3 min read

Online matching platforms are best conceptualized as friend-finders rather than date-finders. Given how difficult it is to find compatible people in the world, it is irrational to rule out the possibility of friendship (or activity buddy, or professional connection) being the optimal path with a cool person you meet online.

Let’s say that, for a particular guy I’ve just met and like, my assessments of the future are:

- Probability of a long-term romantic relationship success = 33%
- Probability of remaining friends after a short-term dating fling = 25%
- Value to me of a friendship with him = 80 (on a scale of 1-100)
- Value to me of short-term dating with no ongoing friendship = 30
- Value to me of a long-term romantic relationship/partnership with him = 100

Let’s visualize my assumed probabilities and values of potential outcomes as a decision tree *[click on image** to open larger in a new tab]*:

If I ignore risk (i.e., probability of success), I might jump to the conclusion that dating is a “better” strategy than friendship. So, let’s calculate the expected values (i.e., probability % * outcome value) of my two strategy alternatives:

Expected value of dating is calculated as 36. Expected value of friendship was given by me as 80.

Therefore, if my preference is **to maximize expected value, I should befriend this guy instead of dating him**. This would be the risk-neutral, probability-weighted, “rational” strategy.

However, if I prefer a strategy of maximizing potential value, I should date him. I have a 50% chance of extracting more value from the connection by dating him than by friending him. I could end up with a high value of 110 (16.8% chance) or 100 (33% chance)….or a low value of only 30 (50.3% chance). This is a risk-seeking, high-Beta gamble for me.

Risk tolerance and decision optimization approach (i.e., strategy preference) is specific to each person in each circumstance.

Now let’s say the guy in question has done his own personal decision analysis as well (by filling out the interactive calculator below!), and he decides his preferred strategy is dating me. He may value friendship less than I do — perhaps because he has lived in this city for a long time and already has plenty of friends. Or, he may think I’m the best thing since sliced bread; so, he sees immensely more value in a romantic partnership compared to a friendship, and he’s perhaps also pretty confident that it would work out between us.

He can’t do much about my decision criterion of preferring to maximize expected value. But, he could still convince me to date him if he provides information that updates my initial assessments of the future. For example, as I spend time with him, I might estimate a higher likelihood of long-term romantic relationship success. As I meet his ex-girlfriends with whom he is still friends, I might perceive an increased probability of maintaining a friendship after dating him (thus limiting the downside risk of getting romantically involved). As I realize he’s an upstanding feminist gentleman, I might update my valuation of a short-term fling. *[See my essay “Game theory of hookups” for detail on how critical a man’s demonstrated integrity is for increasing his appeal to women for short-term intimacy.]* Most importantly, as I get to know him and experience his appealing qualities, I might see him in a different light and come to value a long-term romantic outcome with him more than I do initially. In other words, the more he starts to seem worth the risk and/or represents a lower risk, the more likely I’d want to take the risk. And, of course, this all applies equally to both parties considering the friendship-versus-dating question.

Input your own assumptions in this interactive calculator!:

[forthcoming]

*Input your assumptions:*

- Probability of long-term romantic relationship = __33__ %
- Probability of friendship after short-term dating = __25___ %
- Value of friendship = __80___ [1-100 points]
- Value of short-term dating = __30___ [1-100 points]
- Value of long-term romantic relationship = __100___ [1-100 points]

*Calculated results and recommendations:*

- Incremental value of long-term over short-term romantic connection = __70__
- Expected value of dating = __36___
- Maximum potential value among all theoretical outcomes = __110___

**If you prefer a strategy of maximizing***expected*value, then choose ___friendship_______**If you prefer a strategy of maximizing***potential*value, then choose ___dating_______- Probability that potential-value-maximizing strategy yields better outcome than expected-value-maximizing strategy = __50___ %