12-11-2001, 06:08 PM
It was either here, or in another forum involving asymmetric microprocessors and experimental autopoietic state machines (or somewhere), that somebody asked if he could prove a sequence of numbers was random. Another guy said "never" but then went on to give him some things to try to prove it wasn't random. This reminds me of the stock market, and of cutting my own hair.
Thing is, the few times in this life I held up a mirror while standing with another mirror behind me, and gave myself a haircut, the angle along two axes of the scissors was completely random. I "knew" - in theory - how to reverse-reverse translate the image into nerve impulses, and use visual feedback to put the scissors at the correct angle. Meaning, my model was perfectly accurate, and yet it was still hopeless! Try it yourself!
Which brings me to, like, foreign currency rate fluctuations. Well of course, if your model is wrong, it's going to look REALLY random. But that doesn't mean anything. Being wrong over and over about anything makes it look random. But it can either mean that your model is wrong, or that you just can't get it done. Most often, you have just overestimated your own abilities.
leroy
Thing is, the few times in this life I held up a mirror while standing with another mirror behind me, and gave myself a haircut, the angle along two axes of the scissors was completely random. I "knew" - in theory - how to reverse-reverse translate the image into nerve impulses, and use visual feedback to put the scissors at the correct angle. Meaning, my model was perfectly accurate, and yet it was still hopeless! Try it yourself!
Which brings me to, like, foreign currency rate fluctuations. Well of course, if your model is wrong, it's going to look REALLY random. But that doesn't mean anything. Being wrong over and over about anything makes it look random. But it can either mean that your model is wrong, or that you just can't get it done. Most often, you have just overestimated your own abilities.
leroy