Pressure

As a stat-head, I’m not one that believes in hot streaks or clutch plays. That’s not to say that such things should be banned from the sports vernacular, just that they aren’t skills people have or can acquire. I do believe that people tend to perform better under pressure, be it in basketball or any other activity. I don’t know of a good way to support that in basketball, though. Clutch shooting can always be offset by clutch defense, so it’s not something that comes out in the stats. Free throw shooting can’t be defended, so maybe that can tell us something about how pressure affects players.

     2005 Season
   MOV    FT%  Games
   33+   66.8   171
  29-32  66.4   123
  25-28  66.2   158
  21-24  67.7   285
  17-20  69.0   454
  13-16  69.4   655
   9-12  69.4   859
   5-8   69.4   871

This data above shows how well Division I teams shot free throws last season based on the margin of victory or defeat (MOV) of each game. In games decided by 33 points or more, teams shot 66.8% from the line, and shooting tends to improve as the games become more competitive. This fits my perception of how players handle pressure. During lopsided games, there’s less concentration and free throw percentage is lower than it is during tighter games.

But I conveniently omitted games decided by four points or less.

   MOV    FT%  Games
   1-4   68.0   912

There’s a noticeable decrease in free throw percentage when the games get really tight. You might think that this could be a statistical fluke. Without going through some excruciating number-crunching, just trust me that due to the gargantuan sample size involved (over 30,000 FTAs in each of the 1-4 and 5-8 groups), the difference of 1.4% means something. And if that doesn’t convince you, the same pattern exists in data from the 2004 season and confirms the idea that this is no accident.

     2004 Season
   MOV    FT%  Games
   33+   67.4   164
  29-32  66.5   112
  25-28  67.7   170
  21-24  67.5   282
  17-20  68.2   439
  13-16  69.0   675
   9-12  69.4   797
   5-8   69.6   858
   1-4   68.4   882

So what’s going on here? Are players reaching the breaking point in those really close games and falling victim to the choke factor? I was thinking that, but then this set of data cleared things up.

 MOV    FTA/G
 33+    39.4
29-32   38.1
25-28   39.7
21-24   39.0
17-20   38.4
13-16   40.9
 9-12   42.5
 5-8    42.3
 1-4    38.7

This data is combined for the last two seasons and indicates that there are significantly fewer free throws in really close games. The reason for this is that there are fewer intentional fouls towards the end of the game. And those extra fouls tend to inflate the free throw percentage in the sort-of-close games because the trailing team is forced to foul the winning team’s better shooters. Indulge me for one more set of data here – again this data combines the last two seasons.

          Winners       Losers
   MOV   FT%  FTA/G    FT% FTA/G
   1-4   68.1  21.0   68.3  17.8 
   5-8   70.8  24.2   67.9  18.0
   9-12  71.0  24.8   67.1  17.7
  13-16  71.4  23.1   66.3  17.7

Essentially, all of the increase in free throw attempts in games decided by 5-12 points is from the winners, and they’re the ones that shoot better from the line in those games. That’s because the better free throw shooters end up going to the line at the end of the game. In the very tight games, there is less end-of-game fouling, which allows free throw attempts to be more randomly distributed among all types of free-throw shooters. At least that’s the theory I’m going with.

So in the end, this experiment didn’t teach us much about pressure. Although it deserves pointing out that even in the games decided by four points or less, teams shot better than in the 20+ point blowouts.

What it does show is how much winning can affect the free throw production of a team. Take a team like Utah that was third in the nation in its ability to get to the line. Based on the margin of their wins – they were 15-5 in games decided by 5 to 16 points – they benefited as much as any team from what I’ll term Intentional Free Throw Padding. (See a doctor if you suffer from this.) Subtracting an estimation of the additional late-game free throws would have dropped their rank in free throw rate to around the high teens.

So the effect of IFTP is real but even in the extreme it’s not overwhelming. Ability is much more important to free throw rate than circumstance is.