Corrected for your risk tolerance, of course. Suppose my data is worth $10k, and and there is a 1/100 chance of data loss (making those numbers up). I'd gladl

Corrected for your risk tolerance, of course. Suppose my data is worth $10k, and and there is a 1/100 chance of data loss (making those numbers up). I'd gladly pay $500 for "data loss insurance" (i.e. backups) to avoid the chance of data loss, even though $500 > (1/100)*$10k. Just multiplying by the probability is what's known as "risk neutral" - it works and is a very good approximation when the dollar amounts amounts are small compared to your net worth. It tends not to be a good approximation when the dollar amounts are higher. Of course, a realistic analysis would also account for the fact the the "data loss insurance" itself is probabilistic, but this is a start.