2016-17 NBA Preview: Houston Rockets

The Houston Rockets are doubling down on their three-point attack, defense is an afterthought strategy. Dwight Howard, three-time Defensive Player of the Year, left and they’re bringing in extreme offense-first players. They even brought in famed offensive mastermind Mike D’Antoni. For many people, their defense will be so poor that it’s impossible to imagine an above average team, but analytics-based projections are more favorable. What caused this divide in perception? Will Houston actually have a better season?

“Get your facts first, then you can distort them as you please.” – Mark Twain

2015-16 in review

Dwight Howard was healthy. Houston brought back most of their main core of players from the Conference Finals. They signed Ty Lawson, a point guard with borderline All-Star production. They were poised to have one of most successful seasons in a while, but the team imploded.

Defensively, the team was a mess as they finished 21st in defensive efficiency — they were only above average in creating turnovers. James Harden regressed slightly, a year after nearly winning an MVP, and they fired Kevin McHale after a handful of games. In retrospect, we should have seen at least part of this coming: their win total in 2015 was misleading; they had a point differential that suggested 50 wins and they were “first” in opponent three-point percentage, which is almost entirely luck. Nonetheless, several players had worse than expected seasons, and led to the end of the Dwight Howard era.

Rotation players in: Ryan Anderson, Nene Hilario, Eric Gordon, Pablo Prigioni.

Rotation players out: Dwight Howard, Jason Terry, Ty Lawson, Terrence Jones, Marcus Thornton.

The Houston Rockets retooled the roster, bringing in new starters and key role players. They’ve coveted Ryan Anderson and Nene for a long time and can finally bring them in — they’re just nearing their expiration dates now. Teams often do this, fixating on a player they wanted a while ago and buying them even when the price is low for a good reason. Nene, however, actually out-rates Howard across several metrics on a per possession basis; he just has to stay on the court.

Eric Gordon, like Anderson, is a liability on defense, but both are excellent shooters for a team full of guys who shoot often but without accuracy. Prigioni is a conservative, low-impact point guard who will mostly stay out of the way, even when he’s open; he’s a good fit next to James Harden. Houston lost a few players, but they won’t miss most of them. Ty Lawson was atrocious, and they’ll be better without him. Jason Terry was a useful placeholder, but he played far too many minutes for where he is in his career. Thornton was another unsuccessful guard; they had many. Finally. Terrence Jones, after three years of positive growth and encouraging play, inexplicably had a ruthlessly poor season.

2016-17 projected

More from Nylon Calculus

Explaining that the Rockets will be better than last season should be easy, but you have to navigate some recency bias and this idea that if a team looks bad on defense, they’re very bad — there’s less of a case for gradations of badness here. The Rockets have one of the highest disparities between typical betting lines for win totals and predicted wins from sites like Nylon Calculus, this side of the Timberwolves, The perception effects here are severe.

Firstly, Houston had a staggering number of seasons from players that ranged from “slump” to “colossal wreck.” Out of their 18 players in 2016, 11 of them had a BPM (box-score metric) below -2 — and -2 denotes the theoretical replacement level, which is a production level you can buy for the league minimum on the market. They weren’t all low minute players either; several were significant rotation guys like Ty Lawson, Jason Terry, Donatas Motiejunas, and Corey Brewer. Donatas was dealing with a back problem, which is a real excuse at least. Brewer actually is an example of a guy Houston should worry about because he’s reliant on his athleticism for fast breaks and he looked gassed all season, but they got rid of most of the “below-replacement level” players and the others should improve.

One important concept is that most predictions use several years of data because you get better results that way. Thus, Houston’s 2015 season factors in, but only for their holdovers (James Harden, Trevor Ariza, and Patrick Beverley, mainly.) It’s easy to imagine this team being better because they have been in the recent past. They have a few new parts, but many aren’t essential and Dwight Howard was injured for much of 2015 anyway.

Dwight Howard’s absence is another matter of perception. While he isn’t a respected figure anymore, he’s still a big name and he’s seen as a significant loss by many. But he’s slipped considerably on defense since his Orlando heydays, and he’s a drag on the offense, thanks to his issues with shooting, turnovers, passing, and his post-ups clogging up the half-court. Clint Capela at the very least equaled his production, and he’s much younger and should thus improve. Box score metrics rated them fairly evenly last season, and the team played better with him on the court. Nene, by the way, has a higher composite projection. He’s a good option to have off the bench because his strengths differ from Capela, instead of the duplication with Dwight Howard; it’ll allow them to matchup more effectively.

Overall, the team should indeed be poor on defense, but they’re far from the league’s worst. They still have Trevor Ariza, Clint Capela, Nene, and Patrick Beverley (maybe). The problem is more with depth because there are too few options at about every position. For instance, Beverley is the nominal starting point guard, but he could have a “minor” surgery and will miss the season opener, and then some. Eric Gordon or KJ McDaniels will slide into the other starting guard slot. Prigioni is nearing 40 and can’t be counted on for heavy minutes. The Harden-Gordon experience on defense will be a disaster most nights, and KJ has yet to prove he’s a real NBA player. Then at power forward you have Ryan Anderson, who was one of the worst defensive players in the league. They’ll need Donatas Motiejunas, Nene, and perhaps Ariza available for power forward minutes, but against certain matchups Anderson’s shooting will be too valuable to pass up.

Basketball is a complex game, so the hand-wringing over the defense is missing a driving force: defenses are helped by good offense. If Houston sinks its baskets, then its defense will be in an advantageous position compared to a miss. You don’t need a great cast to be decent on defense either. D’Antoni’s Phoenix Suns were actually near league average most seasons even though they had two liabilities in Steve Nash and Amare Stoudemire. And, finally, just consider this: last season was a disaster with several inexplicably awful seasons from players, but they were still a 0.500 team. It won’t take much to be a little better.

Quick statistic

A while ago, I created the Morey Index as a quick way to describe how a player tended to go for more efficient shots. It’s just the number of free throws and three-pointers attempted divided by total shot attempts, and league leaders are often like James Harden, whose Morey Index nearly approaches 1.0 for several seasons. When you apply it to teams, however, you can see why I named it the Morey Index. Houston has the top three marks all-time, and the seventh highest too.

Actually, Dwight Howard is the name who most dominates the list. He’s on eight of the top ten teams. His high foul rate is well-known, but teams have correctly surrounded him with outside shooters — and in his prime, he helped induce those shots by drawing a lot of attention. Houston still has Harden and a row of shooters, so they should do fine, but they may have a tough time breaking this unofficial record again.

Summary

The Rockets are a joke in some circles, but James Harden is undeniably a great basketball player and a one-man offense. They’ll surround him with a set of role players that will likely not reenact the horrors from last season. Depending on how they use Ryan Anderson, they could have one of the best offenses in the league, and it should offset most of their defensive issues. Remember, just two years ago this team made the Conference Finals, and despite nightmarish seasons for several players last season they still won 40 games. An improvement should be expected.

Win predictions:

Mine: 47.8. A blend of several metrics, including Dredge, with a few other factors considered, like coaching.

Andrew Johnson’s: 48. A combination of PT-PM (a SportVU player tracking metric) blended with RAPM. Two-time reigning champion of the APBRmetrics board predictions contest.

Nick Restifo’s: 44. A simulation using BPM and RPM for player value, which includes game effects like altitude and rest.

Kevin Ferrigan’s: 47. A player projection system with inputs from RAPM, BPM, height, and age.