There was a post this morning asking about how many people disable JavaScript. Then I began to wonder what techniques might be used to determine if the user has it disabled.
noscript blocks are executed when JavaScript is disabled, and are typically used to display alternative content to that you've generated in JavaScript, e.g.
Users without js will get the next_page
link - you can add parameters here so that you know on the next page whether they've come via a JS/non-JS link, or attempt to set a cookie via JS, the absence of which implies JS is disabled. Both of these examples are fairly trivial and open to manipulation, but you get the idea.
If you want a purely statistical idea of how many of your users have javascript disabled, you could do something like:
then check your access logs to see how many times this image has been hit. A slightly crude solution, but it'll give you a good idea percentage-wise for your user base.
The above approach (image tracking) won't work well for text-only browsers or those that don't support js at all, so if your userbase swings primarily towards that area, this mightn't be the best approach.