The article makes it sound as if „200 question“ dating sited like eHarmony were any better, but it’s scientifically proven that there’s no correlation between such questionnaire matches and actual relationship compatibility
As far as I know, eHarmony's claim of the questionnaire and algorithm being based on science was marketing fluff from the beginning [1]. One of the founders was a clinical psychologist, but the only science I've ever seen them cite is just studies that they claim show eHarmony to be more successful than other dating sites (which might have something to do with the company's policy of outright rejecting users who are too depressed or too divorced [2]).
Most important is phermone compatibility, and you can only gauge it by smell and presence.
Online dating is therefore doomed to failure from the get go, because it ignores it completely.
Of course, you can get very, very lucky...
If the woman doesn't smell right (read: irresistible), all the other qualities don't really matter. It happens pretty rarely that a woman smells just so right that your head starts spinning, and that's the only type of girl worth going out with as far as I'm concerned.
I surmise that a big portion of people that hooked up via dating apps are incompatible at a fundamental level and they just don't know better.
Indeed, these kind of apps make dating so … official and formalized. You need a profile, you need to answer questions (which implies knowing what you’re actually looking for), you need to fit into the options offered by the app (which vary depending on religion and politics in your area) and then “the algorithm” trying to match you based on engagement. It’s just not an enjoyable process.
Of course, just a list of profiles may not be helpful either e.g. I met my husband on Grindr but that required a lot of effort and enduring the emotional downsides of the experience. Dating online also deprives one of using all of one’s senses to “feel” if the person(s) in front are a good match.
PS: For some of us, online dating is not always a choice as you can imagine, so even with the downsides, these apps are needed still.
I wasn’t thinking of any particular paper there, but if you look at some papers you’ll usually find the following reasons:
1. The app algorithms are usually black boxes, so who knows whether they’re really using scientific methods
2. Apps often include questions such as height due to customer demand, which are known to be very bad predictors
3. The interaction between the factors is very significant and complicated/unknown
4. People don’t answer truthfully
5. People don’t know what they want
What is known however is that there’s a strong placebo effect initially, i.e. the mere suggestion that somebody is a good match although they aren’t increases the success of first dates, but not in the long term. There’s also research on how these systems can be designed to be addicting, which is surely being done.
One of the strongest known predictors instead is “shared emotional experiences”, e.g. a study about making people walk over a wobbly bridge together. Hence why “running groups” and so on are popular nowadays among people disillusioned by algorithms
