On The Crest
It’s a sparsely populated meme, but lately I’ve found myself stuck on clutch. I know that for many of you, discussions of clutch performance are about as important as a velvet painting of a whale and a dolphin getting it on. But for all it lacks in grand scheme significance, I still find it an immensely curious subject.
Last week we looked at what clutch means in terms of wins and losses, finding that clutch performance explains some of the random noise in a predictive model like Pythagorean Expectation. The conclusion being that executing down the stretch, presents an opportunity to out-achieve the standard established by your overall talent. Today I’d like to turn to individuals and revisit an exercise I did last season. The idea is not to look at a player’s clutch performance as a comparison to expectations or the achievement of any other individual, but rather as a comparison to what they have done across the rest of the season.
The data-set below covers only the players who have seen at least 100 clutch minutes this season. As always, I’m defining clutch as the last five minutes of the 4th quarter or overtime, with neither team ahead by more than five. Except for shooting percentages, these statistics are per 40 minutes. For each category you’ll see how much that statistic changes for each player in clutch situations, compared to their season averages.
By it’s nature, clutch raises the importance of context and shrinks sample sizes, making most conclusions based on clutch data generalizations and assumptions at best. I’m trying to avoid those pitfalls by making this examination purely historical. I am not out say that one player’s inherent skill or make-up is greater or more adapted to perform in these situations that any other’s. It is also not meant to predict how any future situation will resolve itself. It is merely a look back at who has done what this season.
* PPM is a poorly chosen acronym for possessions used per 40 minutes
A few thoughts:
- As I said, I did this same analysis last season. Then, as now, I am struck by how few players take on a larger offensive role and also do it with greater efficiency. 20 of the 49 players we are looking at here averaged more field goal attempts per 40 minutes in clutch situations. Just three, Jason Terry, Danny Granger and Rudy Gay also increased their field goal percentage.
- It’s striking how limited the Thunder appear to be in crunch time. Across the regular season they’ve been a three-headed best offensively, but with the game on the line James Harden takes 9.4 fewer shots per 40 minutes. Most of those shots are gobbled up by Kevin Durant, but Russell Westbrook also shoots more often in close games. Combined those two players average an extra 11.5 possessions used per 40 minutes. Both are talented players but they shoot worse from the field in close games than they do across the entire season. It’s still a legitimate question as to whether their rigid offensive duality in close games make them easier to defend. I mean no slight to Nick Collison, Serge Ibaka and the rest of the Thunder, but given the way their team has approached close games all season, is there any reason an opposing coach would pay defensive attention to anyone besides Durant or Westbrook?
- We can see, numerically, how Chris Paul‘s method changes along with the game situation. The perception that he shrugs off his robes and assumes the mantle of finisher in the last five minutes turns out to be incredibly accurate. In crunch time minutes, Paul averages 2.4 fewer assists per 40 minutes and takes an extra 7.4 shots per 40 minutes. There is the influence of his teammates here as well but I find it interesting that his per 40 minute +/- ratio is 6.9 points worse in clutch situations. His FG% and 3PT% are both lower than in regular game situations as well.
- His shot selection still gives me shivers, but Danny Granger deserves some recognition for what he’s done, particularly in the past few weeks. Over the past few seasons his performance in close games has been inconsistent to say the least. Lately he’s really raised his game, bringing the Pacers right along with him. Granger is taking more shots from the field, the free throw line, and behind the three-point line and has raised his efficiency from all three areas in close games. He’s also increasing his assists and +/- ratio, while cutting his turnovers. TrueHoop found that the Pacers have been one of the best clutch defenses this season, but Granger is doing his best to carry the offense in those situations as well.
- Just Like last season, Kobe Byrant’s numbers mislead slightly. His assists per 40 minutes go up slightly in close game situations, but so do his field goal and free throw attempts. Using more possessions overall means that despite the assist increase, he’s even more likely to shoot than pass as the game gets tight. Before rolls of toilet paper start raining down on my house, I’m not criticizing Kobe, just describing indisputable patterns.
- In the analysis from a year ago, Dirk Nowitzki was the stand-out star. He took on a much larger share of the Mavericks’ offense in crunch-time, taking an extra 3 shots and an extra 12.8 free throw attempts per 40 minutes. He also handed out more assists per 40 minutes and raised his field goal percentage 7.8 percentage points. This season he’s taking fewer shots per 40 in clutch situations, and although he averages more free throw attempts per 40 in close games, it’s a much smaller increase than last season. Most importantly instead of a field goal percentage that goes up in close games, it’s now falling. Dallas was a crunch-time juggernaut last season, especially in the playoffs. Unfortunately Dirk’s struggles explain more than a few of their frustrating losses this season.