> Zork Grand Inquisitor, on the other hand, is way too easy. You can die, but you cannot do anything wrong without dying. That means, no matter what you do, if it doesn't kill you, then it's good.

Or neutral. This game design philosophy is NOT what makes the game easy, it's what makes it playable. Some of the hardest adventure games I've played are designed this way -- because although you can't do anything that will stop you from solving the game, there could be a million possible combinations of things you can do, while only the right thing gets you anywhere. ZGI may have been too easy, but this wasn't the reason.

To me, it is ridiculous, foolish, and downright maddening when an adventure game lets you do something that stops you from solving it, then lets you continue playing without telling you. A lot of the early Sierra games are like this -- do something innocent early on, then 500 hours of play later, find out you have to backtrack all the way back. That's not challenging gameplay, it's nasty and stupid. Player satisfaction does not increase by virtue of the game being manually cumbersome to play. (Because it's not like the puzzles are necessarily more difficult, it's just that you're forced to either maintain a full history of saved games and/or constantly replay pieces of the game you've already solved.)

Consequently, it is my policy for games that show up on AGL (see link below) always be possible to solve no matter what you do, and although you can die, the move that killed you can always be undone. If the games on AGL are too easy, that's because of its menu-driven interface, not because of this game constructing policy.