I have a problem when it comes to game balance and level design, I always end up making it too hard to beat.

There were a few early comments on Crypto Crisis regarding how hard it was, some were even asking if the game was beatable. During the development of our next game, Stefan has mentioned to me multiple times that the AI is too OP; “It’s not fun when it’s too difficult”.

So why did I make everything so challenging?

Through my own personal experience of playing games I concluded that when a game is too easy, it’s not fun.

I now realise that it’s far more complicated than that. It took me a week off in Japan and playing some arcade games to finally figure out where I was wrong.

I played a lot of Initial D (a racing game) while I was in Japan. I realised that when you were racing the AI in story mode, it was REALLY hard to lose. Despite this it was still fun and I kept putting money into the machine.

If the computer was so easy to beat, why did I keep playing? How was it still fun?

  1. The driving/drifting mechanic was the most fun aspect of the game. A lot of the game’s challenge was right here, it was easy to learn but hard to master. The feeling of going around a corner sideways at speed without hitting anything was extremely satisfying.
  2. The AI was there to enhance the experience. It would make it more fun to win by taunting you at the beginning of the game, over taking you mid race (no matter how quick you were going) and then let you overtake it before the race finished. It would make the driving/drifting mechanic more fun by giving you opportunities to overtake it slightly before, midway through or after a corner.
  3. How well you mastered the driving/drifting mechanic was reflected in your race time rather than whether you beat the AI or not.
  4. Losing sucks. At one point I started playing with other people. I won a few times then I got matched against harder opponents and lost 3 times in a row. The point at which I wanted to quit playing the most was after the 3 losses. If I got repeatedly beaten by an AI like that and was not able to progress, I probably would’ve stopped playing.

So how does this change how I approach game balance and level design? The following only really applies in player/s vs AI situations.

  1. The challenge of the game should come from mastering the core mechanics. Doing things well should be satisfying.
  2. How well you have mastered the core mechanics doesn’t necessarily need to be reflected in whether you win or lose in a level. E.g. There could be better rewards for performing better within a level.
  3. A level should not be so difficult to beat as to cause the player to repeatedly lose and get stuck on it.

In a lecture by Canadian psychologist Jordan Peterson, he explains how a bigger rat which could win a play fight against a smaller rat 100% of the time doesn’t always win. The bigger rat needs to let the smaller rat win at least 30% of the time or else the smaller rat will no longer want to play.