This is what the usage statistics (July 1-17) of a specialised, Pagerank 5 weblog on European politics look like after an attack of “referer spam“. Of the top 45 referers*, only 2 are genuine: google.co.uk and fistfulofeuros.net.
Some sites, especially blogs, publish a list of referring websites, and by targeting thousands of websites and blogs, the spammer hopes to get visibility on at least a few of them.
Note: in the contect of a referring web page, the misspelled “referer” is often used instead of “referrer”.
Reports based on the webserver logs, such as produced by Webalizer or AWstats, become harder to interpret. Referer spam is not the only issue: the number of machines hitting your webserver is growing faster than the number of human visitors: the multitude of feed-polling rss clients and rss-indexers, search bots/spiders and even shameless content harvesters.
A good log analyzer will let you distinguish human “visitors” (producing a click trail from consecutive page views or at least the succession of page load and image loads) from isolated page views (that’s what most bona fide search bots produce). You also might enter the arms race with referer spammers by blacklisting them in your log configuration.