I wrote: >say 15 million to one (on the conservative end). And then there >is the factor of say, 2 for the falls that are never registered, leaving >true odds in 40 years at 1.5 million to one make that 7.5 million to one:)