=> 1 merit source has to look to an average of 106 new posts/day or 3,255 new posts/month in order to make the system works as intended.
This assumption is wrong. At least half the posts aren't worth reading at all (check
recent unread posts and see how often the word "bounty" is in the title).
That leaves less than 53 posts per day to read by your math. That's not much at all for an active member.
Update: I forgot the overlap in sources reading the same posts as other sources. Let me rephrase it: I think 90% of all new posts isn't worth reading, and I'll
prove it provide empirical evidence by showing the 10 newest (English) posts from the unread posts link:
1,
2,
3,
4,
5,
6,
7,
8,
9,
10Since sources are humans obviously i doubt they can (not all of them maybe) give rewards properly to everyone deserving it.
I don't think this is a problem. Suppose someone who makes 100 good posts only receives merit on 10 of them. On average, that will be the same for everyone.
Proposition :
It's much easier to add more merit sources (
hint hint).
Why not just add that depth 2 sources ( if your system happened) as merit sources? why do we need to differ them? in my opinion it is better if there are more merit sources rather than making this system more complex.
Exactly! Theymos said this:
By no means is the set of merit sources or their sMerit/month static. I will be adjusting it a lot as needed, especially in the beginning. I wouldn't be surprised if there are 100-200 sources in a year from now.