Gamedev Grievances #3: The Balancing Act

It’s easy to make a game, but it’s a lot harder to make a “good” game. By “good”, I mean something that catches the player’s attention, has a cool concept or story, neat gameplay, and so on. But another really important factor is balancing the gameplay, mechanics and so on to present a reasonable challenge. The game shouldn’t be too easy – otherwise it becomes boring. Nor should it be too hard – that just makes players give up and forget about it. The sweet spot, I think, is to have the game present a reasonable challenge, with a solution that’s achievable but has to be worked for. That’s the sort of thing that gives players a sense of satisfaction at having completed that challenge, and also a desire to keep on playing.

In other words, it’s one big “balancing” act. And it’s a tricky one, too: getting the right difficulty level requires a lot of tweaking and play-testing.

Here’s how I’ve been dealing with it in Ambience so far:


Obviously, when making Ambience I needed to make sure that the enemies the player encountered weren’t too strong, the dungeons weren’t too long, and so on. This in turn required me to have a fair idea of how powerful the player could be expected to be at a particular stage in the game. Which meant, well, lots of play-testing.

It was more than just balancing enemies and dungeons, though – I also had to balance all that against the player’s own stats and rate of experience growth. A few times during play-testing, I found that the player would become too powerful too quickly, making the game too easy. At other times, the player would be left behind too easily while enemies grew too strong. Sometimes this was because enemies just didn’t yield enough experience points, leading to lots of grinding in lower-level dungeons before you could move on to the next stage of the story.

One of my biggest aims is for the game to provide a good challenge without the need for too much experience grinding or item hoarding. Besides the whole trial-and-error process, one factor that really helped for this was automatic saving at the beginning of each area. This maintained the difficulty at all stages of the game, kept the pressure on the player to keep winning, and also established a neat, “every decision matters” gameplay style.


I also needed to balance weapons very carefully as well. For weapons, the general trend was that increased power required increased effort (drained more Special Points), so could be used fewer times in succession. When I released the Ambience demo, one of the things I found was that one of the weapons, the Sabre, was unexpectedly powerful. It had a high power, and was very efficient – much more so compared to many other weapons with similar power output.

Notice the Sabre only consumes 3 SP with each swing, but is still very powerful
An old GIF from the Ambience Demo. Notice the Sabre (equipped) only consumes 3 SP with each swing, but is still very powerful.

This meant I needed to go back to the drawing board and find a way of determining how powerful each weapon was relative to the others, so I could effectively balance them against each other. Some of the factors I considered in doing this were:

  • How powerful is the weapon?
  • How efficient is the weapon – how much SP does it use up with each swing?
  • How accurate is the weapon?
  • Does the weapon boost (or reduce) any of your stats?
  • Does the weapon have a chance of inflicting a status condition?

I put all these factors together to calculate a modified “power per turn” value for each weapon, which I then turned into a pretty chart:

The weapon balancing graph, subject to change.
The current weapon balancing chart, subject to change.

As you can see, I nerfed the Sabre considerably from its original power level of about 9.5, and now it’s a much more reasonable weapon. However, it’s also probably pretty obvious that the Heavy Club also turned out to be far more powerful than I expected, mainly due to its giving the holder a large Defense boost. Thanks to the graph, I now know I probably should nerf the Heavy Club a bit, but not too much – perhaps to be at the same level as the Heavy Battleaxe, a similar weapon.

It’s also interesting to see from the graph how the weapons could be categorized into hypothetical “tiers” based on their power level. But of course, a single number based on a rough algorithm is hardly likely to tell us which weapon is the best in practice. (You guessed it… that means more play-testing.)

Leave a Reply