In <p04330100ba630d2479fa@[203.79.72.8]>, Owen Watson <[EMAIL PROTECTED]> writes: > >The characterisation is accurate, and regardless of whether you use > >the free or > >the commercial databases, you will see false mappings (which largely rely on > >netblock delegations, and do not take into account that larger > >companies and ISPs > > I'm no expert, but isn't this the same problem that afflicts analysis > of web logs for any data? You always have to explain to others about > proxies etc. What you'll always get is a guide, not definitive > answers.
Absolutely. As long as the users of your statistics understand that, fine. The more granular data you provide, the more likely users will be led to believe that the information is very accurate. Which of the statements below do you think people are more likely to trust? * About half of the visitors to our site are from outside the US. * 2.71% of the visitors to our site are from New York City. I suspect statement 2 looks a lot more reliable and trustworthy to many people, regardless of how accurate the data collection mechanism is (could just be a single ISP located in NYC that is responsible for a large portion of traffic). Let me ask the question differently: what are you interested in, where your site visitors reside, or where their internet access point resides. GeoIP and other similar databases can give you the location of the internet access point, and as long as that is sufficiently close for you the mapping will provide useful information. If you want to know more about your visitors, the only alternative is asking them (and yes, you will have some false information that way, too :-)) -- Klaus Johannes Rusch [EMAIL PROTECTED] http://www.atmedia.net/KlausRusch/ +------------------------------------------------------------------------ | TO UNSUBSCRIBE from this list: | http://lists.isite.net/listgate/analog-help/unsubscribe.html | | Digest version: http://lists.isite.net/listgate/analog-help-digest/ | Usenet version: news://news.gmane.org/gmane.comp.web.analog.general | List archives: http://www.analog.cx/docs/mailing.html#listarchives +------------------------------------------------------------------------
