[Update 7-28-11: ] I’ve learned this site needs some upgrading on the backend, and also there was an issue with the game theory archive page–for the meantime: try the following link: game theory archives. I’m working to fix these things, and new posts to resume on August 1.
Today I want to try a fun experiment.
I am going to ask you a set of questions / problems. I have already answered these myself. Your goal is to write an answer that matches mine. For every answer we match, we get one point. The goal is to get as many points as possible.
Are you ready? Jot down your answers for these questions / problems.
1. The choices are ‘heads’ or ‘tails.’ Pick one of these.
2. Pick one of the numbers from the following list: 7, 13, 99, 100, 261, 555.
3. You are traveling around the world and you agree to meet with your partner in a capital city. Before you agree to meet all communication is lost. Which of the following cities would you choose to meet your partner: Rome, Berlin, Paris, New York, London?
4. You know the date but not the time for the previous question. You have to meet your partner at an exact minute. Which time will you choose?
5. Write a positive number.
6. Name an amount of money.
7. You are given 100 dollars. You have to split the money into two piles A and B. Separately your partner does the same thing. If you divide the money into piles A and B with the same amounts as your partner, each of you gets $100. What amounts will you put in the two piles?
What’s the point of the quiz?
On Monday, I discussed the problem of two friends trying to meet up in a bar without communicating. The problem is simple, but the mathematics and game theory are fascinating.
An important facet of the game is the time does not really matter. What matters is that both people pick the same time so they can meet.
Situations like this are known more generally as coordination games. The strategy is to try to match, and often there is a particular outcome that is culturally “obvious” or stands out in some way as a custom.
These are called focal / Schelling points.
Do these exist in practice? Someone went ahead and tried an experiment to ask those questions above to the TED community. It turned out there are many focal points, even across cultures.
Here is what that experiment found out.
Results from the experiment
There’s a 10 minute video detailing the results. Here are a few screencaptures taken from that video.
1. The choices are ‘heads’ or ‘tails.’ Pick one of these options.
Common answer: Heads
My answer: Heads
2. Pick one of the numbers from the following list: 7, 13, 99, 100, 261, 555.
Common answer: 7
My answer: 100
3. You are traveling around the world and you agree to meet with your partner in a capital city. Before you agree to meet all communication is lost. Which of the following cities would you choose to meet your partner: Rome, Berlin, Paris, New York, London.
Common answer: London / Paris
My answer: New York
4. You know the date but not the hour for the previous question. You have to meet your partner at an exact minute. Which time will you choose?
Common answer: Meet at 12:00 noon
My answer: Meet at 12:00 noon
5. Write a positive number.
Common answers: 2 or 7
My answer: 4
6. Name an amount of money.
Common answer: one million (1,000,000)
My answer: one million
7. You are given 100 dollars. You have to split the money into two piles A and B. Separately your partner does the same thing. If you divide the money into piles A and B with the same amounts as your partner, each of you gets $100. What amounts will you put in the two piles?
Common answer: Split into even 50/50 piles
My answer: split into 50/50 piles
How often did you match the popular answer?
I got 4 of the 7 questions to match the common answers–not great, but I suppose over 50 percent is not bad considering the realm of possible answers.
Did you do any better?
The experiment also had one more stipulation: people either played with someone from their own culture or a different one, and they were told of this difference beforehand.
Did the success rate change?
Surprisingly it did not! People were able to communicate across cultures, though they gave slightly different answers when playing people from another culture.
I omitted some of the details to make this post clear. I highly urge you watch this ten minute video for all the nitty gritty details (it’s a bit dry but it’s worth a listen)
Video: The TED Community Experiment on Intercultural Coordination
Another video to watch
Just after posting this article I found another great video about focal points from William Spaniel, who has a great series of videos “Game theory 101.” Definitely check this one out as well:
, so if you buy a print version you can get the Kindle version for free (you can read the ebook using the Kindle app, available on smartphones/tablets/computers).
This is interesting. Comparing your answers, my answers and the common answers I match you and the whole sample in 4 answers each, although there’s one question I match yours and not the common and one where it is vice versa. Funny thing, this.
Q C You Me
1. x x Head
2. – x 100
3. x – Paris
4. x x 12:00 noon
5. – – 1
6. – – 1
7. x x 50
Where does the 1.000.000 money come from? And the 2? It’s amazing how the question for a completely arbitrary number of unspecified currency units results in an answer implying a concept of “wealth” in a “healthy” currency 🙂
I got 4/7 with the average, 3/7 with you.
I tried a game theory approach and chose the first items in the series (number and dollar figure both 1, though I was less confident about these but I was committed to the theory). The only two that differed were the time and the division of money where both were essentially 50% (and got both right).
What is New York doing on a list of capital cities?