Optimizing Your Wife | Hacker News Reader

Optimizing Your Wife

mathpages.com

33 pointssiftrics31 comments

> If a man can expect to meet exactly N eligible women in his life, what strategy should he use to maximize his chances of choosing the very best one?

If you show her this article and she's still into you, that's the right one for you.

I imagine that approach would result in 0 eligible women! :D

You will always marry the wrong person. The trick is to make the best of it.

Corollary: You were always the wrong person to marry. Make the best of yourself, for their (and your) sake.

I wouldn't agree that you always marry the "wrong" person. I will admit that truly happy long-term marriages are exceedingly rare, an educated estimate I have read is that it is maybe about 10%. Then we have the well-known statistic that about 50% of marriages end in divorce. About 90% of those divorces are initiated by the woman, and in most of those cases the man ends up paying the woman while have access to his children diminished. The 40% that don't get divorced despite being unhappy are "making the best of it," as you say. I suspect most of those people are "doing it for the kids." Chris Rock joked that those 40% simply "don't have the courage to leave." Given that 90% of divorces are initiated by women, this would imply that 85% of men do not have the courage to initiate a divorce regardless of how unhappy or dire the circumstances may be.

The "optimal" mathematical strategy described in the article is also known as the "Secretary problem" and is definitely something worth pondering. The TLDR of the post is that, if you believe that you can attract up to 8 candidate partners, then you should pass on the first 3 candidates and then only stop when the next candidate is better than all prior ones. If you think that can attract up to 100 partners, then the optimal strategy is to pass on the first 9 candidates. Now this is just a mathematical model, with lots of assumptions, but the crucial point that it conveys is the importance of evaluating and passing before deciding, if you want to improve your chances of having a happy marriage. Knowing know dismal the actual statistics are, and how great the risks are, it would behoove an unmarried man to proceed with extreme caution when considering legally binding relationships or marriage contracts.

The chances of one getting divorced vary greatly depending on one’s income and ethnicity:

https://journeys.dartmouth.edu/censushistory/2016/03/04/divo...

https://flowingdata.com/2021/05/04/divorce-rates-and-income/

I reckon Wife-Optimizers also tend to be Resumé-Optimizers or Driver's-License-Picture-Optimizers.

Alternatively: Drive safe, work hard, and build a healthy relationship/family.

The drivers license picture optimizers are doing it for vanity. There is no benefit to optimizing that.

Resume and wife optimization provide cumulative lifetime value.

How does this contend with other optimal stopping problems? I just started reading 'Algorithms to Live By' and the very first chapter talks about the 37% rule [1], which feels similar. According to The Dating Problem [2] I should reject the first 37 candidates (when N=100 like in OP's example)

So do I reject the first 9 or 37 candidates?

source: [1] https://medium.com/galleys/optimal-stopping-45c54da6d8d0 [2] https://medium.com/intriguing-algorithms/thirty-seven-percen...

Reject 33 candidates if you want to maximize your likelihood of getting the best candidate out of 100, and 9 if you want to maximize how good the candidate you get is.

I wonder if any big company has ever implemented this in their hiring lol

Doesn't consider that what one wants in a spouse/partner can change over one's lifetime.

I have an algorithm: when choosing anything (shoes, cars, spouse) develop a criterion and search until you find someone(thing) that meets it. Then pass that one up, and look until you find one that is better. Choose that better one.

It increases your statistical chance of finding a 'good' on, with a pretty good bound on search time. And you feel like you did your diligence, so it limits 'buyer's remorse' to a manageable level.

Or it will leave one forever dissatisfied with the choice they made, because they passed on a someone they'd be genuinely happy with because of FOMO.

Shoes and spouses aren't in the same category.

I don't know, I've had some horrible FOMO for shoes. ;)

Being dissatisfied because of a surfeit of choice was where we started. Ok if this algorithm doesn't work for some, they're right where they started so no different?

I liked the Wikipedia write up for this a lot: https://en.m.wikipedia.org/wiki/Secretary_problem

I’ve found knowing the 1/e rule helps a lot with reducing stress when interviewing. It’s knowing that, if I haven’t randomly landed outside the first third, they should optimally say no to me, regardless of performance.

I read the article and came way confused. What's the 1/e rule? You mention something about "the first third", which implies you have a number in mind. How is that chosen? Is it just the available set of choices? For a spouse, wouldn't that be in the billions?

https://www.washingtonpost.com/news/wonk/wp/2016/02/16/when-...

Certainly it's not in the billions unless you spend an impressionable amount of time traveling. Realistically your personal total addressable market is certainly well less than 100k, probably less than 10k. You can subdivide further based on obvious preferences.

The whole thing is assuming you can only date/interview one person at a time. Taking time as finite, the question of N is more about how long it takes you to evaluate a candidate.

Brides.com says that most people date 1-3 years before proposing. Shortest I’ve ever broken up with someone is a few months so that should probably be shortened a bit. CDC.gov says the mean age or marriage is a touch under 30. If that is the first third, our dating age is about 20-50, which is probably a bit too wide if kids are on the table. So I’m going to say N is somewhere between 10-50, with the lower end of the range much more likely for.

Hannah Fry has a TED Talk covers this topic as well: https://www.youtube.com/watch?v=yFVXsjVdvmY

In this framing, it's based on time/age instead of number of candidates, so you don't need to know N.

This person is never going to be married.

They will be plagued for the rest of their life by their inability to discern the absolute value of their life partner, forever wondering if there is somebody 10% more compatible out there.

You say that as if this is a problem that isn't extremely common. Almost everyone in long term relationships stops to think at times if this was the right relationship for them to bind themselves to.

I did a variation on this in high school, where the wife (in this case) would also have to choose you. So it was at what point you stop searching if your partner is also playing the same game. It makes more sense imo since you can't just decide that someone will marry you.

All candidates are by definition, candidates. Someone who you might be willing but they aren't, is not in the set of candidates. IE they have no bearing on this at all.

I have a feeling this strategy will not be a great success with the ladies!

After all, the woman has the ability to choose, too. So the matter may remain rather academic.

Before posting a silly article like this, please consider the effect it might have on women in tech.

I found it to be very interesting from a statistical perspective. The language used does feel outdated, but I quite liked having a real world (tongue in cheek) example to make the math relatable.

Removing the genders (perhaps using the second person 'you'), and using 'partner' instead of wife may be the modern way, and would probably be just as relatable and compelling, while also being inclusive.

I'm wondering what negative effect this article could have for women in tech though?

My partner, a woman in tech, frequently tells me there is one type of person she feels holds her back professionally. It's the type of person that treats her like she is weak and needs to be coddled and protected from anything that could be perceived as threatening. I certainly don't have any real idea on what effect this article or your comment has, but I do think women are plenty capable of not only handling, but maybe even enjoying this article.

OK, suppose it gets flagged on HN and is removed. The article would still exist on the parent site and would (supposedly) keep harming women in tech. What then? Remove it from search results? Or even better, make it a crime to write wrongthink of this sort?

Reply on news.ycombinator.com