How many people see their ideal match and get humbled (or emboldened) by surprising results? How many of these try immediately to change something about themselves?
Since the match pool will never be exactly 1:1, do you do anything at all about the least matchable people?
Would it become possible to gauge whether changed answers are efforts to widen one's net or legitimate personal growth and/or a bit of self-reflection?
If the dating service detects deceptive answering, is it correct to just pair these people with other dishonest users?
Can you gather any feedback from unsuccessful dates and apply a "Je n'sais quoi" Factor to positively/negatively weight their chances when calculating their compatibility with future matches?
If you take this idea to the extreme, can you start to distrust negative date feedback if the feedback provider is shown to be unreliable? (Person A doesn't like anybody, their feedback shouldn't matter.)
Are there detectable signals of someone's Je ne sais quoi Factor in their profile metadata, including profile photo(s) and answering habits? (i.e. People answering the full survey in less than 15 minutes at 3 a.m. are 28% less desireable on average.)
While it's good to grow together with a partner, it doesn't always work out that way. How would such an algorithm be able to determine that an ambitious (broadly: likely to dramatically change) person will not be paired with a long-term apathetic, set-in-stone person? (I do realize they have the, "always working on yourself?" question, but there's so much this question leaves unanswered.)
Is this a ploy to get more applicants to join the government staff, if the pilot remains closed to the public?
(tying together "least matchable people" and "government staff") Do you start monitoring the extreme ends of matchability to see if there's a link between perceived social agreeableness and job performance? What was that again about detecting dishonesty and limiting the pool to government workers?
-----
One detail I particularly liked about early OKCupid was the ability to rank how much someone's answer mattered. For instance, I don't particularly mind someone's exercise and organization preferences, and it's fine if we didn't match because they DO have expectations from me in these areas which I fail to meet. I would hope the Singapore service would implement this as well, although a description of FirstDate in The Straits Times suggests this nuance isn't there.
Another thing the Singapore dating service has in common with OKCupid is at least a few poorly worded questions: "What’s a country that’s ruined all other countries for you?" is akin to OKC's "In a certain light, wouldn't nuclear war be exciting?" You can't split hairs as a Yes/No response in OKC, "sure: I would understandably have tons of negative excitement, a sudden spike in anxiety/dread, instantaneous long-term shift in priorities, etc.; although the use of atomic weapons thoroughly horrifies me," just like you can't explain whether the country that "ruined others" did so via greatness (Spain with its social and expressive population, food, beaches, siestas, and music) or by causing a loss of faith in humanity (Ireland, due to a stolen laptop plus being verbally threatened by a primary school-aged child while I was walking with a baby stroller).
My personal take on the nuclear war question? I just tried to optimize (cast the widest net) and decided, "Anyone willing to misunderstand this one question while also prioritizing and demanding its response is less likely to be someone whose company I value," and left the question unanswered + unimportant to me.
Now if only my local government had a "friends group" service, so I don't end up hanging out with other parents just by virtue of also being a parent and needing to be around them in the same places at the same times...
PennRobotics · · focus · HN ↗
How many people see their ideal match and get humbled (or emboldened) by surprising results? How many of these try immediately to change something about themselves?
Since the match pool will never be exactly 1:1, do you do anything at all about the least matchable people?
Would it become possible to gauge whether changed answers are efforts to widen one's net or legitimate personal growth and/or a bit of self-reflection?
If the dating service detects deceptive answering, is it correct to just pair these people with other dishonest users?
Can you gather any feedback from unsuccessful dates and apply a "Je n'sais quoi" Factor to positively/negatively weight their chances when calculating their compatibility with future matches?
If you take this idea to the extreme, can you start to distrust negative date feedback if the feedback provider is shown to be unreliable? (Person A doesn't like anybody, their feedback shouldn't matter.)
Are there detectable signals of someone's Je ne sais quoi Factor in their profile metadata, including profile photo(s) and answering habits? (i.e. People answering the full survey in less than 15 minutes at 3 a.m. are 28% less desireable on average.)
While it's good to grow together with a partner, it doesn't always work out that way. How would such an algorithm be able to determine that an ambitious (broadly: likely to dramatically change) person will not be paired with a long-term apathetic, set-in-stone person? (I do realize they have the, "always working on yourself?" question, but there's so much this question leaves unanswered.)
Is this a ploy to get more applicants to join the government staff, if the pilot remains closed to the public?
(tying together "least matchable people" and "government staff") Do you start monitoring the extreme ends of matchability to see if there's a link between perceived social agreeableness and job performance? What was that again about detecting dishonesty and limiting the pool to government workers?
-----
One detail I particularly liked about early OKCupid was the ability to rank how much someone's answer mattered. For instance, I don't particularly mind someone's exercise and organization preferences, and it's fine if we didn't match because they DO have expectations from me in these areas which I fail to meet. I would hope the Singapore service would implement this as well, although a description of FirstDate in The Straits Times suggests this nuance isn't there.
Another thing the Singapore dating service has in common with OKCupid is at least a few poorly worded questions: "What’s a country that’s ruined all other countries for you?" is akin to OKC's "In a certain light, wouldn't nuclear war be exciting?" You can't split hairs as a Yes/No response in OKC, "sure: I would understandably have tons of negative excitement, a sudden spike in anxiety/dread, instantaneous long-term shift in priorities, etc.; although the use of atomic weapons thoroughly horrifies me," just like you can't explain whether the country that "ruined others" did so via greatness (Spain with its social and expressive population, food, beaches, siestas, and music) or by causing a loss of faith in humanity (Ireland, due to a stolen laptop plus being verbally threatened by a primary school-aged child while I was walking with a baby stroller).
My personal take on the nuclear war question? I just tried to optimize (cast the widest net) and decided, "Anyone willing to misunderstand this one question while also prioritizing and demanding its response is less likely to be someone whose company I value," and left the question unanswered + unimportant to me.
Now if only my local government had a "friends group" service, so I don't end up hanging out with other parents just by virtue of also being a parent and needing to be around them in the same places at the same times...