2015-16 NBA Preview: Miami Heat

MIAMI HEAT

A year removed from LeBron James, Miami is an awkward team with some seemingly valuable pieces but limited success. The team refused to rebuild, and in turn the roster is a gaggle of veterans and odds and ends. In a depleted Eastern Conference, this is enough to fight for the playoffs. Ultimately health will likely determine their fate.

2015 in review:

In a bizarre season, Dwyane Wade was only partially effective and Chris Bosh suffered a serious illness with a blood clot in his lungs. Prize free agent signing Josh McRoberts played a mere 296 minutes before injuries sidelined him for the rest of the season. But Hassan Whiteside emerged from the Atlantic and leveled entire cities with his wingspan, swatting everything and finishing inside with reckless abandon. Then Miami added a “dragon,” and they played better down the stretch but, alas, it was not enough. The Heat missed the playoffs, presumably as Pat Riley cursed at the heavens for LeBron’s departure and all the injuries the team endured.

Rotation players in: Justise Winslow, Gerald Green, Amare Stoudemire.

Rotation players out: Norris Cole, Shabazz Napier, Shawne Williams.

The Heat will bring back all of their important players while they bring in more talent for the bench. Norris Cole’s role as a backup point guard is no longer needed with Goran Dragic in the fold and Shabazz Napier was similarly surplus to requirements[As the GM who “drafted” him already fled to Cleveland.]. Shawne Williams was a semi-useful shooting forward, but hopefully the Heat get more minutes from McRoberts. Justise Winslow can also soak up backup forward minutes. He was a draft day steal, perhaps on the order of Paul George a few years ago. Gerald Green is still rail-thin with probably the highest vertical in the league, providing the Heat with a high volume three-point shooter off the bench and a transition threat to pair with Dragic. It’s very useful for Miami to have a competent backup shooting guard, of course. Finally, Amar’e Stoudemire can still score efficiently, but he can’t do much else and if things go well, he won’t be utilized often.

2016 Projected

In most universes, the Heat should definitely be better next season than they were last. On paper, based simply on the names they have, it looks like a fairly impressive team, but they were fairly awful last season and some of their headlining players are worse than they seem. The worse you are, the more ground you have to cover to rebound and climb the playoff seeding ladder.

When assessing how good a roster is, you can’t latch onto famous names and make judgment based on the highlights of the past. NBA players age quickly and changes are sudden. Wade is only a semi-functional player at this point in his career, and he’s a liability at times on defense. Bosh is a likely Hall of Famer too, but he’s not an MVP force and he consistently misses a few games almost every season. Luol Deng is a former all-star but is a role player these days. Whiteside is young, but last year was essentially his rookie season and with his history of immaturity he’s not dependable yet. And Amare, of course, is far from his Phoenix heydays.

Whiteside in particular is tough to project. Not only does he have a scattered history of playing against elite competition, his stats profile is singular – there has truly never been a season like his before in NBA history. He combined one of the highest rebound rates ever, beyond only a few Dennis Rodman seasons and one from Reggie Evans, per b-ref, with one of the highest block rates ever, and that was all in conjunction with stellar efficiency on slightly above average volume. The most similar player I could find was a young Artis Gilmore, and it works surprisingly well: Gilmore was an enormous player with some of the most efficient seasons ever and high rebound and block rates[4. Gheorghe Muresan is another good comparison. He’s partially forgotten now, and when he’s mentioned it’s for his 7’ 7” measurement, but at his best he hit the same stats that Whiteside did for the Bullets for a stretch back in the 90’s]. Whiteside’s sheer size gives him an immense advantage in the paint, but he’ll need to learn how to assist NBA players if he wants to be a special player – the talent is there.

Whiteside’s presence over a full season with a healthier Bosh should result in a better defense, but Miami should improve considerably on offense too.  After trading for Dragic, Miami’s glacially paced offense was significantly better, and other pieces like Green and McRoberts should provide more firepower as well. LeBron’s exit changed the team considerably, and they cannot base their team around Wade anymore because of his injury problems and age. In fact, I think they should give him the Ginobili treatment and let him rest for one game of a back-to-back and keep his minutes well below 30 per game. Even if his injuries are truly random and not fatigue based, it’s still giving him fewer opportunities to injure something.

Functionally, with everyone available, the team could be dynamic with lots of slashing from both Dragic and Wade with Luol Deng or another shooter on an opposite wing with Bosh as a pop option and Whiteside as a giant option inside. Josh McRoberts can slide into Bosh’s role too given his range, but he’s also a high post fulcrum and could develop a rapport with some of the team’s options, like lobs to Whiteside – PF/C two-man games are tricky for many teams to defend – or quick cuts or give and go’s with the guards. The pieces fit together well, assuming availability.

The issue with a prediction is that the numbers are tied to past results and outlier cases can derail the stats. There’s also the subjective bias creeping in, as the names are intriguing but the sum of the parts is lacking. There are too many factors, bad and good, seeping into the projection that I wouldn’t want to assume Miami’s a strong eastern conference team anyway. All things in the NBA eventually regress to the mean, and that’s the safest bet here. Miami has a wide range of possible outcomes, and they shouldn’t be tied to a single number.

Quick statistic/graph

Hassan Whiteside had a historic season in a few respects, but one dimension was especially curious. He had one of the lowest assist rates ever, totaling six assists for the season even though he played over 1100 minutes. Few players in NBA history have even been close to his anemic assist rate, but those players were not as involved in the offense as he was[5. Yinka Dare is the only player with a lower assist rate, but he just missed the minutes cutoff.]. You’ll note that in the graph below most players are to the left (i.e. below average) of the league average usage rate line. Whiteside had an above average usage rate, and he grabbed an obscene amount of offensive rebounds. He had the ball frequently in the half-court, and no one has ever matched his assist incompetence. If he doesn’t improve, people will catch onto his blinders-on routine and crowd him aggressively.

Summary

Wade and Udonis Haslem are going onto their twelfth season of playing together. They’ve seen a lot of change in Miami, and this iteration is a strange roster halfway between mediocre and veteran playoff status. Nearly every rotation player has a high variance on expected value this season – Dragic is adjusting to a new team and peaked so late in his career it’s tough to know if his all-NBA season was a fluke, Whiteside needs more minutes and maturity, Wade is perpetually on a bad wheel and aching somewhere, et al – and the team is largely intact and following a dreadful season with a point differential that suggested 33 wins, not the 37 they actually won. Miami is not going to be a serious eastern conference foe for Cleveland, but they’ll at least be easier on the eyes with a handful of interesting pieces.

PBP-Metric[1. This is the initial version of my own metric, which uses a full range of stats collected from play-by-play logs and tested extensively to avoid overfitting.]: 40.0

PT-PM: 40

Nick‘s[2. For a short description, the predictions use regression models and neural networks to apply various stats like BPM, RAPM, and Win Shares to 10,000 simulations of the season game-by-game to select the “best” result.]: 38

Nathan Walker: 35