Monday, April 16, 2007


When trying to determine an outcome by the tossing of a coin, 75% of the population call heads. Is there any way to use this piece of information to your advantage?


Mike Peatman said...

Heads I won, tails you lose seems the only response to this key data.

Anonymous said...

As Vic Reeves quite rightly points out, 82.7% of statistics are made up.

Martin said...

93.6% of the remaining statistics are based on subjective questions asked of small sample of the population, in tests which are not double blind. (read the small print on any cosmetics ad)

Jonathan Potts said...

There's no way that involves the standard "pick H or T and then toss a coin - if you are correct you win" way of using a coin toss, since the event is completely independent of the knowledge of either person playing. But you could try this...

1. Ask the person "heads or tails?"
2. Regardless of what they say, say that you reckon it will be heads.
3. Toss the coin into a box where neither you nor the other person can see it.
4. Ask the person "heads or tails?"
5. Regardless of what they say, say that you reckon it will be tails.
6. Toss another coin into the same box.
7. Repeat.
8. Given your statistic, they will have chosen 3 heads and 1 tail (as an average) but you will have chosen 2 of each. Open the lid of the box to see who was closest. Chances are it will be you.

Another way: Get a whole load of people (say 8) to write down either "heads" or "tails" on a bit of paper and not show anyone. Then ask someone else to write down down how many people they reckon will have written "heads". Most will say 4 (or there abouts). But you suggest 6. Chances are you will win.

Martin said...

Jon, Surely if the bias exists, and you play a fixed level bet against an infinite number of people, always letting them choose, you would win?

What's wrong with the logic here? anything?