Aug 06, 2005 21:49
humph [a sentiment unrelated to the following] --
PLEASE ENDURE THE FIRST PARAGRAPH TO GET TO THE SECOND
so at the first national math convention i went to there was an actuary as keynote speaker. and one night he presented an interesting idea in probability. if you have a model of a situation that provides odds for a given outcome and u get an outcome by trial that seems near impossible based on ur model, then it's more likely that ur model is wrong. for example, the odds that u flip head on a coin is 1/2. now if u get heads 50 times in a row, u can say "wow that's unlikely". if hundred times in arow: "wow that's not right". and a thousand times in a row: "alright, this can't be". so it probably isn't. more likely than flipping heads a thousand times in a row is that ur model is wrong, maybe it's a double-headed coin. maybe the coin is weighted wrong so it comes up heads every time. the bottom line is, it's very very unlikely [near impossible] that u have a fair coin and flip heads a thousand times.
so here's something i came up with mid-discussion last night. for real i want people to comment with what they think of its validity. so if we consider there are two things in the world, good and bad. there are good and bad people based upon a majority of their decisions being good or bad. so as we make decisions we do so based on hundreds of influences and circumstances. now, the odds of someone who is good [or bad] making exactly the decisions based on excactly the right influences and circumstances end up being fairly slim...heck really slim, near impossible. more likely is that rather than choosing from hundreds of influences and circumstances, we're choosing based on one of two things, good or bad - read divinity or damnation, God or Satan.
OK so for real, comment on that.
word.
[also as i had a God discussion last night with an old heathon friend (term of endearment, not me being mean/bigotted)i realized how much i apply math and logic to all my thinking...i think it's cool]