So a solver more or less calculates the optimal way to play hands given a limited subset of the game tree (only a small sample of the available bet-sizes, as th

So a solver more or less calculates the optimal way to play hands given a limited subset of the game tree (only a small sample of the available bet-sizes, as there are an unlimited amount given the game is no-limit) given the opponents' perceived "range" (basket of hands they could have up to this point) and performs in such a way that a player would not be exploitable (e.g. game-theory optimal). In theory this should guarantee that a player would not lose money in the long run, but that also doesn't imply that it would make the player the most money in the long run either. The very best players at the high-stakes nowadays are studying these solver outputs as an abstract theoretical framework and deciding in-game when and by how much to deviate from the solutions.