It depends on what aspect of game theory you're looking at. The basics don't distinguish between long term and short term. The primary purpose is to analyze decision-making reasoning based on incomplete information (such as not knowing what choice the other player is going to make). Then you have iterative versions where you repeat the same decision over and over against the same person. Then you have the evolutionary model, which looks at decisions among large populations over time, where you're not facing off against the same person over and over, but instead facing off against someone with an unknown strategy, and you need a strategy that will work against as many alternate strategies as possible.
Then you start adding in different challenges (eg: the Hawk/Dove problem, the Stag/Hare problem, etc), and add in meta analysis, and the overall challenge becomes very interesting indeed.
As an interesting extrapolation from the process, the basic conclusions reached from most game theory problems is that it would appear to largely support the amoralist viewpoint (cf: Carinthium), which of course suits the Incubators just fine. There's almost no support for any reason for anyone to behave in a 'moral' manner. Greed, then, isn't even a specific weakness; it's just the basic result of trying to get the best outcome.
However, if you take things a step or two further, combining evolutionary game theory with a social decision-making unit of arbitrary depth, you can see where morals must naturally evolve as a fundamental part of the system. The only way the Incubators can evolve without morals (or rather, to not develop beyond the most simplistic rational behavior) is because their hivemind removes all competing social units.