What is a good theological critique of "coincidences"? How do you know that answered prayers aren't merely a coincidence?

[TruthSeeker60]

TruthSeeker60;7664603:

If the hypothesis was that praying for X to happen in observable reality makes X more likely to happen, that is

something that one would expect to see evidence in this world (empirical evidence) for. For example, if one was to claim that intercessory prayer increased the likelihood of survival of cancer patients, that’s something that can be tested empirically because it is a claim that would have an effect on reality if true. How else would you test it other than empirical evidence?

With no way to control for all these variables(bolded below) one cannot formulate a testable hypothesis about praying for X.

There are ways to reasonably control for the fact that some people who pray may be more righteous, pray more fervently, or with more humility, etc. One way is to set up a study by having an entire group of people (say, a church) pray for the entire group of patients who are to receive prayer. That way, even if some of the persons praying had ineffective prayer, the persons whose prayer was of better quality will go towards all (presumably, if the group is large enough, at least some people in that group will offer quality intercessory prayers to god).

 

[warpspeedpetey]

…presumably, if the group is large enough, at least some people in that group will offer quality intercessory prayers to god).

Why would one presume that? How would you even know that enough truly Holy people existed to show a statistically significant result if every one of their prayers were answered? Five of the seven billion people on the planet are not Christians and therefore can’t meet the Biblical requirements for efficacious prayer by definition. Of the 2 billion Christians, even the most committed of us are unsure as to the efficacy of our prayers.

Obviously, then one can’t presume that at least some of the subjects have efficacious prayer simply because of the size of the data set. In fact the data sets needed would be so large that one would need to use the whole of humanity. At that point how could one develop a statistically significant variation from empirical reality? We don’t need a test for that, we can simply look around.

It doesn’t seem that one has a way of knowing if any subject is making an efficacious prayer, and good science isn’t based on assumptions concerning non-empirical phenomenon.