version 0.14:
	symmetric-NE
	general (not necessarily symmetric) NE
	one-shot game

version 0.13: 
   TODO: 
    Vector/array arguments for parameters a, b, initial-window
    New "OptimizeA" method to find the best response of a player
    May be a new class "State" that contains the current a and b...

version 0.12:
    Capacity Variation: Variable percentage and the capacity changes at most +-1 in each round.
    Solved Prince issue: Selecting the scapegoat when > 1 flows have the same maximum window size.
    
16 Parameters:
	0:  ExperimentType (EnumExperimentType)
	1:	BlendNum (Number of repetitions) (value ignored in one-game experiments)
	2:  GameId (base number for all game ids of this experiment)
	3:  Duration
	4:  Pregame Rounds
	5:  Capacity
	6:  VariableCapacity
	7:  Penalty (EnumFlowPenalty)
	8:  QueuePolicyType (EnumQueuePolicy)
	9:  nFlows (Number of Flows)
	10: FlowPolicyType (EnumFlowController)
	11: a (for the n-1 flows)
	12: b (for the n-1 flows)
	13: a (for deviator flow) (value ignored in experiments "for all a")
	14: b (for deviator flow) (value ignored in experiments "for all b")
	15: seed for random number generator
	