Here's my version (script and output included):
https://gist.github.com/bac2da6d69b841eec92a
the base score is the same formula that BoltEyeAm came up with, but just looking at the last 60 days of activity:
Code:
( totalPost(w)(y)+thanksReceived(x)(y)+thanksSent(z) ) / ( daysMember < 60 ? daysMember : 60 )
then to try to make it a little more interesting, I calculate the score of each thanks received based on the base score of the user who sent the thanks. This way thanks that come from users with a higher base score are worth more than those with lower.
Code:
thanksReceivedBonus += thankersBaseScore * z
then I use the same calculations as the base score, adding the thanksReceivedBonus to thanksReceived to get the final acf score:
Code:
( totalPost(w)(y)+(thanksReceived + thanksReceivedBonus)(x)(y)+thanksSent(z) ) / ( daysMember < 60 ? daysMember : 60 )
I take no credit in any of this.. BoltEyeAm, RedHerby, Kradek, and others came up with this fun idea and formulas. Having said that, if any models would like a higher score, we can get on skype and work something out...
I'll implement this into the site sometime this weekend along with a top 100 page.