Nylon Calculus: Warriors-Rockets win probabilities, assist maps, style charts and more

HOUSTON, TX - MARCH 13: James Harden #13 of the Houston Rockets handles the ball against Stephen Curry #30 of the Golden State Warriors on March 13, 2019 at the Toyota Center in Houston, Texas. NOTE TO USER: User expressly acknowledges and agrees that, by downloading and or using this photograph, User is consenting to the terms and conditions of the Getty Images License Agreement. Mandatory Copyright Notice: Copyright 2019 NBAE (Photo by Jesse D. Garrabrant/NBAE via Getty Images)
HOUSTON, TX - MARCH 13: James Harden #13 of the Houston Rockets handles the ball against Stephen Curry #30 of the Golden State Warriors on March 13, 2019 at the Toyota Center in Houston, Texas. NOTE TO USER: User expressly acknowledges and agrees that, by downloading and or using this photograph, User is consenting to the terms and conditions of the Getty Images License Agreement. Mandatory Copyright Notice: Copyright 2019 NBAE (Photo by Jesse D. Garrabrant/NBAE via Getty Images) /
facebooktwitterreddit

Previewing the second-round series between the Golden State Warriors and the Houston Rockets with shot charts, assist maps, offensive style charts and expected win probabilities.

Shot charts

The Golden State Warriors finished off the Los Angeles Clippers in six games on the offensive strength of Stephen Curry and Kevin Durant. Curry averaged 25 points per game on a 71 true shooting percent and 24 percent usage, whereas Durant averaged 35 points per game on a 71.2 true shooting percentage and 32 percent usage. Durant has two different types of play style, a more passive play finisher when Curry is on the floor, and a more aggressive play creator when Curry is off — this can be seen below where Durant’s percent assisted field goals is broken out by Curry’s on/off data.

Despite their offensive mastery, the Warriors did have some defensive trouble against the Clippers. Most notably, the Warriors struggled to contain Lou Williams and Montrezl Harrell. Based on the points per shot surfaces, the Warriors gave up more points per shot at a wide range of location across the court. The ability of Williams and Harrell to score presents an interesting question of how the Warriors will prevent the more potent combination of James Harden and Clint Capella from doing far more damage.

Despite a season’s worth of James Harden offensive brilliance, the Houston Rockets dispatched the Utah Jazz in five games on the basis of excellent defense. Their points per shot surfaces below are predominantly green, indicating that the Rockets performed over their season averages in defensive efficiency. In the regular season, the Jazz had an offensive rating of 110.2 on a 57 true shooting percentage but were held to an offensive rating of 99.2 on a 50 true shooting percentage in the postseason. However, the Jazz shot 26-of-100 on 3-point attempts with no defender within six feet, over ten percentage points worse than the worst team in the regular season. In contrast, the Golden State Warriors have averaged 40 percent on those shots in the regular season and 43 percent in the playoffs. It is unlikely that Houston will be able to survive against the Warriors giving up a similar number of uncontested 3s, so how they make defensive adjustments to that fact will be a crucial point in deciding the series.

— Andrew Patton (@anpatt7)

Offensive style chart

These charts are not meant to evaluate whether an offense is good or bad. They are designed to help illustrate how teams go about the goal of trying to put the ball in the basket. Each team’s offense is evaluated on four stylistic spectrums.

Ball movement is measured with the average touch time for each team, from the NBA’s player tracking statistics. A lower average touch time means the ball is moving from player to player more quickly.

Player movement is measured with a combination of different NBA.com tracking statistics and works out to average distance traveled per 24 seconds of offensive possession.

Pace is measured with the average length of an offensive possession from Inpredictable, a more accurate representation for how quickly a team is working than traditional pace.

Shot selection is measured with MoreyBall percentage — in this case the percentage of a team’s true shooting opportunities that came at the rim, from the free throw line, or on a 3-pointer. It’s a generalized measure but captures something about how much each team hews to the shots that are, on average, the most efficient.

— Ian Levy (@HickoryHigh)

Assist maps: James Harden

James Harden is the most dominant offensive force in the NBA right now — the league-leader in time of possession (9.3 minutes) and points produced (36 points scored + 18 points assisted) per game. He uses his unique combination of effective 3-point shooting off-the-dribble (12 pull-up 3-point attempts per game), a hard-to-contain first step (20 drives per game), and a knack for drawing contact in the lane (11 free throw attempts per game) to keep opposing defenses under constant stress. And all of the attention being paid to Harden opens up a ton of quality shots (53 expected effective field goal percentage) for his teammates. In particular, Harden’s penetration often leads directly to dunks for Clint Capela or corner-3s for the Rocket wings.

Despite his immense capabilities, Harden’s assist charts demonstrate two potential limitations in his passing: 1. On both his lobs and his kick-outs he prefers to pass from left to right — zipping around his array of left-handed hook-, sling-, and slip-passes — because he is less fluid and less accurate passing in the other direction and 2. He rarely makes assists to the top of the key from inside the paint. Earlier this year, the Bucks devised a defensive scheme meant to exploit these weaknesses and downplay Harden’s strengths. Mike Zavagno has done a nice job of explaining the “Harden Rules” that govern Milwaukee’s approach:

1. The on-ball defender holds his hands up and positions himself alongside Harden’s left shoulder to force him to initiate a right-handed dribble (and to discourage Harden from taking a step-back 3 while also priming the referee NOT to call a foul),
2. Once Harden drives towards the basket, a big man is in position to help at the top of the restricted area with his hands up to prevent lobs (and to discourage layups while conceding lower-value floaters), and
3. Additional help defenders slide down to block Harden’s passing lanes to the corners, forcing him to make a less comfortable pass to the wing or to the top of the key.

The plan worked pretty well for the Bucks. They won both regular-season games while holding the Rockets offense in check, allowing just 98 points per 100 possessions. For his part, Harden had fewer assists (6.5 per game) and more turnovers (6.5 per game) than usual in those two games. Moreover, the Bucks did a good job of preventing Harden’s juiciest passes — conceding very few lobs or corner-3 kick-outs.

In the first round, the Jazz tried to implement their own version of the Harden Rules, but they were mostly unsuccessful. In the four Utah losses, Harden racked up 9 assists per game and the Houston offense hummed along with a rating of 111 points per 100 possessions. Zach Lowe and Ben Falk have both provided really thorough explanations of what went wrong for the Jazz in implementing this new defensive strategy. One big problem — especially early in the series — was that Rudy Gobert was helping too early and showing too high in the lane, making it easy for Harden to toss lobs to Capela over his head. This mistake was compounded by the fact that Utah’s other help defenders were a little lost trying to cover up the corner-3 shooters.

Still, there were moments when the plan worked for the Jazz. In Game 4, Utah forced Harden into 8 turnovers while conceding just 4 assists, including one lob and no corner-3 kick-outs. They succeeded in turning some of Harden’s favorite “downhill” passes into second-choice “toss-backs.” As a result, the Rockets managed just 91 points (with an offensive rating of 97) in that game. But, the moments of defensive cohesion proved fleeting and the Jazz eventually lost the series in 5 games.

Now it’s the Warriors turn to grapple with how to stop Harden. Last year in the Western Conference Finals, Golden State sent a variety of long, mobile defenders at Harden: Klay Thompson, Kevin Durant, Andre Iguodala, and Draymond Green all took turns. They let point guards (Stephen Curry and Shaun Livingston) and big guys (Jordan Bell and Kevon Looney) switch onto Harden in pick-and-roll coverages, too. Like the Bucks and Jazz, the Warriors tried to influence Harden towards his right hand but not in the same overtly aggressive way. The Warriors also have their own strategies for sending help to prevent and recover from breakdowns in isolation. Using their more conventional approach, Golden State gave up 10 Harden-to-Capela alley-oops in the seven games series. Harden also assisted on 16 3-pointers although, by my count, only 3 of those were kick-outs from the paint to the corner.

After Utah’s defensive breakdown in the first round, it would be easy to write-off the Harden Rules as a foolhardy gimmick. And — based on how things turned out for the Jazz — I doubt the Warriors will feel obliged to overhaul their defensive scheme before Sunday. But it wouldn’t surprise me to see at least SOME of the elements of the Bucks successful strategy seep into the Warriors’ approach to guarding Harden. Forcing Harden to his right and trying to limit his chances to throw lobs and corner-3 kick-outs will be important points of emphasis in Game 1.

— Todd Whitehead (@CrumpledJumper)

Win probabilities

To project the series, I am using my in-season game projection model. The model is trained off historical game data, and accounts for rest, travel, team strength, and matchup. Since I began using the model to predict outcomes, I have been able to correctly identify the winner in about 70 percent of games and has correctly identified the winner in 78 percent of games in these playoffs.

The Warriors come into the series as strong favorites, winning in 76 percent of the 10,000 simulations. The average length of the series was 5.7 games, with the Warriors given nearly a 13 percent chance to sweep the series. The Warriors struggled more than expected in round one against a very well prepared Clippers team, and the Rockets have certainly been watching and seeing what worked against the Warriors to prepare. The extra two days of preparation could prove extremely useful at the start of the series in terms of game plan for the Rockets. As good as the Rockets are and can be, the Warriors have a good chance to move on to the Western Conference Finals.

Jacob Goldstein (@JacobEGoldstein)