Method I developed for finding good difficulty balance

A game project @frosted and myself have been working on is in the play testing phase, and we've got some really useful and also encouraging feedback. One thing some people have said and is the most encouraging thing we could hear is that the difficulty balance was just about right. Challenging, but felt fair and wasn't frustrating.

Although this is the first time I've been level designing, I think through tons of iteration I've developed a reliable method for getting to a good starting point for game balance/difficulty. I want to share that here.

First, the tools you are working with gotta be simple and balanced. Don't fuss over minutia, it must be simple. Think tic-tac-toe. Games like this survive through the ages because they are simple, not complex. Complex code and game design does not indicate intelligence or sophistication, only a lack of prioritization. So balanced tools are ones where one element cancels another, but each is checked by another. Alone, no element can win without player input. In other words, decision making, or strategy has to be the deciding factor in victory.

In our case, we started from a battle-tested formula that's been featured in a few games already, and just made our own adjustments. This is a great way for the learning developer to begin as you are climbing up on giants shoulders rather than reinventing the wheel.

But our tools are still very simple. Just a basic board game with a few unit types that each move a certain way and have specific attack/defense values. Really bare bones stuff, but it's everything you need and nothing you don't to make a compelling game. So we are trying to do a lot with a little, which is the most efficient way a small/solo developers can work.

Second, using those balanced tools, build your level to be impossible. Then play it many times until you, the designer, can almost beat it if you do everything perfectly. Then trim out factors that cause difficulty until you can just beat the level if you do every move perfectly. That's your 'hardcore' mode. This is what players who play the game religiously will eventually be able to achieve.

For the average gamer though, this isn't acceptable. But getting from hardcore to moderate is pretty simple. Just reduce the difficulty causing factors until, doing each move perfectly, you find the mission to be easy and have all resources left over at the end. This is a good starting point for your play testing, and tweaking from here won't be hard at all. Shouldn't require any major changes.

In conclusion :

1. Start with balanced tools.
2. Build your levels to be impossible.
3. Play your impossible levels until you can almost beat them.
4. Reduce the difficulty causing factors until, doing each move perfectly, you can just beat your levels.
5. Reduce difficulty causing factors until, doing each move perfectly, you can easily beat your levels.
6. Deliver to play testers and fine tune from there! Video recording of gameplay is more valuable than written feedback.


Link to thread about our game (still open to playtesters! help us improve the game and we give you a free copy):
https://forum.unity.com/threads/project-in-6-weeks-lessons-about-scope-and-such.686152/#post-4624090

 

"decision making, or strategy has to be the deciding factor in victory."

And a few more thoughts about this:

This is really where the magic is. Designing the balanced tools is easy enough. Basic math has got your back there. But how do we ensure the decisions player is making are interesting? This is where your soft skills come into play. Even if you are an agoraphobic introvert, you have to understand the mind of the person you are selling to. Easiest way to start is with yourself, of course. If you play games a lot but aren't keenly aware of the happenings in your own noggin' while playing them, level design just isn't gonna work out for you.... I expect. Could be wrong.

But if you know what makes yourself tick, it's not too hard to design a game that you know you would enjoy playing, right? That's a good starting point.

How about testing though? How can you test that you aren't a complete weirdo and others will like what you've designed too? You have to have clear goals, firstly, and second you need data.

Clear goal is a specific reaction or emotional response you are trying to elicit. Data is how you measure if you got that response -- play testing. You need a specific questionnaire but best thing is video recording or being present while play tester first experiences your work.

For instance, in the first mission of our game, I want player to learn quickly that despite the goofy theme, it's a serious game. You gotta be engaged. So first thing I do is introduce a river crossing. Rivers reduce player defense and movement, making them vulnerable. As player approaches river for the first time, dialogue pop up suggest suspense that something might happen, but then dismisses the notion. "Eh, it's probably fine. Let's go."

Watching a few playtesters, I saw that some took this literally and just blazed on through. Other's were more suspicious and went around. But nobody was quite careful enough, and when that first enemy jumps out and nails them in the river, the reaction is exactly what I was looking for. First, they are alarmed. Some laugh, others get upset. But from there on, they become very engaged. Because that loss hurts -- they only have a few units at disposal. From then on out, they play very cautiously, which is exactly what I want. Then, playing that way, they can generally be successful so long as they keep the discipline of staying engaged.

 

@BIGTIMEMASTER
That is some really interesting stuff in your first post.

Especially this bit of advice.

 

thanks I hope it can help in some way on your own projects @BrandyStarbrite