DOs and DON'Ts of Honest Puzzle Game Design

An educational talk on Puzzle Game Design which I gave live on March 21 2015.
Q&A starts at 58:17
The links at the end of the talk are:
Designing to Reveal the Nature of the Universe - Jonathan Blow & Marc Ten Bosch
• IndieCade 2011: Jonath...
Attempting Deep Work - Jonathan Blow
• Video
A Good Puzzle Game is Hard to Build - Marsh Davies
www.rockpapershotgun.com/2015/...
Games mentioned in this talk:
Braid
braid-game.com
The Witness
the-witness.net/
Antichamber
antichamber-game.com/
Manifold Garden (previously Relativity)
manifold.garden
Miegakure
miegakure.com/
Portal
store.steampowered.com/app/400/
The Swapper
facepalmgames.com/the-swapper/
Monkey Island 2: Special Edition
www.gog.com/game/monkey_island...
Quake
store.steampowered.com/app/2310/
DOOM
store.steampowered.com/app/2280/
The Talos Principle
www.croteam.com/talosprinciple/
Duet
• Duet Gameplay Overview
Opposition
www.puzzlescript.net/play.html...
The Bridge
thebridgeisblackandwhite.com/

[@GabrielHasbun]

Good idea of recording your own footage in the car. It is the "poor man" 's sound isolation booth.

[@tomgcolemangoplayer]

Well presented, and useful content. Enjoyed the contrast of one perfect pot vs. high volume of pots to get in the practice. Being creative takes practice. Telling about idea contrast, and how some cannot help but be done. Thank you for publishing.

[@tedbendixson]

Here because of Taiji. Loving it

[@JohnDaniels]

Just started watching the video but so far so good can't wait to listen to the rest of it I'm making a new game called Moto trials for the PS4 this will be my second game and I loved limbo and inside I thought they had some incredible well thought out puzzles and so I started making puzzles and it's harder than one would think but it's really cool after you create the puzzle you have your friends and family play them and get to watch them and see what they do

[@MyAdventureLog]

Great stuff! I could watch talks about puzzle game design for hours. Thanks!

[@mvandevander]

Thanks. I am glad you enjoyed the talk. :)

[@nathanielcolumna5353]

Braid was your favorite game in 2015, did you ever get around to completing the Witness? Mr. Blow's approach to puzzle design is life-changing (in my opinion) and having missed this masterpiece is a shame! Curious to know if you have a new favorite puzzle game now, in 2019. For my money, the Witness blows Braid out of the water, and I love both.

[@mvandevander]

Oh definitely. I would agree with you on all counts there. I actually was lucky enough to be off work when the game came out, and I pretty much did nothing but play that every day until I completed it. I hadnt done that in a long time. I felt like I had somewhat unreasonable expectations but the game still exceeded them. So needless to say it's my favorite game, and much better than Braid. Since we're both fans of The Witness, perhaps you'd be interested in a somewhat similar game that I'm making nowadays called Taiji: taiji-game.com/about

[@Yitzu640]

DOs: -Design Forward -Experiment -Be Minimalist -Use Reprises -Create Sequences -Iterate -Break all these rules... wait what?... DON'Ts: -Design Backwards -Forget to take out the trash -Be Maximalist -Repeat Yourself -Create Labyrinths -Be Perfectionistic! -Listen to me... ARE YOU SERIOUS!?

[@mvandevander]

I am sorry if you feel like the talk was not helpful to you.

[@Yitzu640]

Oh, no, haha. On the contrary, I thought it was amazingly helpful, that's why I wrote it down. I was just adding a little bit of humor at the end. Thanks for the video!

[@mvandevander]

+Markus Villalobos You're welcome. My point at the end was just to not focus so much on the bullet points of the talk and instead don't stop thinking about what works best for your own game.

[@Gnurklesquimp]

I find that that is the best, most universal advice for any type of design. Learn what design philosophies work for to achieve certain goals so you can create a sort of mental model. That way, when you've clearly formalized your goals, you have a much better picture of how to get to the desired design! :D

[@Panj0]

Brilliant talk with some fantastic insights, thanks very much.

[@WadeHenricksen]

Great talk, but did I miss something? Why are you in a car?

[@mvandevander]

Sorry, it had gotten answered before the stream, but I suppose never made it into the proper Q&A. I had to do the talk from a public wifi spot because I don't have internet at home. And it was more convenient to do it from my car since it gave me a good quiet private space to talk loudly at a laptop. :)

[@nobytes2]

Does it really matter?

[@gamgammerz]

It's part of great puzzle design: how do you get out of a car when you've locked yourself in?

[@jethrolarson]

Good material. Helped me solidify some thoughts I had on level design.

[@james-s-smith]

33:00 I disagree that the Nervewrecker level of The Talos Principle is redundant. The first room needs to be there as a warmup to the rest of the level. The last two rooms need to both be there, because if they weren't, you could just jam-swap into the first room, lock the two mines in the first room in the start, escort the middle mine of the second room and lock it in the start, jam-swap into the second room, and then jam one mine and the last door to win. The point of the level is juggling the last two rooms in such a way that you can escort the second middle mine to be locked out of the way and jam-swap into the third room while still under threat of mines; that's why it's called Nervewrecker.

[@stacymitchell1890]

Thank you so much for sharing.

[@aridoba]

Good talk. Thanks for sharing!

[@gustavorios2]

Thank you!! It will help me a lot to make my puzzle game better

[@Dakobah20]

Great video

[@davie5104]

My problem is that I have ideas for puzzle mechanics, and I feel like they have great potential, but still struggle making it work as a puzzle. For example I'm working on a massive metroidvania currently, with 8 dungeons and I'm just really married to the idea (for one of my dungeons) of using these giant sandworms that basically move through an underground labyrinth of tunnels of their own making in looping paths. This creates many one-way paths as you can follow one, but cannot pass one coming your way. But that isn't a puzzle, that's a maze with one-way paths. Their bodies can be walked on as a platform, which could be a solution to a puzzle, they could potentially FOLLOW you at a forked path, chasing you and changing their direction which could be a solution to a puzzle (but if you're already in front of one, it's no longer in your way, so how's that solve anything?) I thought of placing various giant bells that, when shot, ring loudly and drive the sandworms away from it, but again, laying that out just feels clunky and forced. Is this a bad idea, or a good idea that's just missing an essential modification or two? Perhaps the sandworms serve other functions? perhaps they can open new paths by being led to burrow through like, sand-blocked tunnels? (that one feels like it has some potential maybe) How can I even know without brute-force trying to come up with solutions until I give up?

[@mvandevander]

Well, it's hard for me to say for sure whether it's a good idea or a bad idea, but it definitely sounds like you're having some issues in terms of avoiding contrivance. I mean "contrivance" not to suggest that the thing you have created is badly concocted, but in the sense that anything you add into the game is, in some sense, arbitrary. And the ideal is that you get the most "bang for your buck", or you get the most gameplay (or puzzles) with the least amount of engineering. So in that sense, you're feeling frustrated because you feel like the thing you have added is not giving you more out than you put into it. Just based on your description, if the idea is that the sandworms are moving, perhaps it is good to give them a "reason why they are moving", and then as a player this reason can be manipulated or even removed. Perhaps they move towards sound (the player walks on metal or something and this causes the worm to move), or they move towards heat (player is heat, but maybe can start fires), or light (can be shown on screen in a clear way). Or something. I suppose you already suggested something similar, but I think the important thing is to pick _one reason_ that is predictable and fits in with the other concepts in your game. And if you can make the behavior of the worms predictable, then it should be easy for you to build situations where the player can manipulate that. So I guess, in the end, I don't think this is a fundamentally bad idea. It's just about how you execute on that idea. Just try to make it something that the player can clearly build a mental model of the underlying systems and learn to manipulate it. Also, I do see that you tend to be thinking of things that "make sense" or "naturally follow" from the base conceit of the mechanic. This is often a good way to avoid having the game "feel contrived" to the player. You pick things to add that are connected in some way to what is already in the game. There's worms, and worms burrow, so having them burrow through some obstacle and destroy it feels natural. However, I do want to emphasize that in picking things that are connected to ideas already in the game, you should be attempting to pick things that have the least overlap with existing ideas. You want things in your game to generally be orthogonal in their purposes, and not have multiple things that produce the same exact gameplay result. This is why the concept of "the worm moves towards noise" is a good pick, because you can actually do a lot of things with noise while it really just being one new concept for the player to understand. Rather than say, well it moves away from this bell, and then if you want it to move towards something, you have to also contrive something else that it moves towards.

[@DanZaidan]

Thanks for the great talk! I wonder what came to be of Duet? :)

[@michaelros7817]

Great talk, but I have to sincerely disagree with your take on Adventure games (e.g. monkey island). The puzzles a lot of the time *are* the narrative discourse (Insult swordfighting, "How to get ahead in navigation", Social engineering/tricking other characters). In fact, I'd say that this is evident since point and click adventures are the grandchildren of text-adventures. So in other words, if you're stripping the game down to it's core mechanice, you are not stripping it down to simple input commands and fetch-objectives; you would be stripping it down to literally asking the player: "so how would you go abuot this"? Like it's great grandfather, the tabletop RPG :); or from the other branch of the family; the riddle! Nevertheless; great talk; and I learned a lot from it. Thank you for recording and sharing the presentation with us!

[@hunterzayne2730]

Sorry to be so offtopic but does any of you know a tool to get back into an Instagram account? I somehow forgot my account password. I love any tips you can give me.

[@terranceharper6193]

@Hunter Zayne Instablaster ;)

[@hunterzayne2730]

@Terrance Harper thanks for your reply. I got to the site thru google and I'm in the hacking process atm. Seems to take quite some time so I will reply here later when my account password hopefully is recovered.

[@hunterzayne2730]

@Terrance Harper It did the trick and I now got access to my account again. I am so happy! Thank you so much you saved my ass !

[@terranceharper6193]

@Hunter Zayne you are welcome :D

[@jamesunruh3307]

The Talos Principle is my favorite puzzle game. But I agree that the example you gave was a big misstep for the game. Anyways, thanks for the content. Not sure how much of this I could apply to a concept I'm working on, but interesting to think about nonetheless.

[@Tennisers]

also, have you heard of Even The Ocean? The central idea is very similar to your duet prototype idea.

[@mvandevander]

I had not, but I took a look at it. It's slightly similar with the speed/jump dichotomy thing, but definitely a very different game concept. I'll keep an eye on it. :)

[@zissou6928]

+Matthew VanDevander Do you also feel like this topic hasnt been deeply explored enough?? Simply typing in puzzles into google and only seeing kids' toys seems to be a clear indicator no?

[@neuropiano]

What is the game on the slide of screenshots of different puzzle games that has the beautiful, colorful trees?

[@mvandevander]

That is a game called The Witness. It is a game that I am very excited about, and it is about to come out this very Tuesday, January 26th. It will be out on PC and PS4. I would definitely recommend picking it up. Feel free to wait for reviews but really with this game the less you know going in, the better as it is all about surprising the player. Here's a link to the steam page: store.steampowered.com/app/210970/

[@neuropiano]

HOLY COW! It looks incredible. Thank you! The video is really helpful.

[@DStash]

I would think having optional puzzles for valuable in game items would be cool in an adventure game. For example solving a moderately difficult puzzle for an item that increases your health, abilities, or amount of usable items (think golden sun for a concrete example, you often solved puzzles in golden sun for rare creatures that improved your stats and altered your abilities/player class, which combined the joy of solving the puzzle with the joy of acquiring useful items that you could further customize your character with.). Golden sun had puzzles with clearly defined rules, and they often mixed things up very well over time.

[@mvandevander]

I'm not sure, as I haven't played that game. Although I certainly appreciate puzzles for other purposes (like changing pace in a game that has other code gameplay), most of my concrete thoughts are about puzzle games which have the puzzles at the core and rely on those being intrinsically compelling. So either way, I think the issue with puzzles in other types of games are that generally they are not very good. I also think that having a puzzle which is designed to reward you in the rest of the game for solving it can inherently cause the player to devalue the puzzle itself. Anyway, just some thoughts. Thanks for watching.

[@maingimoveyourteleporter]

Puzzle-platformer game designers commonly love to add timers on to the puzzles for meaningless star-bonuses (3 star!) but all that does is pretty much make the game timing-based to all completionists.

[@DStash]

Prince of Persia is a game that comes to mind to refer to what I was thinking of. You solve a puzzle that is inherently difficult (not obvious and just time consuming) and you get a reward that otherwise you could have passed up. The idea is to reward the player for noticing something off of the primary path and rewarding them for exploring. Half life also did this often by having optional hidden items that you could find by platforming or solving optional puzzles. Mass Effect 1 had a handful of actual logic puzzles (Noveria reactor puzzle comes to mind) that rewarded you with experience and to a degree bragging rights as my friends had skipped it. To clarify: Adventure game has optional puzzle -> Player solves puzzle based on clearly defined rules and a logic problem -> Player possibly experiences positive emotion for solving puzzle -> player is further rewarded by either a valuable in game item or something interesting. I can't recall what I got for solving the ME1 puzzle but I recall the puzzle (pretty sure if was experience but I think you got gold as well). Same goes for most metroid prime games, often the solving of a thing is more memorable than the item.

[@mvandevander]

Well, mostly my talk was about games that are entirely focused on puzzles. And within that, I think it's possible to have optional puzzles to great effect, especially if there are some very interesting but very difficult puzzles that you want to include in the game somehow. It gives players the possibility for choosing their level of engagement with the material. I don't really remember the puzzles in Prince of Persia (although, I also have no idea _which_ prince of persia you are referring to). I think it depends on your design goals. If quality puzzles is the goal, the puzzles should be good enough to stand on their own.

[@maingimoveyourteleporter]

yeah, I think he's referring to hidden puzzles, which is distinct to an "extra goal" which seems equivalent to either A) making the game overall more tricky, or B) making the player play the game twice for it.

[@user-fr9sm3it1k]

Great talk! Do you have any advice for coming up with the rules that you want your game mechanics to exist in? Im having trouble coming up with the hook that the rest of the game can sort of exist around

[@mvandevander]

People do often ask me the “how do you get an idea” question. Honestly I’m not really sure how to answer that. Generally I just will have ideas pop in my head sometimes and I don’t really have control over it. Trying to come up with ideas explicitly doesn’t always work. Usually I just do other things that I enjoy and maybe some ideas will come to me because of that. Or maybe not. Knowing whether ideas are good or not usually requires prototyping.

[@Cookingsource]

You can also check for reference a game called Ibb and Obb. It's very focused on puzzle platforming through cooperative play. ibbandobb.com/

[@mvandevander]

Yes. I have not yet played the game, however I participated in a double interview with the designer of that game a while back about the similarities between the two games, so I'm definitely aware of it. :)

[@amidos2006]

Will ur game duet be out sometime?

[@mvandevander]

The plan is to still finish it yes, however I am presently working on a different project. It is very exciting so far. :)

[@randomcommentor]

There's some great advice in here, along with some things i disagree with. "Backwards design", as you call it, is actually seen a lot in very well designed games. Starting with the solution is a really good method for nailing the "lesson" or eureka moment, you want to emphasize with the puzzle. Many puzzles in Baba is you, for example, started out as an idea for a final interaction, the developer wanted the player to make, where the rest of the puzzle is build around facilitating this solution. By categorizing puzzles as 'arbitrary' or 'truthful' in this manner, you're making it seem like this method doesn't stand a chance to the alternatives, which is certainly not the case in my opinion.

[@mvandevander]

I totally agree with you that "backwards" design or "not honest" design can be good. I was never intending to imply that it can't be. In fact, I picked Monkey Island 2 not because I think it has bad design, but because I love that game and it has a completely different design approach. My hope was that me having a great fondness for the game which I chose as a counterexample would allow the analysis to come across as more neutral, but I made some other mistakes that undermined that. I will say that there are many excesses that can happen in that type of approach (google cat hair moustache for a famous one), but that doesn't preclude the possibility of great puzzles using that approach (getting ahead in navigating, for example) So, the lecture isn't meant to say that other approaches don't work. I just intended to establish a focus on a particular type of puzzle design to the exclusion of all others. So, for lack of better names, I called the approach "honest" or "bottom up" or "forwards". None of these terms were ones that I invented, but unfortunately I failed to consider that they imply that the counter approaches are bad, because "dishonest" and "backwards" have obviously negative connotations. Other than that, the DOs and DONts generally stand. I just regret that the takeaway for so many people seems to have been "DO Honest Design, DON'T do Dishonest Design" which is never something I said in the lecture or something I intended. Thanks for watching!

[@user-cn4qb7nr2m]

So at 37:55 when you saying dont repeat yourself, you repeat yourself aren`t you?:)(Example from Talos) Great talk overall, you deserve more attention (i think content is better then in GMToolkit for ex.) - you need to work on your presentation/image.

[@plushloler]

18:40 I don't think calling them "forward" and "backward" design really makes sense. You could design an arbitrary puzzle forward, stacking layers from the beginning until deciding to say that one of the steps is the last one. And when designing a truthful puzzle, I think you can very often do it backwards and have it be good, if you first discover an interaction in the system and then make a puzzle that forces the player to use it, you are starting from the solution, but you still make a puzzle with a new concept that teaches the player about the system.

[@andrew777spencer]

22:45 Thomas Was Alone?

[@0xCAFEF00D]

What's bottom left on the first slide?

[@mvandevander]

That is a game called RELATIVITY. relativitygame.com/

[@0xCAFEF00D]

Matthew VanDevander Yeah I noticed it at the end of the video and did some searching. Looks really cool. Thanks.

[@Frank-ju8qr]

To anyone watching now, the name has been changed to 'Manifold Garden'

[@mvandevander]

Frank Tiemens Yeah. There's an updated link in the video description.

[@Frank-ju8qr]

Matthew VanDevander Oh sorry I hadn't noticed

[@woohale5678]

lol im making a halo 5 puzzle map while listning

[@GunnarClovis]

Was Duet ever released?

[@mvandevander]

I actually just made a short video addressing what happened with Duet: kzbin.info/www/bejne/nJCpXnSVpdSKqLs

[@GunnarClovis]

Matthew VanDevander Thanks for your quick response! I'm glad you're still doing stuff on this channel. I remember watching this talk a few times several years ago closer to when it came out, and I'm nursing a new puzzle concept now and just felt like revisiting this talk Always loved your stuff!

[@mvandevander]

Gunnar Clovis Thanks for the kind words. I'm really trying to find a way to do more stuff. It's difficult to find the time with my day job.

[@GunnarClovis]

Matthew VanDevander I completely understand. I've been trying to (finally) get into KZbin for like five months. I have a ton of footage all recorded and an intro all animated and scripts written, but I want to finish enough videos that I can do regular uploads twice a week in an effort to actually grow, but it takes so much time to prepare that many videos in advance, and it's difficult to find any time.

[@mvandevander]

Interesting, I've never considered that approach. I basically always upload stuff as quickly as I can do it, and don't bother buffering up things. For better or worse, it does let you see how long the process takes me. Haha. I'll be sure to check out your videos when you start putting them up!

[@Tennisers]

i get why you'd want to have this reductionist argument, but I don't really agree with it. If you apply a reductionist argument to even pure puzzle games, you could say that "Oh, there's some some audiovisual feedback relating to some abstract rules, but if you take that away, then you're just pressing buttons sort of randomly!" I understand why you'd want to make that argument with adventure games a little, but games are all in people's heads, and there are different ways of making interesting mind pictures. I know you know that, I just feel like saying that, because I'm annoying :P I enjoyed what I've watched so far of the talk (but I can't watch the whole thing now)

[@mvandevander]

Good good. Thanks for watching. Yeah it's a reductionist argument, but my point was to contrast between two main theoretical types of puzzle design. One is entirely systemic and one is entirely arbitrary. However, most puzzles are some combination of the two.

[@josiahmanson]

It's Mario and Luigi in block form?

[@GabrielHasbun]

55:30 "Story in games[...]" not "Story is games[...]".

[@mvandevander]

Ah well.

[@GabrielHasbun]

I enjoyed the lecture. However I did not watch the parts in which you talked about Braid because I fear you will spoil surprises. I am still playing it.

[@mvandevander]

+Gabriel Hasbun Ah yeah. Haha. Well I don't think I talked about puzzle solutions. But if you haven't played Braid then now is a good time. :)

[@GabrielHasbun]

I am stuck in some of them. I have figured some of them out but executing the solutions is the most challenging task now.

[@mvandevander]

+Gabriel Hasbun Yeah unfortunately there is a bit of execution difficulty in some of the solutions. The Witness should be interesting as a counterpoint to that.

[@dandymcgee]

I wish I could watch this talk.. I really do.. but the high-pitched noise in the background is driving me fucking mad. Please, please, please edit your sound files in the future to remove sounds above a few thousand hertz.

[@mvandevander]

Ah sorry man. Yeah the audio is kind of a mess in this talk. I have never been happy with it. The sound is the fan on the laptop.

[@gamgammerz]

Best puzzle: Portal 1: Using gravity to accelerate you by jumping off the steps into a portal below and out of another portal high up in a wall to fly over a barrier.

[@maingimoveyourteleporter]

Some different terms I've applied or heard applied on things like this are: "Bullshit" puzzles (where you get stuck on a puzzle, look it up, and the puzzle is so arbitrary and asinine that it prompts the "Bullshit!" response.) "Core" of the puzzle (puzzle difficulty is based on cores, only using extra complexity to fill in the spaces between cores) "Depth and Complexity" (stole this from Extra Credits and renamed them, where Depth are cores and Complexity is, well, arbitrary-ness)

[@GabrielHasbun]

main_gi OK, you just introduced a bunch of terms nobody understands.

[@tensevo]

This would have been more congruent if you were in a quiet study room. Who solves puzzles in a car?/??

[@mvandevander]

Mark Freeman Issue is that I didn't have internet at home. So I had to use public wifi around here. And it is too loud in Starbucks to give a presentation. Library might have wifi that could stream but I didn't check, also not sure if they had private rooms so I wouldn't be disturbing people there.

[@PauLtus_B]

Shame that with most big budget games nowadays the story is the main course...

[@GabrielHasbun]

PauLtus B "Coarse"? Perhaps you meant "course"?

[@PauLtus_B]

Gabriel Hasbun ...hmmm I actually don't remember, I think so XD.

[@stephenkamenar]

why are you always in a car? are you homeless?

[@dolgrulit]

Thank you, the talk helped me a lot in designing puzzles for The Talos Principle mod =)

[@GoatOfTheWoods]

you are my hero

[@gamgammerz]

Talos was great. Contains some of the toughest puzzles I've played.