VPN: performance rules

One of my VPN providers has a nifty feature to constantly monitor its network and select the best exit port for my device based on actual network latency. In a way, this is hyper-capitalism in real-time.

The idea is that network latency – how long it takes for data to move across portions of the internet – is the primary measure of value. The higher the latency, the lower the value.

But just to make it more extreme, the program on any given device monitors its latency in transiting to any of the nodes operated by this provider and switches to the fastest on-the-fly. This is the default configuration, but you can instead select specific nodes.

Sometimes the auto-select mode results in non-intuitive connections. The ‘closest’ nodes are regularly not the fastest; Silicon Valley in California is oddly often the fastest place to connect in the world, even though it is usually upwards 1000s of km away. Not that I would normally select any node within the USA, but I do use the default settings rather often.

Because, in fact, speed really is the most important factor to me most of the time, if I can assume there is reasonable anonymity at the exit node.