Monday, January 09, 2006

Mountain High and Valley So Low

Or how I mastered statistical reasoning and disproved global warming.

Some have claimed that mountain heights can be modelled by a statistical distribution, somewhat similar to Brownian motion.

Consider for example a random mountain. Other mountains in the neighborhood should, by this hypothesis, randomly vary in steps up or down relative to it. In particular, there should be a very small probability that the closest eleven mountains are all shorter than it! Details of the statistical reasoning involved are developed in more detail by LM and commenters here and here. However, actual analysis of the mountains near our randomly chosen mountain (Mount Everest) show that all of the eleven nearest mountains are in fact shorter!

This proves that global warming, plate tectonics, and pure quantum gravity are all bullshit.