top of page

Why You Can't Just "Look At The Evidence"

Why gathering more facts often makes you more wrong — and how the search itself is rigged from the start.

Claim

More information does not make people more accurate. Often it makes them more confident in whatever they already believed, which is not the same thing at all.


We tend to assume that a careful, open mind is just a mind that looks at enough evidence. Read more, check more sources, ask more questions — surely the truth wins out in the end.


But reality repeatedly shows the opposite.


People who read widely and dig deeply often end up more certain of their original view, not less, because the digging itself is steered. They are not searching for the truth. They are searching for confirmation, and calling it research.


The mistake is treating "I looked into it" as proof of fairness.

It is not.



ChatGPT Image 8_edited.jpg

Decomposition

Confirmation bias is not one error. It is a small chain of them, each one feeding the next.


1. The question gets rigged before the search starts

  • a belief is already in place

  • the mind frames the question around that belief

  • "Is this true?" quietly becomes "Can I find support for this?"


Once the question is rigged, the search is rigged too. You don't notice, because it still feels like curiosity.



2. Matching evidence gets waved through

  • facts that fit slide in with no resistance

  • sources are not double-checked when they say what we hoped

  • agreement feels like recognition, not persuasion


It feels less like learning something new and more like meeting an old friend.



3. Conflicting evidence gets stopped at the door

  • the same fact, from a source we distrust, gets picked apart

  • flaws in method are suddenly worth mentioning

  • one bad example is enough to dismiss the whole point


"Why does this one feel so much weaker than the other one?"

It isn't weaker. It's just unwelcome.



4. Memory finishes the job

  • supporting details get rehearsed and remembered

  • contradicting details fade fast

  • over time, the past looks like it always agreed with us


By the end, it is not even a choice anymore. The record itself has been quietly edited.



Assumptions

The conventional view — that looking harder gets you closer to the truth — only works if several things are true. They are not.


Assumption 1: More evidence means more accuracy.

It does not. More evidence means more material to sort through, and the sorting is where the bias lives. A bigger pile just gives a biased mind more to pick from.


Assumption 2: Smart people are immune.

Often the reverse happens. A sharp, well-read mind is better at building a case for almost anything, including the wrong thing. Skill at reasoning is not the same as skill at being right — it can just make the wrong belief sound better defended.



Assumption 3: We'd notice if we were doing this.

We would not. The bias does not feel like bias from the inside. It feels like good judgment, because every step along the way had a reasonable-sounding excuse attached.


Assumption 4: Bias is about dishonesty.

It is not, mostly. People genuinely believe they are being fair. The unfairness happens earlier, in which questions get asked and which sources get a second look — long before anyone consciously decides to ignore something.



ChatGPT Image 8. 5. 2026 15_47_12.png
bottom of page