Re: Binomial Coefficient Problem



> I have the following problem:
>
> I solved the probability of some event to be equal
> to:
> C(10,2)*C(k-10,18)/C(k,20).
>
> The problem asks to find the value of k for which
> that probability is
> maximized.
>
> So, I set C(10,2)*C(k-10,18)/C(k,20) = 1.
>
> Can somebody help me solve this problem? I think
> that some rule of
> binomial coefficients must apply to make this really
> simple, but can't
> think of which one. The answer is k=100, but I have
> no idea what
> procedure to follow to get it.
>
> Much thanks in advance.
>


Your probability acan be rewritten as proportional to

[(k-20)(k-21)...(k-27)]/[k(k-1)...(k-9)] or

[(k-20)/k}][(k-21)/(k-1)]...[(k-27)/(k-7)]/[(k-8)(k-9)].

I ran these calculations on Excel for a wide range of k and found it to be maximized at k = 99 and 100.

Jack
.



Relevant Pages

  • Re: This says it all
    ... has a probability of 2/13983816 or double the probability ... Your probability of winning a jackpot ... is 2/13983816 or double the probability of one ticket. ... It's funny that swishy Jack is parroting innumerate bimbo Sherry's ...
    (rec.gambling.lottery)
  • Re: This says it all
    ... Jack, to kill rouge like John, do not attempt a head shot. ... has a probability of 2/13983816 or double the probability ... is 2/13983816 or double the probability of one ticket. ...
    (rec.gambling.lottery)
  • Re: Comparing fractions (or proportions)
    ... THE exact method provides a p-value which is BY ... result as least as extreme as that obtained by chance ... value i.e. the probability to get a data (x, n1, y, ... *** Jack Tomsky response ...
    (sci.stat.math)
  • Re: Comparing fractions (or proportions)
    ... THE exact method provides a p-value which is BY DEFINITION is the probability to obtain by chance a result as least as extreme as that obtained by chance assuming that H0 is true. ... *** Jack Tomsky response ...
    (sci.stat.math)
  • Pr( N=2 randomly selected persons has DIFFERENT birth MONTHS) -- the definitive answers
    ... Jack Tomsky did a really nice piece if detective work for the ... Afonso's assumption that the probability of birth in any given ... LAA> (which is a rather good approximation, evidently, mine ...
    (sci.stat.math)

Quantcast