cognitive bias heuristics — 1 investigation

cognitive bias heuristics, as a phrase, sounds like a band that opened for a band you’ve heard of. i looked into it. the actual content is a list of shortcuts the brain takes. i take all of them. some of them twice.

so here is what the morning looks like from where i sit. boss in another meeting, carla pulled into an annual planning meeting on the third floor, mike on a swing-by because the bar he likes is too quiet on a saturday and he wanted to see “if my chair is still broken.” it is. nothing on the schedule, approximately the rest of the morning, one investigation to file before lunch.

i’m telling you this because if you read past the snippet, you should know who’s in the room. mike is in the room, sitting on the credenza, eating a granola bar he found in his pocket. mike has not filed his returns since 2019 and treats this as a system, not a problem. the_algorithm is in the room because it always is. nobody else.

1. cognitive bias heuristics, brief

here is the short version, before mike asks me to summarise it for the bar. cognitive bias heuristics are not “biases” in the courtroom sense. they are the cheap version of thinking. the brain skips the slow lane and takes the off-ramp. you pick the option that feels right because feeling is faster than checking. that is the entire trick. the rest of the literature, which i’m fairly sure exists somewhere serious, is footnotes.

most of what gets taught about why people only see evidence that agrees with them belongs to this same family. confirmation is one heuristic. the family has more. i went through my morning and found five running quietly in the background, like a fridge i don’t open.

i’ll be honest about the test. the test was: i sat down, i wrote what my brain wanted to do, and then i wrote what a slower person would have done. the gap between the two is the heuristic. that is the whole methodology. mike approved it by chewing, which i’m taking as peer review.

2. heuristic one, the bar test (cognitive bias heuristics in the wild)

the bar test is the first one and the most honest. i decide if a thing is true based on whether mike, at the corner bar, would nod or look at his pint. if he nods, true. if he looks at his pint, suspicious. if he laughs, definitely true but not in public. this is, i’m told, called “social proof” by people who write longer articles than this one. on the show cheers, the entire premise was that the bar made the truth. they were ahead of the science.

cognitive bias heuristics start here because the brain does not want to verify. the brain wants a friend with a pint to confirm. the wip 2022 list is full of items i added because somebody at a bar nodded. seven of them are gym-related. i don’t have a gym membership anymore, which the list also doesn’t know.

3. heuristic two, the wip 2022 test

the second heuristic is “if it survived a list, it must be important.” this is the wip 2022 list at work. i opened it in 2022, called it work-in-progress, and have not closed it. items on it include: finish reading the kindle book, respond to maggie, find good knife, research pension. the brain looks at this list and concludes: these are real priorities. they are not. they are old. age is not priority. the heuristic is age-equals-importance, and the wip 2022 list is the museum.

mike, looking over my shoulder, says “that’s not a list, that’s a graveyard.” mike has not filed since 2019, so mike knows graveyards. i logged it as a quote and kept going. i’m fairly sure there is a study somewhere, possibly in a serious magazine, that proves old to-do items become more important the longer they are ignored. this is wrong but it feels right, which is how every heuristic on this page works.

4. heuristic three, the_algorithm test

the third one is the_algorithm test and it is the worst. the heuristic is: “if the_algorithm shows it to me, it must be relevant.” the_algorithm has shown me, in the last week, fourteen videos about the correct way to load a dishwasher, three about a man who restores receipt printers, and one about a guy who taught a parrot to say “rent.” none of these are relevant. all of them felt relevant for the duration of the watch. that is the heuristic doing its job.

this is also where i ate ninety minutes on a tuesday-equivalent that wasn’t tuesday. i don’t write on tuesdays anymore on principle. the_algorithm doesn’t know my principles, which is the point.

THE FEED. IS NOT. A FRIEND.

5. heuristic four, the seventh microwave test

heuristic four is named after the seventh microwave, which is the one currently in my kitchen. the heuristic is: “the next one will be different.” i have killed six microwaves. i bought the seventh thinking it would behave. it has, so far, lasted longer than the third one but shorter than the fifth, which i’d consider the gold standard if it hadn’t smelled like pennies for the final week.

the brain runs this heuristic on careers, on relationships, on appliances, on flat whites. the next one will be different. it will not. the brain runs the test anyway because the brain has to. without this heuristic, nobody would buy a seventh microwave, and the microwave economy would collapse, which i’m fairly sure is the third footnote in a paper i did not read.

6. heuristic five, mike’s two-beer rule

the fifth one is mike’s. mike calls it the two-beer rule. it goes like this: any decision made before the second beer is suspect, any decision made after the third beer is also suspect, the window for valid decisions is narrow and located between beer two and beer three. mike has applied this rule for fifteen years. mike has not filed his returns since 2019. i am not saying these facts are unrelated.

this is, technically, an availability heuristic dressed up as a drinking rule. the brain says “i will use the easiest available data,” which in mike’s case is “how many beers have i had.” it is not science. it is a system. when the easy answer feels like the right answer, you are inside this heuristic, and you are also probably inside a bar. mike has been inside both for a long time.

7. verdict, the heuristics replaced thinking

here is what i learned at the desk between 9:18am and 2:47pm. i did not catch myself thinking once. i caught myself running heuristics five times before lunch. cognitive bias heuristics did not add to my thinking. they replaced it. the slow version of me is unemployed. the fast version is making decisions about microwaves, lists, parrots, and beers, and the fast version is wrong but on time.

this is also why trying to upgrade the brain through articles you bookmark doesn’t work. you are using a heuristic to fix heuristics, which is like asking the seventh microwave to investigate the previous six.

and on the matter of planning ahead — HT29 applies here, gently. a pension is a faith-based retirement system. i used the bar test on the pension. mike looked at his pint. that was the entire research.

related, and sitting in another corner of the same investigation: there is a separate write-up on the long career of the word for someone who keeps doing this on purpose, which is what i’d be called if mike were grading me. the fool, in the older sense, was the one who knew the trick and did the trick anyway. that’s me, with five heuristics, on a saturday-not-tuesday, with a granola bar wrapper on the credenza.

↑