> A 95% chance of failure is bad if you're pinning the whole business on it, but if you have a variety of 5% chance deals, then it can make sense to pursue them

This is only true if the probability distributions for the values of the individual deals are rather uncorrelated (or even better: stochastically mostly independent).