A Statistical Analysis of the 2-for-1

[HuskerFever]

There have been a few discussions throughout the season pertaining to the "2-for-1" method of ending a half. For those who are curious, here is a statistical approach to analyze the impact that a 2-for-1 has in a basketball game.

 

Assumptions:

1) The inputs for 2pt% and 3pt% come from Nebraska and our most recent opponent, Ohio State.

2) Petteway was used for 2pt (41.7%) and 3pt (43.2%) attempts, as he has the highest percentage in these categories.

3) Smith Jr. was used for a 3pt attempt (41.0%); Craft was used for a 2pt attempt (47.2%).

4) There is an equal chance that the team may attempt a 2pt and a 3pt attempt.

5) 2-for-1 means that Nebraska has the ball, attempts a 2pt or 3pt play, Ohio State attempts a 2pt or 3 pt play, and then Nebraska attempts a 2pt or 3pt play (no FT or any other outcome is analyzed).

 

To save headaches for most of you, I won't dive into the meat of the statistics, but here are the possible outcomes by %:

 

2-for-1 Outcomes:

(+ means NU increased lead by x amount after 1. NU, 2. Opponent, 3. NU possession)

+6: 2.52%

+5: 5.13%

+4: 3.67%

+3: 16.66%

+2: 16.79%

+1: 6.69%

0: 29.08%

-1: 5.06%

-2: 7.71%

-3: 6.70%

 

1-for-1 Outcomes:

(+ means NU increased lead by x amount after 1. NU, 2. Opponent possession)

+3: 12.07%

+2: 12.07%

+1: 5.10%

0: 41.28%

-1: 4.43%

-2: 13.40%

-3: 11.64%

 

 

To put these into laymen's terms:

 

2-for-1:

Increased lead/shrunk deficit: 51.44%

No change: 29.08%

Decreased lead/increased deficit: 19.48%

 

1-for-1:

Increased lead/shrunk deficit: 29.25%

No change: 41.28%

Decreased lead/increased deficit: 29.48%

 

Therefore, by utilizing the 2-for-1 method at the end of the half, you give your team a 51.44% chance to increase the lead/shrink the deficit. By utilizing the 1-for-1 method at the end of the half, you only give yourself a 29.25% chance to increase the lead/shrink the deficit.

 

If appears that you should, statistically, only utilize the 1-for-1 method if you are looking to come out even at half.

 

I was curious when the question of 2-for-1 was asked, so I decided to take a look at the impact based on the numbers. Just thought I'd share this with the HHC family. Hopefully someone finds this useful.

[Huskerpapa]

Ouch...that hurt my head -

[HB]

Damn, Fever, I hope you are able to a lot of money off of those statistical abilities.  That shows some serious game

[tcp]

I'll respond in more depth later when I'm less stoned, but I need to know what kind of data you plugged in. Did you have specific data for short possessions that would qualify as 2 for 1? Most teams keep that data now expressed as percentages for short possessions and  'normal' percentages, and the difference is usually pretty significant. That would seriously skew your outcomes.

 

Oh, and why did you even bother using OSU? Why not just aggregate NUs possession stats based on time, since that's the primary qualitative component of a 2 for 1 possession?

[hhcmatt]

I'll respond in more depth later when I'm less stoned, but I need to know what kind of data you plugged in. Did you have specific data for short possessions that would qualify as 2 for 1? Most teams keep that data now expressed as percentages for short possessions and  'normal' percentages, and the difference is usually pretty significant. That would seriously skew your outcomes.

 

Oh, and why did you even bother using OSU? Why not just aggregate NUs possession stats based on time, since that's the primary qualitative component of a 2 for 1 possession?

 

I'd assume we'd have better odds using anything but OSU

[hal9000]

[Norm Peterson]

Or you could just go the lazy route and say a 2 for 1 gives me probably two possessions instead of just one to end the half.  Score quick and I should get the ball back rather than let the other team hold for the last shot.