One must always try to strive to achieve it… especially a game creator. Many many hours are poured into it from the largest gaming companies out there. Most prevalent in multiplayer games where people will cry if it is not achieved, balance can destroy any sort of diversity in a game.
My mind is cast back to the days of Diablo 2, before skill synergies were brought in. I had tried PVP only a few times, but when I did, my necromancer spent the entire time running from a homing arrow that would one shot me. Later blizzard addressed much of these issues, providing ways for spells to be just as powerful, though this was years after the release.
This game is particularly notable here as PVP was an after thought, though it became so popular that some people would spend their entire game time PVPing. The popularity of it forced blizzard to address issues that were not considered issues at launch, though many of these also applied to PVE.
Concussion grenade jumping became a staple in Team Fortress Classic, so much that obstacle course maps were made around the idea. Not what you would expect any respectable medic to be up to!
Another example which created adverse effects is Team Fortress Classic. My favorite class here was medic, the flag capper. Anyone who has not played the game may be confused here.. yes medic was one of the best classes for flag capture especially in a public game. They were fast-ish and had access to concussion grenades which when used correctly (or incorrectly depending on your perspective) could propel you across the entire map. Having more health than a scout meant your mission was slightly more likely to succeed. Sod the med kit, just have a shotgun out all the way.
The point being, if you do not address these issues, often the game will not end up as the design intended, people will find exploits, people will only do 1 thing that works and people will not enjoy themselves (usually, conc jumping with medic was bags of fun actually). A game design team must strive to balance all aspects, and deal with unintended situations as they are found. This is why a good developer maintains support over a long period of time (the final patch for Diablo 2 came out 2011, over a decade after the original game release).
The game we are making is no exception to this. Already with us three testing the game balance issues are appearing, and I have developed a single style of play that I am no longer straying from. This is for both myself (game design hence balance) and for Matt (he is working on the computer AI) to deal with. The AI needs to force me to try new things, and punish me for sticking to the same tactics, as well as be unpredictable enough that I cannot guess their moves. There should also be no one size fits all play and things that work in one situation should completely fail in others.
You may get some sort of idea from this article regarding what genres we are working within. No official announcements yet but I hope you are beginning to build up a picture. I want to tell you more specifics about our game but we must wait until there is more to tell (don’t want to get the ball rolling too early).