The Golden State Warriors look good. Quite good. On Wednesday they knocked out the Miami Heat as rookie Draymond Green hit a short jumper in the closing seconds after a brilliant cut on a slipped pindown screen. It was a marvelous scoring play, and the Warriors' offense did its job. That's not entirely surprising: the Warriors ranked No. 14 in offense last season despite having Stephen Curry for less than half the season and openly tanking for the last month. Between David Lee, Klay Thompson and any amount of Curry, this is a team with some fine offensive talent.
The remarkable thing about Wednesday's win and, really, the entire season to date for Golden State is that 97 points won the game. The Warriors had the No. 14 offense last year, yes, but they still went 23-43. They still had the seventh worst record in the NBA. The offense isn't a whole lot better this season (No. 12). Almost the entire defense in the stat sheet from 2011-12's 23-43 record and this season's 15-7 mark is on defense.
Yep, a Stephen Curry-David Lee team doing work on defense.
The transformation has been stunning. Golden State was No. 27 on defense last season, and sits at No. 14 after Wednesday's win, according to Basketball-Reference. This was, admittedly, part of the plan. The plan when the team traded Monta Ellis for Andrew Bogut last season. Bogut, who has played a grand total of 73 minutes for the Warriors. Bogut was supposed to be the centerpiece of a revived Golden State defense. Bogut isn't in there, but the defense got better anyway. It's mindblowing.
The new additions to the team are Jarrett Jack (a good defender), Carl Landry (a fair defender), Harrison Barnes (a rookie), Festus Ezeli (rookie) and Draymond Green (rookie). That's a defensive upgrade -- the Warriors had Nate Robinson in Jack's spot last season, so that in particular is a big defensive upgrade -- but nothing monumental. Other than Jack, the Warriors have largely upgraded internally.
Breaking down the key facets of defense show where it's happened. The Warriors were dead last in the NBA in defensive rebounding last season. They are No. 1 this season. Jack ain't helping that too much. When you look at the player breakdowns, you see that Lee has been a much better rebounder this season. In fact, last season looks like a fluke down year for him: his career indicates he's a top-flight defensive rebounder, and he's showing it this season. He's gathering 24 percent of opponent misses while on the court, up from 20 percent last season. To put that in perspective ...
* There are 44 opponent misses in the average 48-minute Warriors game.
* Lee plays 38 minutes per game, so he's on the court for an estimated 35 of those defensive rebound opportunities.
* Last year, Lee would have rebounded seven of them. This year, he's getting 8.5 of them.
* Assuming one point per shot for the opponents off of offensive rebounds, and assuming Lee's teammates wouldn't be picking up those extra rebounds (most of the returning players have static rebound rates), Lee's own defensive rebounding improvement is worth at least 1.5 points per game for the Warriors defense.
That's a big deal. That by itself -- ignoring any other defensive improvements on the team -- is the difference between the Warriors having a 1-point positive scoring margin and having a slight negative scoring margin. And the cost of the improvement (unlike getting Jack or giving more minutes to better defenders over better scorers) is zero. That's the huge thing: this was free improvement.
The other massive change in the Warriors' defense has been its improved shot defense, which is more difficult to explain. Golden State ranked No. 23 in opponent shooting last season, and sits at No. 8 currently. Shooting defense is the single most important factor in team defense: it's twice as important as defensive rebounding and turnover creation. Improving that much in this category is simply huge. (Reducing it to points per game: opponents take about 84 shots per game. In 2011-12, they'd score 83.3 points on those. Remember, this doesn't include free throws. In 2012-13, those 84 shots are getting opponents 80.4 points. That's a massive 2.9 points per game swing.)
A huge factor in that is that Golden State opponents are shooting from lower efficiency ranges: the Warriors are No. 2 in expected opponent field goal percentage, which estimates field goal percentage if a team shot equal to league average at each range. As a result, teams that force more non-three jumpers rate better. Golden State has been forcing a lot of non-three jumpers. Per Hoopdata, the Warriors allow the lowest frequency of shots at the rim. That's the most efficient shot in basketball. Golden State is only roughly league average at defending those shots when they do happen. But preventing them works just as well, provided you can challenge jump shooters.
The Warriors frequently don't have an imposing shotblocker on the court -- Ezeli is the team's only real rotation player with an aptitude for it, and he plays 16 minutes a game. So this is coaching and execution. The Warriors' staff, led by Mark Jackson (who I and others have given a lot of flak in the past 18 months), has designed ways to prevent easy shots at the rim. The players have executed. (Jack figures here: the importance of point guard defense is so undersold. Jack is playing 26 minutes per game, darn near starter minutes.)
Can it hold up? That's the million dollar question. But I find it much easier to believe in defense holding after 22 games than, say, inordinately hot shooting. The Warriors aren't playing over their heads on offense; in fact, Golden State may have some room to grow on that end, especially as Barnes gets his NBA legs up to snuff. It's the defense that has helped the Warriors make a leap, and it's very hard to create a defensive mirage that lasts 22 games. And if they do get Bogut back for good? Man, this team might give someone a headache in the playoffs.
The Hook is an NBA column by Tom Ziller. See the archives.