Ask HN: How are you detecting proxy's and fake users?

Using WebRTC to detect IP addresses leaked by proxy's seems to work reasonably well. It isn't 100% reliable. Using REMOTE_ADDR isn't going to get you anything of value.

With regards to fake users, access statistics cam be effective. Yet, this has to be used with care and it is highly dependent on the nature of your service. If the nature of your service allows it, a verified credit card account through a small charge (and charge amount verification) can work well.

If your service can justify it, there can be explicit ID verification. For example, we have a site where we charge users $99 per year for verification.

How are you handling this, if at all?