alt Hacker NewsA failed cheese usually leaves the attacker so weak that the game is already lost. Cheeses are intended to win the game.
But if you are known to never cheese, your opponent might bet on greedy strategies, sometimes known as "economic cheese": you don't prepare any defense, and skip scouting, to build an overwhelming army all of a sudden at some given time like just after an important couple of upgrades that boost the army (a timing attack).
The "normal" play (economic growth plus scouting) is usually the superior strat, but if your scouting fails to detect a cheese attempt that must be countered with a very specific defense, the game is lost. The occasional cheese keep the players honest so they spend resources in scouting, instead of going greedy.
StarCraft has its own bluffing scheme, that is faking a build so the opponent goes for a specific counter, but actually going for something else.
Replies
This is a good example of a game-theoretic equilibrium :)
> A failed cheese usually leaves the attacker so weak that the game is already lost. Cheeses are intended to win the game.
> But if you are known to never cheese, your opponent might bet on greedy strategies, sometimes known as "economic cheese": you don't prepare any defense, and skip scouting, to build an overwhelming army all of a sudden at some given time like just after an important couple of upgrades that boost the army (a timing attack).
Yes, that's exactly what I was trying to say.
And the optimal cheese frequency is when cheesing has the same expected win-rate as 'normal' play.
> StarCraft has its own bluffing scheme, that is faking a build so the opponent goes for a specific counter, but actually going for something else.
Yes. I didn't say cheese was bluffing. Just that the strategic considerations around cheese frequency are similar to the math for bluffing in poker.