Imagine you are asked whether Mount Everest is higher or lower than 2,000 metres, and then asked how high it is. Now imagine instead being asked whether it is higher or lower than 45,000 metres. The first group will give a lower answer than the second, even though everyone knows the first number is far too low and the second far too high. This is the anchoring effect: a number you have just considered becomes the starting point of your estimate, and later thinking adjusts from it too little.
The classic experiment
In a 1974 paper in Science, Amos Tversky and Daniel Kahneman described a study in which a wheel of fortune, rigged to stop at either 10 or 65, was spun in front of participants. They were then asked first whether the percentage of African countries in the United Nations was higher or lower than the number on the wheel, and next to estimate the percentage. People who saw 10 gave a median estimate of 25 percent. People who saw 65 gave a median of 45 percent. The wheel was obviously random and had nothing to do with geography, yet it moved the answers by twenty points.
Since then the effect has been found for prices, ages, legal sentences, negotiation and even judges' decisions. A 2006 study of German judges by Englich, Mussweiler and Strack found that sentences were influenced by a sentencing demand that they knew had been decided by a roll of the dice.
Why it happens
Two explanations are commonly given and both probably contribute.
- Insufficient adjustment. You start at the anchor and move towards what feels right, but you stop as soon as the answer seems plausible, which is still near the anchor.
- Selective accessibility. Considering the anchor makes facts that agree with it come to mind more easily. If you ask yourself whether a car costs more than 40,000 euros, you start thinking of expensive features, and your final estimate rises.
The effect is stronger when you are unsure of the answer, when you are tired or in a hurry, and when the anchor is within a believable range. It is weaker for experts in the field, but not absent: experienced estate agents were also moved by a list price in a study by Northcraft and Neale in 1987, even though they denied it.
Anchors you meet every day
- The "original price" next to a sale price.
- The first offer in a negotiation, which tends to pull the final price towards it.
- The estimate of a colleague at the start of a meeting.
- The number of a previous question in a questionnaire.
- A suggested donation or tip amount.
How to resist it
- Think of your own number first. Before you look at the offered value, write down your own estimate. An anchor you have already out-guessed has less pull.
- Generate reasons against the anchor. Studies by Mussweiler found that explicitly thinking about why the anchor might be wrong reduces its effect.
- Use several routes. Estimate the quantity from two different directions, for example from the top down and from the bottom up. The anchors in the two routes differ and partly cancel.
- Be suspicious of round, big or precise numbers that are put in front of you before you have formed a view.
Anchoring in the games
Anchoring is easiest to notice in games where a number is shown next to the thing you have to judge. In Closer To, the target number is an anchor for your quick mental sums: people often accept a calculation that looks like the target and skip checking the others. In Geo Distance and Higher or Lower, the first figure you see in the question can pull your answer. If you find that your answers drift towards whatever number is on the screen, try to settle your own guess first. The statistics page shows your general tendency to guess high or low, which is not the same as anchoring but is a useful first diagnosis.
Anchoring is not a sign of stupidity. It is a side effect of a system that is fast and useful, and which uses whatever information is at hand. The remedy is a small habit: notice the first number and then ask what you would have said without it.