types of bias psychology: the short list of mental shortcuts that bend how you read the world. the household ones are confirmation, anchoring, availability, hindsight, and the algorithmic version that runs your phone. each one feels like clarity from the inside. each one is your brain rounding up so you don’t have to do the arithmetic.

1. types of bias psychology, the brief that nobody wants

here is the brief. the brief is that the brain, in any given hour, is running about a dozen shortcuts to keep you from having to think from scratch. these shortcuts are called biases when they go wrong, and they are called intuition when they go right. they are the same shortcuts. the labels are post-hoc. you do not get to know in advance which one you’re using. that is the whole shape of the problem.

the textbooks list somewhere around one hundred and eighty of these. some are real. some are arguably the same one wearing a different sweater. the categories of bias in psychology that show up in your actual tuesday, by my honest count, are five — and a half, if you let me count tom, which i will, because this is my list. the half is tom. tom drives a volvo. that’s relevant.

i am not going to define each one in the way the manual would. i am going to define each one by where it lives in my apartment, because that is the only definition i can keep in my head between coffees. if you wanted clean definitions, you would not have stayed past paragraph two. you stayed. that’s already information.

2. confirmation bias, the household one

confirmation bias is the bias of looking for evidence that proves you right and treating the rest as background noise. it is, of all the types of cognitive bias psychology has named, the one with the highest household penetration. everyone has it. nobody admits to having it. (i am admitting to it. i am also lying about how much of it i have.)

here’s how it shows up in my apartment. i decided, three weeks ago, that the third yoga mat under my couch was “still useful”. the evidence i collected for this was the fact that it had not, technically, decomposed. the evidence i ignored was that i have not unrolled it since 2023, and that whatever is happening on its surface may, at this point, be sentient. the brain treated the non-decomposition as a fan in the stands. the brain treated the non-use as a heckler. the heckler was ejected.

this is the one that builds the rest. it is also the one most people think they have under control. they don’t. i don’t. the long version is here; the short version is, you cannot read your own bias the way you can read a page. you can only read the rear-view mirror six months later, in the shower, when the trains have already left.

3. anchoring bias, the rent one

anchoring bias is when the first number you hear, however unrelated, becomes the gravitational center of every number you hear after it. you are not aware that you are anchored. you cannot become unanchored by being told you are anchored. that is the trick of the anchor.

i’ll give you mine. when i moved into this apartment, the landlord said the rent was a number that, at the time, sounded high. it was not high. it was the market. but because i had been paying less in the previous apartment, every rent number i hear now gets compared, automatically and stupidly, to the previous one. when somebody tells me their rent, the first thing my brain does is the subtraction against my old number. their rent is not a number. their rent is a delta. that’s anchoring. that’s the rent one.

the worst version of anchoring is when the anchor is yours and you don’t know it. mike has not filed since 2019. mike is anchored, by his own admission, on the year his accountant moved away, and every year since has been measured, in his head, against that one. mike is, on this front, lost. mike is also, somehow, not in jail. that’s a different bias. we’ll get to it.

4. availability bias, the microwave one

availability bias is when the brain treats “things i can easily remember” as the same thing as “things that actually happen”. they are not the same thing. memory is a small, vivid sample. the world is a large, dull sample. the brain prefers the small vivid one because the small vivid one fits in the cup.

here is mine. i have killed seven microwaves. when somebody mentions microwaves, my brain does not produce “a useful kitchen appliance most people use without incident”. it produces, in vivid color, the flash from the third one. it produces the smell. it produces dave laughing for nine straight minutes on the phone. (i timed it. of course i timed it.) on the strength of seven personal incidents, my brain has concluded that microwaves are, statistically, dangerous appliances. they are not. i am the danger. the appliance is innocent. the sample is rigged.

this is, by the way, why your phone news feed feels like the world is on fire. it is not the world that is on fire. it is the sample. the feed is a small vivid cup. you are reading it like it is the bowl. i’m fairly sure there is a study somewhere, possibly in a magazine i don’t pay for, that confirms this. i didn’t read it. i agreed with it on contact. that, by itself, is one of the other types of human bias on this list.

5. the algorithm bias, the modern one

the algorithm bias is the new one, technically a sibling of availability bias, but it deserves its own line because the algorithm deserves its own line. the algorithm is not a person. the algorithm is the new draft from the window. it shapes what you see in a way that you were not invited to consent to, and then your brain, helpfully, treats the shaped sample as a representative one.

here’s how it lands. i was served, last week, three videos in a row about a financial product i would never buy, two videos about a stretch for a hamstring i do not have, and one video about a guy who built a chair out of a single piece of plywood. i felt, briefly, that the world had gone insane on these specific topics. the world has not gone insane on those topics. the world has gone normal on those topics. my feed has gone insane. the feed is not the world. the feed is, technically, a mirror with the lighting cranked up.

i’m fairly sure books on tape are cheating, by the way, and i bring this up because the book i have not finished reading is one i bought because the algorithm told me, fourteen times, that i would love it. i did not love it. it sits on the windowsill next to brenda, who is a dead plant. they are, in their respective ways, both monuments to the algorithm and what it knows about me, which is everything except whether i actually wanted any of this. the social dilemma covered some of this, on a different couch, in a more dramatic key. i watched it on the algorithm that recommended it to me. the loop is the species. the species is us.

6. tom-bias, where everything goes better in his volvo

tom-bias is mine. it is not in any manual. you may steal it. tom-bias is the bias by which everything tom does looks, retroactively, like the better version of the thing i did. tom owns. i rent. tom files his taxes in february. i file mine sometime around the threats. tom drives a volvo with seats that adjust in fourteen distinct directions. i walk, mostly, because my bicycle has a flat that is older than some friendships.

here is how tom-bias works. when i compare my life to tom’s, i do not compare the daily texture. i compare the headline. tom’s headline is “stable, married, two kids, pension”. mine is “alive, mostly, and current on most utilities”. those headlines are not on the same scale. but the brain, helpfully, places them next to each other and runs a subtraction. the subtraction is unflattering. the subtraction is also a lie, because tom’s tuesday afternoon is, statistically, not an ad for his life. it is, like everyone’s tuesday afternoon, a slow event.

this is, technically, a flavor of contrast bias, with a side of survivorship and a slice of social comparison. but i’m calling it tom-bias because giving it a name is the only way i know to spot it in the wild. when i feel the subtraction running, i now mutter tom-bias under my breath, and the brain, briefly, lets go of the math. it doesn’t last. but it is something. (reading on a kindle is the same as reading, by the way. tom does that. so do i. one of the few categories where the subtraction comes out flat.)

7. verdict, the list keeps the bias intact

so here is where the list lands. the bias did not leave. the bias is still here. the list, if anything, gave me a more sophisticated way to confirm what i already thought, which is that i, personally, am one of the more thoughtful people who has ever sat at this desk. that is not true. it has never been true. but the brain, at this exact moment, is treating the writing of the list as evidence in support of the conclusion. the brain does not care that the list is about how the brain does that. the brain is on shift. the brain is doing its job.

the practical takeaway, if you are the kind of person who needs one, is this. the goal is not to remove these biases. the goal is to know which one is most likely to be running on a given tuesday and turn its volume down by, say, ten percent. ten percent is the whole win. anybody who tells you they are “unbiased” after reading a list of biases is the most biased person in the building. they have just acquired a vocabulary for self-justification. it is more dangerous than the bias was on its own.

my list keeps the bias intact. it does not solve a thing. it is, by my own honest accounting, an inventory of the rooms i live in. you have your own rooms. some of the floor plan overlaps. some of it does not. (every meeting could be a 3-line email, by the way. that’s not a bias. that’s a fact. but i bring it up because if you’d like to argue with me, you’ll be doing so on confirmation, anchoring, availability, the algorithm, or tom. pick one. i’ll see you in the comments, which i will read, in the spirit of the algorithm, while ignoring the ones that disagree.)

↑