The trying to lose doesn't apply. Indeed one way is to look at fewer scenarios, but it can also work with simpler versions of the valuation model, i.e. they don't necessarily give importance to certain factors that a more experienced player might take into account. Another way is to have a time limit for selecting a play, which directly forces you to look less and select a play that looks good but without taking into account longer sequences. And obviously it is also possible to combine these ways to simulate a less skilled player.