Scoring Rates During Overtime

[Eternal]

Estimates via model vs estimates via time allotment.

Code: Select all

1OT         0.0973 (by time 0.1041)
2OT         0.1957 (0.2083)
3OT         0.2937 (0.3125)
4OT         0.4380 (0.4166)

-Chris

[Crow]

Not sure what the unit is here.

[Eternal]

Fraction increase in total points over points scored in regulation.

-Chris

[talkingpractice]

Can you help to explain the units just a tad more, and I'm quite sure it's my comprehension at fault and not your explanation...

So, 9.7% more points are scored (per minute?) than in regulation, and 19.6% more are scored in 2OT (than in 1OT, or than in regulation?).

[Eternal]

No, just points scored overall. So with an extra 5 minutes of time you'd expect the extra fraction of 5/48 total points scored per side over and above total points scored in regulation in the NBA. In fact it's somewhat smaller (fatigue, slower pacing, weaker subs, etc).

-Chris

[Mike G]

This puzzled me, too, for a couple of reasons. Now that it seems to make sense, I'm offering another view of it:

Code: Select all

OT    all    each   48/5    adj
1   .0973   .0973   .934   .969
2   .1957   .0984   .945   .980
3   .2937   .0980   .941   .976
4   .4380   .1443  1.385  1.437

In 4th-overtimes, I guess all good defenders have fouled out.

"all" is the per-minute fraction of regulation-time points scored in the overtimes. Same number as in the opening post.
"each" is just the average single overtime: .1957 - .0973 = .0984 (for 2nd OT)
"48/5" is the per-minute overtime scoring avg relative to the regulation (48 minute) average.

But a 12-minute period is more efficient than a 5-minute period of play, in general; because there are discrete possessions, and a full 24-second possession is more valuable (efficient) than a fraction of that interval.

In 12 minutes, you may see avg 24 "possessions" by each team. In reality, suppose one team gets the ball with 12 seconds on the clock at the end of the quarter (or overtime).
Suppose further that the efficiency of half-a-possession is about half that of a full possession.

Now we may suppose that a 12-min. quarter contains 23.5 possessions, and a 5-min. OT has 9.5
In other words, while the ratio of minutes is 5/48 -- .1042 -- the ratio of possessions is more like 9.5/94 -- .1004

The difference is small, but it's something. A game that's 100-100 in regulation can expect to see 20 points scored in another 5 minutes, rather than 21.

The final column in the table, "adj" is adjusted to reflect that. It's the previous column *1042/1004
The possessions/minutes "inefficiency" seems to account for almost half of the reduced scoring rates in overtimes, with the arbitrary parameters I used.

[Eternal]

I considered this too - it makes sense that an overtime period would be less efficient, e.g. jump balls slow the game down by some fraction of a minute.

-Chris

[Mike G]

Has anyone studied the efficiency of possessions beginning with, say, 24 to 36 seconds remaining in the period?
A team that has the ball in that situation is often not going to run a standard possession, where you take the first good shot that avails itself.
It's considered a better strategy to shoot late in the shot clock, thus leaving the other team with just a few seconds to try and score.

This means giving your own team a less than optimal efficiency, just to give the opponent an even worse opportunity.
The overtime of 5 minutes may therefore be even poorer in full-opportunity possessions per minute than is the 12-minute quarter.

What's the average scoring in the last 5 minutes of quarters, vs the overall 5-minute average, and vs overtime avg?