I didn't say I wanted all unknown speeds to be slow speeds, I said (twice) that they could be equally distributed (proportionally) to all known speeds. Seems like a reasonable method.
I think all the statistics, when combined; education, horsepower, speed, boat size, alcohol, etc. are consistent. If one were out of skew with the rest, it would be one thing, but they aren't. I honestly have to ask that if they pointed in a different direction, would you be championing them? Would that be a
you would follow?
I'm not sure about the speed unknown thing either. What I really don't understand is the horsepower unknown. Seems like that should be pretty straight forward to determine.
Time to go boating!