I think that the issue of negatives is still an area of confusion.
I think that you mean that you don't deny there are negative results.
Percy means that any analysis should take into account the negative results. They shouldn't be discounted just because you expect them to happen.
Now perhaps Percy might better phrase his point by referring to false positives - and there WILL be some. When you accept that then you have to find ways to tell whether there are any genuine positives in the results - instead of just assuming that some or all of the positive results are due to intervention. Only a statistical analysis - or ideally a double-blind test with a control group - can really do that reliably. You can't just cherry-pick successes and declare that prayer works - not if you want to live up to your chosen handle. You can't even cherry-pick particular occasions like your claimed 100% success.