OMGWAT!

Oh My God, What A Twit!

David Shormann, Ph.D. does not know probability theory.

This is appalling. He's written the "Dive into Math" materials for home schoolers, and I now have grave doubts about whether parents should ever use them.

Pretty strong, right? Well, here's what provoked it:

Here is a definition of natural selection from a biology textbook used in Texas schools:

Natural selection is the outcome of variations in shared traits that affect which individuals of a population survive and reproduce each generation. This microevolutionary process results in adaptation to the environment.

Consider for example a female sockeye salmon in Alaska's Copper River. Let's say she lays 3,000 eggs, and all of them hatch. Now, to keep the population stable, only two of those eggs need to mature to adults and return, which means 2,998 of them will probably not make the return journey and produce offspring. Some will get eaten by birds, others by bears, or maybe even a salmon shark. Some will get smashed against rocks, others may starve. Only two are likely to survive to journey from their birthplace to the sea, then venture thousands of miles, before returning to their birthplace.

Now, do you really think the two salmon that survived to adulthood did so because they were clearly the best suited for the environment? Perhaps, but in reality, there is only a 1 in 3000 chance the salmon with the best set of genes survived to adulthood.

Now, this sounds really good at first glance, but it's an incredibly naïve view of fitness. Fitness, and natural selection, are both probabilistic functions. We're looking at differential success over large populations. The fate of an individual, while it matters a great deal to that individual, doesn't matter that much when we're looking at large numbers.

The system is actually a lot like a casino. From the perspective of the individual, there's a lot of luck involved. He pulls the lever and maybe he hits the jackpot, or maybe not. (Most likely, not.) Some people walk away from casinos considerably richer than they went in. Did that person find the key to overcoming the house advantage? No, of course not. Individual luck creates huge variations in the outcomes for individual gamblers, and these variations vastly overwhelm the house advantage. Even in a game like red/black at the roulette wheel, the house advantage is just under three percent. Over lots of bets, the gambler wins back 48.6% of his money each time, on average. But an individual bet is not an average. It either wins (100%) or loses (0%). The three percent house advantage disappears against the noise.

The casino, on the other hand, doesn't care about individual wins and losses. Those are noise. The casino makes money, even though some people win a great deal, because in the long run, the odds are always with the house, however slightly. Take a small casino with a dozen roulette wheels. Call it 60 people playing, and 50 spins of the wheel per table per hour. If each bet is a dollar, that's an income stream of $81 per hour.

That's what we see with only a one-in-thirty-seven increase in "fitness" on the side of the house.

The argument Dr. Shormann offers to "disprove" natural selection is like saying casinos can only make money if every gambler loses on every bet. If this is the level of understanding he brings to math, I would recommend home-schooling parents look at someone else's math package.