Holding Out for the Last Shot

A week ago, against the Portland Trailblazers, Alec Burks really screwed up the end of the first quarter. That's what Twitter told me, anyway. He screwed up by taking a shot with 6 seconds left in the quarter and allowed the Blazers to hit a 3-pointer at the buzzer.

And yet, I wondered.

You see, I have long thought that the traditional end-of-quarter strategies are questionable. It was probably Jason Hart's magnificent 2-for-1's that really made me wonder if teams are doing the smart thing. It's kind of how I shake my head at end-of-game Hero Ball strategies.

Seriously, imagine this scenario:

Every possession, Mo Williams crosses mid-court and then just stands there until there are 5-6 seconds left on the shot clock. At that point he suddenly attacks, and the Jazz hurry to get a quick (often contested) shot off before the shot clock expires.

Does that sound like a great strategy? Does that sound like the most effective way to score? What would the typical Game Thread be like if this was what the Jazz did every possession?

Yet that's what virtually every team does at the end of every quarter. And the main reason teams do it is to try to prevent the opponent from getting an additional possession at the end of the quarter. Which is funny, because tell me how you'd react if this was your team's typical offensive strategy:

Rush down the court ASAP and throw up a 30-footer within 4 seconds.

But that's the very thing teams are so desperate to prevent.

* * *

Anyway, thinking through this logically wasn't enough. I decided to investigate how effective teams' end of quarter strategies actually are. So I went through the Play-by-Play records at ESPN of every Jazz game this year (up to the Denver loss on Wednesday). I wanted to find out how two main things: 1) how efficiently teams score when holding out for the last shot, and 2) how dangerous a rushed shot at the end of the quarter ends up with a score (the kind of shot Burks enabled that was so bad).

Here's the criteria I used:

Holding out for the last shot:

• The team gains possession somewhere between 16 and 32 seconds left in the quarter. Basically this is enough time to get possession, cross mid-court, and then stand there for a bit before beginning a final play with 5-6 seconds left in the shot clock. I included up to 32 seconds because teams routinely do this last-possession thing, even with more than 24 seconds left. My research showed me that this typically happens up to 32 seconds.
• The team shoots after 20 or more seconds have expired on the shot clock. If the team shoots earlier than this, then they likely didn't hold the ball until the 5-6 second mark, i.e. they didn't do this last-possession dance.
• Because the ends of games are so different (with the fouling game, etc.), I investigated only the first 3 quarters.
• I looked at possessions. So if a team shoots, gets the offensive rebound, and then scores ... that counts as a successful outcome of that possession.
• I counted total opportunities, total points scored, and ended up with a points per possession in these situations.

A final rushed shot

• The team gains possession somewhere between 10 and 5 seconds left in the quarter. More than that and it's not really a rushed shot ... it's actually enough time to run a quick play. Less than this and it's not a rushed shot ... it's a pray-for-a-miracle heave. Something like "Al Jefferson misses 58-foot 3-point shot." (this was my favorite line in the play-by-plays ... Why 58 feet? Why so precise? Why not call it a 60-foot shot?).
• Again, because the end-of-game scenarios often produce vastly different situations, I only looked at the first three quarters.
• Again, I looked at possessions.
• I counted total opportunities, total points scored, and ended up with a points per possession result.

Baseline

The Jazz are an interesting team to look at with this exercise, because it's ORtg and DRtg overall are pretty much the same. The Jazz score 107.0 points per 100 possessions (1.07 ppp) and they give up 107.4 points per 100 possessions (1.074 ppp). This is helpful for this exercise, because I can look at both teams in every Jazz game without adjusting anything.

So that 1.07 ppp becomes a baseline for comparison. Over the entire season, that is the average rate teams score in a Jazz game.

The Questions, Restated

I wanted to know two things:

1. Does holding out for the last shot affect a team's likelihood of scoring?
2. Do teams score enough with few seconds left to justify worrying about giving them time for a hurried shot?

The Data:

1. Efficiency of holding out for the last shot

Qualifying Possessions	Points Scored	Points Per Possession
199	143	0.72

2. Efficiency of rushed shots (4-10 seconds left in the quarter)

Qualifying Possessions	Points Scored	Points Per Possession
86	45	0.52

My Conclusion

I'm not surprised, but it's nice to have numbers that back it up:

1. When teams hold for the last possession they score at a significantly lower rate than the average rate of other possessions.
2. When teams gain possession with 5 to 10 seconds left, they score at an even worse rate ... about half as efficient as other possessions.

Basically teams do two things when they hold the ball to get the last shot of the quarter: 1) they reduce the likelihood that they'll have a good possession and score, and 2) they're WAY too worried about giving up that rushed shot. Seriously, the odds are that whatever shot a team gets off in this rushed situation it's going to be a lousy one.

The more I look at this, the funnier it is. Teams are intentionally using a possession poorly in order to prevent opponents from having an opportunity for one that is even worse.

Teams would be better off just running their offense as usual ... taking a quick shot if a good one is available, but not settling for a rushed one. Even in this scenario—a team gets the ball with 28 seconds left, pushes it up to their side of the court, and then get a quick dunk after only 8 seconds have ticked off—even then, it's not bad. Why? Because 1) you actually scored, and 2) the other team will probably hold for the last shot and thus hurt their own chances at a successful possession.

Funny wrinkles

Interestingly enough, one team seemed to NOT do this hold-for-the-last-shot-of-the-quarter thing: the Heat. Here's an example, at the end of the first quarter in the game on December 22, 2012: the Heat gained possession with 25 seconds left. Ray Allen then scores with 19 seconds left. They could have held on for the last shot, but chose not to. The Jazz, of course, then DID hold the ball for the last shot and did not score.

I didn't find a single instance in the Jazz-Heat games in which the Heat held on for the last shot. And they had opportunities. But they apparently don't worry about it. Which is smart.

The Burks Mistake

The funniest bit of all was this: Alec Burks' HUGE mistake wasn't even a mistake. The Jazz gained possession with 24 seconds left in the quarter, and Burks missed the shot at 3 seconds left. He used 21 seconds of the shot clock. He cut the time down to miracle-heave status ... as is evidenced by the play-by-play at the buzzer: "Wesley Matthews hits 35-foot three-point shot."

Come on.

When you're worried that you gave the other team enough time for a last-second 35-foot shot, you've made Much Ado About Nothing. Is there any other situation in which a 35-foot shot after 3 seconds of gaining possession is a good shot? Of course not.

Seriously, I originally counted any shot taken after gaining the ball with 10 seconds or less. But I discovered that teams score so rarely with 3 or fewer seconds left that it skewed the data. So I threw those possessions out, realizing that the miracle-heaves were nothing to worry about.

But do crappy shots sometimes go in? Sure they do. J.R. Smith has made himself a very rich dude thanks to this truth.

But it's not something to build a game plan around. And it's not something to create an entire strategy to prevent ... particularly not when that strategy includes hurting your own chance to score.

Final Note

Because I've learned that sometimes I have to be explicit with these things. I'm not really talking about Alec Burks here. I'm not talking about Corbin's coaching skills or not. I'm talking about a very, very common view of how to properly handle the end of quarters. It's a conventional view that most teams follow, and one that one of my favorite coaches ever (Sloan) had.

I just happen to think it's wrong.