lanesawyer / gshe

The Ginormous Stackrank of Human Experience
MIT License
3 stars 0 forks source link

Allow users to compare their personal stackrank to the stackrank of others #11

Open lanesawyer opened 9 years ago

Stackranker commented 9 years ago

What this "similarity score" might look like:

A) Identify all the experiences that both Phil and Jane have both rated in A vs. B pairings at least 10 times B) Subtract Phil's ranking for experience X from Jane's ranking for experience X (e.g. Taking a nap for Phil is 1,209 and 1,119 for Jane = 90) C) Sum the absolute value of all the differences (e.g. |-100| + |90| + |-20| = 210) D) Divide that result by the number experiences from A

Repeat for a given number of potential comparison users. To be eligible for comparison, set a threshold for A (e.g. they must both have at least 100 experiences that both have rated in A vs. B pairings at least 10 times).

Then, display the results (e.g. out of pool of 10 compared folks, the 2 people with the highest, and the 2 people with the smallest, similarity ratio). The smaller the ratio, the more similar the other user.

(We could play with the score to make it more easily understandable- e.g. invert such that the higher the ratio, the more similar the other person. We could also divide or multiple by 1000 or whatnot so that the scores vary between a familiar range such as 0 and 10.)

Enable selection of a comparison user (e.g. a friend they want to compare to).

Michal: think about how to make this feature attractive to the user so they understand why the result is meaningful. Add interpretation.

Might help go viral - they introduce a friend to see how similar they are.

Might have application for dating similarity/compatibility. Or looking for a travel companion. Or a companion to explore food. or adventure.