[lutorm]

Interesting, but I thought it sort of built a straw-man argument in that the main premise, that all the delay on a Mars mission is just to lower risk to the astronauts, isn't really substantiated. The calculation that ends up showing that the cost of a Mars mission is a hugely inefficient way of reducing risk to human life assumes the entire cost is to lower human risk. So it's only an upper limit, and there is no way to judge whether it's a useful upper limit.

Besides, the fact that there is a difference between risk to human life and risk of mission success is only relevant if there is a significant probability of mission success. You can only play the game with multiple missions for redundancy if an individual mission has a probability of success reasonably close to 1, otherwise it doesn't buy you much.

Of course, this whole affair assumes that we actually have some hope of a priori estimating the risk of failure of complex systems. I doubt it's possible, and I think that's confirmed by the observed 2% shuttle failure rate compared to what the "acceptable risk" of the mission was supposed to be.

[lutorm]

Sure would be interesting to know what people disagree with so strongly that it warrants putting me in the negatives. This seems like a noncontroversial comment to me.