24 ms·

How about a dating app that explicitly uses Optimal Stopping, the solution to the Marriage Problem/Secretary Problem: > the best way to proceed is to interview

How about a dating app that explicitly uses Optimal Stopping, the solution to the Marriage Problem/Secretary Problem: > the best way to proceed is to interview (or date) the first 36.8 percent of the candidates. Don't hire (or marry) any of them, but as soon as you meet a candidate who's better than the best of that first group — that's the one you choose! https://www.npr.org/sections/krulwich/2014/05/15/312537965/how-to-marry-the-right-girl-a-mathematical-solution https://www.npr.org/sections/krulwich/2014/05/15/312537965/h... The question becomes how many more dates or relationships will you have in your life? 36.8% of what? More importantly, will my soulmate also be on that app? Better chance to find them on Tinder.

Sebb767 5y ago
> More importantly, will my soulmate also be on that app? Better chance to find them on Tinder. You could actually employ that strategy on Tinder, no need for a separate app. Allthough looking for a soulmate using optimal stopping does not seem like a good match, anyways.
nsomani 5y ago
Hmm yeah two points on that What exactly qualifies as a rejection (is it a swipe left or a match + date + unmatch)? Thinking match + date + unmatch is the more reasonable application but difficult to enforce Second I just don't agree that it's the right optimization function, IMO you should max E[quality of partner] rather than max Pr[ending up with best quality partner]