Having Trouble Making a Big Decision? We Can Help

We all face big choices from time to time.  Which college to choose?  Should I break up with my girlfriend?  Should I quit my job?  Should I dye my hair blond?

Sometimes the decisions are easy and obvious.  Other times, no matter how much you think about it, no clear answer emerges.  Your life might be very different depending on what path you take, but you just can’t tell which choice will leave you better off.

If you find yourself in that kind of bind, we are launching a new website today, www.FreakonomicsExperiments.com, that just might be your savior.

At that website, you tell us what your problem is, or pick from a large menu of problems we’ve pre-loaded.  We then take you through a questionnaire that might help you think differently about your problem and resolve your dilemma.  If you are still stumped after the questionnaire, we will solve the problem for you in the most sensible way we can think of.

We’ll flip a coin.   Heads you take one action; tails you take the other.

If you really can’t decide between two actions, a coin toss is as good as any other solution.  Not only will it give you the answer.  It will help our research.  By flipping the coin and following the coin’s advice, you will become a critical member of the Freakonomics research team.  We’ll send you a follow-up survey to learn how things turned out, and you will (with complete anonymity, of course), become a data point in our experiment to understand the outcomes of important decisions.  Believe it or not, we are completely serious about this.

And if advancing science isn’t enough reward to entice you to participate, we will even throw in a little Freakonomics swag at the completion of the study.

But, your participation only helps us if you are really so tied up in knots that you are willing to do what the coin tells you.  If not, contribute to science by sending all your indecisive friends in our direction!

Seminymous Coward

I would suggest your follow-up form include a question asking which resolution the user ultimately went with instead of assuming they followed your advice or coin flip. Then everyone's participation could help, and you wouldn't get (that particular kind of) potentially tainted data.

Seminymous Coward

Further on that note, I'm quite curious whether you think I should e.g. ask for a raise, but I am unwilling to commit to doing so, especially in a sufficiently timely manner, just because you think I should.

Eric M. Jones

In the 1960's Hippies used the "I Ching". Worked for me...maybe. And I still believe it was better than flipping a coin.

Curious factoid: The ancient and venerable 3000 year-old "I Ching" most used in China today is the Wilhelm/Baynes Bollingen Princeton edition. Hey, nobody can read 3000-year-old Chinese.

It's online too.

Matt Bonges

So this post got me thinking...
I'm a high school physics teacher and recently I found an interesting strategy that my students were using for the multiple part of my tests. The students called it "rolling thunder" in which if they got stuck on a question, they would roll there pencil across the desk. Each side of the pencil would have A, B, C, D on it. Since the pencils usually have 6 sides, 2 of the sides would have a roll again symbol.
The reason I found this interesting was I was trying to figure out if by using a tool to choose for them they would pick more randomly and not be influenced by previous answer choices. Do you think this is a better way to guess randomly?

Seminymous Coward

While that will generate random numbers with superior statistical properties, using it is nevertheless a terrible test-taking strategy, even assuming they take the obvious step of disregarding random answers that they can rule out. There are many superior heuristics. Unless you, the teacher, specifically randomize your answer order, the frequency distribution of the correct answer is extremely unlikely to be uniform; it's likely got B, C, A, and D ranked in that order of popularity (for mutually exclusive answers), if not then it's almost certainly due to your happening to have some other unconscious bias which could also be detected after the first test is returned with correct answers marked.

Picking from lists of numeric answers is made easier by noting patterns based on the incorrect answers being the result of common, simple errors like being off by a sign or integer multiple, e.g. from {0, 3/2, -3/2, 3} the most likely answer is 3/2. If similar patterns aren't present, that can be exploited by a student unsure of the sign convention or the like, e.g. selecting the same number with the correct units when the correct number but wrong units is not present amongst the choices or answering x^3/3 + C because all the answers have a "+ C" suffix. Having those variations in addition to "entirely" wrong answers and corresponding variations for them rapidly gets to be impractically many choices. Having only the variant-style answers harshly penalizes minor errors like incorrect signs and can effectively favor intelligent guessing over sloppy or rushed execution of the actual methods.

There are some other quick and dirty probability maximizing rules. Lists that contain "All of the Above" and/or "None of the Above" answers are far more likely to have All or None be the answer, in that order of preference. Answers that create double negatives in combination with the question are unlikely to be right.

As one final note, I would investigate whether your students are using these markings to cheat by communicating answers amongst themselves. In any case, I'd ban such pencils for offering the appearance of impropriety.


Matt Bonges

Thanks for the response! Many of the points you mentioned are why I don't like to use multiple choice as an assessment tool. However since their final examination has a multiple choice section (This is an IB course), I feel its necessary to give them some practice seeing types of problems that would be on the exam.
I'll try and convince these students of better ways of guessing but I'm not sure how successful I'll be. They seem to have a gamblers addiction to guessing and getting the answer right.

I'll be honest I hadn't really thought of the cheating aspect and I'll look into it.


I tend to leave life-altering decisions to my magic 8 ball ;)
On a serious note though: This sounds like an interesting project, and I don't think anything similar has been done. Crowdsourcing research has fantastic potential, but have you thought about how you will ensure data quality? Just curious from a fellow-researcher point of view. What do you expect the follow-up ratio to be? Or is that part of the research?

Bob Dahl

Long time reader / fan. Flip a coin and decide based on your reaction to the flip. Example, I can't decide between A and B, coin flip says A and I feel disappointed. That means I really wanted to do B but it was "unconscious" until the results of the coin flip.

Human Psychology is cool, we are complex and often not rational beings :)


I remember hearing the advice to flip a coin really high. While it us in the air, you will know which side of the coin you want to come up and you have made your choice. If you don't have that feeling, then you truly don't care and the result of the coin flip provides your solution.


I'm not sure if you'll get clean data. People like me who are not quite ready to commit to buying a house or breaking up with someone on the basis of a coin flip will not be in your sample. What kind of people are in your sample? Those who are so ... ok I'll stop. Interesting though.

Bob of Alex

Levitt's "Coin Man" as a tribute to Rhinehart's "Dice Man"?


That is what I thought too.


i cant decide if i should shoot batman or not


I just took the coin toss for volunteering, something that I've been thinking about when my youngest child starts school in August.

I didn't realize that the questions were geared for someone who is ready to act more quickly than I am. That may affect your follow-up data.

Brandon Sorenson

This does not sound like Random Sampling...


You know, I had an idea for an app that does something like this. It crowdsources decisions for you. The inspiration for it is that I'm kind of an impulse buyer, so I would submit my decisions to buy something on the internet and hopefully more often than not people would point out to me why I didn't need to buy something, the idea being that people that don't have an emotional attachment to the decision will be able to make a more objective analysis of it (and in the cases where I just can't make a decision, they can).

Zuriel Redwood

You will see this in practice ion Chinese temples. People ask their ancestors or their Gods specific questions similar to the ones in your project. Then they throw 2 Jiaobei divination blocks on the floor to get the answer. Here is a brief description:


I asked the freakonomics experiment whether I should hang myself. It said "yes."

I think maybe this should not have been approved by human subjects?


I can't decide whether to participate in this study or not... ;-)


I have to decide...

Computer Careers and Salaries