Whose advice should you take?

My wife finds it irritating that I often take a long time to get round to following her recommendations, even though the ones I follow almost always turn out to be successful. This post is about what sort of advice is worth taking – my view of the evidence is, not much – and why a (very basic) bayesian attitude is worthwhile.

Everyone wants their advice to be taken, but almost no one actually wants or ever follows unsolicited advice. That paradox is worth thinking about.

The quality of people’s advice varies considerably depending on how specialist it is.

Even those people who are experts offering expert advice may not be that useful to you. A 2014 meta-study found, ‘deliberate practice explained 26% of the variance in performance for games, 21% for music, 18% for sports, 4% for education, and less than 1% for professions.’

All that practice will give them a view of what you ought to do. But it might not be very relevant. Any advice you are given that generalised from sports or the arts is probably meaningless, unless you really know how to receive that advice, as per Tyler Cowen’s view that the best management books are detailed books about other subjects which you already know a good deal about. That in itself is a theory that corroborates my thesis, which is that most advice is useless unless highly informational and contextualised.

A Yale study found that general life advice based on the fact that people are ‘older and wiser’ tends to be bunk. That’s called ‘passive expertise’. Avoid it.

General advice is crap, but what about expert advice, called ‘formal expertise’? While formally educated people are better at realising they do not have a generally good understanding of things outside their area of expertise, they often overestimate their knowledge of their own subject.

If you are a specialist in something, you tend to assume you are top-drawer, even if you’re only middle-ranking.

As the study says, ‘While expertise can sometimes lead to accurate self-knowledge, it can also create illusions of competence.’

Some of the workplace examples they give of this phenomenon, from previous studies, are sobering:

• Trained nurses overestimate their ability to perform basic life support skills.

• General practitioners do not correctly assess their own medical knowledge.

Life is long and complicated. People forget things. Every GCSE student is more of an expert in their subjects at 16 than 26. The decline doesn’t stop after that.

Think of books as the collective advice we have from previous generations. It has the advantage of not being forgotten. We filter the books through the collective process of recommendation. Books that fall out of circulation tend not to be as useful. Advice ought to follow a similar rule. The longer I delay taking recommendations, the more rigorous the test. If my wife can still remember and be bothered to recommend something a year later, it’s probably a good call. She’s less likely to persist with the stuff that isn’t so good.

This is why most people don’t take most recommendations for films, books, whatever. The marginal cost of giving advice is way lower than of taking it, even if, on the whole, most people should take more small risks and therefore follow more, possibly bad, advice.

Overall, advice seems to be intuitive rather than empirical. One study found that ‘even well-intended advisers often advise others to act differently than they choose for themselves.’

There’s even evidence to suggest that all advice may be completely worthless. In a large-scale study (which has a fairly questionable sampling process) people were instructed to make decisions, including serious ones like breaking up with a partner or quitting a job, based on a coin toss. Turns out, the coin toss worked well.

Levitt found that those who had made a major change, like filing for divorce or leaving their jobs, were more likely to report being happier two months later, and even happier six months later, Levitt says. ​This was true whether they had made the change based on the coin toss or of their own volition. For those who made trivial decisions, such as whether to begin a diet or grow a beard, the coin toss didn’t seem to much matter: They were just as happy whichever path they took.

Then there’s the 80,000 hours project. Their careers advice goes against pretty much all the prevailing wisdom about career choice, including the evergreen stuff about following your passion. They have compiled a detailed career guide based on a review of the major studies. Following your passion is not a good idea.

What we are seeing is a pattern where advice tends to be good when it is based on expertise and relatively fresh. Real advice is closer to knowledge or informed opinion than wisdom. People mostly go wrong with their advice because they try and become the reverend sage who understands life, the universe and everything, rather than simply tell you the things you want to know like the quickest route to the station or how long it takes to do a certain sort of job.

Polonius is the archetype of an advice giver. Everyone thinks his famous speech is a repository of good advice. Really, Laertes cannot wait to get away, Polonius come across like an old fool and Shakespeare is careful to show us that he doesn’t follow his own advice. Rather than grapple him to his soul with hoops of steel, he has someone spy on Laertes. And far from having any friends at court, Polonius is the consummate politician, loyal to no one. His advice is easy to agree with and difficult to follow. ‘Neither a borrower nor a lender be’, is actually extremely bad advice if you think about it.

His advice is so bad because it is so general. Look at the (good) advice Tyler Cowen gives on his blog, Marginal Revolution, often about travelling. It follows a few basic patterns.

1. It is factual. He gives you information about the place; rather than saying ‘see the art’ he tells you which paintings in which galleries.

2. It can be disagreed with and often purposefully contradicts itself.

3. It is rules-based. If you do want to do it it’s easily (haha) done.

4. It is a proposition about the world. Cowen’s first rule, ‘there is something wrong with everything,’ shows you that his advice is a genuine attempt to leave you better able to get informed, rather than just offer comfort (what the person seeking advice usually wants) or signal status (what the person giving advice usually wants).

It is striking that the self-help genre started life as a way of giving people the education they had missed in history, finance, personal economics, politics etc and has morphed into a simplification of (some parts of) the Western philosophy cannon and modern psychology. Advice in that sense has become extremely general and more difficult to follow.

So, my advice on advice is this. Only take advice from those people who are experts, who have recently done something similar and who were perhaps a little unwilling to give the advice. This might mean they simply weren’t thinking about it before you asked them. Spontaneous advice will often be superior to the advice someone has been giving for some time.

Examples are only to take advice about babies from people with babies. (My youngest is 18 months and I don’t give advice about newborns anymore). Get your careers advice from people in broad-ranging industries as well as narrow ones. Only listen to the advice from people one or two rungs up from you. Read great literature and take it more seriously than the people you talk to. Write about the subjects you need advice on.

Look at what people have done, not what they say. As I said in my post ‘What I have learned at 32‘: ‘Your real judgements about the world are your actions.’

We need to be more bayesian. It’s a theme of this blog that, ‘there is no certainty in some topics.’ There is however much certainty in most advice.

Rather than seeking out the general wisdom of our elders and superiors about the way things are, we should have a view of things that is probabilistic, about the way things are right now. We should evaluate it fairly constantly, and keep updating it.

Finance advice is simple and fairly unchanging. Same for child rearing. But you are not unchanging, neither are your circumstances. You need different advice at different times. Keep updating yourself.

Advice is a map and the world is a terrain, as they say at Farnham Street. You will only get so far with advice. The world is what it is and you have to face its shifting challenges constantly.

Life is an evolutionary struggle. We are constantly adapting to the niches we find ourselves in, which are often similar to other people’s but never entirely the same. Don’t look for a well-worn path, look for general advice from multiple sources (fresh, well-informed sources) and then go down the path as best you can.

Oh, and remember, it’s extremely likely that a coin flip is just as good as any advice you will get.

To take me seriously, by the way, you ought to be asking why I think advice exists if it’s such a useless genre. Here’s a extremely compelling answer from the philosopher Agnes Callard.

The myth of advice is the possibility that we can transform one another with the most glancing contact, and so it is not surprising that one finds so much advice exchanged on social media. When people are not fighting on Twitter, they are cheerfully and helpfully telling one another how to live. In that context, advice functions as a kind of small talk or social glue: it helps people feel they are getting along in a space not bound together by any kind of shared weather.

Callard nails the problem with most advice: survivor’s bias. People give advice based on where they got to, not how they got there. As Callard says, ‘The moral of every great person’s story seems to be that they were not trying to retell another’s.’

Generalising from yourself is pointless. Unless you’re just trying to chat.