This is a topic which I haven't seen discussed extensively, but I always have this question in my mind when using a solver.
So basically my question is to what degree are strategies on an earlier street influenced by a finite number of bet and raise sizing options on later streets, and is this a concern while studying flop and/or Turn betsize strategies?
So for example we let the solver use three sizings on the flop (small, big and overbet) but limit the options on the turn/river to only two or one options to reduce complexity of the tree. If we have the nuts on the flop and we can use a small or big size, I suppose a big size is used more often if there is no option of very big bets on the turn and river.
Is this assumption correct and if yes, do we risk getting completey inaccurate results of our flop strategy if we simplify the tree too much?