Page 1 of 1

statistics/probability question

Posted: Fri Sep 18, 2015 10:31 am
by Portia
I'm theoretically participating in a writing competition. The weekends it takes place have been particularly busy, though, so I didn't submit.

Here's how the math works:

There is a given number of writers in your bracket. In mine, there are 29. The top 15 stories get points corresponding to their placement: 15 for 1st, 14 for 2nd, 1 for 15th. The same process is repeated for the second round. The placements are added and determine whether you advance.

I got 0 because I didn't submit.

My question is, could we use probability to determine the possibility of advancing? I would arbitrarily say that your first round results are 60% predictive of your second round, with the rest being random. So it should be unsurprising if a 2nd place writer gets 1st place in this next round, though it's not impossible for a 0-point writer to vault into 3rd place for the second round (though rarer).

Anyone want to run the math on this? My gut feeling is a 5% or less chance of advancing, but not 0%.

Re: statistics/probability question

Posted: Fri Sep 18, 2015 10:41 am
by Portia
I've been entering this contest about once a year for the past five years, and by far the best-received story I ever submitted for it was about Mountain Meadows. I'm quite proud of it. If anyone wants to read it, I'd email it to you. Start up critiques again!

Re: statistics/probability question

Posted: Fri Sep 18, 2015 12:52 pm
by Shrinky Dink
How many points do you need to advance?

Re: statistics/probability question

Posted: Fri Sep 18, 2015 1:04 pm
by Portia
Shrinky Dink wrote:How many points do you need to advance?
There's not a set number. It's the top 5 combined scorers, regardless of the actual score. So realistically, even if the current top 5 ALL get 0, that means I'd HAVE to be the top scorer to be in the running, right?

Re: statistics/probability question

Posted: Fri Sep 18, 2015 2:09 pm
by Violet
Current top 5 all got 0? You'd have to be the top scorer, and the other people scoring in your top 5 would have to be 0 or other low scorers.

Re: statistics/probability question

Posted: Fri Sep 18, 2015 6:45 pm
by Portia
Violet wrote:Current top 5 all got 0? You'd have to be the top scorer, and the other people scoring in your top 5 would have to be 0 or other low scorers.
Yeah, can we generate semi-random scores and average them?

Re: statistics/probability question

Posted: Mon Sep 21, 2015 8:51 am
by Portia
Here's a probability question and answer for you:

(Q) What's the probability your Comcast Internet will be down when you need to submit?
(A) 100% SUCKAS

Re: statistics/probability question

Posted: Mon Sep 21, 2015 10:05 am
by Digit
Portia wrote:can we generate semi-random scores
At Random.org, they use atmospheric noise to generate their random numbers.

Re: statistics/probability question

Posted: Mon Sep 21, 2015 12:01 pm
by Portia
Digit wrote:
Portia wrote:can we generate semi-random scores
At Random.org, they use atmospheric noise to generate their random numbers.
Nice. I've used that site before but only for super-simple tasks, like determining whether a round of Dominion should include Colonies.