General Forum

Re: secret prizes

Craig — Tue, 04 Mar 2008 15:34:01 GMT

9 LEE:
Can't Compute.. Can't Compute.. Self Destruct !!!

Lee, where do you find all these pictures you attatch to your posts.

Craig

Re: secret prizes

Jandyt — Tue, 04 Mar 2008 15:01:28 GMT

9 LEE:

it has indeed come up with the same number twice on two occasions.
Lee

Oh good, I'm about due a trip to Eastbourne!

Andy T.

Re: secret prizes

9 LEE — Tue, 04 Mar 2008 14:58:44 GMT

Can't Compute.. Can't Compute.. Self Destruct !!!

Re: secret prizes

tournedos — Tue, 04 Mar 2008 14:48:11 GMT

9 LEE:
Correct. the number generator is asked to draw 10 random numbers between 1 and the latest member number.

I hope this is implemented properly, and not like "rand() modulo 700", which would weigh the beginning of the number range

(No, I'm not serious. But I once came across a vote count program, which would use an operation like the above for resolving ties. And this was in a technical university...)

-mika

Re: secret prizes

Craig — Tue, 04 Mar 2008 14:45:02 GMT

9 LEE:

britops:
Is it not true the random number generator could draw the same number for all ten prizes? As I understand it, it doesn't remove the number from the draw after it has been drawn.
Correct. the number generator is asked to draw 10 random numbers between 1 and the latest member number. It is not asked to generate unique numbers, and although i'm sure the chances are slim (i'm sure someone will come up with the odds) - i have tested the generator numerous times and it has indeed come up with the same number twice on two occasions.
Lee

Well that is how I would expect a random number generator to work. No I don't know the odds either.

Craig

Re: secret prizes

9 LEE — Tue, 04 Mar 2008 14:18:33 GMT

britops:
Is it not true the random number generator could draw the same number for all ten prizes? As I understand it, it doesn't remove the number from the draw after it has been drawn.

Correct. the number generator is asked to draw 10 random numbers between 1 and the latest member number. It is not asked to generate unique numbers, and although i'm sure the chances are slim (i'm sure someone will come up with the odds) - i have tested the generator numerous times and it has indeed come up with the same number twice on two occasions.

Lee

Re: secret prizes

Daniel — Tue, 04 Mar 2008 13:48:26 GMT

tournedos:

P = (1 - (10 / 700)) ^ 12 ~= 0.84 (or of course, simply the opposite of the one in six you mention).

1 - 0.84 = 0.16 ~= 1/6

Re: secret prizes

Craig — Tue, 04 Mar 2008 13:46:04 GMT

britops:
Is it not true the random number generator could draw the same number for all ten prizes? As I understand it, it doesn't remove the number from the draw after it has been drawn.

That I also how I believe it works.

Craig

Re: secret prizes

Tue, 04 Mar 2008 13:44:08 GMT

Is it not true the random number generator could draw the same number for all ten prizes? As I understand it, it doesn't remove the number from the draw after it has been drawn.

Re: secret prizes

Craig — Tue, 04 Mar 2008 13:41:12 GMT

What I think the OP was saying is. You would expct members to win for a second time lese and less as the draws go on. Due to the fact there are a greater number of members each draw as more people join as gold members. Hope that made sense.

Craig

Re: secret prizes

joeyboygolf — Tue, 04 Mar 2008 13:09:50 GMT

Snap!!

Re: secret prizes

9 LEE — Tue, 04 Mar 2008 13:06:29 GMT

I don't even get the initial question!!

Re: secret prizes

Stonk — Tue, 04 Mar 2008 12:53:26 GMT

Figures? What figures! All I can see is ASCII. I used to be able to draw a car using ASCII.

Actually, I did try to work out the odds too but my wife caught me staring at my fingers and counting out loud then warned me about how much it hurt last time.

Re: secret prizes

Medogsfat — Tue, 04 Mar 2008 10:16:51 GMT

Is Stephen Hawking in da house?

Yes that's exactly the figures I got too......... cough........ cough......

Chris.

Re: secret prizes

tournedos — Tue, 04 Mar 2008 09:13:36 GMT

So that means that we have about 700 gold members?

In that case, the probability of someone not winning at all during a year is

P = (1 - (10 / 700)) ^ 12 ~= 0.84 (or of course, simply the opposite of the one in six you mention).

Calculating further, the probability of a member winning exactly twice during a year is

    P = (10 / 700) ^ 2                   (two winning months)
        * (1 - (10 / 700)) ^ 10           (ten non-winning months)
        * ( 12! / ( 2! * ( 12 - 2)! )))   (the winning months can be selected in this many ways)
        ~= 0.012

...or little more than one percent. The odds for winning three times in a year would still be better than one in six thousand.

Yes, I have a boring day at work.