So recently I've been making my way through Cave Story, which is, for those not familiar with it, a homemade 2-d platformer made by one Japanese dude all by his lonesome. It is a fantastic game, and well worth playing (especially since it is free).
Now, as you get towards the of the game, things get pretty difficult. The final cave, in fact, gets extra difficult if you are on the path towards "Good End." I happened to know I was on the path to the good ending, because a friend had warned me when I first started playing that there was one specific action I had to take to put myself on that track, and I took it. Finally, after multiple tries, I managed to beat the final cave and was all set to tackle the final bosses.
After a few attempts to beat the final boss, I took a trip to good ol' GameFAQs to look for some advice. As I was reading up on things, I discovered something rather dismaying: despite my earlier actions, I was still not going to get the good ending to the game. The reason for this, it turned out, was that in addition to the one arbitrary action I had taken earlier, there were a series of other, even more arbitrary actions I had to take! I had not taken these actions, and since I was already at the end of the game, I couldn't go back and do them.
This frustrated me, of course, as making your way through an entire game, only to be told that because you didn't do something seemingly insignificant hours earlier, you are only beat the game in the "bad" way. Now, I understand Cave Story, as I said, was put together by one guy who was not a professional, so I can forgive this little flaw. It got me thinking, though, of many other experiences I'd had in the past with "real" games that share the same flaw.
Just because they are cute doesn't mean you shouldn't shoot them in the face.
The first one to pop to mind was Silent Hill. I remember years ago getting towards the end of the game and encountering Cybil at the carousel after she's been evil-fied. I said to myself "there's got to be a way to save her!" Though, as it stood, I couldn't figure it out. Nothing I had in my inventory helped, and filling her full of lead didn't really seem to be doing the trick. After a while, I decided there must not be a way to do it, so I shot her down and moved on with the game. As you may have guessed, this of course put me on the road to BAD END. It turns out, as I discovered later, that there was a way to save! I had to, many hours earlier, in a small room, not only examine a nondescript puddle of liquid on the floor, but then also decide to scoop up some of the liquid in an empty bottle! A black puddle in a dark room in a game full of random puddles of blood and other liquids on floors, and I was supposed to decide to check out that specific one more closely? Totally frustrating.
The way I see it, designing your game to have paths to good ends and bad ends isn't necessarily a bad thing, but you have to do it well. First of all, making the "bad end" the default is sort of like punishing your player. I just made it all the way through your game, and instead of rewarding me, you are basically telling me "sorry, try again but this time do something else!" That might be ok in Bubble Bobble, but at the end of a 15 or 20 or more hour game? No way.
Secondly, make sure the player can easily see the reasons for why they get the ending they do. Not examining a puddle is not a good reason to give someone a bad ending, nor is anything else involving arbitrary, insignificant actions. Make it clear about what choices will have consequences. If I walk into a burning castle, and the princess is trapped across the room, if I'm given the choice to save her or not, and I choose "No," that is something I am fully aware will probably have long-term consequences! If I get the bad end after that, at least I know I've earned it!
Now, if you'll excuse me, I have to head to YouTube to look up the good ending of Cave Story...