While I had created this topic I have mentioned I have tried to find out but could not find out calculation system. So, I do not need to know how to search as I already know. I need to know how to calculate. If you know then you can reply. If you don't know then no problem.
This should do:
The trust score numbers are now slightly different:
- The first number is the trust score.
- The second number is the number of unique users who have given that person negative feedback.
- The third number is the number of unique users who have given that person positive feedback.
- The fourth number was removed.
I also completely changed the trust score algorithm to this:
if there are no negative ratings
score = 0
for each rating, oldest to newest
if this rater has already been counted
continue
score += min(10, round_up(months since rating))
else
score = unique_positive - 2^(unique_negative)
if score >= 0
start_time = time of first negative
score = unique_positive since start_time - unique_negative since start_time
if(score < 0)
return ??? (orange)
move score to range [-9999,9999]
return score
This algorithm is a little slower than the previous one. Post here if you think you see extra slowness due to this change. Maybe I need to add extra caching to compensate.
Also post here if someone has a trust score that seems wrong.
I was going to change it so that everyone with 0 trust had orange trust, but I decided that this looked bad and changed it back.
Trust system is a little complicated to understand first, so you're not the only one who thinks that. There were different versions of trust system, and different types of it too. I believe these were the ones:
- Scammer(alert) system(?)
- Old Scammer Tag
- Default Trust part 1(before the algorithm change).
- Default Trust part 2
