4 ms·
Yes, this was my first thought. If I was seriously stressed for time, then I'd agree with idea of optimizing for overall percentages. But if you really want to
by msluyter 17y ago
Yes, this was my first thought. If I was seriously stressed for time, then I'd agree with idea of optimizing for overall percentages. But if you really want to optimize the chances that a specific person will reply to you, then it seems as if longer is better. Still, I'd like to compare a short, witty, well-crafted, personally tailored message to a longer, more-generic one. (Not that that could be measured statistically...)
- roundsquare 17y agoOn that note, it'd be interesting to somehow mix this data in with (self reported) happiness of people who end up together. Not sure how... maybe happiness (1 - 10) vs length of original message.