Row Matching?

[Rayn]

Ok, this my sounds a little odd so I will tyr my best to describe it.

Now I have a forum, and it uses a post table that looks like

postid, threadid, username, userid, dateline, title, post

I want to be able to match up similar groups in rows so a example might look like

23, 4, Bob, 1, 1043813198, etc..
65, 4, Dan, 2, 1043814685, etc..

Now since we are on a forum you might see people that will post one after the other. I want to be able to get a output to see the highest amount two people do that so the final output would look like

Dan & Bob have 500 posts together

The big thing is that pre searching for names in the query cannot happen, I am only trying to find the top to people that post together..

I can takes a guess that threadid and time closeness are the ones that will make the matching.. I just dont know how to..

I hope any of this makes sence

thanks for your time
- Rayn

[tomhath]

Should be something like this, untested:

select p1.userid, p12.userid, count(distinct threadid)
from posts p1
inner join posts p2
on p1.threadid = p2.threadid
and p1.userid < p2.userid
group by p1.userid, p12.userid