Why proc gen works - The Promise

Procedural generation is a fantastic tool to create huge, possibly infinite amounts of content for a game. With a strong set of rules, you can generate places, characters, items or narrative. However, within development there is often (very well justified) pushback on developing these complex systems, instead of putting that effort into hand made content.

I believe that in many cases, while proc gen isn’t always as practical or magical as it sounds, it creates a Promise with players that justifies the cost. Let’s review!

I’ve worked on a lot of projects at early stages, both during my time as an indie and also when I was at Roll7 working as prototype designer. Each time when working with a team and considering procedural generation, there was pushback, generally from more technically minded developers.

This comes from expected playtime vs predicted development costs.

In the minimum viable content article we made an educated assumption that players will generally want at least 2 hours of content per $1 spent. So let’s build from the knowledge in that article for our systems driven puzzle title that we will sell for $10. Each puzzle takes ten minutes, so we will need 120 puzzles to fill that time. We’re confident that the Systemic Synergy in our design supports that playtime, but soon after players will start to get bored.

We then in theory have a very simple calculation to do. How much does it cost for a designer to create 120 levels, vs how much does it cost for a programmer to create a system for 120 levels worth of content, with a designer creating the rules data. You can easily see here how two ‘how long is a piece of string’ type questions with different stakeholders isn’t getting resolved easily, and ends up being done on gut feel.

This is where The Promise comes in.

The Promise of Infinity.

The Promise is that players who want to can keep playing for a very, very long time. While we know that in reality, we only need those 120 puzzles to satisfy the majority of players, and that the game will get boring soon after, there is a huge amount of value from a sales perspective to players if you are appearing to offer that much more potential playtime.

Generally, I would advise against making content for the top 1% of players (for premium, low cost games) that no one else will see, but that isn’t quite what we are doing here. By using procedural generation and developing rules to make sure that the 120 puzzles are a great experience, and basing our costings around that expectation, we’re selling the dream to players that there is a vast sea of content to explore.

This doesn’t come without risks, but in an ideal world, players will play for 20 hours exploring all of the mechanics we have designed to keep that time engaging, feel satisfied with the amount of money that they paid for the game, and then drop off at various rates.

But, each of these will have been influenced by the promise that the procedural generation offers. Maybe this will be their forever game?