This week's mailbag features your questions on Dwight Howard's fit in Charlotte, playoff reform, and more.
-- Jeremi H.
Yes, I think that's fair to say. Obviously, Steve Clifford was an assistant to Stan Van Gundy for those Magic teams, and there's a great deal of stylistic similarity between their systems in terms of emphasizing 3s, limiting turnovers (the Hornets and Van Gundy's Detroit Pistons finished first and second in turnover rate last season), controlling the defensive glass (they were first and second in the opposite order there) and staying home on shooters rather than gambling to force turnovers (Charlotte ranked 22nd and Detroit 23rd in forcing turnovers last season).
Of course, the biggest difference between those teams is why Howard might not work out any better offensively with the Hornets than in his recent stops. Just 6 percent of Charlotte's plays last season were finished with a shot, trip to the free throw line or turnover generated by a post-up opportunity, which ranked 27th in the NBA. By contrast, during Howard's prime, Orlando ranked fifth in post-up percentage in 2007-08, second in 2008-09, third in 2009-10 and second in 2010-11.
Naturally, personnel is part of that. The best post-up player on last year's Hornets was ... Frank Kaminsky, I guess? Suffice it to say, there wasn't a prime Howard on the roster. But there isn't a prime Howard now, and his comments to ESPN's Adrian Wojnarowski and Lee Jenkins of Sports Illustrated suggest Howard still hasn't come to terms with that fact.
"You have often stated that a college player's free throw percentage is a far better predictor of 3-point accuracy in the NBA than college 3-point percentage. However, someone like Marcus Smart shoots over 80 percent at the line and is one of the worst shooters in the league. Does NBA free throw percentage not predict 3-point prowess or is Smart just an outlier?"
-- Brett Hullihen
First, just to be clear, college free throw percentage is not a far better predictor than college 3-point percentage; the two are pretty close but free throw percentage came out marginally better the way I studied the issue.
Looking specifically at Smart, it's true this same relationship doesn't hold once players enter the NBA. I looked at players with at least 100 3-point attempts in consecutive seasons. Their 3-point percentage in the first season has a correlation of about 0.4 with 3-point percentage the second season, which isn't good but still much better than the correlation with free throw percentage the first season (0.15).
When you put both 3-point percentage and free throw percentage into a regression to predict 3-point percentage the following season, 3-point percentage makes up the bulk of it. So even for a player with percentages as mismatched as Smart, incorporating free throw percentage only improves his 2017-18 projected 3-point percentage from 32.9 percent to 33.4 percent. (With the benefit of additional years, it becomes clear Smart is unlikely to reach even that lower projection; my SCHOENE projection system forecasts 30.6 percent shooting.)
So the interesting question here is why free throw shooting doesn't have the same predictive power in the NBA than it does for college projections, and I'm not really sure. The fact that the distance is changing could make raw shooting ability more meaningful going from the NCAA to the NBA, but Ken Pomeroy has found predictive power for free throw percentage to year-to-year improvements among college players. So it might have more to do with the development of young players.
If we limit our NBA sample to players 22 and younger, the correlation between free throw percentage in year one and 3-point percentage in year two does improve about 50 percent to 0.22. So there might be a little more reason for optimism about Smart's shooting, but with each passing year it becomes less likely he'll translate his free throw accuracy to the longer distance.
Who are the historical long-two takers who would have been most helped by taking a step back? #peltonmailbag— Kevin Garrity (@kg_cagey) September 22, 2017
I tried to come up with a way to answer this question statistically but didn't come up with anything satisfying, so instead we'll go case by case. The first two players I thought of when I read this question were two I watched as Sonics growing up: super sixth men Eddie Johnson and Ricky Pierce.
Pierce was an 87.5 percent career foul shooter who shot accurately enough from the midrange to shoot nearly 50 percent from the field for his career (49.3 percent), yet he attempted just 920 career 3s and shot them at a 32.2 percent clip. Johnson embraced the 3 a bit more, particularly in his later years, so he attempted a more reasonable number (1,679).
Perhaps the ultimate example is someone who got namechecked in this column last week: Jeff Malone, who like Pierce shot a high percentage (48.4 percent) and was a fine free throw shooter (87.1 percent). Malone attempted just 321 3s in his 13-year career, fewer than 31 players attempted in his final NBA season (1995-96) alone.
"I was brainstorming ways to fix the playoffs and came up with something that I haven't really seen being mentioned. We often see the idea of eliminating the conferences, but then it gets shot down because of travel concerns, and because then it would get very complicated to balance all these teams' regular- season schedules without considering conferences. So, I propose the following: Instead of ranking 1-16, you could still do eight teams in each conference, but a non-playoff team from the opposite conference with a better record could replace a team with a worse record."
-- Michael Cohen
I tend to think of this as the MLS solution, because the soccer league used a somewhat similar format earlier this decade. Technically, the MLS called the last four spots in its playoffs "wild cards," allowing teams from either conference to participate but not be seeded higher than playoff teams in the other conference. Either way, that resulted in Real Salt Lake and the Colorado Rapids winning the "Eastern Conference" playoffs in 2009 and 2010.
I like that this method gets the best 16 teams to the playoffs without dramatically altering their structure in terms of travel or conference rivalries. As Michael noted in the full version of his email, in 2008 the 48-win Golden State Warriors stayed home while five teams from the East with weaker records made the playoffs. That's worth trying to avoid.
That noted, there are questions about this idea. My biggest concern is that under your proposal, teams would have huge incentive to tank their way out of their own conference's bracket and into the other whenever the conferences are lopsided.
There's less incentive if the teams from the opposite conference automatically go to the bottom of the line. (In 2008, for example, the Warriors would be seeded seventh in the East this way instead of fourth.) But that format would mean more travel and better opponents for the top seeds in the weaker conference than the lower seeds in the same conference, and that's not particularly fair. So I'm not sure there's a good solution here.
Unfortunately, we've decided not to do them starting this season. I understand many readers will be disappointed, and the player profiles was perhaps my favorite thing to write, so I'll miss them too.
But they were also far and away the most time-consuming writing on my plate, particularly with Bradford Doolittle moving to the MLB beat and unable to assist, so with the shortened offseason we decided to focus instead on regular analysis throughout the summer months.