> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Stephen Turner
> Sent: 25 April 2001 17:20
> To: [EMAIL PROTECTED]
> Subject: Re: [analog-help] Hosts vs Visitors
>
>
> On Wed, 25 Apr 2001, Iain Hunneybell wrote:
>
> > I know and understand the hosts vs. visitors arguement well and
> the analog
> > 'How the web works' document is excellent at explaining such issues to
> > others keen for 'visitor counts', however...
> >
> > I've seen other discussion of analog vs. WebTrends and I've
> been doing some
> > work to validate WT results with analog (I've even got close answers!).
> > However, looking at the 'Distinct Hosts Served' figures vs. the
> equivalent
> > WT 'Visits' and 'Unique Visitors' counts, I find analog gives a
> > significantly (60%+) higher count than WT does for 'Unique Visitors'.
> >
> > This surprises me. I'd have thought that a strict 'Hosts count'
> - which I
> > presume means unique IP - would give a _lower_ count than a WT
> 'visitors'
> > count as WT might count as two or more 'visitors' the same IP
> if separated
> > by a significant amount of time (like >30 mins).
> >
> > Anyone any ideas on why an analog 'Hosts count' should come out
> higher than
> > a WT 'visitors' count?
> >
>
> Look for discussion of "AOL" in docs/webworks.html. It's possible for one
> person to send requests from many hosts, as well as many people from one
> host.
Yeah I read the 'webworks' document and thought it an excellent explaination
:-) I've passed it on to others with whom it's sometimes a struggle to
explain these concepts. It's a very useful doc and I wasn't aware of the AOL
behaviour, but...
The issue I've got is the WT count is _less_ than the analog count and my
expectation would be for it to be higher. If the issue was AOL user sessions
coming through as a high number of different IPs (as explained in
'webworks') then I'd expect to see a much higher WT count as (say) 10 AOL
users will generate (say) 200 different IPs (unique hosts) and with WT then
counting use of an IP after (say) 30 mins as being a new visitor it ought to
be counting a much higher number of visitors. analog would simply count '200
distinct hosts served', but WT (I'd expect) would see that as maybe '1000
visitors' as those 200 IPs get shuffled and reused by different users over
time.
>
> Of course, I can't say if that accounts for it all. You might like to test
> them both with small invented logfiles to see if they're both
> reporting what
> you think they should be.
What do you mean by inverted?? Reversed time sequence?
Many thanks again for your help :-)
.../Iain
>
> --
> Stephen Turner http://www.statslab.cam.ac.uk/~sret1/
> Statistical Laboratory, Wilberforce Road, Cambridge, CB3 0WB, England
> "Your account can only be used for a single internet session at any one
> time and for no more than 24 hours in any one day." (NTL terms of use)
>
> +------------------------------------------------------------------------
> | This is the analog-help mailing list. To unsubscribe from this
> | mailing list, go to
> | http://lists.isite.net/listgate/analog-help/unsubscribe.html
> |
> | List archives are available at
> | http://www.mail-archive.com/analog-help@lists.isite.net/
> | http://lists.isite.net/listgate/analog-help/archives/
> | http://www.tallylist.com/archives/index.cfm/mlist.7
> +------------------------------------------------------------------------
>
+------------------------------------------------------------------------
| This is the analog-help mailing list. To unsubscribe from this
| mailing list, go to
| http://lists.isite.net/listgate/analog-help/unsubscribe.html
|
| List archives are available at
| http://www.mail-archive.com/analog-help@lists.isite.net/
| http://lists.isite.net/listgate/analog-help/archives/
| http://www.tallylist.com/archives/index.cfm/mlist.7
+------------------------------------------------------------------------
