Working on it...
I have a few suggestions. Namely with the minimum ops to be qualified. Given, I like the idea of it being above 3, but 8 seems a bit too much. Now since the "ranking" requires only videos that have 8 or more ops to be qualified, I think that leaves less then 700 videos qualified before one even takes the top 10% of those.
It seems a bit harsh to cut the qualified videos from ~2,500 -> ~700. I'd suggest something around 5 or 6 opinions. I'd say that this would still take care of the problem that was rampant with the previous format of 3 ops. This would probably populate the top 10% with about ~100 videos which I think is pretty accurate compared to 50 - 60. I won't name videos, but I can think of plenty of great videos that aren't making the "cut" since the qualified videos "pool" is less than 700 and thus aren't within those Top 60 or so videos.
Oh and no, this post is not motivated by my own videos. Mine are from the top 10% currently
. I just think a lot of GREAT videos aren't on there and thus wouldn't get the exposure they deserve.
Another suggestion would be to go with the old format of 3 opinions to qualify and this new Baysian equation. That would definitely take care of the 3 op wonders because of the weighted rank (WR) = (v ÷ (v+m))
× R + (m ÷ (v+m)) × C in the equation. This would obviously have the same # of entries in the top 10% (~250 videos), but the weighted average will make it so only the good videos (with more opinions) would populate the list.
I guess what I'm getting at is that the # of qualified videos isn't that many and thus the "top 10%" is more like the "top 5%" since it's so picky. At the same time, I realize why it is since the other problem is 3 op wonders. Thus, I think lowering the op requirement back to 3 AND having this Baysian equation will kill both problems. This would lead to a good 200 or so videos that deserve to be listed as such.
I hope this makes sense. I guess it's really up to you Phade to see which one works out the best and I do appreciate your work.