In Spamfire 1.x, the spam score of a message was a cumulative total of the filter scores for that message. Multiple filters could combine to create spam scores less than zero and higher than 100.
In Spamfire 2, the score is computed by the new Bayesian filter and can never be less than 1 or more than 99. Learn more about How Spam Scores Work in Version 2.
There are still Rule Filters in Spamfire 2, but they do not affect the Spam Score. See How Rule Filters Work in Version 2.
In Spamfire 1.x, you could set a score threshhold. Everything below you threshhold would be considered good and everything above your threshhold would be considered spam. This feature is no longer available in Spamfire 2.0. Messages with a score between 1 and 50 are always considered good. Messages with a score between 51 and 99 are always considered spam. However, as you correct mistakes that Spamfire makes, the new Bayesian filters will automatically adjust future scores so there is no need to change the threshhold.
In Spamfire 1.x, identical messages always get the same spam score. In Spamfire 2.0, each message re-inforces the Bayesian filter and causes the spam scores to get more extreme. The first time you receive a particular spam message, it may appear as a borderline message with a score close to 50. But if you mark the message as spam, future messages from that spammer will be recognized and will receive higher scores. The longer you use Spamfire, you will see fewer Borderline messages and more Definite messages. See the icon legend for more about what the different scores mean.