Choice paralysis

March 2nd, 2007

Today I watched an excellent Google lecture on the paradox of choice – why more is less, by Barry Schwarz (who’s book I’ve just bought). I’m in the middle of designing an interface where this issue is particularly relevant, so finding this vid (via governomics) was perfectly timed. It’s fascinating.

In a nutshell, the more choice we’re given, the more likely we are to be paralysed and not make a decision. For my own benefit, here are my notes (much of which is quoted verbatim):

  • Too much choice either causes paralysis, or induces people to make poor choices by forcing them to use a simplifying strategy to make their decision. If, by chance, people manage to make a good decision, they feel worse about it.
  • In any device there is a trade-off between capability and usability – the more features you add, the harder it is to use. No one has figured how to maximise of both in a single device.
  • To purchasers, in prospect capability seems way more important than usability. In practice, the reverse is true
  • People have already learned this, but nonetheless, the next device choice is still driven by capability – perhaps because we aspire to know how to use it, even if experience tells us we will never get it to work – because of talent or temprement or time
  • Additionally, in a highly capable device (i.e. less usable), it takes longer to learn how to do the basic things.


Lots of choice makes people unhappy for four reasons:

  • Regret. With any choice that is not perfect, it’s easy to imagine that there is an alternative that would have been better. This reduces the satisfaction you get out of choices. Anticipated regret prevents you from making choices at all.
  • Opportunity cost (everything suffers from comparison). A choice is generally not the best in every single respect – you make trade offs to get something suitable within budget. The more options there are, the more likely you will be to identify attractive features in options that you don’t choose. Cumulatively they subtract from the satisfaction you get from the decision you made, and the thing you choose is never going to feel as good as if the choice set was smaller.
  • [Aside: it’s all of the things we want to do and like to do, that create the sense of time pressure, not the time pressure created by chores.]
  • Escalation of expectations. When there are lots of options, expectations are raised. When there are lots of options we expect perfection. When expectations are high, even good experiences leave us feeling like we have failed. There’s no way that adding options to people’s lives will do anything other than raise expectations, but the likelihood of actually choosing perfection is very small, so we are disappointed by even very good choices.
  • Self blame. You make a choice, and it’s a good choice – but it’s not as good as you expected it to be. What went wrong? Whe there’s little choice it’s obvious – it’s the World’s fault: “they don’t make what I want, I did the best I could”. With lots of options there’s no excuse: “it’s my fault if I choose something that is anything less than perfection”.

Everything was better back when everything was worse. When things were worse [when there was less choice] people’s expectations were lower and it was possible, occasionally, to have an experience that exceeded expectations. In modern society I don’t think it’s possible for something to happen that is better than we expect it to be, because we expect everything to be perfect.

Maximising vs. satisficing – different decision making styles.

Maximising leads to unhappiness, as you have to examine all the options to properly maximise, but cannot because it’s an impossible workload. The result is unhappiness because if you only looked a bit harder you would have found something better. Satisficing is choosing something adequate and not looking back.

Maximisers do better, but feel worse

The psychological costs of choice very quickly outweigh the benefits of wide choice as the number of options increase.

A way out?

  • Agent / principle seperation – where an agent is someone making the choice for you (e.g. estate agent) and the principle is you. This short-circuits many of the negatives because you don’t torture yourself over the decision.
  • Paternalistic libertarianism (not a contradiction in terms): Organise the options so that if the user does nothing (is paralysed), they almost certainly get what’s in their interest anyway. In a world in which people are more and more likely to take no action, organise the space so that when they do nothing, good things happen. Therefore, think long and hard about how you organise the options.
  • Or, as system designers, create decision trees. [not easy]

Comment on this post

You must be logged in to post a comment.