Of all the testing I have seen, even the old British proof house tests and the later Russian tests, they are statistically meaningless. Joe is correct in this. The sample is always too small to derive any valid conclusions. My job requires dealing with statistical samples and models, so I have some idea what I am talking about.
I also sent barrels to Zircon, both damascus and fluid. He finally asked me to stop sending them.
I would be happy to find out what the damascus barrels were actually composed of. I know we have some historical information, but I would like to see some confirmation.
As far as predictive models, the base data set is too small to be meaningful. For this type of model, you need data extended over time. Consider that in a single year, Belgium alone produced over 156,000 barrels. Extend that over multiple countries, over several decades. Remember you have to gather data on both damascus and fluid. Knowing the individual history of each gun, maker, barrel quality, etc would be mandatory. Certainly you need a way to determine that no blockage was involved in the failures. Other potential failure scenarios would have to be identified, verified and included in the data.
The complexity and required sample size become very large... Finally, who has the facilities, time and money to build and maintain the inevitable database that will be required?
Pete
