the secrets gimmick
or: some casual thoughts on secrets in video game/puzzle design, in which i frequently use scare quotes, and am unsure of where periods go in and around them. british english has different rules to american english and frankly, i don’t remember either.
The game Animal Well is an obviously brilliant video game. It takes up about 30MB of hard drive space, each hand-coded byte of which is going to be genre-defining in the world of metroidbrainias for the next several years. When it came out, I was unable to take my hands off it. And that’s rare for me with challenging games: I tend to find excessive challenges to my dexterity to feel oppressive and stifling. It is also exemplary of “the secrets gimmick” which is this nebulous game design concept that’s been on my mind for a while. When I refer to this secrets gimmick, I’m not thinking about old-fashioned cheat codes, or things you have to Google. I’m thinking of things you’re expected to explore to discover, and are somehow indicated as collectible. Explorables might be a better name. If you’ve played Animal Well, well, you probably know they’re a key part of the experience. And I think they’re a thing that is surprisingly important to my enjoyment of video games—partially because I really love when games give you the freedom and trust to figure out things for yourself—but perhaps even more because when secrets are poorly done, I absolutely hate them. Animal Well is an example of a game that I could easily hate for this reason… and just about don’t, because it does enough things right. And I want to casually tell you about that about it. Some of what I’m saying is common knowledge among game designers, and some is my probably incorrect thoughts. I’ll interleave these two things without discrimination.
Secrets in video games have been around forever. From the warp zone in Mario, to old Zelda’s famous secrets (that I still won’t spoil because everyone should play old Zelda), and from things hidden behind walls in Deus Ex, to the devious tricks in Animal Well, I don’t think it’s fair to describe secrets as something new or game-changing for video games. But I do think it’s reasonable to claim that they used to be more of a gimmick than an actual mechanic, or key part of the game as in Animal Well. And as a mechanic, secrets are attractive to the game designer, and perhaps even more so to the modern indie developer, because you can make your video game feel better by packing it with supposed secrets that the player “discovers”. This makes the player feel better about their play, as they feel more independent, and less like a horse that’s been led to water. It makes them feel different than the average player, special, and smart. For these designers, adding secrets becomes this cheap and easy way to make your video game harder and “better”. The lowest form of this is just to take all the map markers you had for chests, and… get rid of them. They’re secrets now. Now add a hidden chest counter in the pause screen, and you’re done. (And hey, while you’re at it, scrap all the instructions and text, to make the game more “mysterious!”)
And despite my slightly facetious tone above, I genuinely believe that this is often a win-win for gamers and developers. Video games often serve as an aesthetic for exploration and discovery—especially thinky puzzles and metroid-brain-ias, which are all about combining simple rules to discover the emergent properties of a logical system—and secrets are one of the key tools to build this aesthetic. Jon Blow has several excellent talks on this, which in the spirit of discovery, I’ll let you discover yourself. But overall, I think games should have more secrets. I think the secrets in A Monster’s Expedition are fucking fantastic... So keep that in mind as I complain for the next 1000 words.
The problem with secrets arises on both sides of the developer-player divide: developer are taking the gimmick too far, and players are taking the gimmick too seriously. A lot of games are starting to become “about” the secrets, and games with secrets get this cult following. We are told by developers and players alike that we haven’t “truly experienced” Animal Well unless we’ve experienced “all three layers”—the final of which insanely requires every egg in the game to even reach. We are told that Braid only really makes sense if you find all the stars. And… this just sucks. Because it turns out the usual player loves a bit of secret-hunting, especially if it’s something they can do on the side while playing the main game, without substantially altering their playstyle. But this archetypal player also hates doing repeated linear searches of the game’s world—an obligatory thing in the post-games of many games, frankly including Animal Well—in order to say they’ve even roughly “fully beaten" the game. (Again, this is not new. Check out the history of Shadow of the Colossus for people taking secret hunting too seriously. But I think it’s becoming more common.) This effect compounds when previous secrets are needed to find other secrets as in metroidbrainias, as when you do this, you’ve just guaranteed that multiple passes are required, as your players desperately try to topologically sort a DAG without knowing the edges. And you better bet the average player isn’t going to know to mark every slightly odd thing they see, just in case it happens to become a thing they can interact with later on. Now, the word “average” or “usual” is important here, largely because of the existence of an '“other” here. There is a contingent of video game players that has the terrible triad of too much free time, knowing exactly what to do always, and loving to brag. They could be called vocal, and they are certainly a minority. These people eat this shit up, and then gate-keep about it online. The Stanley Parable phrases this whole paragraph far more succinctly than me: “DID U GET THE BROOM CLOSET ENDING? THE BROOM CLOSET ENDING WAS MY FAVRITE! XD”
And, the rest of us, those who miss the broom closet ending… well, we often end up having to google it and feeling shit about ourselves. Phrased from the other side of the divide, as game designers, we’re currently not designing for the average, or even the average smart player. We’re designing for a player that wants to brag. We’re designing for the player who wants to feel that they’re playing a “smart” video game, simply because it elided some useful hints in the post-game. Honestly, with very few added hints, Animal Well’s entire post game (which most people barely touch without google) could be drastically simpler, and in my opinion better. Still fun, but not requiring intractable brute-force search or the internet. And many games are actually confident enough in their game design to do this: to give you a secret-finding tool that genuinely reduces the problem from an O(world size) search to an O(secrets count) search. But then, in the public eye, that game seems to lose some mystique, some myth, that Animal Well seems to have. It loses the factor of “this game is an iceberg and I’m so smart for finding what’s underneath.” But frankly: that mystique is a gimmick—and not even that smart of a gimmick—and one that is usually making a pretty large number of players feel bad about themselves.
And this is where Animal Well compares unfavourably to another fantastic game released in 2024: Balatro. When Balatro makes you feel good about yourself, it’s for finding a cool (perhaps secret!) interaction between cards that made you crazy money. It’s truly independent, truly creative, and everyone will experience it differently because of the randomness. When Animal Well makes you feel special, it’s for your 18th search of the cat area, during which you just happened to stumble upon that hidden area that no-one else on Discord stumbled upon yet. And this is okay, until a large fraction of games become about this experience, not about the actual game proper. That’s not fun. And games not being fun is the problem.
So, how do we do secrets well? Well, here are some ideas. Part of the issue is cultural: people like to show off, especially the type of (usually) male player that likes to play so-called “smart” games. And I don’t really know how to deal with that. But there are some design ideas that designers and developers can keep in mind, and here are a few of them. There will be some spoilers below. And there are of course many exceptions to these rules. I’m just giving general ideas.
Don’t hide the existence of secrets. Secrets are secrets because they are hard to find. Not because the average player doesn’t know they exist. At least tease them, or make one kinda easy.
Braid does this wrong. The Talos Principle does this okay. Animal Well does this well.
Don’t make secret-hunting drastically different than the rest of the game. A common mistake is often these games require a drastically different play-style to the main campaign, and a fair amount more time and brute-forcing of possibilities.
Braid (usually) does this wrong. Honestly, Braid does a lot wrong.
Animal Well and A Monster’s Expedition do this right. The secret hunting is essentially the same puzzle-solving mechanic as the rest of the game. Indeed, some people would say A Monster’s Expedition’s secrets are actually just bonus puzzles!
Tunic is a counter-example to this rule. This game’s post-game is awesome, despite not being like the rest of the game. (But honestly, that’s partially because most of Tunic is a souls-like and souls-likes are usually bad!)
Provide testable hypotheses. Don’t make it random guessing. Make it so you can notice something, guess a mechanic that might let you exploit it, and test that hypothesis. The less dexterity involved, the better.
A Monster’s Expedition does this right. Braid also does this fairly right.
Make your secrets actually something the player should be proud of.
Tunic does this right. It’s hard, but rewarding. The secrets are real puzzles.
Provide secret-hunting tools. Or provide the broad location of secrets.
Braid does this wrong.
Animal Well does this very right.
A Monster’s Expedition does this very right, through the fog.
Don’t make the secret-hunting tools a secret.
Animal Well does this wrong. There’s a tool behind a certain boss that is both very useful, and very easy to miss.
Don’t make secrets irreversibly missable.
Braid does this wrong. Indeed, the most accessible secret in the game is irreversibly miss-able.
Don’t make secrets require massive time grinds.
Braid and The Witness do this wrong. Though they kinda invented this idea, so I’m okay with it.
Make secret progress trackable.
The Witness and Animal Well do this right.
Talos Principle does this wrong.
Keep the mystique. You don’t need to ditch the aesthetic and mystique to follow the other principles here.
Animal Well does this right. The Witness does this very right.
Don’t make core parts of the experience depend on secrets. Okay, it’s fun to provide a little big of new dialogue or story to reward hardcore players. But don’t hide the best ending behind 100% progress. That sucks, unless you’re Undertale.
Baba is You is an interesting case here. I think it’s secrets are clever and telegraphed enough that this is okay. But I still think it could be a little bit easier to progress in the late game without finding clever things.
Have Easter Eggs that aren’t secrets! By Easter egg, I mean it’s not collectible. I mean it’s untracked. I mean it doesn’t matter to progress. I mean no-one is going to claim you have to find this to have “truly played”. It’s just a cool thing to see.
The Witness does this very right.
Don’t gatekeep your own players.
Jon Blow does this wrong.
If you have to pick a game that truly does secrets right, it’s The Outer Wilds. But it does secrets well enough that I’m not even sure that you can fit it into the box, or apply these rules to it. It’s a game that’s only about secrets and exploration, and stupidly clever secrets at that. But for the rest of us human beings who can’t make perfect games, maybe we can try and follow some rules to make our games more fun, and perhaps a little less “smart for the sake of being smart”.