Unfortunately, yet again, millions of Americans failed to find the elusive answer to the annual March madness bracket problem. The problem is straightforward. Pick each of the winners of 63 NCAA college basketball games. I don’t have any personal knowledge of this (!) but I hear that gambling is involved. I’m shocked…shocked to hear that gambling is involved.

The sports section of the LA Times had a brief article calculating the odds of getting the entire tournament right—all games all brackets. The odds according to the Times are 9,223,372,036,854,775,808 to 1.

That’s 9 quintillion, or 9x 10 ^ 18… a pretty high number. I wonder what the odds are that 10 particular specified combinations would come up over a ten year period? That must really be something unusual. As the Times points out the odds of filling all brackets perfectly are greater than winning the lottery, being struck by lightning and bitten by a shark all on the same day.

Of course the amazing thing about this is that one such event out of all the possibilities is absolutely sure to happen. Every year, time and time again.

Now these odds are so long that, according to Dembski the result should actually be evidence of design. Maybe God picks the winners after all.

In truth, the NCAA bracket actually is a pretty good analogy for evolution. There is a structure; the brackets can be analogized to the environment, which changes just as the draw changes on particular wins. There is no goal in the sense that NC has to win. The only goal is that “some” team will win. That’s all that evolution does. Some teams have gotten deep enough into the brackets in a constantly changing environment to be those teams of flora and fauna that we see today. The evolution tournament continues, so we can’t be sure what tomorrow’s winners will be or what they will look like.

Er, Dembski's "universal probability bound" is one in 10^150, rather larger than your piddling 1 in 10^19 (approximately). Of course it's easy enough to demonstrate events with lower probabilities than Dembski's UPB, but picking the winners of all games in March Madness isn't one of 'em.

Posted by: Jon Fleming | April 05, 2005 at 01:25 PM

Well of course the number of molecules of gas in a typical room is on the order of 10^24 or 10^25, so the possible configurations of molucules run to the tune of 10^(1000000000000000000000), so the odds of picking the current configuration of atoms in this office is 1 in 10^(1000000000000000000000). The point is that some configuration is picked, we don't know what the sample size is necessary to generate life on Earth, so 10^150 is just a possible way to look at the probability of what we see (and it's a pretty flawed chain of reasoning anyhow see:

http://www.talkorigins.org/indexcc/CI/CI113.html

evolutionary processes are selected, which makes them less random than gas molecules or basketball games. :)

So, evolution can pick a winner, even if it is a touch difficult to predict the winner.

Posted by: Jason | May 04, 2005 at 10:56 AM