This week's mailbag features your questions on the next Victor Oladipo, the Warriors' bench and NBA draft projections.
You can tweet your questions using the hashtag #peltonmailbag or email them to peltonmailbag@gmail.com.
"Is there a statistical way to project next year's Victor Oladipo or what a third option would do if they had a first option's usage rate? And, would there be a way to assess the market value of that increase?" -- David
I suspect that there will be a lot of articles searching for the next Oladipo over the next eight months, and the simplest answer is that very likely nobody will match the transformation he has made from OK starter with the Orlando Magic and Oklahoma City Thunder to All-Star with the Indiana Pacers.
Among players who changed teams after using 20 percent of their team's plays or more since the ABA-NBA merger, the increase in Oladipo's usage rate is the largest on record.
James Harden and Tracy McGrady followed somewhat similar paths, but the growth in their usage rates wasn't as sizable as Oladipo's, and both were younger at the time (Harden 23 and McGrady 21; Oladipo is 25).
Typically, the rule of thumb is that players become less efficient as they take responsibility for creating a larger share of their team's offense. This isn't easy to show year over year because, naturally, players who increase their usage aren't chosen at random -- even when they change teams, they usually do so because they prove capable of handling the larger load. This is true even in Oladipo's case; while going from Oklahoma City to Indiana naturally was going to increase his usage to some extent, it was certainly possible that Myles Turner was instead going to emerge as the Pacers' go-to guy.
So I don't think we should expect players to become more valuable simply by moving to a larger role. More often, this plays out like Harrison Barnes, who went from the fourth option on offense for the Golden State Warriors to the first or second option for the Dallas Mavericks. Despite maintaining an admirable amount of his efficiency, Barnes hasn't rated much better in Dallas than he did in Golden State. (His best season by my wins above replacement metric remains 2014-15, when Barnes had the lowest usage rate of his career.)
Back to Oladipo and why we shouldn't be expecting a repeat: In terms of improvement from year to year in player win percentage, the per-minute component of WARP, Oladipo is not only the most improved player this season but one of the most improved players in modern NBA history.
Because Oladipo had rated somewhat better in Orlando, posting a .525 player win percentage in 2015-16 before being traded to Oklahoma City, some bounceback was to be expected. Still, the nature of his development is extraordinary.
My SCHOENE player projection system tries to account for variance by looking at how much the 50 most similar players at the same age improved from their baseline (performance over the previous three seasons, adjusted for age and regression to the mean) and applying it to the player's baseline performance. The best projection this yielded for Oladipo this season was a .609 player win percentage. He's beating that by nearly 40 points.
"I was wondering if the Warriors' bench production is in line with its production in recent years? It feels like starters carry a heavier burden this year." -- Michael
The website HoopsStats.com has production by reserves separated out, and Golden State's performance this season is in fact pretty similar to past years.
The past couple of seasons, Golden State's reserves have been incredibly efficient, posting true shooting percentages far better than league average (.557 this season). The key driver of that efficiency this season has been David West, whose .626 true shooting is far and away a career high. West's excellent play has helped compensate for a downturn from Andre Iguodala, whose own true shooting has dropped from .624 to .533 this season (though up to .697 since the All-Star break).
I think Iguodala's play before the break was a big reason the Warriors' bench hasn't felt as effective this season. But even in terms of on-court net rating, the reserves have been as good as ever. According to NBA Advanced Stats, Golden State has outscored opponents by 8.1 points per 100 possessions with at least one reserve on the court, as compared to 7.4 last season (and just 5.1 in 2015-16).
If there was one variable you could include in your draft project model that you don't currently have data for what would it be? #peltonmailbag
- Brian Frederick (@bfred34) March 10, 2018
Maybe this is too abstract for your answer, or too unrealistic, but a measure of a player's capacity and willingness to improve. My projections are built on the assumption that knowing how good a player is entering the league and his age will tell us how good he's going to be in the future. On average, that's the case, but even if we could project with 100 percent accuracy how players were going to play as rookies, that would still leave us with huge uncertainty about their development.
Many of the league's best players, notably two-time MVP Stephen Curry, have bloomed late because they've improved their skills once in the NBA far more than their peers.
Houston Rockets GM Daryl Morey mentioned something similar when asked about future advances in basketball analytics during this year's MIT Sloan Sports Analytics Conference.
"If we could properly forecast, I would say, three things, it would be a huge advance -- at least for us, maybe other teams are better," Morey said. "One would be what will they do the minute they get a lot of money. I have not been able to predict that. Some guys handle it, some guys don't.
"Two would be do they have the self-awareness of where they're not as good as they need to be, meaning do they understand there's a gap between them and Chris Paul or James Harden or any of these great players in the league? And then three, what are their habits to improve that gap? We're trying to get really good at predicting that and we've got a great scouting department, but we're hoping to create an edge so it's not enough to be good, we have to be better than our competitors, which is where it gets really challenging."
"Amazon's HQ2 is going to be a game-changer for some city. How could this impact the NBA in terms of fan base or free agency for smaller markets? Lots of NBA towns are on the list of finalists: Toronto, Indianapolis, Chicago, Denver, LA, Dallas, Miami, Atlanta, DC, New York, Boston, Philly." -- Hank
If HQ2 ends up in one of the bigger cities on the list, I'm not sure it will have an impact noticeable at the NBA level. A smaller NBA market could be more interesting. I still doubt it would affect free agency much -- a few players, notably Iguodala and Kevin Durant, have mentioned the value of playing near Silicon Valley, but only after the Warriors became contenders. More likely, it would increase the corporate base that is crucial for premium tickets and sponsorships. Among the finalists, Denver and Indianapolis seem most likely to benefit from that.
Another interesting outcome would be Amazon putting HQ2 in a city that does not currently have an NBA team but aspires to get one. Already, Pittsburgh is one of the largest cities in the country without an NBA team, and it's possible an influx of the young tech workers who fit the demographics of NBA fans could make adding a fourth pro sports team viable there. Northern Virginia is currently a bigger media market than the smallest NBA cities (New Orleans and Memphis) and Virginia Beach expressed interest in building an arena to relocate the Sacramento Kings before they were able to build a new arena in Sacramento.
Now, none of this is to say that if a non-NBA market gets HQ2 it should expect a team to follow. After all, Amazon's current headquarters are located in far and away the biggest American media market without the NBA: Seattle. While Amazon's growth has expanded Seattle's economic base in a way that makes the city a more attractive home for the NBA than it was when the Sonics moved in 2008, that has yet to translate into the league's return.