Is It Tested?
When I am developing something new, I generally will slow-pull 5 samples. Then I give a strength rating that is lower than the weakest result I found.
I use 5 samples because I have been told that many UIAA and CE standards require 5 samples. Likewise, I use 12mm round-stock carabiner for testing, because I have been told many standards specify 12mm pins or carabiners for pull testing. These standards are hard to find, behind paywalls. I try to learn as much as I can about them and test as similarly as feasible.
I also slow-pull test equipment in various use configurations, i.e. hitch cords tied around ropes in various hitches, anchor slings choked around objects, etc. I have tested many dozens of devices with my deep-bury splices and SnakeSplices this way. None of the deep-bury splices have ever been the failure point, even when loaded beyond 90% of the MBS of the cord.
I have not bothered to test the deep-bury splices on their own. Samson Rope rates its cordage with the splices that they specify, and I splice according to those specifications, then rate products according to their specifications. There has been independent verification of these splicing methods from many sources, not least of which is Ryan Jenks of Hownot2.
Developing the SnakeSplice involved testing about many eye-to-eye cords made out of one size, then testing five made from a different size cord, then a whole other round of development for the loops. Along with a bunch of cords tested with friction hitches.
Sometimes people order custom products, where I haven’t tested the material used that way. For these we generally discuss what data I do have that is relevant, and if the customer and I are both comfortable with how certain we are about strength, we go ahead.