> The fact that Tragedy of the Commons is a subset of game theory defection problems does not change the fact that its defining requirement is the existence of a "Commons."

The fact that the term has long been widely accepted as a label for the general class of problems, OTOH, does.

(Arguably, the structure of the payoff matrix in the class of problems implies an economically real if intangible “commons” in any problem in the class, which is probably why the label is used that way.)

Ok, what is the Commons he is talking about?

Presumably the 'Commons' here is control over the government. However, the example is a trivialization of reality and that matters, in the same way that assuming all economic actors are rational matters (they aren't). It's fun when we can use simple laws of nature to make predictions accurately. Unfortunately, it doesn't work in the field of economics\politics almost ever.