In recent years the question whether relative gains matter to states or not has been the centre of the debate between neorealist and neo-institutionalists. In 1988, Grieco introduced the relative gains argument into (neo-institutional) game-theoretical analysis ("amended prisoner's dilema). Duncan Snidal took the model and worked out the analytical consequences stating that cooperation between two states is not as difficult as suggested by neorealists if two actors cooperate facing a (number of state(s) outside the cooperation. Snidal' s model is restricted to two actors cooperating His claims though were very general. As I have shown in my MA dissertation, Snidal's conclusions are wrongs for public goods in an n-actors case when states care about relative and absolute gain (which was snidal's and Grieco's starting point). The results of my MA dissertation suggest that the situation is more CO mplicated than realised thus far. The modelling and discussion has to be taken further. There arc several interesting situations that cannot be appropriately represented by Snidal' s model or mine. We have to think of big groups of cooperators whereby actors might have the option to join another sub-group of cooperators. We have to consider situations where all states of the system cooperate for excludable goods or for divisible goods etc. We have to work out the analytical arguments for all these situations for varying motivational assumptions (ie. only relative gains matter, only absolute gains matter, both matter) to achieve a typology of situations which we then examined empirically for example by using comparative statistics evaluating the analytical results against each other (to name a few prominent examples GATT, European Union, NATO).