While a positive random effect can only benefit the player who controls it, there are other varieties of random effects. Negative randomness means the effect can actually be bad for the controlling player - for example, if their card or character (like Ogre Brute) randomly targets the wrong enemy. Where positive random effects needed to be balanced by putting a card below the power curve, negative random effects should be offset by raising a card above the power curve so the player feels like they're getting a reward for risking it. There is also reciprocal randomness, where the random effect applies to all players in the game. Hearthstone's Spellslinger card generates a random spell for both players, but is not a well-designed example because its ability to generate any spell in the game for either player creates far too large of a delta of randomness. Winning or losing in that case doesn't rely on the player's ability to adapt, but simply on luck of the draw. A card like Mechanical Yeti which generates similarly powered advantages (or perhaps disadvantages) for both players does a better job of generating small but impactful results for both players. Ultimately, randomness in game design should not aim to create big moments - as those can make it feel like the game tilted unfairly - but should instead be focused on creating engaging moments for all players involved.