The Golem – One Month On

On October 11, 2020, in Uncategorized, by bateleur

The Golem has been out for just over a month, so now seems like a good time for a blog post about how it's all gone.

The best thing about the game has been that – unusually for me – I still like the game and the ideas behind it after release. What's been really interesting to see is that the game's central idea, that players be forced to think about solutions rather than discovering them experimentally, has proven to be much more radical than I would have expected. A number of players have commented that they hadn't seen this approach before, which considering the huge number of puzzle games that already exist is quite surprising.

There is a significant downside to this kind of puzzle, though. I was aware of this before launch, but: the game is hard. This wasn't intentional as such, it simply arises from the fact that thinking about things is hard and humans – even very clever ones – just aren't very good at it. The problem being, of course, that if a game is hard enough it limits who can play. But there is an upside to it too: when a player sits down, thinks about a level and beats it they know they've really achieved something. They come away feeling good about their success and also often with an appreciation for the puzzle design itself.

Speaking of completions, exactly one month after launch the first player that I know of outside the playtest team completed the game! Before launch I wasn't completely sure anyone would do so, but I'm very happy about it. In fact in terms of player progress generally I'm pretty satisfied with the shape of the curve. Not only is there no one difficulty spike, but no clear patterns are emerging in terms of where individual players get stuck, either for short or long periods.

I learned something about puzzle design from watching players and reading their comments. It is well known amongst puzzle designers that good puzzles should not contain things irrelevant to the solution. This seemingly absolute rule can get complicated in practice. If a room is ten columns wide by the leftmost is never used, should it be shrunk? If a wall contains a bump, should that bump be removed if it is not a part of the solution? The Golem adopts an unusually relaxed stance on this issue and focusses on two goals. First, that minimalism should be avoided where it might play the role of a spoiler; giving away the solution in a manner less interesting than letting the player think about the problem. Second, that puzzles should be based around a single core idea and that all other puzzle elements should exist to provide the best possible presentation of that idea. A consequence of this which seems to have caused surprise is that some puzzles even support solutions that leave a block completely unused. Heresy! But in fact there is a simple and important reason for this: if a player has understood the key idea behind a level then I want them to complete it. I don't care if there's a clever way to do it with two blocks. I'll happily give them three if it means that the solution becomes less fiddly to execute.

Of course, there are dangers in extra material. The first is that the player might find an unintended solution that bypasses the key idea. This happened a lot during playtesting and a few more have been found (and fixed) since release. And the other major problem is that extra material can introduce distractions: things which appear to be possible paths to a solution, but which lead nowhere. Such distractions are bad, so care must be taken to avoid them.

Here's what I learned: extra material is not the main cause of such distractions. Not by a long way! Both broken solutions and nasty distractions are caused by levels which give the player lots of power and lots of freedom. Which is a considerable design challenge, because to some extent powerful tools and freedom to use them are fun!

The Golem has some failings in this regard, but they're not nearly as bad as they might have been thanks to an early decision I made to take playtesting seriously on this game. For previous puzzles I've always gone with the usual approach of letting a bunch of people play a beta version for free, listening to their feedback and looking at the analytics. This time, I hired two A-list solvers to go through the entire game finding problems. They did an outstanding job and I would recommend this approach to anyone trying to make a puzzle game of this complexity. (And just a reminder that the game had a separate prototype phase, which also helped a lot with understanding how to build good levels in the first place.)

Just a few more things…

First, given that I'm selling this game a few people have been confused by my description of it as a "recreational project". What this means, in practical terms, is that the game doesn't make enough money to pay my bills. Indeed, after a relatively successful first month it has nearly paid for its own development costs assuming I pay myself zero dollars. My expectation is that future sales will leave total profit from the game somewhere in the low four figures. This isn't any kind of problem, it's just the nature of the games business where niche games are concerned. That's why most of the work on The Golem was done during evenings and weekends.

Second, one thing I didn't expect but maybe should have done is that the puzzle community – particularly other game developers – have been immensely supportive of the game in all kinds of ways. Around 27 years since releasing my first commercial game I'm still not over that feeling you get when people turn out to like it, but what's possibly even better is when you get the feeling or being part of a creative community who are collectively trying to get better at making things. The Golem arose from my wanting to say something about puzzle design. So now I have I get to watch other designers taking the best bits and using them to make even better games, which I will one day get to play myself.

Finally, the single most tweeted thing from The Golem wasn't any level, it was the little passage of text on the About page. In particular, the last two lines "Not everyone will overcome all the puzzles, and that's OK. No matter how great or small your accomplishments, to take time to think about a difficult problem is a beautiful thing." Puzzle friends, I'm glad you feel the same way. ♥


2 Responses to “The Golem – One Month On”

  1. Joe S says:

    About halfway through and loving The Golem. It ranks up with my favorites of Baba is You, Stephen's Sausage Roll, and Snakebird. As with those games, I'm amazed at how you can design such elegant puzzles. Would like to also design my own puzzle game in spare time, though I question how well I could create puzzles with difficult yet non-busy, non-trial-and-error solutions. Thanks for sharing some of your design and test process in this blog.

  2. bateleur says:

    Thanks Joe! I'd definitely recommend giving design a try if you can find the time. Even if you're not 100% satisfied with the results the first time round the process of design itself can be a lot of fun.

Leave a Reply