Reputation Calculation Algorithm

Sat. March 13, 2010 - 50% Discount Special begins.
All ratings below are discounted 50%, making everything better.

The CryoNet code attempts to compute the likelihood of spam from a particular poster by calculating the typical percentage of ratings in each category ("Offtopic", "Interesting", etc.) for his/her postings, multiplying those values by the "spaminess" weight for each category, and adding the results. The current "spaminess" weights are:

Boring => 0.6    Excellent => 0.05    Flamebait => 0.8
Funny => 0.4    Good => 0.15    Informative => 0.1
Insightful => 0.1    Interesting => 0.1    Normal => 0.3
Offtopic => 0.9    Poor => 0.75    Redundant => 0.65
Poor_Subject_Line => 0.4

Additionally, there is an "Abuse" score of 1.75, which exceeds any possible probability, but is useful in situations for which a score greater than 0.9 is needed to distinguish egregious cases.

Thus, a person whose posts always rate 100% "Informative" will have a probability of spam of 0.1 whereas someone whose posts always rate 100% "Flamebait" will have a probability of spam of 0.8. The reputation is just 1 minus the probability of spam. The "Informative" poster thus has a reputation of 0.9 whereas the "Flamebait" poster has a reputation of 0.2. Here are the consequences of a poor reputation:

An easy way to read (and rate) all messages for a digest is to click on the link for rating the entire digest, which is included before the first message of a digest. When the messages are accessed that way, it is easy to read all the messages, including those that were removed entirely from the digest email.

It is unfortunate that such measures need to be taken, but it would be even worse to let a few individuals abuse CryoNet without constraint.