Re: RickVV Re: [GTh] Re: The Scientific method
- At 04:33 PM 5/2/2010, kurt31416 wrote:
By random, I mean how close to algorithmically random the statistical sample is. Algortihmically random being random with the lucky rolls removed, what you approach if you roll the dice a lot.
This is not a dice roll problem. By your framing, this would be a die with only two faces, and that makes it a coin toss.
...Flipping a coin doesn't work logically, Bob, because the probability of a saying being a Mark saying in Thomas isn't 50/50.
Most statistics can handle this, using appropriate probabilities for each category (parallel/not parallel)
And, as Mark sayings are used up in your dice rolling, the odds change, depending on the luck of the dice in the first part. It's a more difficult puzzle than it first appears.
Here you make a good point, but wind up hoist on your own petard. This is the reason your dice analogy won't work, but it is a good point against my coin toss analogy. At the risk of provoking our list moderator's complaint against technicalities, this is covered in statistics by "sampling without replacement." But I shall eschew further discussion of the statistical details.
I hope that addressed all your points, if I missed one, let me know.
Richard Hubbard raised a series of excellent questions for you to ponder, regarding definition of terms.
- Hi Bob,
Well, if the equation says 99% of the time, if random, there wouldn't be a sequence that long, I'll bet on it.
Sayings, sub sayings, sentences, words, or individual characters.
And since the probability of a Mark (or anything in general) isn't 50/50, I just don't see your point about preferring a coin toss vs. dice. A fair dice/coin produces statistically random numbers, and if you roll/flip them a lot, it approaches maximum algorithmic randomness/maximum entropy/maximum complexity/largest compressed file size.
Richard Van Vliet