To be honest I was hesitant when I placed that bullet there and maybe I should just take it out. Some time ago I had changed quite some email with the Google Analytics group about this issue, them stating that the 302 redirection (because it cuts off the referrer) would get you the self-referrer-problem. It didn’t make sense then and to think about it even more now it still makes no sense.

As for the cookie deletion I guess they can be guilty for self-referral problem but I would assume only isolated (unless complex situations like you describe). It’s a valid point though, yes.

Thanks for the comment.

In the slideshare-presentation you have a bullet-point under own domain in referral-reports, that says that 302-redirect might be a reason.

Is this right? I would mean that 302/301-redirects does the same thing, from a GA-tracking-POW, the server-message is just different.

I’m not sure though.

One thing I’ve noticed in the self-referral-problem, is that its often caused by the GA-cookies being reset or lost – this have been the case in many big websites, which had a lot of internal cookies and such. Browser limit of cookies has also been a reason.

Yes, I totally agree with you. Thanks for the heads up.
Untill now I was using filters to take out the IP of the development team so I didn’t have problems with this issue yet, but you are totally right. A simple filter should take care of the hostname issue which will keep you safe even from websites cloning content from your website together with your analytics tracking code.

This error can occur if someone is using your GA ID in their tracking code. This really messes your data up by inflating it

It happens more than you’d expect: when a client is running sub-domains for testing/development eg test.domain.com, using your GA ID, and you’re tracking domain.com!