[sparklite2]

Would you be willing to forego a controllable test scenario for very high rates of predictability?

[mad_as_he$$]

My general position on meteorology is that if one has the same success rate as great baseball hitters they are considered successful. 0.300 is good enough for both fields. ;-)

[Alberta's Child]

For practical applications, sure. But that's not quite the same as "scientific" as I would define it.

This is why the engineers among us build margins of safety into everything we do.