Thursday, December 5, 2013

The Value of Contesting Shots

The value of contesting a shot
“The idea is not to block every shot. The idea is to make your opponent believe that you might block every shot.” -Bill Russell

Bill Russell is one of the all-time greats in basketball. His shot-blocking ability is legendary and the NBA has never seen a player like him since. The heart of Russell’s quote is about effort: the daily grind of contesting every shot. 

Until recently, we had no dataset that could separate the grinders from the slackers, the everydayers from the spotlight-seekers. Dwight Howard is widely considered to be the player most similar to Russell, one of the best defenders in the NBA. 

But Dwight is no Russell. Despite the high rate of blocks, he doesn't contest shots nearly enough. And the difference between not contesting a shot and contesting it is huge. To examine how huge, we go to the numbers.

Vantage tracks every level of shot defense, including contested and pressured shots. Contested shots are defined as being within three feet of the shooter while getting your hand up. Pau Gasol contesting a shot:

video

Vantage defines pressured shots as being within three feet of the shooter but with no hand up. Dwight Howard pressuring a shot:

video

As we will see in the following analysis, the difference between getting your hand up (pressured vs. contested) can make a huge difference, especially near the basket.

Previous research has shown that the difference in FG% between a contested shot and pressured shot is more than 10%. However, when we dig deeper and look at specific shot locations, we find that the difference in FG% between a contested shot and pressured shot in the low post (basically shots close to the basket) is about 20% (about the difference between Austin Rivers and LeBron James!).

Moreover, we also find that shots near the basket are contested or altered on only 37.6% of the attempts. From here, we can start to calculate the value of a contested shot vs. pressured shot. Given that difference in FG% of about 20%, we can expect the value of a contested shot vs. pressured shot to be 0.4 points per possession near the basket or about 40 points per 100 possessions. Forty points may seem like a lot, but it’s like shooting bunnies near the basket while being wide open.

Let’s examine the difference between Dwight and Pau in addition to three other players -- Amir Johnson, Kosta Koufos, and Taj Gibson -- and estimate how much impact contesting a shot near the basket can make.

Contested/altered shots near the basket
PCF- Player Contest Frequency, how often a player contests or alters a shot near the basket
Opp FG%- Opponent FG% when the player contests or alters a shot near the basket
PCF +/-: PCF minus the league average contest frequency on shots near the basket
DPS/100- Estimated Defensive Points Saved per 100 possessions on shots near the basket

Pau Gasol doesn’t block a ton of shots and the perception is that he’s a soft player who doesn't always play hard enough on defense. Yet Pau Gasol contests shots at a much higher rate than Dwight. Dwight may block a ton of shots, but out of the five, he contests the lowest percentage of shots near the basket. Undoubtedly, this is due to a disappointing 2011-12 season where trade rumors may have affected his play plus an injury riddled 2012-13 season (not coincidentally, the Lakers ranked 18th in defensive efficiency last year while the Magic were 13th in defensive efficiency the year before).

Regardless, Dwight contested shots marginally higher (40.7%) than the league average of 37.6% as reflected in column three (Opp. FG%). Perhaps most surprising is the fact that Amir Johnson, Pau Gasol, and Kosta Koufos all contested shots at a higher frequency than both Dwight and Taj, who both finished in the top 30 in blocks per chance.

FG% on contested shots doesn’t vary much between players in this five player sample. Given Dwight’s wingspan and athleticism, if there was a noticeable difference you would expect to see his Opp. FG% to be lower than Pau Gasol or Amir Johnson. Intuitively, then, FG% has some of the quality of BABIP in baseball: once the ball leaves the bat or the hand, whether it is caught or the ball goes in the basket is random.

Here’s why it’s important: if we make the assumption that FG% doesn’t vary much on contested/altered shots for different players, we can estimate the defensive points saved on contested shots per 100 possessions vs. if those same shots were pressured. The DPS/100 column shows the PCF minus the league average (which is just the difference between the frequency with which the player contests shots vs. the league wide rate), multiplied by the expected value difference per possession between a contested shot and pressured shot (the 0.4 value from earlier in the article). In the DPS/100 column, we can really see the value of getting your hand up and contesting shots for different players. 

Think of it this way: if Kosta Koufos didn't get his hand up on those 15.08 extra shots per 100 possessions that he contests versus the league average (PCF +/- column), his team would be expected to give up 6.25 more points per 100 possessions. That’s the difference between winning and losing most games. Getting your hand up and contesting a shot near the basket can make an absolutely massive difference.




No comments:

Post a Comment