There has been some discussion here about the probability of LENR being real. To muse on this a little, somewhere JedRothwell said there have been (approximately) 4500 positive F&P type PdD electrolysis experiment replications.
Let us imagine that of these experimenters, 1/3 were very careful scientists, 1/3 were normally careful scientists, and 1/3 were poor scientists (a reasonable supposition I think) and we accept that 90% of the careful scientists results were correct, 50% of the normally careful scientists were correct, and just 10% of the poor scientists were correct we end up with 1350 + 750 + 150 correct experiments = 2,250, correct experiments. That is exactly 50% of the total number of experiments performed, but most of those were performed by very careful scientists.
On the whole this shows that LENR is real and proven on the balance of probabilities, because of the 2250 'rejected on imagined grounds of quality or rigour' experiments 1500 (2/3) were carried out by 'less careful or poor' scientists.