> It would be interesting to do a serious analysis of the fossil discovery distribution to see if we can make a better estimate of the earliest emergence of hom

> It would be interesting to do a serious analysis of the fossil discovery distribution to see if we can make a better estimate of the earliest emergence of hominids based on empirical estimates of fossilization and discovery probabilities. Surely someone's done this? It's reminiscent of the tank problem.