Just how good are the Warriors without Kevin Durant, who has missed Golden State's past five games because of a calf strain as the team enters the NBA Finals?
At some point as the Warriors ripped off a five-game winning streak to close out the Houston Rockets and sweep the Portland Trail Blazers, the statistic began circulating that Golden State had lost only once in the past 30-plus games that Durant sat out and Stephen Curry played.
That record is now 31 of the past 32 games with Curry and without Durant dating to March 2017, a .969 win percentage that would be equivalent to 79 wins over an 82-game schedule. How have the Warriors been so effective without Durant? And how likely are they to keep it up facing their toughest test yet, with Durant not expected to return for Thursday's Game 1 of the Finals against the Toronto Raptors?
Let's take a closer look at how Golden State has played without Durant to help answer those questions.
How much better without Durant?
It's pretty indisputable from the record that the Warriors have played better without Durant, which is not the same as being better without him. But the record might slightly overstate just how much better Golden State has played when Durant has been sidelined.
First, there's no compelling reason to throw out the first six games the Warriors played without Durant after signing him, during which they went 3-3. That brings their overall record without Durant in games Curry has played to 34-4 (.895), as compared to 163-46 (.780) with both MVPs in the lineup.
Two additional factors even things up. Golden State's schedule has been a little more difficult when both Curry and Durant play than when Durant has been sidelined. Also, the Warriors' plus-9.9 point differential in games both have played would typically translate into something closer to an .825 winning percentage than their actual .780 mark.
Add that all up, and when you adjust Golden State's net rating for the opponents' season mark and the location of the game, the Warriors have played 12.4 points per 100 possessions better than an average NBA team in games Durant has missed and Curry has played. That's indeed better than their plus-11.3 adjusted net rating in games with both stars, but the difference is no longer quite as dramatic as it appears.
Golden State weaker on offense without Durant
One of the lingering questions about the Warriors in Durant's absence is how they replace his efficient 26 points per game -- 34 PPG in this year's playoffs before his injury. Well, part of the answer is they don't entirely. Golden State's offense isn't as potent without Durant.
Using Basketball-Reference.com's game logs, which do not track team turnovers, the Warriors have posted a 115.5 offensive rating in games both Curry and Durant have played the past three seasons. With Durant out of the lineup, that has dropped to 112.9 points per 100 possessions. (For comparison, the Warriors' 2018-19 regular-season offensive rating without team turnovers was in between: an NBA-best 114.5.)
It's also not true that the ball moves better with Durant out, at least in terms of producing assists. Golden State has assisted on 68.5 percent of its field goals in games Curry has played without Durant. When they both play, that mark has been ... 68.5 percent.
To the extent that the Warriors are still scoring at an elite rate without perhaps the NBA's best individual scorer -- their offensive rating in games Curry plays and Durant does not still would have ranked second in the league this season behind the Rockets -- that's largely because of the Splash Brothers (Curry and Klay Thompson) and Andre Iguodala.
Curry's usage rate has massively increased without Durant from 30 percent of the team's plays in games both are available to 34 percent, higher even than he posted in his 2015-16 unanimous MVP season (32.6 percent). Given that larger role, it's remarkable that Curry's true shooting percentage has slipped only from .643 with Durant to .625 without him. As a result, Curry's scoring per 36 minutes has jumped from 27.6 in games Durant plays to 31 in games he has missed.
Thompson's usage has also increased from 24 percent of Golden State's plays when Durant is available to 28 percent without him. The metronomic Thompson has seen virtually no change in his true shooting, down ever so slightly from .579 to .576.
As for Iguodala, often Durant's replacement in the starting lineup -- though this season, he was already starting in the playoffs alongside Durant as part of the smaller "Hamptons Five" unit -- both his usage and his efficiency have jumped in games Durant misses. Iguodala has gone from using 11 percent of the team's plays to 13.7 percent, while his true shooting percentage has skyrocketed from .569 to .635, largely because of 38 percent 3-point shooting without Durant as compared to 31 percent accuracy with him.
Despite his strong Western Conference finals performance, Draymond Green is the one Warriors mainstay whose play has noticeably suffered when Durant is sidelined. Green's true shooting percentage has dropped from .548 with Durant in the lineup to .503 without him, meaning his points per 36 minutes have barely increased (from 10.9 to 11.5) even with increased usage.
Durant's absence compensated with defense, consistency
It's self-evident that if Golden State performs worse on offense without Durant and better overall, the explanation must lie at the defensive end of the court. Indeed, the Warriors' defensive rating has dropped from 105.0 in games both Curry and Durant play to 99.7 with just Curry. Adjusting for opponent quality, Golden State has held teams 5.8 percent below their usual per-possession output without Durant, which would have been best in the league this season by a wide margin. (Milwaukee led the NBA with defense 4.5 percent better than league average.)
I think this makes sense for a couple of reasons. First, as good as Durant can be defensively when engaged, replacing him with Iguodala is an upgrade at the defensive end.
Second, the Warriors' toughest foe -- particularly this season -- has been complacency. Without the margin for error provided by Durant's presence, Golden State has to be more locked in. That might be difficult to maintain over a full season, but for the stretches of games Durant has missed, it's possible for the Warriors to return to the kind of elite defense we saw on a more regular basis earlier in the team's championship run.
We also can show Golden State's superior consistency without Durant by another method. Let's look at the Warriors' game-by-game net ratings, adjusted for opponent quality and the location of the game. As a percentage of games played with Curry and without Durant and with both players, here's how those fall into various buckets.
This analysis suggests, unsurprisingly, that Golden State's ceiling is much higher with Durant. Just once in the 29 games he has missed and Curry has played have the Warriors posted an adjusted game score rating of 30 or better. That has happened a full 10 percent of the time when both Curry and Durant play.
At the same time, when Durant plays, Golden State is also prone to negative adjusted game score ratings, which don't always translate into losses -- it's possible to play at a below-average level and still beat a weak opponent, particularly at home -- but are obviously much more likely to do so.
With Durant sidelined, the Warriors have far more often fallen into a narrower band between a game score of 0 and 30, doing so in 84 percent of their games as compared to just 71 percent of them when Durant plays. That level is easily good enough to beat lesser foes, but it might not be sufficient in the NBA Finals.
Given the Raptors' regular-season ratings, Golden State would need an adjusted game score of about 10 or better to win in Toronto -- something the Warriors will have to do to take the series without home-court advantage. Intriguingly, Golden State has fallen short of that mark three times in the five games Durant has missed this postseason.
That's partially because the Rockets had a worse net rating in the regular season than the Raptors, and your mileage may vary on whether that's a reasonable assessment of the true ability of those two teams. Certainly, the Warriors will face a more formidable challenge in the Finals than they did from Portland in the conference finals, when they needed to come from deficits of at least 15 points to win each of the final three games.
Golden State's remaining core players are certainly capable of beating Toronto even if Durant does not return during the Finals. The Warriors' record without Durant is proof of that. Still, a deeper look suggests that for Golden State to peak against the Raptors, Durant's return would be beneficial.