Like most fans, I have been incredibly disappointed with the start of the season for the Rockets. The expectations were so high after last year's WCF appearance and I assumed that the drop off would be minimal with the change of a few tertiary players. So far, that hasn't been the case, but I wanted to take a deeper look if those changes were real or not. Yes, we've gotten worse defensively, but also significantly regressed offensively. A lot of the focus on the Rockets struggles is on the defensive regression, but the offensive regression has been almost as egregious. Bad offense leads to easy transition offense and less half-court sets. Offensive efficiency in the league is up 1 point per 100 possessions, notable but does not explain the full situation. However, fair to say that 1 out of the 9 points per possession regression for the Rockets is due to the league improving overall offensively. This graph also shows that there's a lot of scatter and randomness in the data, and we should expect things to regress to normal over time, as the data from last year's full season is more compact and this year's data is more varied. Part of the problem is rebounding The Rockets have regressed from one of the best defensive rebounding teams in the NBA, rebounding 75% of missed shots, to one of the worst defensive rebounding teams in the NBA, rebounding 70% of their missed shots. The difference has accounted for 2 points per game worse second chance points (13.8 vs 11.9), so let's call that 2 more points out of 9 accounted for. With our main rebounders returning this season, I think this will fix itself over time as the front court gets healthier (hopefully) The Rockets are also giving up more points off turnovers Last year, the Rockets gave up 15.8 points off turnovers, this year it's up to 18.2. The Rockets turnover percentage is actually the same this year versus last year, so that might be randomness or may be an indication that the Rockets are having more live-ball turnovers. That explains another 2 and a half points of our 9 points of defensive regression. With CP3 and Harden doing the majority of the dirty work on offense and the iso-heavy style we play, I imagine this may regress to more normal figures. 61.6 points per game allowed in the paint is astronomical The Rockets are last in this figure in the league and last year the Rockets were third worst in the league with 48 points allowed in the paint. Last year the Rockets allowed 63.9% shooting within 5 feet of the rim, but this year they're up to 69%, second worst figure in the league. Last year the worst team in the NBA allowed 64.9%, so this figure promises to go down as well, with one caveat... The freedom of movement rule changes appear to be having an effect league-wide and on the Rockets This year the Rockets are fouling opponents 23 times per game, up from 19.5 last year. Last year league average was 19.86, but this year it's up to 22.5. Therefore, the difference is largely attributable to league wide trends, but the Rockets year-over-year trend is worse than the league. The Rockets this season are allowing 22.2 free throws made per game, versus 14.6 last year, again turning from a team amongst the best to amongst the worst. That's a full 7.5 points per game difference. Perhaps the Rockets are more susceptible to these fouls by facing frequent mismatches due to frequent switching, or perhaps the new personnel getting minutes is more foul prone to lack of discipline and athleticism, but either way it's been a major factor. The Rockets free throw defense is partly to blame (hi Matt Bullard) The Rockets opponents are shooting 84% free throws, second worst in the league, up from 74.6% last year. Since there's no such thing as free throw defense, let's throw out the difference and call our free throw issue more like a 6 point regression. Therefore, we can pretty much explain the difference in defense from last year to this year with overall league efficiency bump, spotty rebounding, the freedom of movement rule change, more fouling overall, and unlucky rim and foul shooting percentages. But wait, we haven't blamed anybody yet... Yes, results have varied by player, byt the data seems very noisy. A lot of critics have feasted that the Rockets are +11.3 points per possession worse on defense when Melo is on the court. It's hard not to notice that he struggles to contain a drive in isolation. However, Melo is not alone in this extreme stat. Isaiah Hartenstein and Gary Clark are +15.2 and +21.2, understandably. Less understandably, Chris Paul and MCW are +10.5 and +6.5. Lineup data is notoriously noisy at this time of the year with such small sample sizes. Consider this, when Melo is on court vs off court, opponents eFG% is the same, our defensive rebounding has been significantly better (80.8% vs 73.6%), and our blocked shot percentage has been higher. Maybe it's just noise. The Rockets most played lineup is also our starting lineup and our best defensive lineup: ".C. Paul, .P. Tucker, .J. Harden, .J. Ennis III, .C. Capela". Would you be surprised to hear that our defensive rating from that lineup is a comical 135 points allowed per 100 possessions? In conclusion, the sky may not be falling. We may ultimately regress on defense, but a large portion of the issues feel fixable with the personnel we have.
