Rationally speaking, is a presentation of New York City skeptics dedicated to promoting critical thinking, skeptical inquiry and science education. For more information, please visit us at NYC Skeptic's Doug. Welcome to, rationally speaking, the podcast, where we explore the borderlands between reason and nonsense. I'm your host, Masimo. You can with me, as always, is my co-host, Julia Gillard. Julia, what are we going to talk about today?

Nosmo, today are topics a fun one. We are going to be discussing the world of online dating.

Online sex, no what? No online dating, online dating than last time you thought our naturalism episode was about people naked on the beach and this time you think are perfectly innocent. Episode on dating websites is about sex. No. Right. Online dating.

OK, fine. We'll talk about dating. So what about online dating? You don't sound so grumpy. Yeah. So we are going to talk about sort of the booming business of online dating and matchmaking websites and what researchers have learned about how well these websites actually work, like how well they end up matching people with, you know, others who are good matters for them. And you know what kinds of strategies people have learned for using these websites effectively, how to use the algorithms and tweak your profile to actually find that soulmate or that, you know, one night stand made as the case may be?

Well, it worked for me.

I did. That's anecdotal evidence. Find a one night stand made possible. No, no, no.

It worked for me, as you know, long term relationship. Yeah. Oh, it sounds well. OK, Cupid, as a matter of fact.

Oh, yeah. OK, Cupid is is a big favorite among the smart intellectual nerds that I know many coastal cities if it fits the profile.

Yeah. I you know, I should probably note for people for whom this isn't common knowledge that online dating is like a huge thing at this point. It started out I mean, I guess as recently as ten years ago, it was still kind of fringe or kind of niche.

Actually, the first site to open was Match.com, and that was in 95 and five.

Wow. I know. Not that.

And then eHarmony, which is the the probably the second largest one at this point, was launched in 2000.

Yeah. Yeah. And now I was trying to Google around for statistics on just what percentage of the American population nor the American single population uses online dating websites. And I found a smattering of different statistics, but they're all like roughly around 30 to 50 or 60 percent of people who are single had been have tried online dating.

Yeah, it's only getting I think there was an article in USA Today last year that was putting online dating as at this point as the second or third most frequent way in which people go out on a date or meet or meet of people.

I also read that one in five new relationships that are formed in the last few years have been via an online dating site.

That's right. Which which is a pretty large percentage, actually, if you think it is, although it must be.

I don't know this for sure, but my hunch is that this is pretty unbalanced in like urban areas versus rural or maybe coastal versus middle of the country, because I like basically everyone I know in my social group. I would just assume that they have an OK Cupid profile. I would ask them about their Cupid profile the same way you ask someone about their Facebook profile. It's just assumed. And yet I have run into people who who if online dating comes up, they sort of laugh.

They're like, who would do that? And I sort of gawk at them similarly. Like, who would not do that? That's right. So there must just be a divide that's like averaging out to forty percent, you know, not something like that.

But there's also, I think, a distinction, of course, between different countries and even within the Western Hemisphere. You know, my my own experience is between the United States and Italy, for instance. And in Italy, I've seen through my one of my brothers who just got married recently, but but until recently, he was actually dating, you know, in a variety and meeting people in a variety of ways, one of which was online. And that's what we talked about it over the years.

And it's interesting. That is his impression, of course, again, entirely anecdotal was that the attitude was changing over time. Slowly but surely. There is you know, a few years ago in Italy, you would have been looked at as an odd quirk of, you know, a human being if you were actually using online dating. But now it's getting more and more acceptable, although I don't think as acceptable as it is in in large cities in the United States and who knows one who knows where in the rest of the world is, you know, sort of the western world.

One side you can look to to see like growing normalcy, normal growing acceptability is how how often do you see people in their profiles say things like, you know, I can't believe I'm doing this or, you know, I hope no one finds out I'm on this side or like in one of the it's often profiles will have a prompt. That's like what's one thing that's or what's the biggest secret you're willing to confess? And it used to be that a lot of people would write.

The fact that I'm on an online dating site, as if that weren't an insult to everyone else reading their profile, who is also, of course, on an online dating site, but I feel like that's much less common now. It wouldn't it doesn't occur to as many people that this would be something to be ashamed of.

All right. Well, there was a I just found that some data about, again, the percentage of people using these these systems before we get into the more detail of what they actually do and how they work and all that.

But, for instance, Match.com apparently is responsible for about 12 percent of all the dates in the United States.

Do you see 12, 12, which is somewhat, you know, pretty high? eHarmony apparently is responsible for about five percent of all weddings, which, again, it's a rather small percentage, but it's a significant number of people hired to me.

Yeah, five percent.

And there was a survey done in 2010 10 based on 12000 people sampled, which was performed by the independent research firm.

And that survey found that one in six couples who got married said they met online.

So that's that's pretty high, actually. Oh, yeah. And they would have had to meet several years ago or at least a year or two ago if they're married now. So and because the rate of increase is so high, the sad thing is even higher now, like relationship starting now, that would lead to marriage. All right.

Another thing, though, I think at some point we should talk about, of course, is that the as you know, some states particularly, OK, Cupid actually uses its own data, which are millions and millions of data points to find out interesting things about the people who use it.

Oh, yes.

And I think on my mental list of things to talk about, I found out that those are some interesting statistics about about that, unfortunately, sort of reaffirmed a lot of stereotypes. Oh, OK.

Yeah. Well, let's talk about the then. Well, what statistics do you think reaffirm unfortunate stereotypes?

Well, for instance, let's see here. One of the things is that the things that comes out regularly is, in fact, that there is quite a bit of racism out there, believe it or not. And then we find the actual data. So, for instance, the people of Lockerbie did this interesting analysis, which, by the way, they tend to do, as far as I can tell, these are not peer reviewed papers, of course, now just published on our blog.

But they are very careful about, you know, sort of statistically taking into account a variety of variables so, you know, filtering out things by possible confounding variables. So this is these are not raw data. This is actually somewhat sophisticated statistics. I'd like to see somebody doing a thesis on this thing.

But anyway, the interesting thing, I thought in terms of sort of racism and so the general conclusion is that racism is alive and well among their own members and that the way they reached this conclusion is they did this really interesting comparison.

So as you as you know and but we should perhaps explaining explain to our listeners. So, okay. Cubitt has a really complex questionnaire.

You can answer, you know, if you want to hundreds of questions about all sorts of aspects of your life. And then there is, of course, a matching algorithm. A lot of these these websites do have matching algorithms. And so they give you a percentage scoring. When you when you look at a profile of another person, there is a percentage that tells you, according to this complex algorithm and depending on how many answers you're given to the questionnaire, what your expected percentage of match is.

And it goes, you know, it can go up as high as 99 percent.

So what they did was to look at the match percentage. It was predicted by the algorithm.

And then they looked at whether people answered people of different genders and different ethnicities, answered queries from from other people in whether they responded to the message, right?

That's right. Whether they responded to the message and the basically the null hypothesis was that they should respond to the messages essentially in proportion to how closely matched they are. That is, regardless of race or, well, gender, it's not regardless, but regardless of race. In other words, that that the match, the matching level should be the predictor. And they tested this hypothesis first as a as a control by looking at the percentage of matches against the percentage of replies by Zodiac sign.

OK, so they looked at, you know, at this theological component. Right. So and sure enough, they found that zodiacal signs had absolutely no effect on the response rate. That is, the response rate was essentially what you would expect from the matching percentage, regardless of today callsign. So people look at the matching percentage, then respond to the matching percentage without taking into consideration the so like they sign of the prospective mate.

That is the good news because that means that, OK, Cupid users tend to ignore essential information about, you know, the Zodiac. People aren't Zodiac's. That's right. I hear you say cancer. Right now, the problem is, on the other hand, that when they did the same analysis comparing the match percentage by race, the results, unfortunately, were very different. And they have these really eye-catching sort of ways in which they graft the the results.

But essentially, it turned out that there is a major discrepancy between the percentage, the match and the response is that the number of responses you get by race and that varies whether it is a female sender that is, in other words, is the male that has to respond or a male sender, in which case is the female.

That response we're not talking about in this case on gay relationships and the surprises that they found evidence for a significantly higher degree of racism, that is, which is in this case is measure simply, as I said, as a as a discrepancy with the expectations from the match percentage, a higher percentage of racism among females than males.

And they have some really women are less likely to respond to a message from someone who's not of their race than men are to respond to a message from a woman who's not of their race.

That's right. So, for instance, some of the numbers are kind of interesting. So if you look at and also divided by, of course, again, further by ethnicity. And let me find the data. So they also add they also asked actually they had the questionnaire and they asked, you know, would you strongly prefer to date someone of your own skin color or racial background? And they compared the answer of that to the generic question, is interracial marriage a bad idea?

Now, here's the interesting thing. If you ask people, is interracial marriage a bad idea? The overwhelming majority, more than 95 percent, respond that it is not a bad idea.

That is, that interracial marriage is actually acceptable and that is regardless of ethnicity or gender. OK. Mm hmm.

But if you ask the same people, would you strongly prefer to date someone of your own skin color, then the results are completely different.

And it turns out that the answer, the yes, for instance, is 54 percent of white females answer yes.

Oh. And as opposed to say only and I'm using the word only in quotation marks here, 40 percent of white males, the lowest percentage of response. Oh, yes. Response is that is to the question again. Would you strongly prefer to date someone of your own skin color racial background? The lowest percentage responses are among Asian females, only 18 percent. Black males, 11 percent, Indian males, 17 percent.

Those those are the lowest percentages, but the highest percentages are among whites, which is interesting.

Yeah. So that actually does sound like one of the stronger OK trends or occupied data analysis, because they are they're implicitly controlling for any sort of other like socioeconomic conflating factors or something like that, because if those mattered to people, they would be taken into account. One would assume in the match percentage. All right. There are other trends, analysis that made me want to like, you know, jump up and down and tear my hair out because of be, you know, the fact that they weren't taking into account confounding factors like there is one post they did about how if you're a man, you can get more responses from women.

If you post a shirtless photo of yourself and actually know someone who acted on this knowledge or on this information and posted a photo of himself without a shirt. But of course, what trends was actually measuring was response rates among men with shirtless photos versus response rates among men without shirtless photos. And, you know, guess what kind of man is more likely to post shirtless photo of someone who looks good? So. Oh, surprise, surprise. Entirely free of confounding factors.

Oh, and then they also did this analysis where they were there were they use this is evidence of people lying in their dating profiles, which I'm sure is rampant, but their evidence was kind of weak. They showed a graph of reported male height on OK, Cupid, and they showed that men on occupied report themselves as being taller on average than men in general and the US population. But I would expect confounding factors to be really influential there, because, for example, I if I had to guess, I'd say, OK, Cupid is more white, like disproportionately white compared to the US population, but whites are disproportionately taller than other ethnicities.

So I'm not sure that I would expect men on OK Cupid to actually be Representative Hitwise of Americans.

Yeah, that's a good point. Now, about the D. There are some data actually out there about lying on these and these profiles. Yeah, and they're actually kind of interesting.

First of all, the percentage of people who lie is apparently not as high as as as.

Seem to think, how do they measure that they don't ask people if they lie? You think that that's not a good way to go about it? I didn't say that. Well, I don't I don't know exactly here the methodology.

But this this this survey, for instance, they found out using a particular site and this site is is geared toward extramarital affairs.

So you would think that they're more likely to lie. Exactly right. Well, or not.

Because, in fact, if if the site is is openly for extramarital affairs, you don't need to lie.

I mean, you know, if if somebody knows if you're there, presumably you want to. So that's true.

But still like the kind of person who's willing to lie to their partner, you know, and I have an affair is also probably more willing to lie about their height or age or income or something like that or whether they have an STD or something.

Right. That's possible. But anyway, the results of these of these sort of very interesting. So it turns out that the largest percentage of liars, and that's 11 percent of the members of this particular group is I.T., an engineering workers.

And it's followed by the second one at eight percent is financial industry. So people who are common I.T. engineering or financial industry are particularly likely to lie about about their profile. Now, the other interesting thing I found was that actually Americans are more likely to lie than than people in the U.K. on their profile by by a significant percentage. So it's like so there are there are variations of the actual numbers are 53 percent of U.S. online daters lie and as opposed to 44 percent of UK, their U.K. counterparts.

Now, the numbers are high. First of all, they're actually not as high as I think the crew numbers are, but they are much higher than one would like them to be. Right. Right there. There is a grand tradition in the U.K. of sort of brutal, self-deprecating honesty in a precursor to online dating. The London Review of Books, personals, you know what I'm talking about there? The personal ads in the back of it's called the London Review of Books and such a thing.

So, yeah, yeah. They're just famously self-deprecating, like, you know, overweight, unappealing spinster in search of someone to listen to her blather at night. Please call this number. So maybe maybe that tradition accounts for the difference and it could be.

Well, here's a study from and again, I use the word study generously in this case.

But but I think it's pretty well, then it's kind of interesting. This is also from the occupied. Datasets and this this answers the question of why is it that certain profiles get so much more attention than others, even though you might expect that that would that would not be the case. So bear with me for a second. This is about female profiles, but they found similar results in four male profiles. So this is not a matter of objectifying anybody in particular.

It's just the data that I have at about the female profiles.

And the question in particular was there seems to be certain categories of women who are ranked on average as attractive or good-Looking or whatever as another category. And yet the first category gets a lot more responses than the second one. So in other words, that the higher than average, better looking women tend to be seems to be belonging to two different groups, to the different peaks, one of which gets a lot of responses and one gets less so.

And if they have a term for that, it's the curse of being cute.

Go on.

Now, here's what they did. So they look, first of all at the raw data. And sure enough, they they showed that there are there is a clear bimodal distribution basically of of women who have these these equivalent scores in terms of attractiveness.

And so they started looking at, well, what is it exactly that make what makes one group different from the other?

It turns out that, of course, as you know, statistically, there's more than one way to get an average result, right? If you just look at the average, it doesn't tell you anything about the distribution of the results. Right. It turns out that when they turn their attention to the distribution of the results, they they found out that there is the two groups of women have very different sort of spreads of their their responses in one group.

Again, the average in this case in both cases was high. But in one group, it had a distribution and were skewed to the right. So there was a large number of positive responses, you know, no, let's say that were talking about a scale from one to four where one to five where five is most attractive.

Both groups have an average of four, but one group has a peak around four.

And then the long tail to the left, the other group has two peaks, one around four and the other one and one on one.

That is there are there are some women who, although on average they score very high. Some people really like them and some people really, really dislike them.

And those are the women actually, surprisingly, do better in their profiles.

And it's not clear exactly why that is this whoever wrote these article actually puts it in terms of it has an explanation or she has an explanation in terms of game theory.

And the basic idea, which is an interesting hypothesis, is this, that if you look at a profile of a woman or a man and you think that that person is above average and a lot of people are going to be probably writing to her, you are less likely, you're less motivated to write because you think that your message is going to be swamped by a bunch of other competitors.

But if you think that that person is interesting to you and yet it has some characteristics that other people may dislike. And I'll give you an example in a minute.

And the idea is that that actually is a better bet for you to to to write so that other things being equal these do you like these to women or men equally, but one has some kind of characteristic that you think makes it him or her less attractive. Your best bet is to write to the second one rather than to the first one.

And the typical example here is women or men who have who show tattoos on on their profile photos.

That's divisive because that's divisive. Some people really dislike it.

And so if you like a person who is, you know, sort of above average, cute, interesting looking and all that, but you like her or him as much as a similar sort of person, except that the latter one has a tattoo. You're much better off in terms of success rate to write to the second one, the one with the tattoo, then not.

And so this is practical advice, so I I remember this blog post, I actually wrote a response to it on my own blog. I kind of disagree with their analysis. So they like another way of putting the this results is that you get you know, people get rated one to five on physical attractiveness and getting a for rating is negatively correlated with getting messages. And getting a five is positively correlated with getting messages like for each additional four or five you get.

Right. And so their explanation is like, yeah, if someone gets a four, they assume that like other men are like she's attractive, but also other men are going to be into her. So I won't bother. And then, you know, five is like, yes, other men are into her, but she's so attractive. I'm going to try anyway. But so the way the attractiveness rating system works is that you rape people one to five.

And if you rate them a four or five and they also rate you a four or five, then occupied sends you both a message saying, hey, this person is interested in you. And I happen to know that a lot of men at speed dating events or in online dating will just rape everyone. Yes. Or like everyone a four or five, because they just want to like carpet bomb strategy so that any woman who's interested in them will get notified.

Good point. So you would expect that like even the less attractive women, a lot of men would write them off for just to like, you know, have a chance of getting contacted by them, whereas the women who would get made of five would actually be attractive. So that would be why I would expect, of course, negatively correlated, testable hypotheses.

I mean, the thing I find interesting is, of course, that that there are probably, in fact, different strategies at play in these, you know, in the online dating game, just like probably there are similar strategies, quite frankly, in offline dating game.

Right. I mean, I, I never actually had the experience of going to a bar to try to pick up somebody. But I assume that it's a similar thing. I mean, you you hedge your bets, you look around and you see, you know, not only what you're interested in, but also what your chances you think are of succeeding. And, you know, you can do that consciously or unconsciously.

But the thing is, certainly there are different strategies at play. And I'm sure somebody could come up with with, you know, algorithms that can compete for the best but the best outcome tradeoffs.

Yeah, I mean, didn't John Nash and yeah. The mathematician and a beautiful mind give a game theoretic analysis of which woman to approach it? That's right. I'll leave it at that. You can go rent the movie.

Now, this is some some data also that I think are interesting about from these dating sites that are interesting about people's behaviors other than, you know, how do they pick their mate. But sort of self reported, of course, these are all self reported.

So by several positive behaviors, some of which are interesting because the people again, typical, it's OK, Cupid does these these kinds of analysis, they glean some interesting insights into their data by bypassing the correlations between between different completely superficially, completely different questions, such as, for instance, let me give you one that I thought was curious, the chance that a woman has of achieving an orgasm and whether she exercises or not.

Now, the two questions apparently have nothing to do with each other.

So it's hard to imagine that, you know, somebody would actually, you know, consciously lie in that sort of stuff.

But when they pass the date, the data, it turns out that the two curves. So if you plot, you know, the chances of achieving or not achieving an orgasm versus age, the two curves are perfectly equivalent, essentially parallel, so that it becomes more and more it's more and more difficult for young women and for better and for older women.

So it turns out that it's almost a flat line between the ages of 30 and 50. And that's where you are.

And this is a side note, right? What generation you come from here is is like totally entangled with what your ages. So it's not necessarily correct to assume that for any given woman as she gets older, she will have a harder time achieving. That's right.

So but anyway, apparently the easiest time actually is between about the age of 30 or late 20s and the age of early 50s. But the thing is, if if a woman does not enjoy exercising that the entire curve is shifted up, meaning that it has she has much more difficulty achieving orgasm.

So there if you enjoy exercise, there you go.

And apparently things are going to be better for you. I found some other interesting statistics.

There are some that I'm not sure that we want to get into it. But let me let me see here. Oh, yeah, that's right. This one was kind of interesting. And it has to do with masturbation. You're on a roll. I am.

So, first of all, the one question was, what are the one statistic was? What are the odds that somebody would masturbate today? Turns out that if you are if you use Twitter every day, you're much more likely to masturbate than if you don't.

I'm not sure at least be honest about it. Right. I'm not sure what that tells you about Twitter users. But but here's the more interesting part I think about this, these statistics.

So they also divided the they ask how many people, you know, the way they looked at the claim that the people never masturbated.

And it turns out that there is a really interesting distribution by religion.

So it the claim that you're never masturbated is much higher among Jewish and Muslim women particularly, and it goes down a little bit of buy into Buddhist, goes down significantly.

Christians, atheists. That's right.

Christians, atheists and agnostics are about the same and they're the lowest, meaning that they're they don't claim that they never masturbated. That's that's as far as women are concerned.

As far as men are concerned, there is less variation and the claim is a percentage lower. So it's more men, I do admit, to have masturbated. But even there, the surprising thing is there is that Christians and atheists are the lowest, meaning that they admit to masturbating more often.

Agnostics are a little higher there, as high as India and Muslims, which is, you know, I don't know what it is about atheist on the one hand, nasty on the other.

But either they practice differently or they admit things differently. That's why you said the word masturbated enough now that it's just lost all meaning to me. Do you see the same word again and again, just like a collection of sound? Oh, yeah. So. The one other OK trends result that I wanted to mention before we move on to all the rest of online dating, that isn't OK Cupid data, OK, is that they had this way of finding out information about your date that you're interested in, but you can't really ask about politely on the first date.

So like maybe you're interested in whether your first date is the kind of person who would be willing to have sex on a first date. But you can't ask them that straight out. No, but probably would be counterproductive. But you can ask them a question which, OK, Cupid has shown to be highly correlated with willingness to have sex on the first date. And one of the highest correlations across genders is do you like the taste of beer answering?

Yes. Make someone or should raise your confidence by 60 percent that they're going to want to have sex on the first date. Wow.

And for men, there's an even stronger correlation, which is in a sense, wouldn't it be really exciting to have nuclear war that would raise your confidence 83 percent that that the man is going to want to have sex with you on the first date? Don't ask me why.

That is what I call actionable information. Yeah.

OK, so what about you? OK, you be right. So you you mentioned comparing online dating to sort of a more traditional like meeting people at a bar, you know, some kind of some kind of non online context. And I wanted to talk about the research on whether online dating is actually helping people make better matches than they would make for themselves if they were meeting people in person first. So there's a bunch of research on this. And some of the most intriguing to me has been the research on whether people's stated preferences for what they want in a match actually correlate with the people they successfully match with.

OK, because there's a lot of research suggesting that those correlations are really weak when people state their preferences about things like income or interests or personality types. That these preferences actually do not just fairly predict who they're going to actually under a successful relationship with and in particular in online dating, some researchers have raised the concern that the kinds of things that you tend to look at and online dating, aside from the person's appearance, are the sort of easily quantifiable things like income or years of education, height, things like that.

And there's a growing body of research suggests that people make worse decisions when their attention is called to just the easily quantifiable things. So, for example, I think it was conmen and some other researchers did a study where they asked people to choose a car to buy and half of the subjects were asked to just go with their gut and the other half were asked to rate the cars on various dimensions like mileage and price and safety. Right. And then go with the car that scored the best in the rating system.

And when they followed up, when the researchers followed up with people later, they found that the people who had used the scoring system of decision making were less happy with their choice. And the interpretation there, which I think is quite plausible, is that focusing just on quantifiable metrics means you're not taking into account things like how happy do I feel when I look at this car, which of course, should be a pretty huge component of how glad you are that you made that purchase?

Yeah, it seems like there's a lot of room for that effect to be in play with online dating as well. That's interesting.

So let me make a couple of comments there. First of all, this this reminds me actually of a general problem in completely different fields. So when I was a practicing evolutionary biologist and ecologist, one of the problems in the ongoing discussions there was that on the one hand, you want to do, you know, as quantitative type of analysis as possible and you want to go out and measure characteristics of different organisms and correlate them with environmental characteristics and so on and so forth as rigorously as possible.

So you get these huge data sets and all that. And but a common objection to that approach was precisely what you are saying that is. Yes.

But now the problem is that now you're biasing your sampling, your understanding of, you know, a population of plants or animals or whatever or a group of species living in a community.

You're biasing it on by by focusing on the stuff that you can easily quantify. There's a lot of stuff that it may be equally, if not even more relevant, but it's more qualitative or it's more difficult to quantify. And so you simply ignore it.

And essentially you're letting your tools, your analytical tools dictate your research agenda and with results that are unpredictable. I mean, you know, because you don't know what is in it. You're missing, how much you're missing, how much off of the big picture you actually are.

So that's actually an interesting general question. That is, how do we balance quantitative versus qualitative information, and especially when, of course, it comes to human behavior. And, you know, there's been a trend for many years now, even in the social sciences, to focus more and more on quantitative data analysis, because that makes the thing more scientific, quote unquote. But there is a you know, generally speaking, that's that's a good idea. But there is that problem that you just pointed out.

The other thing than the other comment was actually more anecdotal, and that is when I was using online dating services, I would, in fact, look, of course, some of what I would consider my crucial variables, like especially in specialty level of education, for instance, you know, political leanings or religious leanings are usually ignored the income level.

But the thing is, all of those were simply the way I used them, was simply to narrow down the field dramatically.

And then within that narrowed field, which, of course, it also depends on whether the person is responding or not. But obviously, it's not entirely after you. Oh, it's a plus if they actually respond to you.

Well, and so, I mean, it's not like picking up a car, picking up a car or a pair of shoes. You know, here's a situation where the car can say, no, I don't think I'm going to be doing it, but I don't want you to drive me like that.

Assuming, of course, that a significant subsample of that actually does agree to go out on a date with you, then at that point, all bets were off.

That is, once you're out, you are natural on person to person date. All of those are the statistics essentially matter? Not at all, because at that point, what becomes relevant is, in fact, your gut feeling. You know, do I relate well with this person? Am I enjoying the conversation? You know, is there chemistry? And I mean, really chemistry as in, you know, do you feel sexual attraction to that person or not?

And those things are you know, you simply cannot glean from a profile or from any quantitative data that I know of.

And you can't you just can't do it. Well, I have some actionable advice of my own. All right, so as I was mentioning this like lack of correlation between these measurable factors and people's actual happiness with the person when they end up dating has led a lot of commentators to say that the algorithms, the matching algorithms these sites use are kind of pseudoscientific. I mean, a lot of them don't release their criteria for matching anyway. Right. Although you hear dribs and drabs, like I think it was one site, it might have been eHarmony or maybe it was perfect.

Match.com said that they use one of the things they take into account in their matching algorithm is the ratio of index to ring finger length, which is supposedly a marker that's coming to one level.

Yes, that's chemistry. And and it's based on the theories of Helen Fisher, who is a cognitive scientist who has done actual research on on human sexual behavior and from a neurobiological perspective.

And then she has become a consultant to Match.com. And with Match.com, they studied that this offshoot called chemistry. Yeah, that's right.

Oh, OK. I didn't know that background. But regardless, there's like a lot of uncertainty about what actually correlates with happiness in the match. So the thing that I advise you to do is to make the matching algorithm work for you. So I guess this applies mostly to, OK, Cupid, which is the only site I have any experience with. Back when I was on OK Cupid, I would essentially force the matching algorithm to be a good predictor for me because I know, as I mentioned, a bunch of people on OK, Cupid, and I would go to the profiles of the people who I knew and really liked and I would look at what our match percentage was.

And then I would really answer some questions to increase our match. So nice. And then I would look at the people who I knew, who I didn't like, or at least didn't like as a potential, you know, romantic match. And I would look at our match percentage and then I would answer some of my questions until our match percentage went down. And if you do this enough over time, your match percentage will actually become a better predictor of how much you're going to like someone.

Wow, that sounds like a lot of work. It was fun work.

Well, there is another thing that I want into another aspect of this whole online dating system that that I want to bring up. And actually, we very, very briefly touched upon a few minutes ago, which was which is the ethical aspect of all of this.

I mean, I think I really do think that that there is plenty of room here for a number of theses on on on the phenomenon of online dating, because there's a lot of data out there that people can use.

But those things could have different, different, different focus. Right. One can look at the, you know, actual efficiency of these algorithms. We can look at at how people self describe and how much there are racist or not in practice as opposed to the theory and so on and so forth, as we've seen. But one of the things that it's interesting is clearly there is an ethical dimension. So we mentioned that a surprisingly or maybe not so surprisingly, according to your opinion, a number of people actually lie on their profiles.

I mean, that never occurred to me. You know, I I don't know.

That's the lie, really.

It never occurred to, you know, and it never did because and here's why. It's not because I'm a particular virtuous person. It's because I approach the dating always in for the purpose of finding a long term relationship.

And in that case, I don't want to find myself in a situation of starting a relationship with with another human being. And then, you know, six months later, a year later, when things have come, they actually become serious.

Oh, by the way, I lied about my age.

I'm actually, you know, three years older, are five years older or whatever it is where I lied about my weight or something like that.

That seems like it would put me in a really bad situation in terms of, you know, so can I trust you now? What else did you lie about? So, no, it really never occurred to me. And the same thing never occurred to me for practical reasons, although I've seen people doing it to say, put a picture, an older picture, a younger picture of a younger self, because, again, you know, that picture may work immediately, meaning that people might might look at it and say, oh, yeah, for you know, for somebody with this age, that guy looks good.

But then you actually have to go out and meet these people.

And presumably you're not going to bring your, you know, with you a sort of an instant algorithm that makes makes makes you look younger when people look at you. Therefore. Right.

Therefore, there is an immediate sort of negative payoff to this thing. So anyway, the point is, it never occurred to me, but obviously does occur to a lot of people. And I've actually encountered people who I found out later lied and lied about on their profile.

I would I would love to actually ask them what their strategy there was, if they had a strategy, like did they just think that the force of their personality was so charming that once they got the person out on the date, it would end up, you know, not mattering that they weren't as young or thin as they were in their photo. Or, yeah, apparently, I mean, I don't know, I don't have many theories about it, right?

I mean, I can see why for if you're interested in short term dating now, just in a casual, casual date, then then it probably wouldn't matter that much, I mean, because you're there to attract the person for the short term. But if you're actually interested in anything in the long term, it seems like a completely self-defeating strategy to me. And yet, apparently, there's plenty of people, as I said, both in terms of anecdotal evidence and in terms of, you know, statistics that do it.

But there's also a broader dimension to this. So this is about the ethics of individual profiles. So individual people and what they say about themselves.

But then there is sort of a systematic issue with with sites that, as I mentioned a few minutes ago, actually are devoted exclusively to things like cheating.

You know, so there's there's things like there's a site called Discreet Adventure, the Ventures, Dotcom or Merrett Secrets Dotcom or Ashley Madison Dotcom. These are all for they all catered to married people who want the next marital affair. And, you know, one of these people that runs these websites was actually asked about the ethics of setting up this thing. And his response was interesting.

He says, well, he says that he merely facilitates infidelity.

It doesn't encourage it.

And I thought that was such a subtle distinction that this guy must be a lawyer or have must have talked to a lawyer or it I mean, to me, that's that's an interesting dimension of the whole thing.

I mean, that that we actually live in a society now where a behavior that is frowned upon normally and I think for good reasons, you know, I'm not I'm not being prude here, understand? I mean, I don't I don't have any particular point.

Is the betrayal and the deception. Exactly. Exactly. Right, exactly. It's not the promiscuity is not a number of people you go out with. That's your business as long as everybody's aware of it. But if you're lying to somebody, then I think that is to my way of looking at it. That's the way where you where you're crossing an ethical line. And now we're in a situation where you can actually, you know, do that for free essentially or for a very small monthly fee.

And and it creates this interesting is, you know, do we actually want a society where we not only condone these things, we actually encourage and make it much more easy to do?

Yeah, I mean, we could do a whole other podcast on, like moral blame or culpability and that sort of fuzzy spectrum where you're like doing the thing yourself or you're just letting it happen and not stopping it, or you're like, you know, handing the gun to the person, but you're not actually telling them to do the shooting or you're leaving the gun out where, you know, they could take it. And, you know, it is actually unclear to me where it makes sense to draw the line and say, like, you deserve blame on this side and not on that side.

No, I agree. I agree. I like many ethical questions.

There is there is probably a you know, the line is in that somewhere in the sand that there is there's a continuum there. But that doesn't mean that there's at some point you're clearly point beyond that line.

It's true. Well, I'm glad that we we moved beyond masturbation and for state sex into broader questions of scientific methodology and moral culpability.

Got to keep this podcast respectable, you know. Yeah, well, all right.

Well, on that respectable note, we are just out of time. So let's wrap this up and move on to the rationally speaking, PEX. Welcome back. Every episode, Julie and I pick a couple of our favorite books, movies, websites or whatever tickles our rational fancy. Let's start as usual with you.

Thanks, Massimo. My pick is a book. I'm about halfway through it at this point, but enough to recommend it. It's called Seeing Like A State How Certain Schemes to Improve the Human Condition Have Failed by James C. Scott in about 10 years ago. And it's all about these situations in the last century where, I don't know, maybe 200 years where sort of top down theories or planners like architects or government officials or like political writers have thought that they understood the way that the world works and that they could improve things, like they could improve the way the way people lived physically, like the way the cities were organized or apartment buildings were organized, or that they could improve political economic systems by redesigning the means of production, things like that.

And the book looks at all the ways things go wrong. And it's essentially like one big warning against overconfidence and against a basically a warning about the difficulty when you're using essentially explicit analysis and failing to take into account all of the sort of tacit knowledge that's just been developed over time organically in societies that have sort of self organized and how hard it is to see that from the top down. Nice. Yeah, it's really interesting. Well, my pick, on the other hand, is an article that appeared recently in eight or Dotcom, and we'll post, of course, the link and it is entitled Do Different Kinds of Alcohol Get You Different Kinds of drunk.

And apparently that's a very well, you know, wisecracked people claim they do. Yeah, a lot of people think that they do that it does.

And it turns out that it's there's no evidence of it. And in fact, there's pretty good reason to think that it doesn't. That is, of course, different ways of imbibing alcohol are much more effective than others. Right. So if if you you know, the National Institutes of Health has a something called the standard drink and the Southern drink is a way to make comparisons in terms of alcohol content between different drinks.

And it turns out, for instance, that a 12 ounces is regular beer is equivalent in terms of alcohol content to five ounces of table wine, and both of them are equivalent to one and a half ounces of brandy.

So obviously, it's not just about the total like percentage of alcohol times the volume.

It's about you know, the idea is that if you change the. Yes, it is the percentage of alcohol per. That's right. So.

So which means, however, that if you don't pay attention to the equivalent to the to the equivalency, to the standard drink and then of course you're going to get drunk, are easier more easily on, you know, with shots of vodka or with brandy than you are with beer. But that's it. That's entirely a matter of different quantities of actual alcohol. You imbibing. It's not different. There is no qualitative difference between alcohol. In fact, it according to the article, recent shows that alcohol effects on things like motor and cognitive functioning depend on, you know, there are individual differences, but these differences are related to the specific types of effects that the drinkers expect that is regarded.

You know, it depends on what you think it's going to happen. You know, in general, those who expect the least impairment, for instance, are in fact at least impaired. And those who expect to be most impaired, they are most impaired. And in fact, the interesting thing is that the same relationship is observed with respond in response to a placebo.

That is, if you if you they think that you give people non-alcoholic beer and wine and then they respond in the same way as if they had actually had alcohol.

Yeah, it's pretty entertaining.

Now, all of that said, there is only one caveat, which is it is possible that there are differences between different kinds of alcohol in what are called congeners. Congeners are a sort of byproduct of fermentation and distillation processes. These are small substances like acetone and, you know, asters, acetaldehyde and that sort of stuff.

And those, in fact, are different for different kinds of alcohol. And it may be that some people, of course, respond. You know, they're more or less sensitive to to to some of these byproducts. But I've of that most of the research seems to show that it doesn't matter if you're drinking brandy or beer, you're going to get drunk exactly in the same way, in exactly the same quantity of alcohol.

Well, no, Massimo, sometimes people refer to things as like good cocktail party trivia, but this one really is a good cocktail party trivia. All right. This concludes another episode of rationally speaking. Join us next time for more explorations on the borderlands between reason and nonsense. The rationally speaking podcast is presented by New York City skeptics for program notes, links, and to get involved in an online conversation about this and other episodes, please visit rationally speaking podcast Dog.

This podcast is produced by Benny Pollack and recorded in the heart of Greenwich Village, New York. Our theme, Truth by Todd Rundgren, is used by permission. Thank you for listening.