Serverside detection of full or partial hits to block spammers?

1 page + 3 images = 4 hits, 1 request = spammer...

[JAB Creations]

I've seen a few common UAs fail to request files I know a normal human being would download (without knowing it such as a script) when they go to certain pages.

I'm wondering if there is a method one could implement to detect if an IP fails to request linked files yet goes on requesting other pages.

I'd like to ultimately block users (who claim a false UA...working on a script to cover that aspect) and fail to request linked files to be added to a IP blacklist.

[2by4]

Sounds pretty much impossible, if I visit your website, with a standard browser, after the first visit, all your images, css etc are cached in my browser. Unless I have a static IP, which few people have, my next visit will be from a different IP, but I won't be requesting any of the cached material, just the html page.

A more profitable activity is to search WebmasterWorld for php bad bot scripts, there's a few threads on that, those work quite well.

Serious bots tend to work from rotating ip addresses anyway, the bot masters are way ahead of you in this game, you have to catch them in the act, which means catch them trying to access pages blocked in robots.txt, but then all they have to do is have their bots obey robots.txt, tough game to win long term.

[Dijkgraaf]

Have a look at [danielwebb.us ].

I think I'm allowed to post this URL here, if not do a search for Bot trap and you will find other examples.