Conflict between experimental and theoretical probability results

[Akshay_Anti]

Hello!

Recently, I read abt some experiments related to probability which involved tossing of coins.It listed the experiments performed by following-
Comte buffon:- 2048 heads from 4040 tosses.
J E Kerrich:-5067 heads from 10000 tosses.
Karl Pearson:-12012 heads from 24000 tosses.

My question- Theoretical probability states the result should be approximately 0.500. Okay, agreed! But why only number of heads is greater than number of tails and not vice versa?

Thanks in advance

 

[jashua]

It may happen accidentally, or the coin used may be biased. But notice that for the last experiment the bias is very low. In the limit, we expect the result be 0.5 as you agreed.

 

[Akshay_Anti]

exactly... as the no of trials increase, the value tends to 0.500

but when i searched for documentations on net, i found only 1 case when no of occurences of tails was larger.. And that too after 15 cases in favour of heads... Any idea why?

 

[jashua]

Then I think there may be a little bias as both sides of a coin are not completely the same.

Notice that in probability, the experiment "flipping a coin" assumes that there is no bias.

 

[Akshay_Anti]

true that... but all coins being biased in a similar fashion that heads turn up more than tails? That's not just co-incidence...

 

[ssd]

true that... but all coins being biased in a similar fashion that heads turn up more than tails? That's not just co-incidence...

1/ None of the coins are biased.
2/ That's just co-incidence.

These four figures indicate nothing. Are you sure that at all stages the of all those four experiments Heads were ahead of tails?