Puzzle Game Magic Secrets

In this 2019 GDC session, My Dog Zorro’s Brett Taylor approaches the art and science of puzzle game design through the unstoppably cool lens of human cognition.

GDC talks cover a range of developmental topics including game design, programming, audio, visual arts, business management, production, online games, and much more. We post a fresh GDC video every day. Subscribe to the channel to stay on top of regular updates, and check out GDC Vault for thousands of more in-depth talks from our archives.

[keithg]

For those who don't know, a ten second break like what he does in this video will help you rest AND remember his lesson.

Spacing is a technique to learn anything faster. While studying or consuming content, giving yourself a ten second break where you focus your mind on something else will give your mind time to quickly reinforce the knowledge so you can better store it in long term memory.

A break, a joke, a transition, a restatement as a metaphor... there are many ways to add these breaks in naturally or do them yourself while studying (like looking away from a video and scrolling the comments).

Some people do this naturally every couple minutes, and most of everyone else struggles to remember things.

[Brindlebrother]

I play a lot of puzzlers, indies, and platformers. Linelight is one of the most dazzling and fun games I've ever played. This game made me feel emotions for those little colored lines. The piano music in the game is fascinating too.

[ThomasAndersonPhD]

For the 10min version of the talk, jump to 50:00 for the recap.

Here's a recap of the 7 lessons:

* Simplify (remove noise)
* Cut pointless levels
* You (the designer) are an expert player (so if players say the puzzle is "too hard", listen)
* Keep 'action' and 'puzzle' separate
* Make solutions unambiguous (avoid solutions not feeling right, avoid trial-and-error, and make impossible actions unambiguously impossible)
* Player Trust (by default, players expect that they don't need agility to solve puzzles; if one puzzle requires agility, players will assume many puzzles require agility)
* Exhaustive Design vs Fun (consciously decide whether you are exploring every possible iteration of the mechanics -or- focusing on the fun levels that grow from mechanical interactions)

[Lunareon]

So many puzzle games fall victim to so many of the problems listed in this talk. Thank you for pointing out the limits of human cognition. Also, water is the best!

[nhandles]

I was not ready for the level of charisma the speaker possesses

[맹현서-jn]

정말 좋은 강연 감사합니다. 강연의 내용뿐만 아니라 구성 까지도 듣는이를 고려하여 준비한 티가 나네요. 마치 퍼즐게임을 푸는듯한 발표였습니다.

감각으로만 익혀 왔던 레벨디자인 기법이 정돈되어 많은것을 배우게 되네요. 제가 들은 GDC강연중 최고였습니다.
Thank you so much for the great lecture. Not only the content of the lecture but also the composition of the lecture is obvious in consideration of the listener. It was a presentation as if solving a puzzle game. And I liked this person's humor.

Level design techniques that have been learned only by senses have been arranged and taught me a lot.
It was the best GDC lecture for me.
(I used a translator)

[ezraszanton]

Woah this guy loves metaphors

Seriously great talk, wish I'd watched this before I started making puzzle games but at least now I'll have it for future ones!

[programaths]

30:00 Another solution is to prime the player by having him make that exact jump in a corridor at the very start of the level. He will do the jump (not other path, only way to go), without even thinking about it. Then he will see that the other jump is the same and will do it, without even noticing he was primed. Priming is a really good tool and it can make people feel really clever while part of the solution was given to them.

In Super Metroid and Metroid Dread, they lock player in areas using skills they don't have yet and also created loops to funnel the player. So, when the player explore, he has a lot of chances to land on the desired path, unless he is tacking extra steps not to. Also, it foster a sentiment of being lost and being able to fin his way. There is one human on Earth, though, who is immune to hint and won't shoot ceilings 🤣 (Even when the game design push you to do so)

[chofmann]

poor sudoku got such a bad reputation. handcrafted sudokus - especially those with extra rules - are quite interesting

[thomasreichert]

Thanks for the tips, and thanks for giving hope! How a mortal human could have created the puzzles of a game like Baba Is You was mystifying. I guess the ingenuity was in creating the mechanics that then "thought" up the puzzles.

[randomrandom]

23:38 "Knowing the solution and not being able to execute it" that's got to be the worse feeling in a puzzle game, that's basically what made me quit Angry birds back in the days. I would know how I'm supposed to be doing it, but with the bad mobile controls, I would miss like 1 of the 3 birds that I absolutely need to land then have to start over. It's the opposite of fun, like I know what I'm supposed to do, I just can't do it.

[Kjhgar]

This was grand thank you :)
Very clarifying for a newer developer

[ben_burnes]

This is a huge list of super valuable material. Wish I knew this when I was making puzzle games a few years ago, had to find most of it out on my own hah

[sherafgames]

Really love that point about noise.
It's something I do a little bit in my level design, but I couldn't properly articulate it.
Thanks!

[programaths]

49:41 That depends on how your mechanics combines and if you want allow multiple solutions or not. That's where the "problem space" come into play.
I designed IQ Test items and more often than not, when the item is checked, shortcuts are found or there are confound and sometime, a second or third solution.
Obviously, in those item you are probing something very specific. Like the number sense, relative size perception, ability to isolate information or process multiple variables at the same time (hardest ones as you've to ensure people can't eliminate solutions using each variable in isolation, also the ones most people can't "read").

Related, The Witness and Metroid Dread lock players in some areas to limit the search space (subset of the problem space). In The Witness, it's gated by knowledge. In Metroid Dread, it's gated by abilities. And those things are mostly invisible to player and even relief it by reducing his search space a lot.

[shreeshayelameli]

What a lovable dork. This guy's great. Awesome talk!

[voltcorp]

lesson 7 really connects to one of the better lessons in that famous mark rosewater talk. interesting doesn't mean fun. having a design space doesn't mean everything in there is a functional possibility.

[aa-xgdi]

Linelight is one of the game I love the most!

[thephilosopher]

I really like the concepts he explained regarding Working Memory. Its actually a concept that could be used to teach ppl about why practice makes perfect (6:55), especially within difficult games that have large learning curves.

[chaostrottel_hdaufdutube]

I loved the “oh *sight* yeah….” On 3:50