The Love Prognosticator
This scientist John Gottman, who I mentioned in my previous entry, says he can predict with 90% accuracy what will happen in a particular relationship in 3 years after observing the couple’s physiology and behavior during conflict resolution. This impressed me at first, but now I think it’s just a matter of mathematics. Interestingly enough, the title of the essay is “The Mathematics of Love.” Well, let’s crunch some numbers.
Let’s say I set up a booth at the county fair. I will have a big sign that reads, “Donkey Boy: Love Prognosticator.” It will be a hit. People will come in droves!
Here are the rules. I’ll charge each couple $20 and will predict where they will be in 3 years. If I’m wrong, they can come back in 3 years and I’ll give them $25. If
I’m right, I keep the money. I’ll have to deal in volume if I’m going to make big bucks, so I can’t wait around to observe conflict resolution. The county fair should be a happy place, so who wants to create conflict anyway? Instead, I will ask 3 multiple choice questions of each person, which will get to the root of the issue.
(1) If you were a dog, you would want your partner to be (a) a snausage treat, (b) a 5 lb rawhide bone, or (c) a fire hydrant
(2) Would you be (a) more, (b) less, or (c) equally attracted to your partner if he/she had a mullet? (If your partner already has a mullet, suppose it was bigger.)
(3) Do you like soup? (a) Yes (b) No or (c) Refuse to Answer
Using the answers to these questions, I think I can get pretty close to 90% accuracy. No matter what people respond, I will tell them that they will break up in 3 years. Let’s assume that a third of the people that come in are married and the rest are just dating. Given that half of marriages end in divorce (and most sooner rather than later), I’ll get about half of 33%, or 16.5% correct. The dating 66% remains. I tried to find some numbers on this, but I couldn’t. So, I’ll just have to make some stuff up. Let’s assume that the average person dates 15 people before they get married. For all these relationships that don’t turn into marriage, then, I’ll be right 15 out of 16 times. But remember, half of the people that get married will get divorced, so it’s 15.5 out of 16. That gives me another 64% correct. Point is, after asking people about their favorite dog toy, mullets, and soup, I can be right about 80% of the time.
All that fancy “pyschologizing” by Professor Gottman is only good for another 10%. Granted, some failed dating relationships will last longer than 3 years and some failed marriages will last longer than 3 years, so my actual power of prediction will be slightly less than 8 out of 10. Still, you get the idea. Of course, it’s also based on the assumption that 1 in 3 people that walks into my booth will be married. This is debatable. Most questionable though, is my assumption that the average person dates 15 people before marriage. This number may be higher or lower, perhaps significantly so. My accuracy rate may actually be greater than 8 out of 10.
Granted my method is of questionable validity, but it took my 5 minutes to put together, so give me a break. It does, however, cast some doubt on the statistical significance of Gottman’s predictions. Maybe his 90% accuracy says more about how people have relationships than how good his method is.
Let’s say I set up a booth at the county fair. I will have a big sign that reads, “Donkey Boy: Love Prognosticator.” It will be a hit. People will come in droves! Here are the rules. I’ll charge each couple $20 and will predict where they will be in 3 years. If I’m wrong, they can come back in 3 years and I’ll give them $25. If
I’m right, I keep the money. I’ll have to deal in volume if I’m going to make big bucks, so I can’t wait around to observe conflict resolution. The county fair should be a happy place, so who wants to create conflict anyway? Instead, I will ask 3 multiple choice questions of each person, which will get to the root of the issue.
(1) If you were a dog, you would want your partner to be (a) a snausage treat, (b) a 5 lb rawhide bone, or (c) a fire hydrant
(2) Would you be (a) more, (b) less, or (c) equally attracted to your partner if he/she had a mullet? (If your partner already has a mullet, suppose it was bigger.)
(3) Do you like soup? (a) Yes (b) No or (c) Refuse to Answer
Using the answers to these questions, I think I can get pretty close to 90% accuracy. No matter what people respond, I will tell them that they will break up in 3 years. Let’s assume that a third of the people that come in are married and the rest are just dating. Given that half of marriages end in divorce (and most sooner rather than later), I’ll get about half of 33%, or 16.5% correct. The dating 66% remains. I tried to find some numbers on this, but I couldn’t. So, I’ll just have to make some stuff up. Let’s assume that the average person dates 15 people before they get married. For all these relationships that don’t turn into marriage, then, I’ll be right 15 out of 16 times. But remember, half of the people that get married will get divorced, so it’s 15.5 out of 16. That gives me another 64% correct. Point is, after asking people about their favorite dog toy, mullets, and soup, I can be right about 80% of the time.
All that fancy “pyschologizing” by Professor Gottman is only good for another 10%. Granted, some failed dating relationships will last longer than 3 years and some failed marriages will last longer than 3 years, so my actual power of prediction will be slightly less than 8 out of 10. Still, you get the idea. Of course, it’s also based on the assumption that 1 in 3 people that walks into my booth will be married. This is debatable. Most questionable though, is my assumption that the average person dates 15 people before marriage. This number may be higher or lower, perhaps significantly so. My accuracy rate may actually be greater than 8 out of 10.
Granted my method is of questionable validity, but it took my 5 minutes to put together, so give me a break. It does, however, cast some doubt on the statistical significance of Gottman’s predictions. Maybe his 90% accuracy says more about how people have relationships than how good his method is.
posted by Donkey Boy at Sunday, January 22, 2006
3 Comments:
the average person dates 15 people before getting married!??
i've gotta get moving.
Yeah, I don't know about that 15 people number. I mean, the average age of first marriage nationally is 26.7 for men and 25.1 for women. I think those people are falling way short. Then again, I've ended up in a bunch of fairly long relationships and rarely date someone for a short time.
I think you'd also have to offer people more than $25 in three years. Nobody would make that deal.
I agree, though, that you can generally assume any two people that are together will eventually break up (especially if you limit yourself to non-married couples, which should be the only people interested in this little game anyway).
the average man marries at 26?? good lord. i'd totally going to have to ho myself in '06 to meet the quota! i'm a relationship person too. we'll have to rely on people like wilt chamberlain to pick up the slack.
Post a Comment
<< Home