I can't assign a probability to tire B, but I can (and did) examine the probability that I would get no flats in 2000 miles if the actual probability were some value P. I didn't crunch the numbers, but I suspect that if tire A had only gotten 3 flats, for instance, then the math would have told me the results were inconclusive.

In both cases I was using the same model tire front and rear. They were used on different, but very similar bikes.