Designing The Golem

On September 10, 2020, in Uncategorized, by bateleur

The GolemThe Golem, my block-pushing puzzle game, was released today on Steam and I want to talk about the design of the game and about puzzle games generally and how they've changed over time and where I hope they're going next.

My previous blog post (over a year ago!) was about the open playtesting process for The Golem, which was a surprising success. At the time I speculated about the possibility of turning the game into a small commercial project. I did indeed end up working on it in my free time as a side project, but the arrival of the worldwide COVID-19 pandemic in 2020 disrupted my work enough that I decided to put my main project down and work on The Golem full time for a couple of months to get it ready for launch this year.

As mentioned previously, my initial motivation in designing The Golem was to create a puzzle where the solution could not be found by fiddling around with the puzzle pieces and simply stumbling upon solutions. This being a very negative thing, I eventually focussed more on a more positive goal: that each level should be built around a single key idea and solving the level should be primarily about discovering that idea. This proved to be a very powerful way to design puzzles, but not entirely without controversy.

Before I talk about why this approach is good I want to talk about what you lose. There are puzzles one can design in The Golem which involve a precise set of moves with multiple pieces interacting in very complex ways that are challenging to find, but which cannot easily be described in terms of a single, core idea. These are what I think of as "micro" puzzles (although they need not be physically small), with the idea-based puzzles I favour being in the "macro" category of puzzles which are more about strategy and do not in general require precise moves. There is nothing wrong with micro-based puzzles. Some players enjoy them and The Golem's ruleset definitely supports them well. There are none in the game because of my rule that all puzzles must have a core idea.

So why is this approach good? The main reason is because it eliminates so many weak level designs before they're even created. I've noticed a tendency for puzzle games to often contain levels which functionally duplicate previous levels or even levels which are little more than doodles in the game's level editor. Whilst it's possible to defend this sort of thing I'm personally not a fan since I feel it's disrespectful of the value of the player's time.

Another reason, perhaps even more important, is that occasionally one comes up with an idea before it's even clear if such a thing can be designed at all. Each level of The Golem began as a single sentence in a text file. In two or three cases it took literally hours of design (once all the iterations and bugfixes were taken into account) to turn that one line concept into a working level. If I'd made levels by sitting down and opening a level editor I don't think those levels would ever have been made.

The Golem is a difficult game, but not intentionally so. Or at least, it's intentional in the sense that I've been aware of it throughout the project and have chosen to leave it this way. It's not intentional in the sense that I don't think a more approachable version of the game could be as good. I don't mean this as some kind of puzzle-solving elitism. It's simply that the game is about thinking deeply about things and humans aren't very good at that. During playtesting my levels were repeatedly broken by playtesters. A few of these breaks were, of course, stupid errors on my part. But most were because the complexities of my own game were just as challenging to me as to anyone else. Equally, I also found it very enjoyable to watch replays of ingenious but unintended solutions enacted by playtesters, watching for the moment when some clever play unravelled all my intentions and left my level in ruins! A puzzle game is a dialogue between designer and player, but when a level breaks it is the player who speaks and the designer who listens.

It is interesting to me to examine the changes in puzzle games over time, both in terms of the games themselves and the attitudes of the community who play them. Many of the early puzzle games I loved had no undo function, whereas now I won't even bother to play a game without undo in most cases unless it's very short or it's a prototype and the designer plans to add one later. Likewise I used to expect puzzle games to be mostly filler, with maybe only a tenth of the levels being both good and original, whereas now I quickly get bored of repetition and busywork (and to their credit most modern designers avoid it).

More subtly, modern puzzles seem to have a much better awareness of player experience. Who is the game for? What does the target audience actually enjoy? The Golem is a puzzle game for people who already play puzzle games and it makes no effort to appeal to a wider audience. That's OK as far as it goes, but I think the most interesting unexplored design space in puzzle games lies in making games which bring the experience of thinking about interesting puzzles to a wider audience. I'm not talking here about casual games. I'm talking about games which offer serious challenge, but do so in a highly approachable way, welcoming inexperienced players and teaching them.

Another game which released today – A Monster's Expedition (Draknek and Friends) – takes new steps towards that goal. If I make another puzzle game in the future it will very likely be something inspired by that approach to design. The Golem is a love letter to the puzzle community, but maybe it's time to share our fun with everyone?


2 Responses to “Designing The Golem”

  1. GenghisPawn says:

    Just found the blog, and enjoyed reading up a bit. I've got a post in the Steam Forum for The Golem with some feedback in it, if you're interested.

  2. bateleur says:

    Ah, cool, didn't realise Steam wouldn't notify me of comments. Have responded on Steam, but many thanks!

Leave a Reply