There is a big difference between controlled replicable double blind studies to evaluate effectiveness and anecdotal evidence from biased observers. Bob