I know personally I've had instances where I just keep adding new mechanics to a game in order to medigate the issues and well... it ends up never working out (most of the time). In practice stripping the game of unsure mechanisms seems to work best. Starting from scatch and finding out what you REALLY want and NEED to make the game a success.
So I guess my plan for this topic is for you, the community, to share your experience with this "Beating a Dead Game" mentality. Successes, failures, current issues. Spill the beans people!
Very cool design...
It goes to show that we designers can have such a narrow scope when creating games. We make up scenarios in our heads of the final product but miss the ascetics.
My experiance with this lead me to two years of development with core mechanics terribly broken.
The biggest issue for my game Battle of the Aces Was the fact that you could create your own deck of cards that all had varrying classes they were put into. These classes then had prey and predators (Cards it could kill and cards that could kill it). There were a few problems with this. First, there was the possibility that one deck could completely counter another. Now there were abilities and such that added depth but the object of the game was to gather the most points at the end which was dependant on the number of cards you killed!!
Broken.
Plus, even the end game was messed up. You needed the most points but to end the game you needed to bring your Ace (a card found in every deck) to you enemies territory. But that didn't mean you won. It was just to end the game. So its possibe that you could end the game without actually winning it.
Dumb.
Then I just started adding things. Giving a bonus for ending the game. Multiple end game scenarios. The whole deal.
Finally I canned the class system and created a whole new deck creation/combat system which is yet to be tested.