winning and losing streaks

[Ganchrow]

Hidden patterns? Number 4.669 appear to be one.
http://www.imho.com/grae/chaos/chaos.html

NO doubt about it. The Feigenbaum Constants, just like pi or e, or any of a whole host of physical contstants are pretty interesting. But they were all first "discovered" in a manner very at odds with pseudoscience -- namely by careful, studied, and ultimately peer reviewed examination of repeatable real-world phenomena. That's not pseudoscience. That's science.

Logically, I have no argument with the gambler's fallacy. But intuitively, I know there is something missing there. Hard to explain. It's like there has to be an intersecting point where the graph of isolated events (flips) meets that of 'stringed' events. Both are happening. An event is isolated, but it is also part of a string. Probably makes no sense, but that's where I'm going. To insanity and beyond! lol.

Now this. by contrast, is pseudoscience. Tell you what -- come up with a falsifiable hypothesis and I'll write you a computer program to test it. But I suspect that it's coming up with the hypothesis that's going to be the hard part. That's the very hallmark of pseudoscience -- the inability to formulate testable and falsifiable criteria for either success or failure.

But I am serious. Formulate a hypothesis. Come up with something testable. Come up with with a clearly enunciated statement as to what results would need to occur for the test of the gambler's fallacy to be deemed either success or failure. And if we can agree on the test conditions I'll test it.

And who knows? Maybe I'll wind up changing my whole world view as a result.