Pelton mailbag: How much can summer league predict stardom?

Garrett Ellwood/NBAE via Getty Images

This week's mailbag features your questions on the effectiveness of playing with multiple point guards at the same time, changing NBA position names, the value of spending early in free agency and more.

You can tweet your questions using the hashtag #peltonmailbag or email them to peltonmailbag@gmail.com.

"What can summer-league top draft performances indicate about their future stardom? Given Kyle Kuzma's overall game in the summer league, is it fair to say he is the biggest draft steal so far?" - Amir

This year, more than any I can remember, there has been great interest in just what summer league means, with many readers revisiting my 2013 piece on the topic. With the help of RealGM.com's incredible database of summer stats, I figured it was time to revisit the question with a much larger sample than the single summer's worth of data I used then. And it turns out one of my key takeaways was the result of a small sample.

I noted then that performance in summer league during 2012 was much more predictive of 2012-13 NBA performance for rookies than for veteran players, and was able to concoct a plausible explanation in terms of the effort level of the respective groups. But looking at the larger data set, 2012-13 was an outlier.

Over the 12 years for which RealGM has summer-league stats, performance by veterans in summer league dating back to 2004 has actually been slightly more predictive (with a correlation of .26 between my per-minute rating in summer league and the following regular season for veterans and .20 for rookies). And rookie performance in summer league has been nowhere near as good of a predictor as the first-year projections I compile throughout the draft process (.30 correlation). So summer league tells us something ... but not nearly as much as everything we know going in.

By contrast, the takeaways regarding which stats translate well from the summer league to the regular season held up well. Rebounding, assist rate and block rate all tend to translate well. Shooting percentages have little predictive power, and steal and turnover rates tend also to be somewhat misleading.

Applying that standard to Kuzma's summer league performance tempers the excitement somewhat. Naturally, much of his value was derived from 48 percent 3-point shooting, and while Kuzma's 50 attempts were relatively high for a single summer league (only P.J. Hairston, with 61 in 2014, has attempted more in a single summer league), that's still a relatively paltry sample overall. Kuzma's strong assist and block rates for a power forward figure to be more meaningful going forward.

I wouldn't call Kuzma the steal of the draft at this point -- the limited evidence so far points to Dennis Smith Jr. as the leader in the clubhouse for that title -- but I would say I'm much higher on Kuzma's chances of becoming a productive NBA player than I was based on his college career.

Well, I think one issue there is it's not at all clear whether Rajon Rondo is a good point guard; that certainly wasn't the case most of the 2016-17 regular season. Setting that distinction aside, I think the bigger concern is that a good point guard often becomes a mediocre wing when asked to play that role.

Specifically, as much as we may want to talk about "positionless" basketball, it's impossible to avoid the fact that there's only one ball to go around. The point guard doesn't necessarily need to be the lead ball handler, but there can be only one such player on a given team. Furthermore, on-ball and off-ball is a significant distinction in terms of skills. Both Rondo (36.7 percent of the time the Chicago Bulls were on offense last season, per SportVU tracking on NBA.com/Stats) and Jrue Holiday (33.6 percent of the time the New Orleans Pelicans were on offense) played primarily on-ball last season in a way that won't be possible when they're paired in the backcourt.

Rondo's poor shooting makes him a liability in an off-ball role, and while Holiday is better suited to play in the 2-spot, he has still been below average as a catch-and-shoot jump shooter according to SportVU data. Plus, moving Holiday from defending point guards to trailing shooting guards neutralizes much of his defensive value. So I think he's no longer nearly as valuable as a wing as he is at point guard.

Speaking of positions, I think there are limitations to the Stevens taxonomy, as he discussed in a recent article by Kareem Copeland of the Associated Press. The ball handlers/wings/bigs categorization does not reflect the modern game better than the traditional positions, where players tend to fill different roles on offense than defense. Take last season's Houston Rockets backcourt, where James Harden served as the ball handler on offense but rarely defended opposing point guards, a job assigned to Patrick Beverley. How do you categorize them?

I also tend to think that lumping the two frontcourt positions together diminishes the significant differences in defensive responsibilities for power forwards and centers. Rim protection is much more important for centers, whereas power forwards have to be more mobile to defend stretch-4s on the perimeter. Stevens has been blessed with interchangeable big men in Al Horford and Amir Johnson, allowing him to choose matchups based on the strengths and weaknesses of the opposing frontcourt. However, that won't be the case as much this season with Jae Crowder and Marcus Morris playing extensively at power forward.

I'm being a little too harsh here; there is no such thing as a perfect way to categorize all player roles at both ends of the court. But that's sort of the point. No matter the breakdown, it's important to recognize it as an abstraction as opposed to a fixed definition.

"Is there historically more value in winning early summer (say first week) free agency signings or in the later? Feels like later signings more often outperform their contract, but would you be able to quantify? Obviously there are the LeBrons/Currys/Durants that will outperform any contract allowable under the CBA, but do they outweigh the Hardaway and Gallo deals that happen every summer?" - Dan Ambrosio

In the past three years, I've kept a record of which day of free agency each contract is agreed to (not necessarily signed because of the moratorium) so as to answer this question. Since I've got only a couple of seasons, we can't necessarily evaluate the entire contract, but we can look at how the value compared in the first season.

Here's a first cut with the salary above the veteran minimum players received, the wins above replacement player (WARP) they provided the following season, the value that translates into based on the cost of each WARP in free agency and the net difference between those two.

It's clear here that there is value in waiting. While the contracts agreed to within the first 24 hours of free agency came out OK, the rest of the first week saw heavy overpayment before money started to dry up and teams got good value. As you note, however, max players can throw off the analysis. So let's take a look at what happens when we exclude them.

Now it becomes clear that teams get even worse value in the opening hours of free agency than during the rest of the week. With the exception of a handful of pre-negotiated deals with restricted free agents eager to lock in their money (e.g. Draymond Green and Khris Middleton), early contracts have been the worst. That makes sense logically. To get a player to avoid looking elsewhere, teams have to offer more than the player realistically could possibly hope to get anywhere else. And such deals tend to be subject to an extreme version of the winner's curse.

"I was checking stats on Michael Jordan and I just realized he has won all four NBA awards for individual excellence. ROY, MVP, DPOY and Finals MVP. I think he's the only one to do it too as Larry Bird and Tim Duncan never won a DPOY, David Robinson never won Finals MVP and Kareem Abdul-Jabbar was too old when they started handing out DPOY trophies.

"Of the current stars playing, only LeBron James and Kevin Durant have a chance of matching this feat, though both are unlikely to win a DPOY anytime in the future. Kawhi Leonard already has 2 DPOY, a Finals MVP and is a possible future MVP, but he missed out on ROY. When you look at the current crop of rookies and perhaps even go back to the ROY for the last 4-5 seasons, do you see anyone with the potential to match Jordan's haul of silverware?" - Faisal Shah

Interesting. The Rookie of the Year-Defensive Player of the Year combo seems most difficult: Jordan and Robinson are the two players who have pulled that off in the modern era. And indeed, that figures to trip up the Rookie of the Year winners currently in the NBA.

Given the way big men tend to (correctly) be favored in voting, Karl-Anthony Towns seems to be the only current Rookie of the Year with a realistic chance of winning Defensive Player of the Year, and even that would require huge improvement given where Towns was defensively last season.

Looking ahead, I don't see any obvious candidate. Maybe someone like Marvin Bagley will replicate Jordan's feat, but there's nobody who projects to be talented enough on both ends of the floor to both earn the league's best rookie award and eventually become its best defender. So Jordan might be alone in this regard for a while.