4 ms·

I agree they are a flawed currency, I don't agree on your solution. Not only is 10 an arbitrary number, but it sounds like a lot less than some people would li

by eterm 13y ago

I agree they are a flawed currency, I don't agree on your solution. Not only is 10 an arbitrary number, but it sounds like a lot less than some people would like to give, and a lot more than most people will give any way. Why not have the best of both worlds by using a PageRank-alike system to determine the weight of any given like. That way, if someone is spamming likes they won't affect any one item very much compared to someone who likes only sparingly. You can go the whole hog and have the system where heavily liked users are more powerful than non-heavily liked users also.

r00fus 13y ago
So how would you weight them? If you assume the 10 likes/day "baseline", then would that indicate that if you moved above this average, your likes would be weighted down? Let's face it, a "like" is not a "currency" - it's a reward system that simply exists to virally promote the rubric from which it spawns - ie, likes are a metric for social network to show their investors user engagement and activity. That users somehow derive value or worth from getting likes is purely incidental. I mean, what's a "like" without a good link, message or photo to accompany it? Vague and useless.
na85 13y ago
I found the opposite: 10/day seemed way too high. I can't think of anyone other than my baby-boomer relatives who go around Liking things willy-nilly. I've probably Liked 15-20 things, total, since the Like feature was introduced.
robotmay 13y ago
I have considered more complex systems (i.e. dynamically altering the value of recommendations) and it's something that's likely to feature in the future. However I needed a simple solution for now whilst the site is developing, and I took a rough stab at it! I'm very likely to adjust the limit (if not alter the system in the future) based on how it's used, but I'm waiting for more data before I make that jump :)

by eterm 13y ago

r00fus 13y ago

So how would you weight them? If you assume the 10 likes/day "baseline", then would that indicate that if you moved above this average, your likes would be weighted down? Let's face it, a "like" is not a "currency" - it's a reward system that simply exists to virally promote the rubric from which it spawns - ie, likes are a metric for social network to show their investors user engagement and activity. That users somehow derive value or worth from getting likes is purely incidental. I mean, what's a "like" without a good link, message or photo to accompany it? Vague and useless.

na85 13y ago

I found the opposite: 10/day seemed way too high. I can't think of anyone other than my baby-boomer relatives who go around Liking things willy-nilly. I've probably Liked 15-20 things, total, since the Like feature was introduced.

robotmay 13y ago

I have considered more complex systems (i.e. dynamically altering the value of recommendations) and it's something that's likely to feature in the future. However I needed a simple solution for now whilst the site is developing, and I took a rough stab at it! I'm very likely to adjust the limit (if not alter the system in the future) based on how it's used, but I'm waiting for more data before I make that jump :)