It’s an NBA truism that defense wins championships. Anyone who watched this year’s finals was reminded incessantly that the Celtics and Warriors were #1 and #2 in defensive efficiency in the regular season. Even casual fans can tell you that 19 of the last 20 championship teams were top-ten defensive teams. But there is less to those facts than meets the eye.
One issue is that an arbitrary ranking cutoff can be misleading. Sometimes the 10th best team is a lot better than the 11th best team, in other years the difference is rounding error. But the bigger problem is that there are simply too few championships in modern NBA history to establish any reliable pattern. Even going back 20 years is a stretch, given changes in rules and playing styles, and the difference between winning and losing a championship can be as idiosyncratic as a tweaked ankle or a shot rimming out.
A more fruitful approach is to recognize that winning a championship requires winning four consecutive playoff series, and to look for factors that consistently contribute to playoff success. That approach increases the sample size by a factor of 15, facilitating systematic analysis. And it leads to a rather surprising conclusion. In the modern NBA, offense contributes more than defense to winning in the playoffs.
Put aside the conventional wisdom and look at the facts. Over the past decade, teams that scored more points per 100 possessions in the regular season won 68 percent of playoff series (102-of-150). By comparison, teams that allowed fewer points won just 58 percent (87-of-150). (I’ve used offensive and defensive ratings from basketball-reference.com; nba.com has slightly different numbers because they estimate possessions differently.)
This simple comparison takes no account of how much better a given team was on either side of the ball. It also takes no account of homecourt advantage, which skews results in favor of higher-seeded teams even when they are evenly matched. Taking simultaneous account of those factors (using multiple regression analysis) provides a more reliable estimate of the relative impact of offensive and defensive efficiency.
The results indicate that, in a playoff series between two evenly matched teams, the team with homecourt advantage wins about 64 percent of the time. That’s a huge advantage, belying the notion that a good team can coast through the regular season without worrying about playoff seeding.
After allowing for the impact of seeding, every additional point of regular season offensive efficiency increases a team’s probability of winning a playoff series by 4.3 percentage points. That estimated impact is about 50 percent greater than the corresponding impact of an additional point of defensive efficiency, 2.8 percentage points. Put another way, it requires 5.1 points of additional defensive efficiency, but only 3.3 points of additional offensive efficiency, to compensate for homecourt advantage in a playoff series. In recent seasons the difference has been even larger...
Is the outsized impact of offensive efficiency a fluke attributable to some highly uneven first-round playoff matchups? No. Ignoring first-round series entirely produces an even larger estimated disparity in impact. Even focusing solely on the conference finals and finals suggests that offensive efficiency is substantially more diagnostic of playoff success, though of course the estimates get noisier as the sample size shrinks.
Even with 150 playoff series, there is about a 10 percent chance that the greater impact of offensive efficiency is illusory. If that’s too much statistical uncertainty to warrant retiring the conventional wisdom, we can turn from analyzing playoff series to analyzing individual playoff games, further increasing the sample size more than five-fold. The results are qualitatively similar; offensive efficiency contributes an estimated 43 percent more than defensive efficiency to winning playoff games, and the probability that the difference is a statistical fluke falls to less than one in a thousand.
In 2021 the Nets had the most prolific offense in NBA history, averaging 118.3 points per 100 possessions in the regular season. But their defense was mediocre, allowing 113.8, and they lost a tough second-round playoff series to a better defensive team, the eventual champion Bucks. A case of defense winning in the playoffs? Not really. Those Bucks were hardly a defensive powerhouse, just 10th-ranked in regular season defensive efficiency. The real story of the series was that the Nets’ offense, beset by injuries to two of their three stars, managed just 107.3 points per 100 possessions.
That summer, Sean Marks beat the bushes for defensive-minded bench players, trading for Jevon Carter and signing DeAndre’ Bembry, James Johnson, and Paul Millsap as free agents. But the team’s defensive efficiency improved only slightly (from 113.8 to 112.8) while the offense took a big hit (from 118.3 to 113.6). By the end of the season, Carter, Bembry, and Johnson had all been waived, and Millsap had been shipped to Philadelphia as a throw-in in the James Harden trade.
The primary return for Harden was Ben Simmons, a first team All-Defensive player in each of his past two seasons. If Simmons can get back to that form, the team defense should be markedly improved next season. But the bigger question mark is on the offensive end. Kevin Durant and Kyrie Irving are two of the league’s premier scorers. Despite all the Nets’ troubles last season, the team averaged a blistering 120.7 points per 100 possessions (+8.5) when they were on the court together. (That’s including their disappointing playoff minutes.) But if they are traded, or sit out, or slow-walk through the season, where will the offense come from? No one else on the current roster has ever averaged even 20 points per game, and the guy who has come closest, T.J. Warren, hasn’t played in over a year.
Without KD and Kyrie, next season’s Nets may be a fun, tough defensive team. But in today’s NBA, offense wins in the playoffs.