I got an idea for a new algorithm that selects saves for FP;
first of all, it would only pick saves posted from the last week.
Then out of all the saves posted sooner than a week ago,
it would compare the rate at which people upvoted and rank them by fastest upvote rate.
I'm not sure about how to add downvotes in the mix, mabye like the average +1 per minute,
counting downvotes.
For example:
Save 1 gets 10 upvotes and 2 downvotes in 4 minutes = on average 2 upvotes a minute.
Save 2 12 upvotes and 1 downvote in 4 minutes also = on average 2.75 upvotes a minute.
Therefore:
Save 2 with 2.75 +1's a minute it ranked number 1, with the Save 1 at 2 +1's a minute ranked number 2.
Dissappearing after a week:
To prevent saves like Destroyable City Five (not that I have anything against it) from holding FP for ages.
UPDATE:
What about adding views into the mix where the amout of views is added to the average votes to calculate a comparision value? That would make it based mainly off of the amount of views a save got, with the votes having some effect.
It's votes/time, maybe put it back at 7 votes to get fp like it was in 2012, I remember the backlash of changing it to 3 votes or 4. It would decrease the chances of a copied save getting enough votes, plus a person who is asking for votes. They will still get votes unless reported right away.
The reason why that's still up is probably because so users can see TPT History.
I'll try think of a system that balances votes and views with necessary limitations.
Nevermind, the best idea I have is comparing the (upvotes - downvotes + views) / minutes of each save, with a one week limit.