of Mikal Bridges or others who were available at #10. Thus the Sixers received value almost equivalent to two #10 picks in return for just one, a 2-for-1 payoff.
The calculations for the first part of this argument are based on the point systems that allocate value to each draft pick based on the historical performance of the players selected at that spot. With the help of my friend Andy, I found and employed several different systems, all of which delivered approximately the same result. The numbers I’ll show here are for a point system that appeared on ESPN a while back; in a piece I can no longer find but that I believe was written by the excellent Kevin Pelton. Some selected figures from the list:
1 4000
2 3100
3 2670
5 2240
10 1720
15 1240
20 980
30 470
40 240
50 140
60 50
Now, let’s get some caveats out of the way.
- No such system is perfect, these are estimates of value based on the best efforts of the folks devising the systems. Different assumptions, e.g. about the quality of specific players or about how many years of a player’s career should be included in estimating the value the drafting team received by taking him, will lead to different conclusions. Most fans who take a quick look at such models think the low picks are overvalued; they immediately want to offer an enormous passel of second-rounders for the next LeBron or AD. This objection cleverly combines several ruses. First, players like LeBron are special; of course he is worth more than 4000 “points” because in general the #1 pick doesn’t net you a LeBron. Moreover it’s hard to think clearly over long periods in the future, and of course fans tend to create trades where the benefits come today and the costs down the road (see below). Suppose you ran an average team, say Detroit or the Clippers. And say you could trade your next 20 second-round picks (around 200 points each) and what you’d receive is the first pick in the 2028 draft. You might say yes, but it hardly seems like a flat-out steal; it feels pretty fair to me. Your mileage may vary of course.
- Each draft is unique, but for simplicity I will assume that complication away. So, the fact that this draft was viewed by many observers as a strong one is ignored, as is the possibility that 2021 is a “super-draft” with two years’ worth of star players as a result of the NBA possibly changing the one-and-done rule.
- Economists and financiers use the term “discount rate” for the extent to which people would rather have something today than in the future. This is not just a result of inflation; if I offered someone a free cup of coffee every day for the next 100 days, or the same coffee every day for 103 days in a row starting a year from now, some would choose the former and some the latter. Most of the time, most fans, coaches, GMs and owners prefer a draft pick this year to one in the future. I am ignoring this effect here. Of course as with all the effects I’m ignoring, you can tweak the analysis yourself to adjust for as large an effect as you believe in; indeed if you do so please post in comments as I’ll be interested to read it!
- In addition to drafts generally being unique, in this specific case the Sixers technically traded not a pick but the rights to an actual player, Mikal Bridges. I really like Bridges and have written extensively about him. But for purposes of the first half of this discussion, I’m going to treat his value as being that of a generic #10 pick.
- Draft position is based not only on final record but on the lottery results, I don’t think modeling that would have a large impact so I didn’t do so.
- Other stuff I’m not thinking of at the moment; let’s call a halt to the caveat section and move on to the analysis!
Other than the above, the key assumption needed here is a probability distribution as to where Miami will pick in three years. Since three years is a long time, a reasonable assumption is that we literally know nothing about how good the Heat will be. In such a case we use the so-called “uniform distribution” which assumes the team is 3.33% likely to select first, 3.33% to select second, ... and 3.33% likely to select 30th. In that case the expected value of the pick will simply be the average of the values of all 30 picks, which is 1444. Since the 12th pick is valued at 1500 and the 13th is at 1400, that means the expectation under this assumption is that the pick’s value will be between the value of picks 12 and 13. Our first guess might have been that the pick would be valued right in the center of the draft, between picks 15 and 16. But that’s not so, because the gap between picking first and second is much larger than the gap between picking 29th and 30th, and similarly for 2-3 vs. 28-29 and so forth. A 50-50 chance of selecting at 4 or 26 is worth a lot more than a guarantee of picking 15th. In truth I thought this effect would be even larger than the model suggests; that is, I thought a pick that was equally likely to be anything from 1 to 30 would be worth just as much as pick 10. But, according to every model I looked at, the answer was not 10 but instead between 12 and 13.
One last step is worth taking. What is the point differential between pick 10 and the 1444 average value of the 1-30 random assignment? Since pick 10 rates 1720 points, the gap is 276. Consulting the chart, we see that 276 points is a value between that of picks 37 and 38. So under this set of assumptions, we traded, effectively, a high-middle second for Zhaire Smith.
Now, of course one can make more conservative assumptions. One could argue the following:
- Miami was a slightly above-average team this year and team quality has some stickiness
- The team is well-managed by all-timer Pat Riley and well-coached by the skilled Eric Spoelstra
- Miami is a warm and glamorous city star players are eager to play in
- Perhaps most important, every year a bunch of teams tank part of all of their season to obtain draft position, but Miami will not be such a team in 2020-21 because they won’t have their own pick; this cuts the likelihood they pick very early in the draft.
Consequently, one might argue that Miami should be expected to be better than average in three years. To represent this, I did the following: I assumed Miami was twice as likely to obtain picks 16-25 than they were to be a below-average team selecting 1-15 or a near-champion team with one of the last five draft picks.
Long story short, this adjustment gives the Miami pick a value around that of pick 13 and increases the difference between #10 and the pick we received to something like pick 33 or 34 rather than 37 or 38. Of course we can go the other way too, Miami arguably doesn’t have a single player who is top 10 at his position, yet they are over the cap, so there is a case to be made that expectations for the 2020-21 performance should be modest. If we tweak the model to assume Miami is below average we can push the “difference pick” value down to the middle of round 2. And of course more extreme assumptions, in either direction, about Miami can lead us to other conclusions. For myself, I’d say that anyone who thinks they know much of anything about how good a team like Miami will be in three years is kidding themselves, so I believe in a broad probability distribution. But I would downweight picks 1-5 a bit based on the tank factor, so I think a reasonable estimate is that the draft value the Sixers gave up was equivalent to a pick very early in Round 2. Or, to say it differently, the price we paid for Zhaire Smith is about the same in value as the extra second-rounder we picked up from Detroit in return for pushing our 2018 second-rounder into 2019, something we wanted to do anyway. Combining the two trades, we gave draft capital (#10 and #38 picks) for equivalent draft capital plus Zhaire Smith.
So, what about Zhaire himself? In order to believe the “doubling” story from the headline, we need to believe his value is about as high as that of Mikal Bridges, who was both the consensus and actual #10 selection. Is that reasonable?
I do not pretend to be a college basketball expert or an expert forecaster of draftees. And obviously every team can look at their selection at e.g. #16 and find a Big Board that had that player at 10 and say they got value for nothing by taking that guy. Nevertheless I am going to argue that Zhaire Smith really does have #10 value as a draftee. My original plan was to make that argument a part of this essay, but it is expanding in length to the point where it makes more sense as a stand-alone piece, which I will complete and post soon. So I will end here with the following statement: the Sixers obtained Zhaire Smith for a net cost of close to zero in draft capital. Moreover, it is my belief that Smith himself is worthy of the 10th pick in even a strong draft like this one. Thus it is my considered view that the Sixers’ draft-night maneuvers effectively doubled the value of their #10 selection. Hat’s off to Brett, Mark, Ned, the ownership group, the analytics team and everyone else who contributed to this coup.