About the Anti-Spam Tool League Table

To understand the data presented in these tables read Understanding Spam Filter Accuracy. Briefly:

 
The Spam Hit Rate is the fraction of spams correctly identified
The Ham Strike Rate is the fraction of ham messages incorrectly identified

Both numbers are expressed as a decimal number to 4 places. For example, a Spam Hit Rate of .9911 means the filter successfully handled 99.11% of spams; a Ham Strike Rate of .0090 means that 0.9% of hams were incorrectly identified as spam.

Sorted by Spam Hit Rate Feed of ASTLT by Spam Hit Rate

Click to expand/hide data table.

Sorted by Ham Strike Rate Feed of ASTLT by Ham Strike Rate

Click to expand/hide data table.

Sorted by Review Feed of ASTLT by Review

Click to expand/hide data table.

FAQ

 
Where's POPFile?

All the data in the league table comes from published reviews of anti-spam software. I don't know of a published review of POPFile that contains enough data to calculate the Spam Hit Rate and Ham Strike Rate. I'm not going to add my own figures for POPFile because that would be biased. If you know of a review that covers POPFile please email it to me.

 

As a chart

The y-axis shows the Spam Hit Rate from .3000 to 1; the x-axis shows the Ham Strike Rate from .5000 to 0 (note that the x-axis is in descending order left to right). The best spam filters appear in the top-right hand corner.

Hover on any cross to find the filter name; click any cross to read the associated review.

.3
.4
.5
.6
.7
.8
.9
1
.5
.4
.3
.2
.1
0

As a chart (only "good" filters)

The y-axis shows the Spam Hit Rate from .9000 to 1; the x-axis shows the Ham Strike Rate from .0100 to 0 (note that the x-axis is in descending order left to right). This chart only shows spam filters that have good performance (i.e. filter greater than .9000 of spam and lose less than .0100 of ham).

Hover on any cross to find the filter name; click any cross to read the associated review.

.9
1
.1
0
Copyright (c) 1999-2008 John Graham-Cumming