Hacker News new | past | comments | ask | show | jobs | submit login
How to Make a Great Puzzle Game (thepuzzle.press)
161 points by keyle 23 days ago | hide | past | web | favorite | 57 comments



Since I am mentioned in the article I have a correction to make that I think is pretty important.

The article says that I like to find ideas by “playtesting” then goes on to describe playtesting as a form of user testing.

Whereas there is some definition of the word playtesting that might be applicable to the way I design puzzles, it is not remotely the one in the article. In fact I think too much “playtesting” as described here will make your game boring, and will degrade your skill as a designer, and over time make you a boring designer. It is important to avoid this.

If you playtest too much it means you don’t really know what you are making or don’t have confidence to determine what playing it can be like. Even if you don’t have these skills or this confidence yet, you should be working on building them. Certain high skill levels of design will not be accessible otherwise.

The amount of playtesting we did on The Witness was very small — about 5 or 6 days over the course of 6.5 years, on a game that takes 50-100 hours to play through.

Whereas I do sometimes get puzzle ideas from bottom-up exploration of a space by myself (which is the definition of “playtesting” that would make sense here), in fact the vast majority of my puzzle ideas just spring to mind after I have built a sufficient understanding of the possibility space. So these are more like two distinct phases of design, and you might switch between them when designing one game. But I think most of the actual puzzle ideas have to come at least semi-intentionally, or the game will not feel very strong.

You have to know where you are trying to drive the truck.


Hi Jonathan, I'm the author of the article and I definitely agree with you. At the time of my writing of this article I had just begun creating puzzle games myself and I just tried to pull together all I could find on the topic.

I stepped away from all this for awhile now to actually create instead of just doing research and found that many of my initial ideas were flawed. What you mentioned is incredibly insightful and I hope you'd allow me to directly quote you and edit my post so that others who visit will be able to better understand the whole scary but wonderful process of puzzle game design.

Goes without saying, but I am a huge fan of your work and you as a person so I'm incredibly honored by your response. That said, I hope that I have not mistakenly misdirected the audience by putting words into your mouth in the form of advice. Please let me know if there are other portions of this that you find incorrect.

Either way, thank you for all you've done!


Hey Chris. Love the posts. Lots of insight.

A bit offtopic, but could you remove the fixed widths of your #content (.baby-container) at screen width below 767px ? 300 pixels is just too narrow for the content. Maybe set width to auto and set margin to 15px for a better responsive experience? :)


Yes! I am pretty inexperienced with web dev so I just changed the theme entirely but I really hope its better for anyone on their mobile phones now. Thanks for bringing it to my attention!


This thread is why Hacker News is one of the best sites online today.


This is a really insightful point, design wouldn't be learnable if it always relied on discovery through testing.

Compare to giant companies like Twitter, where I have heard engineering & product managers say "Every idea has been thought of, it's only a matter of testing them one after another." What an intellectually bankrupt idea.

It is the most experienced product designers who can just look at something and tell you if it is good or bad, before trying it, that big corp product undermines the most.


> It is the most experienced product designers who can just look at something and tell you if it is good or bad, before trying it, that big corp product undermines the most.

that's because for every product designer that can tell good from bad, there are 10's of pretenders who dont know what they're talking about.

The stakeholders don't have the time or expertise to vet (otherwise, they'd be product designers themselves!). Thus, they need objective verification. This is what turns into the 'matter of testing the idea'.


I call it the method of creating self-feedback. Everyone starts off by using hunches, "taste", and the opinion of the crowd or authorities.

But to get beyond that - and getting beyond playtesting, data-collection and other brute force feedback mechanisms entails doing so - one has to systematically devise original feedback signals - pass-fail checklists, writers' rubrics, dependency graphs, workout logs and so on. Once you have the framework in hand and have validated its philosophical premise to some degree, you can feed your work and ideas through it and it automatically corrects you and guides the ideas away from major fallacies.

Of course, some amount of experimental data helps, but the huge amounts we have grown accustomed to have a habit of producing a lot of busywork - a product tuned and tweaked and prodded to improve this or that short-term KPI or audience appeal metric - without striking at the critical issues that one can probe with a bit of followthrough on self-feedback.


It maybe a worldwide communications ecosystem is a different problems space from a throwaway videogame.


And on the opposite side you have Nintendo, known for some of the most beloved games in the world. They have the "Mario Club" and playtest extensively. They test often and continuously. They can test every day if they want. Adjust the jump height 10%, ask the Mario Club if the game got better. In this way they iterate on game play far more than just about any other game company I know of. The biggest benefit is they get feedback early. Most game devs wait until their game is "presentable" which basically means they've already spent too much time making what they've made that they can't change anything except minor tweaks. Nintendo gets feedback early and often.


"Maximizing your audience" != "strengthening your game"

Audience maximization is almost literally regression towards the mean to maximize profit. It implies cutting corners ; but some of these corners are actually making your game stronger, they're giving it its personality.

Think of "Dwarf Fortress", "brogue", or even "Minecraft". What kind of feedback do you expect to get from most players? Better graphics? Real-time play? Would it make the game stronger?


Minecraft was basically driven by user feedback from the very start, when it was originally released it was a tech demo.

For Dwarf Fortress, I believe most players would want a more understandable user interface and a tutorial, and I honestly believe that would make it a better game.

One general result from designing games is that players tend to be good at knowing what needs improving, but very bad at knowing how to improve it, so it's a good idea to see where they get stuck or frustrated, but then ignore their advice on how to fix it.


Are you sure that pressuring the designer/dev to build what players ask for but not what inspires him will make a better game? Just because players want it doesn't mean the dev can deliver it. Let modders make the skins and tutorials.


This is a misunderstanding of what Mario Club is - QA company spun off from the old Seal of Approval. It's on the side of technical testing than design.[0]

The idea you're exploring might be more akin to the Mario Cram School[1] which involved getting staff across all of Nintendo, not just designers, creatively exploring Mario development. But straight up assuming that Nintendo's approach towards design makes it a thumbs-up dismisses notions people have that newer Nintendo games overuse tutorials to the point of infantilising the audience.

[0] https://www.destructoid.com/nintendo-seal-of-quality-becomes... [1] https://www.youtube.com/watch?v=XBXZ5lqdZQ8


I wonder about that. Do you have any additional reading on Nintendo's playtesting?

Although their games are mass mass market, they still feel opinionated. Their designs have a certain Nintendo-ness to them that's persisted for decades in almost all their games. Even when they partner up with a third party their design shows through.

If they had extensive playtesting early on I would think their uniqueness would be eroded. It obviously hasn't. Most recently Nintendo set the bar for what an open world game should be like with their take. Not just in quality but by applying their method they elevated the genre.


Not Nintendo, but Valve has some very interesting playtesting commentary in Portal 2, if you play with commentary on.


> If they had extensive playtesting early on I would think their uniqueness would be eroded.

Authors also have editors and 'test readers' for their novels. That doesn't erode uniqueness. Why should this happen for games?


Nintendo's games are intentionally meant to address as large as audience as possible though. While I'm sure Jonathan wants to sell a lot of games, I think he cares more about making it great for a smaller subset than making it palatable to a larger subset. To put it another way, Nintendo's games are incredibly bland compared to something like The Witness. I think Nintendo makes great games, but they're not nearly as unique an experience as you'll find from some indie developers.


Me thinks maybe you haven't actually played Nintendo games if you believe they are bland compared to a game with 650 repetitive puzzles that would have been just as interesting (or non-interesting) as a 2D game on a phone. Yes, I get there is a context in the game and a big reveal (and I enjoyed the game) but I don't agree that 650 puzzles is less bland than my time spent in Breath of the Wild.

I also don't agree that "Indie" = "better" or "unique". AAA companies, especially Sony and Nintendo have released amazingly creative games. Just a few from Nintendo include Pikman, Animal Crossing, 1-2 Switch (cow milking), Cubivore, and there are several others. Meanwhile 99% of "indie" games are skinned versions of old games who's best examples are often Nintendo's. How many "Metroid"vanias are there? How many Smash Bros clones?

As one more final point, the games made at big companies are made by actual people who poor their creativity into the games they make. Indies don't have a monopoly on creativity. Being AAA != can't be creative nor does it mean can't take risks on new ideas. There are plenty of examples of new ideas from these companies.


It’s unfair to say that The Witness can be reduced to a stack of simple 2D puzzles.

No spoilers, but almost every set of 523 main puzzles in The Witness has a twist that requires a full 3D environment in really interesting ways.

The +135 +6 bonus puzzles require a 3D to 2D transition, but only ironically.


How many sets of 523 puzzles are there?

Aren't almost all the puzzles 2D graph paper mazes?


Each different area on the island requires some part of the surrounding environment to solve the puzzles. The ones that are just 2D maze drawing really are just warm ups for when the mechanics get twisted by the 3D world around you. Good luck!


I didn't say indie games were better, nor did I imply that most indie games are better than most AAA games.

But I strongly disagree that the most creative or the best game will ever be mainstream. When is the last time a AAA game came out like Baba Is You? Celeste? Hollow Knight? Those are games I play and replay because they're so enjoyable and memorable. Nintendo makes great games but I don't feel any of them are quite the same experience as the best indie games I've played.


Celeste is a Mario-like, Mario was mainstream. Hollow Knight is a metroidvania, Super Metroid and Castlevania were mainstream. The Witness is a Myst-like, Myst was mainstream. I like the indie scene as much as the next guy, but it's 90% based in nostalgia.


I chose Celeste and Hollow Knight for being the best in the genre, not for creating a genre. The creative game I mentioned is Baba Is You.


>especially Sony and Nintendo have released amazingly creative games

I mean, what does that even mean? If we use the analogy of movies, where the medium as an art form is more established: Pixar movies are creative, but they can still be bland (most of them are anyway). Yes, some Nintendo games pushed games as a medium forward, but I don't remember any Nintendo or Sony game that pushed in the direction Braid or The Witness did. Very few games do actually. "Games" games are more like toys. I enjoy playing with them, but I don't step away from them and think deeply about the experience I had. And I wish more games would strive for that kind of experience.

>Meanwhile 99% of "indie" games are skinned versions of old games who's best examples are often Nintendo's. How many "Metroid"vanias are there? How many Smash Bros clones?

Compare that to what the parent comment said

>but they're not nearly as unique an experience as you'll find from some indie developers.


> Pixar movies are creative, but they can still be bland (most of them are anyway).

By a matter of definition, I would expect any creative media to not be bland. Even if every other part of it is boring the exploration of the creative ideas in it should be stimulating.

> I don't remember any Nintendo or Sony game that pushed in the direction Braid or The Witness did

I'm not sure what direction you are referring to. I don't think that either has really influenced gaming as a medium. The closest game I can think of as a comparison for Braid would be Portal and I think it also didn't really push any boundaries for games.

The Witness is still a little young, but I don't think its likely to lead the way for other games either.

On the other hand, I think the original Metroid and Legend of Zelda radically pushed forward what games can do. I think both of them would make you think deeply about the experience of playing them if you played them in their time.


Zach Barth from Zachrtonics famously never solved the final SpaceChem level. IIRC the interviews, it (and some other levels in that game) was done by inspiration, and playtesting was 'check out if this is reasonably solvable'.


Indeed, while I found Witness difficulty very uneven, with lack of episodic structure like many other games. It lacks guidance so you stumble around until you find things - but it's not actually truly open world as there are many knowledge gates. Which you will regularly bounce off.

That does not make it worse than other puzzle games, just annoying in parts. Especially its huge reliance on perception as opposed to actually thinking and figuring things out, but then it's not quite a puzzle game. It's more of a big escape room game which does have puzzles in it. There are others on the market, few are well known.

(Even something like Antichamber is more of a puzzle game than Witness, yet it is still an escape room rather than a pure puzzle game.)

A relevant contrasting comparison would be to something like The Talos Principle which was heavily playtested including some puzzle redesigns, reordering, some having been dropped. Puzzles were also automatically checked to be doable in desired and a few atypical ways.

The other design principles apply though the same - minimalism, tutorial principles and difficulty scaling.

Antichamber has the guns and blocks and physics, Talos has the few tools and pseudophysics, Witness has varying puzzle mechanics but one input with tons of output.

The difference is the degree of things being hidden. Puzzles with hidden pieces tend to become adventure games. It does not matter if you have to carry the pieces or if they're already in place.


> just annoying in parts. Especially its huge reliance on perception

I don't understand this attitude for a game that is called The Witness.


> I think most of the actual puzzle ideas have to come at least semi-intentionally, or the game will not feel very strong.

This right here is a really underrated statement. It's why I think the designers behind WPC-style paper puzzles (Toketa, gmpuzzles, etc.) consistently produce works of higher quality than most levels in most puzzle video games. Working on paper forces you to design with intention rather than design by playing around with objects on a grid.

Intentionality is one of the most important aspects of good puzzle design. Reminds me of a 2001 quote from Nikoli editor Nobuhiko Kanamoto: Computer-generated Sudoku puzzles are lacking a vital ingredient that makes puzzles enjoyable - the sense of communication between solver and author.


Thanks for The Witness. I don’t play games at all anymore, and haven’t for years but I’m really glad I took the time to work through The Witness. It’s an absolute work of art. (as was Braid - literally the only other game I’ve played in years)


i've been told that 'Outer Wilds' is also a good game. (not to be confused with 'Outer Worlds', which is not a great one).


Outer Worlds isn't yet released, it will be available on October 25.


I started to read this and thought to myself "This sounds like Jonathan Blow" and then you mentioned the Witness and I finally tore my eyes away from the text to the username, and lo and behold!

Quick question - how do you gain this understanding of the possibility space without bottom-up exploration? Is it a matter of having good influences (and who might those be)? Or some other reason?


I think it’s just, after you have designed enough things, you learn to think like a system and you know what a lot of the possibilities are before you play those specific corners of the system.

That said, you can always be surprised, and it’s delightful when this happens, so you don’t want to get on an egotistical high horse and decide you really know the system without playing it. The actual experience of the actual system is your reality check.

But, the more experience you have with that system, the more you can see where things might go (unless it’s intensely chaotic, thus unplayable).


Very true. The Witness is in my top one PC games, precisely because everything seems to be there for a reason. It's extremely well done!


I had to do a double take on your name.

You are a genius and a god among us. Thank you for making the deepest and most beautiful games I have ever enjoyed.


I don't know why I am being downvoted. Braid and The Witness are two of the best games I have ever played. And I've been playing games for 30 years now.

Your tech talks are also awesome.


Flattery, in these circles, is not seen as continuing the conversation forward. I also imagine that those who are predisposed to dislike Jon will downvote you out of spite.


Chip's Challenge is the all time great.

https://archive.org/details/chips_challenge_windows_3.x


If you listen to the Apple time warp podcast interview with Chuck Somerville”. He talks about chips challenge/ chucks challenge /chips challenge2 and the licensing from the current owners. It’s actually kinda a fun listen about Apple ][ development.

http://appletimewarp.libsyn.com/apple-time-warp-podcast-epis...

For those that like reading more than listening ars summarizes it

https://arstechnica.com/gaming/2015/06/how-an-early-90s-wind...


Oh wow!! I played this game when I was little but never knew what it was called and never found it when I searched (from vague memories). Thanks!!


Love it. Thanks for finding this blast from the past.


Oh my god this runs in my browser! There goes my week


one of my favourite puzzle game creators is Everett Kaser games... http://www.kaser.com/

HoneyComb hotel being my favorite!

The games have been around for a long time, and seem to be updated for modern platforms. But the website is still a blast from the 90s, and the graphics are joyfully 90s amateurish :)

One of the interesting things he did was invent his own language, and virtual machine for doing his games.... http://www.kaser.com/kint.html


Honeycomb Hotel took a short while to figure out how it works (in particular I didn't realize the puzzle isn't considered solved until all walls are placed), but it's quite addictive, thanks for that.

Any others of his that you would recommend?


Thanks for the recommendation! I downloaded it on iOS and have been really enjoying it.

The controls took a few minutes to get used to, but the included documentation was superb.


This collection of games is quirky and fascinating.


For those interest in designing great puzzle games, I would suggest reading the following thesis on evaluating fun in puzzle games.

https://www.researchgate.net/publication/27476077_Automatic_...

(update: Non paywalled version https://pdfs.semanticscholar.org/e401/51cbec7bdcc65b59fff53c...)

Cameron Browne (the author), suggest many way to evaluate your puzzle, and this was used in his further research as a way to automatically generate fun and interesting puzzle games.

http://cambolbro.com/games/index.html


This article talks more about what a great puzzle game is and less about how to design one. it's a bit disappointing!


Myst was probably my favourite puzzle game. :)


It really didn't age well if anyone is getting that itch. The atmosphere is phenomenal but the puzzles can be extremely frustrating. Back in the day, that wasn't a problem because exploration felt like a privilege. Now, the scenery doesn't captivate like it used to and you just want to know how to solve the stupid thing.

Portal 2 is a great brain-teaser if anyone is looking for a constructive use of their time tonight. I recommend the Nyskrte NYS maps if you want a real head challenge.


The Witness, mentioned up-page (at time of writing—will probably stay there, given the author) is great and a bit Mystish, though with a very different sensibility re: puzzles. There's also The Talos Principle, which is kinda Myst+Portal.

I didn't like Portal 2's official single player campaign as much as the first Portal, but the two-player coop mode is excellent, and a rare example of multiplayer puzzling.


If anyone has that itch they can turn to Obduction and Quern : Undying thoughts, both recent and both very good.


My only problem with Obduction is the loading; oh my god the loading. I know they wanted to keep the original Myst on Windows 98 feel, but a 30 second to 2 minute load between scenes is killer and why I haven't finished it.


missing: how to make any money from said game.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: