What exactly is being counted/compared?
The datum extracted for each newsgroup is the number of articles accepted since the last sample.
To gather this data, no header or overview information is collected (that would require too much overhead.) The high article number is just taken, which enables larger sample sets and low impact connections. (Note that it is possible (and often very useful!) to compare content in a newsgroup, article for article. See What other services and reports are available?)
Servers which don't assign article numbers in order of lowest unused article number cannot be sampled reliably. These servers are very rare.
The reported numbers are simply the number of articles accepted.
For example:
-----Daily average arrivals for----
last last last last
1 day 2 days 3 days 7 days
------ ------ ------ ------
45 groups in comp.*
Expected: 3201 2942 3255 3546
Actual: 3061 2951 3220 3538
Percent: 96% 100% 99% 100%
This means the number of articles accepted for "Actual" in 45 newsgroups matching comp.* are:
- 3061 articles in the last 1 day.
- In the last 2 days, there were 2951 * 2 articles in those newsgroups. (The number
reported is the average of the last two days.)
- In the last 3 days, there were 3220 * 3 articles in those newsgroups.
- In the last 7 days there were 3538 * 7 articles in those newsgroups.
View an example report
How are the expected values derived?
Doesn't crossposting affect the counts?
Why doesn't spam make meaningful comparisons impossible?
What if some sample newsgroups are not on the server?
What is considered "excellent" or "good" or "poor" performance?
Can you expound a little on how the GPA is derived?
Newsrate Report Format: What does "n/a" mean?
How do I signup?
Up to The newsrAte RKT
Up to newsrAte home
Up to Mib Software home
Copyright 1998, Forrest J. Cavalier III, Mib Software
INN customization and consulting
Email: newsrate@mibsoftware.com
Voice: 570-992-8824