Probability and Statistics

 How many games to decide who's better?

 From: Stephen Turner Address: sret1@cam.ac.uk Date: 21 March 1997 Subject: Re: Statistically speaking.. Forum: rec.games.backgammon Google: 33328454.1DD8@cam.ac.uk

```Dodd wrote:
>
> I am sure that this question has been asked many times before, but since
> I just found this newsgroup I would like to ask it again.
>
> Statistically, how many games of backgammon must two people play before
> one can definitively say they are the better player?

Unfortunately, it's not quite as simple as that. It depends how far you are
apart after those games. The standard way to set it up as a statistical
problem is to play a pre-determined number of 1-point games -- let's say
100. If you are the same standard, there's only a 2.5% chance that you will
win at least 60 of them, which is enough to reckon that you are probably
the better player.

1) If you are only a tiny bit better, this test is unlikely to pick up the
difference. In that case, we say that the test isn't very powerful.
Conversely, if you are much better, the test will usually pick up the
difference, and it's said to be a powerful test.
2) There's a thing called a "sequential test" whereby you don't have to
play the full 100 games if someone is way ahead after (say) 50 of them.
This is often used in medical experiments, for obvious ethical reasons. But
it's rather complicated to work out what the criterion for stopping and
deciding should be with a sequential test.
3) I assumed 1-point games. The real skill is in the cube, and I've also
neglected gammons. Unfortunately, it's not possible to work out an exact
statistical test for money games without knowing what proportion of each
type of score you would normally get between two equal players. For match
play, however, you're OK: if you play 100 5-point matches, instead of 100
1-point games, and just worry about the winner not the final score,
everything else works the same way.

That's probably more than you wanted to know. But I wouldn't want to

--
Stephen Turner    sret1@cam.ac.uk    http://www.statslab.cam.ac.uk/~sret1/
Statistical Laboratory,  16 Mill Lane, Cambridge, CB2 1SB, England
"This store will remain open during modernisation. We apologise
for any inconvenience this may cause" Topshop, Cambridge
```

### Probability and Statistics

Average game and match length  (JP White, Dec 2000)
Average luck of each roll  (Timothy Chow+, Mar 2013)
Average luck of each roll  (Jørn Thyssen+, Feb 2004)
Calculating winning chances  (Douglas Zare, June 2000)
Chance of rolling x doubles in y rolls  (Raccoon+, July 2007)
Chance of rolling x or more pips in y rolls  (Tom Keith, Feb 2004)
Clumping of random numbers  (Gary Wong, Sept 1998)
Counting shots  (Koyunbaba+, June 2007)
Counting shots  (John Little+, Mar 2007)
Distribution of points per game  (Roland Sutter, June 1999)
Distribution of points per game  (Stig Eide+, Sept 1995)
Expected variation in points after a series of games  (Achim Müller+, Feb 1999)
How many games to decide who's better?  (Stephen Turner, Mar 1997)
How often is too often?  (Gary Wong, Oct 1998)
Losing after bearing off 14 checkers  (Daniel Murphy, July 1999)
Number of games per match  (Jason Lee+, Jan 2005)
Number of rolls to enter x checkers from bar  (Michael Depreli+, Mar 2011)
Visualizing odds  (Daithi, Mar 2011)