エイミー・ウェブ: 私がオンラインデートを攻略した方法

ＴＥＤのホームページへは

をクリックしてください。

直接ここ

で観ることもできます。

Amy Webb: How I hacked online dating

小さいことは気にすんなッ、主題と主張をつかもう！

TEDではアルゴリズムがよく話題になりますが、はじめはよく意味が理解できませんでした。やっと最近、なんとなく判ってきたような気がします。→　NHK Super Presentation

【話題】愛のアルゴリズム

【時間】 17分27秒

【要約】

１．なぜ失恋の連続なのか？

　「選り好みするのを止めなさい」とおばあちゃんに言われた。理想の相手と結婚して、自分が育った環境と同じ家族を持とうと思ったら、5年前に結婚していなければならなかった。

２．理想の男性と出会う確率

　150万人都市フィラデルフィアで、私の理想の結婚相手の男性は35人しかいない。

３．オンライン・デート

　アルゴリズムに基づいたマッチングがされる。昔はコンピュータではなく仲人さんがやっていた。応募するためにプロフィールの欄を埋めていくのが面倒だったので、履歴書をそのまま写し書きした。

４．最悪のデート

　履歴書をただ写しただけのプロフィールにもかかわらず、たくさんのデートの申し出があった。そのうちの一人とデートして、高級レストランで食事をしたら食い逃げされた。それからはデートの度に、相手の男がどれだけ最悪かデータをとりはじめた。

５．問題は私自身にあり

　理想の相手がみつからないのは、アルゴリズムが悪いのではない。データを入力する自分自身に問題があるのだ。プロフィールを入力するときに、心の底から正直になれるひとは少ない。結婚相手を選ぶのに、表面的な質問ばかりを並べているサイトにも問題がある。

６．自分で作る新しいアルゴリズム

　サイトの表面的な質問ではなくて、自分自身が結婚相手に対して求めることがらを72ポイント書き出した。これに優先順位をつけて、採点方法を決めた。合計点が700点以上あればメールを返し、900点以上あればデートする。

７．競争相手を変数に取り込む

　やっと理想の相手をみつけたと思ったら、彼の方からは良い返事がもらえなかった。競争相手に負けたのだ。そこで、市場の競争相手について研究することにした。私が結婚したいと思う理想の男性の架空のプロフィールを捏造して、アプローチしてくる女性のデータをかき集めた。このデータを分析して、最強のプロフィールを作り上げたい。

８．大切なこと

　①近づきやすさ、親しみやすさ

　②タイミング

　③写真

９．ついに理想の男性に！

　現在の夫に出会うことができた

10．愛のアルゴリズム

　愛のアルゴリズムは、自分自身でつくるもの。好きなだけ選り好みすればいい。結婚相手に限らず、やりたいことや仕事を探すときも同じだ。

【語彙】

burn down ：燃え落ちる

spectacular ：壮大な

picky ：選り好みする

monogamous ：一夫一妻

cohabitate ：同棲する

variable ：変数

serendipity ：

avid ：熱心な

in the meantime ：ところで、そうしている間に

tremendous ：途方もなく大きな

peer ：仲間、身内

bump into ：偶然ぶち当たる

Judaism ：ユダヤ教

explicit ：明白な

per se ：それ自体

Cosmo quizzes ：

monetization ：

white-table-cloth ：

regale ：楽しませる

shove ：押し込む

empirically ：経験的に

correlations ：相関関係

reference ：言及する

kinky ：変態の、倒錯した

confront ：直面する

superficiality ：表面的なこと

reverse-engineering ：

amass ：蓄える

shul ：ユダヤ教の礼拝堂、シナゴーク

wrangle ：口論する

tier ：層

threshold ：しきい値

Catfish-style relationships ：

scrape ：かき集める

qualitative ：定性的な、質的な

quantitative ：定量的な、量的な

hallmark ：顕著な特徴

non-specific ：特定しない、あいまいな

courtship ：求愛

optimize ：最適化する

fabulous ：すばらしい

arctic ：北極圏

matzoh balls ：マッツォ・ボール、ユダヤ料理

fluffy ：ふわふわした

【transcripts】

So my name is Amy Webb, and a few years ago I found myself at the end of yet another fantastic relationship that came burning down in a spectacular fashion. And I thought, you know, what's wrong with me? I don't understand why this keeps happening.

So I asked everybody in my life what they thought. I turned to my grandmother, who always had plenty of advice, and she said, "Stop being so picky. You've got to date around. And most importantly, true love will find you when you least expect it."

Now as it turns out, I'm somebody who thinks a lot about data, as you'll soon find. I am constantly swimming in numbers and formulas and charts. I also have a very tight-knit family, and I'm very, very close with my sister, and as a result, I wanted to have the same type of family when I grew up.

So I'm at the end of this bad breakup, I'm 30 years old, I figure I'm probably going to have to date somebody for about six months before I'm ready to get monogamous and before we can sort of cohabitate, and we have to have that happen for a while before we can get engaged. And if I want to start having children by the time I'm 35, that meant that I would have had to have been on my way to marriage five years ago. So that wasn't going to work.

If my strategy was to least-expect my way into true love, then the variable that I had to deal with was serendipity. In short, I was trying to figure out, well, what's the probability of my finding Mr. Right? Well, at the time I was living in the city of Philadelphia, and it's a big city, and I figured, in this entire place, there are lots of possibilities. So again, I started doing some math. Population of Philadelphia: It has 1.5 million people. I figure about half of that are men, so that takes the number down to 750,000. I'm looking for a guy between the ages of 30 and 36, which was only four percent of the population, so now I'm dealing with the possibility of 30,000 men. I was looking for somebody who was Jewish, because that's what I am and that was important to me. That's only 2.3 percent of the population. I figure I'm attracted to maybe one out of 10 of those men, and there was no way I was going to deal with somebody who was an avid golfer. So that basically meant there were 35 men for me that I could possibly date in the entire city of Philadelphia.

In the meantime, my very large Jewish family was already all married and well on their way to having lots and lots of children, and I felt like I was under tremendous peer pressure to get my life going already.

So if I have two possible strategies at this point I'm sort of figuring out. One, I can take my grandmother's advice and sort of least-expect my way into maybe bumping into the one out of 35 possible men in the entire 1.5 million-person city of Philadelphia, or I could try online dating. Now, I like the idea of online dating, because it's predicated on an algorithm, and that's really just a simple way of saying I've got a problem, I'm going to use some data, run it through a system and get to a solution. So online dating is the second most popular way that people now meet each other, but as it turns out, algorithms have been around for thousands of years in almost every culture. In fact, in Judaism, there were matchmakers a long time ago, and though they didn't have an explicit algorithm per se, they definitely were running through formulas in their heads, like, is the girl going to like the boy? Are the families going to get along? What's the rabbi going to say? Are they going to start having children right away? And the matchmaker would sort of think through all of this, put two people together, and that would be the end of it. So in my case, I thought, well, will data and an algorithm lead me to my Prince Charming? So I decided to sign on.

Now, there was one small catch. As I'm signing on to the various dating websites, as it happens, I was really, really busy. But that actually wasn't the biggest problem. The biggest problem is that I hate filling out questionnaires of any kind, and I certainly don't like questionnaires that are like Cosmo quizzes. So I just copied and pasted from my résumé.

So in the descriptive part up top, I said that I was an award-winning journalist and a future thinker. When I was asked about fun activities and my ideal date, I said monetization and fluency in Japanese. I talked a lot about JavaScript.

So obviously this was not the best way to put my most sexy foot forward. But the real failure was that there were plenty of men for me to date. These algorithms had a sea full of men that wanted to take me out on lots of dates -- what turned out to be truly awful dates. There was this guy Steve, the I.T. guy. The algorithm matched us up because we share a love of gadgets, we share a love of math and data and '80s music, and so I agreed to go out with him. So Steve the I.T. guy invited me out to one of Philadelphia's white-table-cloth, extremely expensive restaurants. And we went in, and right off the bat, our conversation really wasn't taking flight, but he was ordering a lot of food. In fact, he didn't even bother looking at the menu. He was ordering multiple appetizers, multiple entrées, for me as well, and suddenly there are piles and piles of food on our table, also lots and lots of bottles of wine. So we're nearing the end of our conversation and the end of dinner, and I've decided Steve the I.T. guy and I are really just not meant for each other, but we'll part ways as friends, when he gets up to go to the bathroom, and in the meantime the bill comes to our table. And listen, I'm a modern woman. I am totally down with splitting the bill. But then Steve the I.T. guy didn't come back. (Gasping) And that was my entire month's rent. So needless to say, I was not having a good night. So I run home, I call my mother, I call my sister, and as I do, at the end of each one of these terrible, terrible dates, I regale them with the details.

And they say to me, "Stop complaining." (Laughter) "You're just being too picky."

So I said, fine, from here on out I'm only going on dates where I know that there's wi-fi, and I'm bringing my laptop. I'm going to shove it into my bag, and I'm going to have this email template, and I'm going to fill it out and collect information on all these different data points during the date to prove to everybody that empirically, these dates really are terrible. (Laughter) So I started tracking things like really stupid, awkward, sexual remarks; bad vocabulary; the number of times a man forced me to high-five him.

So I started to crunch some numbers, and that allowed me to make some correlations. So as it turns out, for some reason, men who drink Scotch reference kinky sex immediately.

Well, it turns out that these probably weren't bad guys. There were just bad for me. And as it happens, the algorithms that were setting us up, they weren't bad either. These algorithms were doing exactly what they were designed to do, which was to take our user-generated information, in my case, my résumé, and match it up with other people's information. See, the real problem here is that, while the algorithms work just fine, you and I don't, when confronted with blank windows where we're supposed to input our information online. Very few of us have the ability to be totally and brutally honest with ourselves. The other problem is that these websites are asking us questions like, are you a dog person or a cat person? Do you like horror films or romance films? I'm not looking for a pen pal. I'm looking for a husband. Right? So there's a certain amount of superficiality in that data.

So I said fine, I've got a new plan. I'm going to keep using these online dating sites, but I'm going to treat them as databases, and rather than waiting for an algorithm to set me up, I think I'm going to try reverse-engineering this entire system. So knowing that there was superficial data that was being used to match me up with other people, I decided instead to ask my own questions. What was every single possible thing that I could think of that I was looking for in a mate?

So I started writing and writing and writing, and at the end, I had amassed 72 different data points. I wanted somebody was Jew...ish, so I was looking for somebody who had the same background and thoughts on our culture, but wasn't going to force me to go to shul every Friday and Saturday. I wanted somebody who worked hard, because work for me is extremely important, but not too hard. For me, the hobbies that I have are really just new work projects that I've launched. I also wanted somebody who not only wanted two children, but was going to have the same attitude toward parenting that I do, so somebody who was going to be totally okay with forcing our child to start taking piano lessons at age three, and also maybe computer science classes if we could wrangle it. So things like that, but I also wanted somebody who would go to far-flung, exotic places, like Petra, Jordan. I also wanted somebody who would weigh 20 pounds more than me at all times, regardless of what I weighed.

So I now have these 72 different data points, which, to be fair, is a lot. So what I did was, I went through and I prioritized that list. I broke it into a top tier and a second tier of points, and I ranked everything starting at 100 and going all the way down to 91, and listing things like I was looking for somebody who was really smart, who would challenge and stimulate me, and balancing that with a second tier and a second set of points. These things were also important to me but not necessarily deal-breakers.

So once I had all this done, I then built a scoring system, because what I wanted to do was to sort of mathematically calculate whether or not I thought the guy that I found online would be a match with me. I figured there would be a minimum of 700 points before I would agree to email somebody or respond to an email message. For 900 points, I'd agree to go out on a date, and I wouldn't even consider any kind of relationship before somebody had crossed the 1,500 point threshold.

Well, as it turns out, this worked pretty well. So I go back online now. I found Jewishdoc57 who's incredibly good-looking, incredibly well-spoken, he had hiked Mt. Fuji, he had walked along the Great Wall. He likes to travel as long as it doesn't involve a cruise ship. And I thought, I've done it! I've cracked the code. I have just found the Jewish Prince Charming of my family's dreams.

There was only one problem: He didn't like me back. And I guess the one variable that I haven't considered is the competition. Who are all of the other women on these dating sites? I found SmileyGirl1978. She said she was a "fun girl who is Happy and Outgoing." She listed her job as teacher. She said she is "silly, nice and friendly." She likes to make people laugh "alot."

At this moment I knew, clicking after profile after profile after profile that looked like this, that I needed to do some market research. So I created 10 fake male profiles. Now, before I lose all of you -- (Laughter) -- understand that I did this strictly to gather data about everybody else in the system. I didn't carry on crazy Catfish-style relationships with anybody. I really was just scraping their data. But I didn't want everybody's data. I only wanted data on the women who were going to be attracted to the type of man that I really, really wanted to marry. (Laughter)

When I released these men into the wild, I did follow some rules. So I didn't reach out to any woman first. I just waited to see who these profiles were going to attract, and mainly what I was looking at was two different data sets. So I was looking at qualitative data, so what was the humor, the tone, the voice, the communication style that these women shared in common? And also quantitative data, so what was the average length of their profile, how much time was spent between messages? What I was trying to get at here was that I figured in person, I would be just as competitive as a SmileyGirl1978. I wanted to figure out how to maximize my own profile online.

Well, one month later, I had a lot of data, and I was able to do another analysis. And as it turns out, content matters a lot. So smart people tend to write a lot -- 3,000, 4,000, 5,000 words about themselves, which may all be very, very interesting. The challenge here, though, is that the popular men and women are sticking to 97 words on average that are written very, very well, even though it may not seem like it all the time. The other sort of hallmark of the people who do this well is that they're using non-specific language. So in my case, you know, "The English Patient" is my most favorite movie ever, but it doesn't work to use that in a profile, because that's a superficial data point, and somebody may disagree with me and decide they don't want to go out with me because they didn't like sitting through the three-hour movie.

Also, optimistic language matters a lot. So this is a word cloud highlighting the most popular words that were used by the most popular women, words like "fun" and "girl" and "love." And what I realized was not that I had to dumb down my own profile. Remember, I'm somebody who said that I speak fluent Japanese and I know JavaScript and I was okay with that. The difference is that it's about being more approachable and helping people understand the best way to reach out to you.

And as it turns out, timing is also really, really important. Just because you have access to somebody's mobile phone number or their instant message account and it's 2 o'clock in the morning and you happen to be awake, doesn't mean that that's a good time to communicate with those people. The popular women on these online sites spend an average of 23 hours in between each communication. And that's what we would normally do in the usual process of courtship.

And finally, there were the photos. All of the women who were popular showed some skin. They all looked really great, which turned out to be in sharp contrast to what I had uploaded.

Once I had all of this information, I was able to create a super profile, so it was still me, but it was me optimized now for this ecosystem. And as it turns out, I did a really good job. I was the most popular person online.

And as it turns out, lots and lots of men wanted to date me. So I call my mom, I call my sister, I call my grandmother. I'm telling them about this fabulous news, and they say, "This is wonderful! How soon are you going out?" And I said, "Well, actually, I'm not going to go out with anybody." Because remember, in my scoring system, they have to reach a minimum threshold of 700 points, and none of them have done that. They said, "What? You're still being too damn picky."

Well, not too long after that, I found this guy, Thevenin, and he said that he was culturally Jewish, he said that his job was an arctic baby seal hunter, which I thought was very clever. He talked in detail about travel. He made a lot of really interesting cultural references. He looked and talked exactly like what I wanted, and immediately, he scored 850 points. It was enough for a date.

Three weeks later, we met up in person for what turned out to be a 14-hour-long conversation that went from coffee shop to restaurant to another coffee shop to another restaurant, and when he dropped me back off at my house that night I re-scored him -- [1,050 points!] -- thought, you know what, this entire time I haven't been picky enough. Well, a year and a half after that, we were non-cruise ship traveling through Petra, Jordan, when he got down on his knee and proposed. A year after that, we were married, and about a year and a half after that, our daughter, Petra, was born. (Applause)

Obviously, I'm having a fabulous life, so -- (Laughter) -- the question is, what does all of this mean for you?

Well, as it turns out, there is an algorithm for love. It's just not the ones that we're being presented with online. In fact, it's something that you write yourself. So whether you're looking for a husband or a wife or you're trying to find your passion or you're trying to start a business, all you have to really do is figure out your own framework and play by your own rules, and feel free to be as picky as you want.

Well, on my wedding day, I had a conversation again with my grandmother, and she said, "All right, maybe I was wrong. It looks like you did come up with a really, really great system. Now, your matzoh balls. They should be fluffy, not hard."

And I'll take her advice on that.

エイミー・ウェブ「ネット婚活で成功する方法」