On not seeing luck

It's often said that people misperceive skill and luck, for example by saying that a team is on form when it has merely had a run of good fortune. A new experiment at the Autonomous University of Barcelona shows that this error is even worse than we thought.

Jordi Brandts and colleagues got a group of students to predict a sequence of five coin tosses, and then selected the best and the worst predictor. They then asked other subjects to bet on whether the best and worst predictor could predict another five coin tosses. The subjects were told that they would bet on the worst predictor from the first round, unless they paid to switch to the best predictor.

82% of subjects paid to make the switch.

But of course, there is no such thing as an ability to predict the toss of a coin. Most subjects, then, saw skill where there was only luck. And, what's more, they were willing to spend good money to back this daft opinion.

These people weren't just idiots plucked from the street. They were fourth year finance undergraduates at one of the best universities in Spain.

Now, a common complaint about laboratory economic experiments is that they lack external validity; what's true of small stakes experiments with students might not apply to the real world. But I'm not sure this concern applies in this case.

For one thing, this result corroborates Powdthavee and Riyanto's work, which has found subjects in Thailand and Singapore willing to pay for the non-existent ability to "predict" the toss of a coin.

And for another thing, there is a vast industry which owes its existence to people believing there is skill where there is (for the most part) only luck. I refer of course to the fund management industry. There's good evidence that actively managed funds generally do badly; for example, in the last five years most UK all company unit trusts have under-performed a decent tracker fund. Despite this, investors spend billions on fees for such managers. This is consistent with the experimental evidence showing that people see skill where none actually exists.

Now, you might object here that if people lose money by buying stupid they'll eventually wise up. Such a view relies upon a silly Econ 101 misunderstanding of how markets work. We know from the work of people such as Andrei Shleifer (pdf), Bernard Dumas and Bjorn-Christopher Witte that financial markets do not necessarily select against the stupid and in favour of the smart. And we also know - from the long history of quack medicines (pdf) in the US - that the demand for bad products can be inelastic with respect to failure.

Comments

[Rob]

"They were fourth year finance undergraduates at one of the best universities in Spain."

This isn't saying anything good about the quality of the education that they received. Seriously, I know we all have cognitive biases, but if you've spent four years doing a finance degree at a top university and you still don't know that coin tosses are random, what hope is there for the future of Spanish finance? (One might say that this tells something about its present and recent past, I suppose).

[Anonymous]

Yeah, those students must have studied basic probability at some point during their course. Could it be the experiment was rigged in some way so as to achieve a desired outcome?

[Szczepan Stachura]

It wasn't rigged. See page 2 & 6 of this short article.

Jaw dropping.

[WNY-WJ]

This is the type of research that will ultimately pay off in economics - not more complicated calculations based on silly assumptions.

More behavioral studies will confirm what everyone already knows: people do not behave like they are "supposed" to in many economic models. Garbage in => garbage out.

[Luke]

Here's an interesting thing (to me). When they did it under strict conditions with none of the students able to talk to each other (the result reported above) 82% paid for the switch. I thought that might be a bit unfair - no wisdom of crowds, people might make cock ups under "exam conditions" that make you nervous. But they did it earlier on a more casual basis with more students, and 87% paid for the switch.

[Torquil Macneil]

It is precisely the fact that these are advanced students that makes the results suspicious. They are likely to know what results their professor wants (the kind worth publishing) and be highly motivated to deliver it. There doesn't seem to be any attempt at blinding or controlling.

[Ralph Musgrave]

I'd pay 1p to switch to the successful coin tossers on the grounds that they might be cheating, and I'd want to share in the illicit profits.

[Napoleon]

But is he lucky?

[richardarnatt]

I've always seen the managed funds market (and stakeholder pensions for that matter!) as a sort of feudal sharecropping. The people in the 60th percentile and above willfully handing their money over to those in the 99th percentile.

My question, then, is why haven't we seen more competition in this market? Fund supermarkets like Fidelity and Hargreaves Lansdown put barriers up for small-time investors such that funds are the "least bad" use of their money, and these habits are likely to persist many years down the line.

The little guy, as ever, is denied a piece of the action.

[solarrman]

There are many examples where intelligent people don't make the best use of data. By labeling these people as "the best predictors" we seem to bias ourselves in their favour even when we know that the odds are 50:50 on a coin toss.

There's a fascinating book "Thinking, Fast and Slow" by Daniel Kahneman, Nobel prize winning economist which explores this. Our emotional link with people can sometimes blind us to the statistics.

[charlieman]

@Torquil Macneil: "It is precisely the fact that these are advanced students that makes the results suspicious. They are likely to know what results their professor wants..."

More likely, the study sample was recruited from fliers on a corridor wall or the web equivalent. For every psychology student, there were two mates looking for beer money.

Is it valid to use psychology students for such experiments? I reckon that psychologists will have had a go.

[cf]

If you look at the study, the amount at stake was only 10 euros.
It'd be interesting to have different groups where the amount at stake varied and see if the behavior changes when it becomes a more significant amount.

When I was a student, the economic experiments were a great way to pick up cash because the people running the experiments found that this was an important variable and one could make more than us$100 in an evening if one lucked into a higher payout experiment.

[Alex]

If you make perfekt amounts of loops with your coin then its no luck. That's skill. ;)

[horn]

Do you really think an article in the FT is 'evidence?' The real evidence, as shown by Ibbotson, and others, is that managerial skill is persistent both for Hedge and Mutual funds. [Both good and bad]. Hedge funds easily beat the relevant benchmarks. The benchmark is not the S+P for 99% of hedgefunds.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=292866

Seth Klarman has no skill? Tepper? Mandel? Soros? Elliott? Peter Lynch was just lucky? John Neff beating the S+P by over 3% for 36 years - more luck? Please.

French doesn't even believe in CAPM anymore.

[Isomorphisms]

Would like to hear more from you on the external validity of economic experiments.

I recall in my economic experiments CLASS in uni we read papers that were supposed to account for this. They would measure the effects with some expensive method, then measure with a cheap method, and show the robustness between the two. Then we can maybe extrapolate from the cheap method (e.g. extra credit points) to the expensive method (e.g. cash) in general.

[Simple Machine]

On the other hand, individually irrational decisions like this can work out as highly intelligent group behavior.

For example, when people have a bad customer experience like having their flight delayed, they probably over-attribute it to the carrier being poor, rather than just having been unlucky. Clearly, they're using far too small a sample size to make an informed decision on the quality of the carrier generally.

But across the market, you'll see that this irrational behavior results in flyers displaying remarkably good group decision making about which carrier to use. Probably even better than if people were deciding according to large scale scientific analysis.

[1:22.5 Figures]

We're a group of volunteers and opening a new scheme in our community. Your site provided us with valuable information to work on. You've done a
formidable job and our entire community will be grateful to you.

[isabel marant sneakers]

Wow, what a quality it is! Since mostly YouTube video clips have no fastidious feature, however this is truly a good quality video.