The javascript based trackers aren't going to follow visitors who have JS disabled. There are a number of privacy plugins etc. that block the google image/javascript, so I generally see slightly lower numbers from that tracker as compared to my actual server logs of pages served. That's going to be the gold standard to compare against, if you can access it from your host.
Looking at the posted image, I think the reason for the HUGE discrepancy is probably bot traffic. Bot traffic doesn't usually register on google (because bots don't bother to do anything but load the page, and don't parse javascript/tracking images), but will count in some tracking systems that use server logs.
Trust analytics to track your real human visitors.
