If 99 labs reproduce the results and 1 doesn't, my guess would be that the one lab screwed up. Realistically, once you get a handful of independent reproductio

If 99 labs reproduce the results and 1 doesn't, my guess would be that the one lab screwed up. Realistically, once you get a handful of independent reproductions, the odds that something is an error or a fraud drop to basically nil.