Wed, 23 November 2011
The Show Notes
Mentioned in the Show
Geo's Music: stock up!
Sign up for the mailing list: Write to Geo!
A reminder that the new portal to the Geologic Universe is at GeorgeHrab.com.
Get your George HrApp here. Thanks to Gerry Orkin for the design and engineering.
Ms. Info sez, "OMIGOSH YOU GUYS- HOTEL ACCOMMODATIONS! However as of November 23rd, we're waiting for the room block to go live. Should be November 28th, 29th at the latest. Maybe Friday, the 25th. We'll keep you posted."
About the randomness issue: I took a statistics course in college a few years ago, so here go (goes?) my two cents. Random numbers/events which occur with the same probability are called uniformly random (which most people consider "truly random"). Non-uniform randomness would still be random, but skewed. For instance, if 100 decimal fractions are generated (by a computer) between 0 and 1 (such as 0.38256 etc) and they are all multiplied with each other so that we get 10,000 multiplication products, those 10,000 values will no longer be uniformly random, but still random. The average of the 100 numbers would be at 0.5, but the average of the 10,000 values is at 0.25, so roughly 5000 of the 10,000 numbers would be less than 1/4, which no longer counts as "truly random" to a lay person, but is still technically random. The Wikipedia page about non-uniform random number generators is titled "Pseudo-random number sampling", although that page probably won't tell you very much. Two more things: 1. Look up "German tank problem" on Wikipedia, it's fascinating in a geeky way and shows that dry, boring math like this can be very useful. 2. If we truly randomly pick a number from 1 to 10,000 (perhaps during a raffle) and it turns out to be 3 or 998 or 500, that number is still truly random, just as likely as 361 or 814, but we dismiss it as being artificial or not really random, and probably pick a new one thereby ensuring that our choice is not really random because the numbers 1 or 10,000 have a lower chance of being accepted. Humans aren't very good random number generators.