Wednesday, August 29, 2012

Prayer - The Self-Reinforcing Delusion Engine

Self-Reinforcing Delusion engine
Delusion Engine v0.1!
This is my first drawing that barely
corresponds to my description below.
My next video - "Arguments Against Evolution - Pot Luck #1" is going up on the 1st, but I'm already working on my next video - "Prayer - The Self-Reinforcing Delusion Engine". I'm in the conceptual stage, where I'm trying to figure out what I'm attempting to convey.

As an aside, one of the items I covered in the current video is the reliability of radiometric dating. I was actually surprised by how much objection there is to this concept. I may have to focus on that sometime.

Anyway, I've been bouncing the concept of the "prayer engine" around in my head for awhile. This will be my first formal written version of it. I find composing small essays is good for organizing thoughts.

Self-Reinforcing Delusion Engine (SRDE)


The purpose of this video is not to necessarily argue that you are delusional. It's to try to make the point that your conviction in the power of prayer is possible even if prayer has no power. It's to explain how it is that prayer can appear to work, even if there's no god answering those prayers.

In short, it comes down to a combination of psychological factors.

The Individual Parts

The Prayer SRDE consists of four basic stages:
  1. The trigger of the next prayer
  2. Results Evaluation
  3. Memory Storage of Results
  4. Efficacy Audit
From #1 through #4, and back to #1 for another cycle, this engine iterates stronger and stronger with each pass. Here now are the descriptions of each stage.

Stage 1 - Trigger of the Next Prayer

This is the most simple stage of the cycle. It's when you are triggered to pray, and drop to your knees. Whether you're trying out prayer for the first time, or it's a regular habit, the engine cycle starts here.

If we're comparing this engine to a car engine, it would be roughly equivalent to the spark plug.

Stage 2 - Results Evaluation

Results Spectrum and Thresholds

Prayer Results Spectrum - before and biased
After you've prayed whatever it is you're praying for, the next step is seeing what happens.  The results are considered and placed into one of three categories:
  1. Success
  2. Neutral
  3. Failure
Each of these categories has a threshold for whether a particular prayer result falls into one, as opposed to another.

For instance, let's say that you've prayed asking for a raise at work. If you indeed get that raise, that would be clearly a success. What if you didn't get that raise, but instead got something else? Let's say that instead you won money from a scratch-ticket - $2000, and you used it to pay off some credit card debt.

It wasn't a raise, but it certainly freed up a fair amount of interest that you were losing per month, so it's kind of like a raise, right? Depending on how open your interpretation of events are, that might qualify as a success too, even though it's not exactly what you asked for.

The effect this liberal interpretation is that you're now accepting a much broader range of possibilities that qualify as successes. The ratios of thresholds between the three categories is now more lopsided towards the success category.

Likewise, if a result does not appear to match your belief that prayer works (more on this later), you have ways of explaining them away. Do any of these sound familiar?:
  • God answers prayers. Sometimes he says "no"
  • God works in mysterious ways
  • You can't understand God's plan
  • God had other plans for me
What you're doing here is essentially dismissing failures. You're pushing them from the "failure" category into the "neutral" category. You've rationalized why the failures aren't really failures. This shrinks the threshold for failure quite a bit.

With these two threshold alterations, successes appear to be bountiful, and failures are rare - and that's just from a biased interpretation of results.

Those rationalizations might make sense to you, particularly in your doctrine where we're not allowed to test God, however, I would point out that none of these rationalizations are testable or falsifiable. They're merely unevidenced assertions that, from an external observer's perspective, appear to be weak excuses to cover up problematic areas of the prayer model. Are you sure that you aren't just rationalizing? How could you tell?

Different Ways of Counting Success

There are six basic ways that a prayer can appear to work, even when it's not.

1. Coincidence

For some reason, many theists dismiss coincidence out of hand. The fact is, there's so many potential coincidences that can happen on a daily basis, that it'd be quite odd if none occurred. The vast majority potential coincidences are unknown to us until they happen.

Do you believe that Paul the Octopus is psychic? If you believe coincidences don't exist, then that would be a likely conclusion. This octopus correctly predicted the results of 8 games of Germany's matches in the 2010 FIFA World Cup.

"How else other than psychic powers do you think Paul predicted the outcomes?", you ask, "Do you think it happened by [laughs] chance?"

Yes. Why not? The one thing you must understand about humans is that we are staggeringly pathetic at probability. Without the discipline of math to objectively analyze events, we'd be left with our intuitions, which fail miserably.

That's because we're so often not in possession of all the facts, or all the possibilities, or all the nuances of every cause/effect that can influence the outcome.

The probability of someone correctly predicting the outcome of 8 games by flipping a coin is 1:256. That's not too bad of a chance of getting it right.

Now, multiply that against all the other people who were predicting the games in different ways.
Now, multiply that against all the different games going on in that year for that country.
Now, multiply that against all the different countries.
Now, multiply that against all the different years (say, the past 20 years).

The occurrence of someone guessing all the games correctly isn't just likely. We should actually be utterly swamped with them... and only through mere chance. Paul the Octopus just happened to be the one predictor that "won" - and we aren't aware of all the failures of others. An animal failing to predict game results is not news worthy, so we don't hear about it. But since you hadn't considered the full scope of events, your assessment of probability was inaccurate. This is how we fail at probability assessment.

Is the result of your prayer chance too? How could we tell? Well, we could test it. We could increase the sample size to help rule out statistical noise.

"But you can't test God", you reiterate. Right. We're not allowed to determine whether your supposed answered prayer is coincidence or not. That's convenient.

2. Praying for something that's likely to happen

The first way dovetails into the second - are you praying for something that's likely?

If your wedding is tomorrow, and you check the weather report, and it says 20% chance of rain, and you decide to pray that you don't get rained on - and you don't - is that really answered prayer?

How often do you pray for things that are unlikely? How often do those prayers come true? Have you noticed a direct correlation between praying for unlikely things, and the prayers going unanswered? If so, why do you think that is?

Many theists, with myself having had this mentality as well, want to be humble about prayer. God isn't a genie - he isn't a vending machine. I found myself praying for things that were fairly likely in the first place, as more of an augmentation... "streamlining" the possibility that my day would go well.

Unfortunately, this has a way of exaggerating the "success" threshold I was explaining earlier. Now even more prayer results count as successes.

3. Making it happen

Dovetailing still with the previous way, the third reason why a prayer can appear to work is that you made it happen.

If you pray for a hamburger, then you cook yourself one, does that count as answered prayer? This example may be a bit absurd, but that's for contrast's sake. Let's consider a much more common example.

You want a raise - badly. You pray for a raise, and you get the raise. Does that count as answered prayer? Have you considered that the prayer was not the cause? This is where the phrase, "correlation does not equal causation" applies - or, if you want the Latin phrasing - "post hoc ergo propter hoc" ("after this, therefore because of this") - as the name of the logical fallacy.

Now, it's not wrong or particularly irrational to deduce that flipping a light switch causes the light bulb to turn on, but you must allow for the possibility that you are incorrectly perceiving the chain of events.

When you were praying for a raise, were you also working hard for a raise? It might have been that the hard work is what paid off, instead of the prayer. You were doing both at the same time, and for some reason you decided to attribute the success to the prayer instead of the hard work. That's where the correlation manifests even if it's not causation.

"God helps those who help themselves", you interject. 

There's that rationalization again. By mere decree of an untestable unfalsifiable statement, you've chosen the lesser likely possibility (prayer) over the more likely possibility (hard work). Do you do this a lot? 

We can gauge how likely each possibility is through precedent probability - we can actually objectively and empirically evaluate that hard work results in greater success, whereas we can't test that prayer works at all without the "You can't test God" rationalization invocation.

The fact remains, you can be doing multiple things at once to achieve a goal - some reality based, some prayer based. How do you know which approach was the successful one?

We have term for this category - "Self-Fulfilling Prophecy."

4. The Placebo Effect

Still yet dovetailing, is a concept relating to "making it happen", though on a more subconscious level - the placebo effect.

If a doctor gives you a sugar pill, tells you that it's a headache-relieving medicine, you may believe him and relax, knowing that your pain is about to go away. That relaxation causes the headache to subside. In the end, you were essentially tricked into making the result come true, instead of the supposed medicine doing anything.

Strangely, this is also one that theists frequently dismiss out of hand. In reality, prayer can have the same effect. 

If you're praying for a good job interview tomorrow, if you strongly believe in prayer, that may give you more confidence, which makes your job interview go better.

If you're trying to quit smoking, and you pray for strength to get through the withdrawals, that may cause you to think you have more willpower to succeed, and it ends up succeeding.

Many of these things are possible without a god answering prayer at all. Depending on how strongly you believe in prayer, it can take you quite far.

How can you distinguish between your prayer result of self improvement being the placebo effect, as opposed to God answering that prayer?

I forgot, you aren't allowed to test it.

5. External Data

A big part of why you believe in prayer is because you've talked with other people who believe in prayer. You compare results and you come to a consensus that prayer works.

You retort, "I know all kinds of people who say that prayer works - we can't all be wrong!"

What if they're doing what you're doing? What if they're liberally interpreting results, and what if they're only telling you about the successes? How frequently are you, as a believer in prayer, having discussions with others about the failure rate of your prayer? Do you hand out statistical analysis reports on the success rates and margins of error, or do you only talk about the times it worked?

You're only hearing the success stories from your friends. Psychology has a term for this - "Availability Heuristic".  You're making judgement about the efficacy of prayer based on a quick tally of the information available - even if it's heavily biased information.

Have you heard extraordinary claims from others? They prayed for cancer to go into remission and it did, or some other very unlikely situation? It goes right back to Paul the Octopus - you aren't aware of the true failure rate, but only the successes, multiplied by potentially tens of millions of people praying. Eventually, someone is going to pray for cancer to go into remission, and it will, just by mere statistical probability. That will be your go-to example of prayer working, even if it's cherry picked from untold millions of people.

If they're doing what you're doing, then you can have an inflated confidence in prayer, even if there's no God answering prayers.

6. Misunderstanding of events

You may have misinterpreted something as being far worse than it was, prayed about it, and since it was mostly benign of a problem, went away on it's own, and then you consider that a success.

For instance, the woman on the right could have mistaken a random non-cancer lump for a tumor. She wasn't in full possession of the facts, and ended up misdiagnosing a situation - one that wasn't extraordinary at all.

Basically, your prayer supposedly solved something that didn't exist in the first place.

Stage 3 - Memory Storage

Human memory is incredibly fallible, to the point where eyewitness testimony in court cases is regarded as very low quality.

The Bouncer

Would it surprise you to know that you've forgotten most of your life? Do you remember what you were doing 5 years ago today, as though it were yesterday? What about for the month? We remember broad generalizations, generic imagery of places and people, but the vast majority of the memory of our second-by-second life is gone. Most of it wasn't particularly important anyway. The important stuff you remember, right?

This is a key component to the engine. 

Your brain is utterly bombarded with information. Every pebble - every stick - every leaf that you walk past when strolling through the woods is registered at the moment. Everything you look at, or smell, or hear, or feel - the brain processes and is aware at the instant.

5 minutes later, that data is gone - gone forever from your mind, with only the broadest generalizations remembered. Yet, that sign you passed on the trail, that had a smiley-face spray painted on it, may linger in your memory for years. Why is that?

Your brain has a bouncer. It's kind of like the "bouncer" at the entrance to a club - his/her job is to only allow important/valid people into the club, and everyone else is turned away.

Your brain has finite storage capacity, and has to be selective about what's remembered. This bouncer is the primary mechanism for that task - only store what's deemed important. It's not perfect, but it works fairly well. You really didn't need to remember every second of sitting at a desk working on a paper 5 years ago for 8 hours, did you?

Note what I just said - it allows data that is deemed important. There's no law of physics which dictates what information is important or not - it all comes down to you, and what you think is important. 

Earlier I was talking about your thresholds, between success and failure, shifting, and how the "failure" category of prayer results are minimized. Many of what would otherwise be considered failed prayers were essentially relabeled as "neutral".

Do you know what you've done by relabeling failed prayers as neutral prayers, by excusing them with "well God works in mysterious ways" or "God answered - he said no"?

You've deemed those results... unimportant.

Those unimportant prayer results try to get into your memory, and the bouncer takes one good look at them, and shoves them aside. In end, you won't remember most of them.

If you firmly believe that Toyota trucks are reliable, and then come across one that can't start, you're more likely the dismiss the anomaly as "Well it must have been abused or sugar in the gas tank, or something" - it'll be dismissed and the original belief will continue unabated and undeterred. Any Toyota trucks that match your preconceptions about their reliability will be remembered, and those that don't will be forgotten.

This overall effect also has a name in psychology - "Confirmation Bias"

I'll get back to this later, but I'll point out now that this is absolutely critical to the SRDE.

Imagination Inflation

Imagination Inflation is another psychological phenomenon where, over time, one can modify one's memory of past events to exaggerate or even manufacture memories.

A simple example can be the ever-increasing fish-size story. Have you ever met someone who tells a fishing story repeatedly, and each time the story is told, the fish has grown in size slightly? They're not lying exactly. That's how they remember it this time.

This topic can be quite extensive, where people can build memories of family gatherings even if they weren't there, but because they've heard other family members talk about the gathering repeatedly.

Like confirmation bias, this also can have a profound implications on what you think you remember about the efficacy of prayer. You may be remembering the prayer working better than it actually did.

Do some research into the fallibility of human memory. You'll be surprised.

Stage 4 - Efficacy Audit

At the end of the day, metaphorically speaking, you'll look back at your past experiences with prayer. You'll look at how well the current results have panned out (Stage 2), and look at how well it's worked in the past (Stage 3).

Since you only remember the successes, and you combine that with all the filtered success stories from other people, you conclude that prayer does seem to work fairly well. In fact, with the most recent success, you are even more confident in the power of prayer than you were yesterday.

What do you do now? Well, you pray again - thus starting the next iteration of the cycle.


I mentioned that there are four stages, but there is another component - one that heavily influences Stage 2 - the "Bias Feedback Loop".

It's a connection between Stage 4 and Stage 2. I've been touching on the idea that your belief in the power of prayer influences what you'll accept as successful results of prayer. This is where that feedback comes from.

10  The more you believe in the power of prayer...
20      ... the more allowance you'll make for a result to qualify as a success.
20      ... the more you're willing to dismiss a failure as a failure.
30      ... the more successes and the less failures will be stored in memory.
40  ... the more you will then believe in the power of prayer.
50  GOTO 10

That, ladies and gentlemen, is the Self-Reinforcing Delusion Engine.

Like any car engine, it'll reach a level of maximum performance, and it'll keep going until the cycle is either interrupted by honest critical thinking, or the fuel runs out (the person dies). Sadly, the latter is more often the fate of most people.

It's possible that your prayer is actually being answered by a god that has not been demonstrated to exist. Of course that's possible. It's also very possible that the perception that prayer works is a combination of psychological factors only.

"But I know prayer works", you bellow, "I've prayed for things that couldn't happen and they did! I prayed for a new car and I got one! Prayer has always worked for me! I mean, I don't always get what I want, and sometimes things don't go my way, but God works in mysterious ways! You can't test it!"

Congratulations, your delusion engine is at full steam.

Have you thought about the implications of what you're saying? Why can't we test God? Whenever we do, we find that the results are indistinguishable from statistical noise. You say that we need to believe on faith that he can't allow evidence of his existence - but prayer could be that we have psychic powers and a god isn't required - so that bypasses the evidence for God part. The connection between God and prayer is tenuous in the first place, anyway. 

If you're right, shouldn't that mean that we need to ban all prayer testing? Is God is fine with allowing people to suffer, not helping them, just so a tenuous connection between prayer and God can't be empirically justified - is he really that petty? Shouldn't we be preventing testing if that's the case? Wouldn't it be harming massive amounts of people to test prayer?

All these rationalizations fail to patch up the gaping holes in the prayer model. Have you thought this all the way through?

Oh, you aren't allowed to. You aren't allowed to doubt.

You're just supposed to believe, no matter what - no matter the evidence - no matter the logic or reason - no no matter how silly it appears to everyone else.

No comments:

Post a Comment