The problem with having "rational" age cutoffs, political litmus tests, etc. is that very few people are within any fixed set of metrics on multiple dimensions.

 A blog post from A young woman speaking about her date and dating philosophy by  Tyler Cowen May 29, 2021 at 3:01 pm in Philosophy

Her other nonnegotiable is quarantine behavior. She was happy when she found out [male name redacted] takes safety seriously, interacting with only a small pod of people and limiting travel. “That showed me we had similar values,” she said. “Being caring, empathetic — and also believing in science and CDC regulations and guidelines.”

No, I am not interested in giving you a link or in identifying anybody by name.  The point is this: this is one of the very best paragraphs ever written in helping you to understand the Blue State reaction to the pandemic.

Interesting at several layers.  One is of course the difference in Blue State and Red State reality assessments.  

But most the interest is in the comments.  Zaua focuses on the social aspect and the so easy to condemn preciousness, naivety, indeed, foolishness, of the young woman quoted.

Zaua  2021-05-29 15:26:39

+118 -7

The full article is an interesting glimpse into dating today. The woman was looking for a guy who would check all her boxes, apparently found a guy who did, with similar interests and values, rated the date a 4/5, and still turned down a second date because she thought the age difference was too much (she is 23 and the guy 27).

However, Sure, goes all informational, mathematical and data oriented.  My preferred mode.  There was a long discussion based on his response:

Sure 2021-05-29 18:07:04

+86 -0

In 1950 US Airforce measured 4,063 pilots against ten dimensions (e.g height, chest depth, neck circumference). Of those, precisely zero managed to be in the middle 60% of all ten dimensions. If you wanted to select for the middle 60% for just three characteristics you excluded 96.5% of pilots. And this was in a population that had been pre-selected for being of average size.

The problem with having "rational" age cutoffs, political litmus tests, etc. is that very few people are within any fixed set of metrics on multiple dimensions. Our dater wants a Democrat, so you just canned 67% of the population. She wants someone "talented at what they do" so let's generously say that she wants somebody in the top third of their profession's bell curve. Well there goes another 67% of the remaining third. She wants somebody, presumably within 4 years of her age, so we are tossing another 90% of the male population. Assuming she wants a guy who is single (goodbye 59%), one who is "romantic" (say top 50% of men for romance), and "cute" (say top 33% on physical attractiveness) we are now looking at something like 0.07% of the male population as meeting just the most basic criteria that could be reasonably inferred from her statements.

Maybe you think the bell curves on these non-physical traits are more highly correlated? Quite possibly, and that makes it worse. After all, if there are a pot of talented, romantic, and cute guys, what are the odds that they are not disproportionately coupled?

Having mandatory check boxes (i.e. anything you are willing to call it quits over alone) is utterly toxic beyond maybe three. And those three had better be things that are nearly immediately obvious and reasonably common.

Because the other thing to recall is that the guy likely has his own set of criteria. If our male suitor were a tenth as picky as this girl's top line numbers we would be expecting something like like a 0.00005% chance that any date will work out. At 100 dates per year we would expect a successful match in 2040 years.

Yes, good searching mechanisms can bring that down, but reality is that "high standards" are typically utterly toxic. If you cannot manage to have a good relationship when you are in "different stages of life", I shudder to think of how you would handle concurrent crises (e.g. your spouse gets diagnosed with cancer and you get laid off).

What actually works is not finding the one rare gem who is already perfectly cut to your precise requirements. It is being flexible and willing to make sacrifices for your relationships. I foresee many fruitless dates before our dater(s) finally get truly reasonable and even odds they fail to do so before they miss out on many of the coupledom experiences they desire (e.g. inexpensive fertility, low hassle travel opportunities, mutual support during career launches).

"We are now looking at something like 0.07% of the male population as meeting just the most basic criteria that could be reasonably inferred from her statements."  As Sure drives home, given that coupling is a symetric process, when we take into account his criteria, the probability of a date leading to some steady coupledom is essentially zero.  Until one or both grow up and reality inflicts a better awareness that you get 1-3 criteria to choose from if you have any realistic chance of hitting it off.  That's where both wisdom and character kick-in.

This unconscious destructive decisioning is prevalent in many areas.  We are always looking for the good, honest, electable, competent, bright, easy-going, and effective politician to deliver us from bad policies.  But as the Founding Fathers recognized, if you are waiting for perfect, you wait forever.  You have to anticipate that participants in government will be imperfect, sometimes astonishingly imperfect.  So you have to build in the checks and balances which allows even the imperfect to contribute.

The inclination toward infinite branching shows up polling all the time.  For any given attribute, you need to a perfectly random and perfectly representative sample of respondents.  This is virtually impossible to achieve.  When you deal with a large population, such as the US, the minimum number of random and representative participants will be roughly 300 to answer a single branch question.  

But what polling firms do all the time is not have random and not have representative samples.  And they then compound the issue with infinite branching.  Instead of "Are you better or worse off than four years ago?" they keep branching.  How many African-Americans think they are better or worse off?  You are now at the second branch.  How many African-Americans in Florida think they are better off?  How many African-Americans in Florida in the middle income quintile think they are better off?  You are now at branch four.  For the answer to meaningful, your sample size needs to be very roughly 95,000.  All random and representative.  

Of course, that is inordinately expensive which is why public opinion polls rarely provide useful information.

Other items raised in the comments includes the issue of trait preference self-deception/trade-off ignorance.  It is like the old manufacturing adage: faster, cheaper, better - pick two.  You want a high income mate who can spend a lot of time?  Pick one.  You want someone who is respectful, assertive, and tender?  Pick two.  A spontaneous but reliable partner? . . . 

This is similar to trait preference transience.  In other words, what you think you want at eighteen may not be what you are looking for at twenty-eight.  There is a lot in a person baked in through DNA and not easily changeable.  There is certainly some flex, but no one is a tabula rasa.  If you have a long list (more than three) of must-have attributes, the greater is the probability of an emerging mismatch over time.  

Of course there is the usual blather in the comments, but this post certainly generated a lot of slivers of insight.