Agree...there are some problems with this model as well. It
presupposes plans of action that have not been carried out. Nash
Equilibrium is based on the assumption that two or more actors
should not deviate from their strategy, so as to gain optimal
payoffs.
The strategies for both opposition and ruling party are different:
The PAP would wish to fully dominate parliament (i.e. no
opposition), while the opposition has to increase or at least
maintain its current representation in parliament. As is, the odds
are already heavily stacked against them.
Taking your assumption that the Yes/No represents decision, the
outcome for PAP = Yes, and Opposition = No, the outcome will still
invariably lead to stronger PAP dominance, notwithstanding what
little the opposition can actually do about it. (I refer to the top
right box)
A policy of no-contest would mean that the status quo remains. This
remains questionable...Potong Pasir remains under opposition, but
just barely...
Perhaps what we need to employ in place of this is a series of
mixed game strategies...
Nash Equilibrium...sometimes referred to as Prisoner's Dilemma, is
not feasible in this case, since the results would still favour the
dominant group. Further, the chances that one side would 'defect'
are very high in this instance.
My 2 cents.. 
