How valuable are lottery picks in draft trades?
This week’s draft mailbag features your questions on trading lottery picks with big contracts, predicting prospects’ shooting abilities and evaluating one-and-done players.
“Rumors suggest the Memphis Grizzlies are considering using the No. 4 pick to get out of the Chandler Parsons deal and the Denver Nuggets are thinking similarly with No. 14 and Kenneth Faried. That seems like a steep price for the No. 4 pick but reasonable for No. 14. Do you have a way to assess the amount of cap space a team would have to take on to acquire certain picks?”
— Jonathan Dennis
Over the next two seasons, I project Parsons’ production to be worth about $4 million, while he’ll make more than $49 million in salary. So the net value of Parsons’ contract is about minus-$45 million.
The financial value of draft picks is no longer what it was a few years ago because of the way the current collective bargaining agreement increased the salaries for first-round picks. This year’s No. 1 pick will make an estimated $8.1 million as a rookie (the exact value will be determined by where the 2018-19 salary cap comes in), nearly 40 percent more than Ben Simmons made as the No. 1 pick two years ago ($5.9 million).
As a result, using the method I outlined here, I estimate the net value of an average No. 4 pick as $19 million, not nearly enough to match the negative value of Parsons’ contract. Even if Luka Doncic falls to No. 4, his net value there ($36 million) is not quite enough on its own.
Memphis could even things up by agreeing to take on another player whose contract offers negative net value. For example, Wesley Matthews (net value minus-$16 million) for Chandler Parsons and the rights to Doncic scores as a positive move for the Dallas Mavericks. Or if the Grizzlies want to stay in the mid-lottery, Parsons plus Doncic to the Orlando Magic for Bismack Biyombo, Jonathon Simmons and the No. 6 would be reasonable value.
Faried’s smaller salary ($13.8 million) does not offer nearly as much negative value, projecting as about a $9 million overpay. The No. 14 pick rates as more valuable than that ($11.5 million), so the Nuggets could reasonably expect to get something else (say, a pick in the middle of the second round) in a trade.
“I know you have a soft spot for Kevins, but I’m confused about all the recent hype for Kevin Huerter. He looks the part of knockdown shooter, but he’s a middling free throw shooter (75.8 percent on 99 attempts this past year). Isn’t free throw percentage a strong predictive metric when evaluating future 3-point shooting performance in the pros? Has any middling college free throw shooter ever become a top 3-point shooter in the NBA?”
— Ben Z.
Sure! Mike Miller is perhaps the best example. He shot just 72 percent on free throws at Florida and wasn’t much of a 3-point shooter there (35 percent) yet is in the NBA’s all-time top 25 in both 3-pointers made and 3-point percentage. Brandon Rush (73 percent foul shooter at Kansas) has also made better than 40 percent of his career NBA 3s, as did Matt Bonner (74 percent free throws at Florida).
While free throw percentage is a predictor of NBA 3-point percentage and possibly even a slightly better one than NCAA 3-point percentage, the regression I put together using both to predict NBA 3-point percentage still shows Huerter developing into a strong shooter. His projected long-term NBA 3-point percentage is fourth-best among players in Jonathan Givony’s top 100:
While these are useful to compare prospects to each other, it’s worth cautioning that the 2015 draft picks in that piece have generally shot much worse than projected in the NBA so far. So don’t expect several 39 percent 3-point shooters from this class.
What numbers best translate from college to pro performance. I know steals and rebounds, but what about catch and shoot shooting or finishing at rim, etc #peltonmailbag
— David Locke (@Lockedonsports) June 18, 2018
Steals tend to be a good indicator but don’t necessarily carry over from college to the NBA as well as blocks and assists, two categories that usually show wide divides between the very best and worst players.
To show this, here are the correlations between my rookie projections — based on college stats — and what players actually accomplished in the NBA, with a minimum of 500 minutes played or 100 shot attempts, where one represents a 1-to-1 linear relationship and zero means no predictive power at all. See the first column of numbers:
For comparison’s sake, I’ve added the year-to-year NBA correlation I found for those stats when I looked at them in 2012.
A lot of what we’re really seeing here is the stability in these stats year to year, rather than stability from college to pro. (Note that these projections include up to three seasons of data, as well as a regression factor, so the actual year-to-year correlation is surely lower, but they’re all on a level playing field.)
Considered this way, it becomes clear that usage rate in particular varies unpredictably from NCAA to NBA. So too do a player’s free throw attempts and turnover rate, as well as his 3-point percentage — which makes sense given the change in distance. Block rate, assist rate and offensive rebound rate stand out as translating particularly cleanly from college to pro.
In addition to these box-score stats, I happened to look at finishing for the skills projections in our Grade A mock draft earlier this week using shooting percentages at the rim from Hoop-Math.com and NBA shooting within three feet from Second Spectrum.
The year-to-year correlation there (.241) was nearly as low as 3-point percentage, and adding multiple seasons of data didn’t help much. Intriguingly, two of the three players who improved their finishing most from college to pro in that span were Donovan Mitchell and Jayson Tatum, helping explain why they exceeded expectations as rookies.
Was thinking about the abolishing of the one and done rule and curious as to whether it’s just superfluous now since we have better info on prospects. I guess, what I’m asking is: are GMs better at evaluating young guys than they used to?
— RMJ=H (@rmj_equals_hero) June 18, 2018
I’m not sure the evaluation argument ever held much water as a rationale for the age limit. Although NBA teams had a difficult time figuring out how much to value high schoolers relative to college prospects — they were undervalued at first, then probably overvalued in 2001 before the market corrected — I don’t think there’s any indication that their rankings of prep prospects relative to each other were any worse than for collegians.
Last year, Michael Lopez posted an analysis of draft efficiency by league and year that showed the NBA Draft order actually has gotten slightly worse at tracking the best players from the draft during the one-and-done era. Now, it’s possible that the stylistic divergence between college basketball and the NBA has become so severe in that span that teams would be drafting even worse if they had to incorporate high schoolers, but that’s certainly not a point in favor of improved predictability by excluding them.
As you point out, between the proliferation of all-star games and USA Basketball getting better at identifying top prospects and bringing them in the fold for youth-level competition, teams now have more opportunities to scout players before they reach a college campus than they did when the age limit was enacted. And the existence of AAU statistics that allow us to evaluate players against similar competition allows us to rate high schoolers statistically, something that wasn’t really the case back then.
(Intriguingly, when I added stats from the Nike EYBL to my projections, they alone projected wins above replacement player in the NBA nearly as well as college performance did for the sample of players who have already reached the league.)
So yes, I think there’s reason to believe that any gap between evaluation of prep prospects and collegians that existed 13 years ago is even narrower now.