Ignore spiders in page count ...

//grab user agent
$user_agent = (!empty( $_SERVER['HTTP_USER_AGENT'])) ? ↵
$_SERVER['HTTP_USER_AGENT'] : "";        

//check for crawler
$is_a_crawler = preg_match("/(?:bot\b|spider|slurp|crawler|↵
ultraseek|webcopier|\bwget\b|asterias)/i",$user_agent);

if($user_agent && !$is_a_crawler) 
{
	//SQL (pseudo code)
	//-----------------
	//UPDATE _Table_ 
	//SET count=count+1 
	//WHERE ID=ID-OF-PAGE
}

↵ denotes wrapping

The very simple explanation of the PHP...

1. grab user agent
2. check to see if agent is "spiderlike"
3. increment count variable

It's probably not perfect ... but will you closer to TRUER numbers.

Source file: pagecount-nospider.txt