Confirmation bias
Confirmation bias is the tendency to seek and favour information that supports what one already believes.
People do not weigh evidence evenly. They reach for what fits what they already believe and explain away the rest. That is confirmation bias.
Confirmation bias is the tendency to seek, interpret, and remember information in ways that confirm one's existing beliefs, while discounting or ignoring evidence that contradicts them. It is among the most pervasive and consequential of the cognitive biases.
Three ways it works
Confirmation bias operates at several stages of thinking. In search, people look for evidence likely to support their view and avoid sources likely to challenge it. In interpretation, ambiguous evidence is read as supporting the prior belief. In memory, confirming instances are recalled more readily than disconfirming ones. The belief is thus protected at every stage, and the person experiences this as simply weighing the facts.
Why it is so dangerous
The bias is dangerous because it feels like rational inquiry from the inside. A person gathering only confirming evidence sincerely believes they are following the facts, when they are in fact curating them. It hardens opinions, deepens disagreements, since two people can grow more confident in opposite views from the same evidence, and lets bad decisions survive long after the warning signs appear. Many forecasting and intelligence failures are confirmation bias at scale.
Designing against it
Because the bias is invisible from within, the defences are external and procedural. Deliberately seeking disconfirming evidence, assigning someone to argue the opposing case, stating in advance what would change your mind, and inviting genuine outside challenge all work against it. Organisations that prize dissent and the devil's advocate are, in effect, institutionalising a defence against confirmation bias, which left alone produces comfortable consensus and costly blind spots.
Confirmation bias is the reason that more information does not always produce better beliefs, and can entrench worse ones. Guarding against it means treating the search for disconfirming evidence as a discipline, because the mind, left to itself, will reliably build the case for what it already wants to believe.