A bat and a ball cost one dollar and ten cents. The bat costs one dollar more than the ball. How much is the ball?
Ten cents. That answer showed up before you finished the question, and it is wrong. A ten-cent ball makes the pair a dollar twenty. The ball costs five cents. In the results the book reports, more than half of the students at Harvard, MIT and Princeton gave the ten-cent answer.
Nobody in that group lacked the arithmetic. They lacked a reason to check. An answer arrived, it fit the wording, no alarm went off, and the search stopped. Kahneman's whole book turns on that small non-event. The feeling of being right is a feeling about how smoothly a thought went, and smoothness is a poor witness to truth. The question is when that feeling deserves a second look, and who should build the second look in, because you will not remember to do it yourself.
Kahneman's two characters
System 1 is everything that happens fast and on its own. You see a woman's face and know she is angry before you know you have looked. System 2 is the slow, effortful part that can hold a rule, follow a sequence and say no to a habit. Ask it for 17 times 24 and you can feel it clock in. He is explicit that these are not compartments in the skull, only a way to separate producing an impression from inspecting it.
Inspection is optional. System 2 could check every impression and mostly does not, because checking costs attention, the scarcest thing you have. Usually the shortcut is fine. The ten-cent answer is what it looks like when it is not.
Knowing this does not fix it. Look at the Müller-Lyer illusion, two lines with fins pointing in and out, and measure them. They are equal. You can learn the rule and give the right answer for the rest of your life, and one of the lines will still look longer every time. Judgment works the same way. Naming a bias lets you correct the answer. It does not make the intuition go away.
The gorilla you did not see
Attention is a budget, and Kahneman measured it in the eye. People heard a string of digits and had to say each one back plus one, or plus three, in order. Their pupils widened as the load built, peaked at the hardest moment, and shrank the instant they answered or gave up. Small talk barely moved them.
Then he added a second task. While people did the digit work, letters flashed on a screen and they were asked afterward whether a K had appeared. Near the start or the end of a trial they caught it. At peak effort they missed it far more often, with their eyes open and pointed at the screen. The invisible gorilla study made the same limit famous. People counting basketball passes fail to see a person in a gorilla suit walk through the game. Their task has spent the attention that noticing would have needed.
Load also weakens the checker. In the studies the book reports, people holding digits in memory were more likely to pick the tempting snack and to accept a superficial judgment about someone. A busy mind checks less.
Why easy feels true
People who had been shown the phrase "the body temperature of a chicken" a few times were later more willing to believe that a chicken's body temperature is 144 degrees. The repeated fragment made the whole claim feel familiar, and familiar felt true. Of two false statements about Hitler's birth year, people more often believed the one printed in bold.
Kahneman's name for the feeling is cognitive ease. Repetition, clear print, a good mood and a fluent sentence all make a thought glide, and a thought that glides gets waved through. Two words, "bananas" and "vomit," can produce a full little story of spoiled fruit, a wrinkled nose and a bodily flinch in under a second. The story always comes with a cause attached. Albert Michotte's viewers watched one square slide toward another and saw it launch the second, though they were only watching geometry. Association runs ahead of judgment and hands System 2 a finished picture with no audit trail.
The picture is always complete, whether or not the evidence is. Kahneman calls this "what you see is all there is." People who heard only one side of a legal dispute were more confident than people who heard both, and they knew the account was one-sided. Less evidence produced more certainty, because fewer facts fit together more neatly. Ask how many animals of each kind Moses took onto the ark and most people say two, without noticing that the ark was Noah's.
Confidence is a feeling produced by coherence and ease, not an estimate of the chance of being right.
Where the second look belongs
The answer to "when should you check" cannot be "always." Constant suspicion would burn the attention it is meant to protect, and fast judgment supplies most of what people are good at. A check earns its cost in recognizable places, when the stakes are high and hard to undo, when you are busy or tired, or when an answer to a hard question arrives instantly, the way ten cents did. It also earns its cost whenever the evidence is one-sided, tiny or suspiciously tidy, and whenever a first number or a polished presentation is doing the persuading.
Education can teach the trigger instead of pretending the trap will vanish. Let students write down their first answer to the bat and ball, then solve it with a rule, and grade the noticing as well as the result.
Work can change the order of operations. Kahneman found this out grading exams. Reading one student's answers in sequence let the first essay color the rest, and every student looked reassuringly consistent. He switched to grading everyone's first question before touching the second. The marks got messier, and he trusted them more, because the tidiness had been an artifact of the halo. Score one dimension at a time, collect opinions before the meeting, and let the overall impression come last.
Technology has to treat attention as a design constraint. An alert in a side channel during peak workload is invisible in practice, and a warning on every routine action becomes one more stream to ignore. Show the evidence separately from the polish, with the sample size and the alternatives.
The feeling does not say where it came from
A fast, confident answer can come from real skill, from sound reasoning, or from a fluent shortcut, and the confidence feels identical in all three cases. Real estate professionals in one study denied that a listing price had moved their valuation while it visibly did.
That is the whole case for building the second look into the situation instead of the person. Give the checker something intuition lacks, a base rate, a comparison case, a rule, and put it where the decision happens. The ball still costs five cents. The trick is arranging your day so that somebody, or something, asks you to check.
Three things to keep
Confidence measures how well a story fits together, not how much evidence is behind it.
Write down your first answer, then check it with a rule. The intuition will stay.
Score one thing at a time. Let the overall impression come last, if at all.