This is looking right to me.  It's not really twice as long for the
second light switch cycle as the first.

Thanks,

-- 
Raul

On Tue, Sep 11, 2012 at 1:04 PM, John Randall
<[email protected]> wrote:
> I get the expected value 10417.7 days for the 100-prisoner problem.
>
> The n-prisoner problem has expected value E n , where
>
> H=:+/ @: % @: >: @: i.
> E=:* (<: + H@:<:)
>
> The time is a sum of two geometric random variables per newly selected
> prisoner, summed over all (but one) prisoners.
>
> Best wishes,
>
> John
>
> Pierpaolo Bernardi wrote:
>> On Tue, Sep 11, 2012 at 12:06 AM, Roger Hui <[email protected]>
>> wrote:
>>> V. good.
>>>
>>> A related question for the 100 Prisoners problem, is the expected
>>> number of rounds before the prisoners go free, if each prisoner has
>>> the same chance of being picked in each round.
>>
>> Running a simulation, without much thinking, I get around 10400 days.
>>
>> I'm eager to learn the proper way to compute this number, though.
>>
>> Cheers
>> P.
>> ----------------------------------------------------------------------
>> For information about J forums see http://www.jsoftware.com/forums.htm
>>
>
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to