What happens during crunch time

Thumbnail

Click for a full-sized interactive version

The above collection of charts shows how the Houston Rockets perform on four metrics during “crunch time,” defined as the last five minutes of a game in which the Houston Rockets are behind by five points or less, or tied.

Technical details

On the x-axis are seven different periods of time. The first one, called “all,” shows performance in a metric for the entire game. The next six shows a team’s performance in a metric only when the team is behind by five points or less, or is tied, with a certain amount of time left in the game. The blue line represents the Houston Rockets. The orange line represents all teams. So, if we look at the first chart, it shows how the Houston Rockets and the rest of the NBA perform on defense in general, and how that performance changes when the game is close and less time remains in the game.

Important limitations

  • Because not all games are close, the sample size of the games (and possessions from those games) as one moves to the right in each graph becomes very small. 
  • The NBA average is an unweighted average. Each team’s performance is considered equally when calculating the average, regardless of how many crunch time possessions and games the teams played. For this reason, the NBA’s offensive and defensive ratings do not mirror each other.

Observations

All teams seem to do worse during crunch time. Considering this is when teams need to score/get a stop the most, it’s interesting that whatever it is they are doing is actually less effective than what they do during the rest of the game. The Houston Rockets are even worse during crunch time compared to NBA average. In each metric, the Houston Rockets see a greater decrease in performance relative to their overall performance than the NBA average.

The Houston Rockets offense, in particular, sees a stunning decrease in performance. In both offensive rating and true shooting percentage, the Houston Rockets are higher than the rest of the NBA over the course of entire games, but plummet during crunch time, dipping more and more below the NBA average as closely contested games near their conclusion. While the sample size is admittedly small, in the final thirty seconds of games in which the Houston Rockets are down by five or less, the offense is operating in a way that would extrapolate to 37 points per 100 possessions (compared to 85 for the entire league), with a true shooting percentage of 18% (compared to 38% for the entire league).

That the NBA’s assist ratio decreases as less time remains suggests that heroball is alive and kicking. That the Houston Rockets assist ratio decreases at a greater rate than the NBA’s as less time remains suggests that heroball is employed readily by the team. Since the offensive rating and TS% of the team also decrease over the same time periods, it might be time for the Houston Rockets to rethink their crunch time tactics.

View this discussion from the forum.

This entry was posted in essays and tagged . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.
Login to leave a comment.
Total comments: 6
  • smeggysmeg says 3 months ago

    love to see the rockets run some sets in crunch time, remember reading some a few years back about how good the hornets (now Pels) offense was in crunch time because they ran offences and team ball rather than iso heroball (formally known as Isomotion under Ruddy T)...... elevator doors for Haren would be sweet (refuse to use the D anymore till he tries and or learns off the ball positioning on Defense) pretty sure in OKC they used to run him off the ball alot, remember one with bigs setting screens on each low block, ball swining and Haren cutting off screens and then into pnr.....even lakers with non shooter Marshall run some nice double pnrs (one screener rolls and another pops).......maybe in the offseason we need to add an assistant with spurs experience on his resume....... a step too far right now but some spurs motion would work great for this teamhttp://www.grantland.com/blog/the-triangle/post/_/id/64414/the-5-elements-of-the-effective-spurs-offense

  • shirtless says 3 months ago

    Any way we could see "change in assist ratio" on one axis and "change in offensive efficiency" on another with each data point being a team? I am really curious to see which teams tend to go "hero ball" during crunch time and whether those teams tend to have better or worse off. efficiency.

    Thanks!

    That's an interesting idea. Let me see what I can put together. Thanks for the suggestion.

  • bluemars says 3 months ago

    Any way we could see "change in assist ratio" on one axis and "change in offensive efficiency" on another with each data point being a team? I am really curious to see which teams tend to go "hero ball" during crunch time and whether those teams tend to have better or worse off. efficiency.

    Thanks!

  • thejohnnygold says 3 months ago

    Dwight's free throw shooting is efficient enough to sustain that approach (most of the time). Plus, if we start it early enough--say 5 minutes before the end of the game--they only have so many fouls before they have to start plugging in bench guys which gives us another advantage. The trick would be to run a hurry-up style offense that gets down court and into the pick n roll quickly and forces the defense's hand.

    I made the pick n roll suggestion a bit tongue in cheek, but if they did it and they foul us I won't mind.

  • Ostrow says 3 months ago

    The problem with that is, Dwight goes to set the screen, the big guy grabs him and Dwight is shooting free throws.

  • thejohnnygold says 3 months ago

    Perhaps we save the pick n roll for crunch time--like an ace up the sleeve. Dwight can't complain about that and it will keep Harden happy and steer him away from his current trend. Couldn't hurt to try it.