Are Ideas Cheap? In Praise of Strong Ideas

Breath of the Wild, Xenoblade Chronicles and the Power of Good Ideas

If you work in the game industry you've probably heard many "ideas are cheap" variations. Ideas are a dime a dozen. Everyone has lots of ideas — it's the execution that matters. Insert your own pablum here.

In this blog I'll grapple with this idea on a conceptual level, then examine two specific examples that rebut it on a practical one: cooking in Breath of the Wild and the look-into-the-future mechanics in Xenoblade Chronicles.

"Ideas are a Dime a Dozen"

At the risk of being snide let me begin with the following: if you believe that ideas don't matter — that ideas are cheap and easy to come by and that on paper all ideas are equal — then you must also necessarily believe that this idea, the one you're currently espousing as some fundamental wisdom, is also a dime-a-dozen idea of no consequence.

Of course I grasp the intent of "ideas are cheap" rhetoric. If you visit any forum for aspiring game developers you'll inevitably run across a person who has no practical skills and is looking for a team to implement their brilliant idea, which is nearly always of the "X+Y" or "X in space" variety. But the problem with these people isn't that ideas are weak, it's that their ideas are weak — shallow and pedestrian.

Execution is what Matters?

A mediocre version of Tetris is better than the best version of Columns.

A first-year CS student could make a version of Tetris better than the best Columns. People loved the Gameboy version of Tetris, with its green and black low-res display. If ideas are cheap and execution is what matters why is any middling execution of Tetris better than the best Columns or Klax or Hatris?

At GDC2017 Nintendo showed off a 2D "NES-style" proof-of-concept of Breath of the Wild. Why bother making it? The prototype looks fun but if execution is what matters and the prototype version is wildly different in execution from the real version what's the point? They proved that this particular execution worked but then threw it out, so why bother? You could argue that the prototype helped them refine their ideas, but if ideas don't matter why bother refining them?

Is Nintendo just terrible at making games and reliant on a nonsense process?

But Don't Just Take My Word For It

This slide from Ryan Clark, taken from here, is self-explanatory and hard to argue with, so I'm not going to elaborate on it much. (Side note: I generally loathe game development business and marketing talks but Clark's are quite good) Instead I'll talk about this in a specific context: Rogue Legacy vs Full Metal Furies, both from Cellar Door Games. According to the devs Full Metal Furies is a "pretty massive failure", and while I'm not going to say that the game idea is bad it definitely seems to whiff on "has great hooks" and "will be easy to promote." The Steam trailer promises "a unique twist on action RPGs" but even after reading forum discussions and interviews with the devs I'm still not sure what that refers to.

That's in stark contrast to Rogue Legacy, which had an immediately obvious hook that helped it stand out in a sea of other Ghouls and Ghosts / Castlevania-style PC games.

Is the execution of Full Metal Furies significantly worse than Rogue Legacy? Same question but sub in Tacoma vs Gone Home or Nidhogg 2 vs Nidhogg. You'd think the followup effort would benefit from increased experience and capital after a successful previous title and thus be executed better. And in practice that seems true enough: Tacoma and Nidhogg 2 are more technically and graphically sophisticated than their predecessors.

Battle royale games are tearing up the market. Is anyone willing to claim that H1Z1, PUBG, Fortnite, Radical Heights, etc, are all finding success not because of the idea behind them but because they are well-executed? (Editors note: I began writing this quite a while ago!) According to Steamspy Radical Heights has 1 to 2 million users, and it's been in development for 5 months and has mostly temp assets. Is it better executed than Lawbreakers?

Just today I read an article about a "triple-I" publisher who described their signing strategy as looking for high-concept games that will stand out in the market. So even if you don't believe that ideas are important the people doling out money think otherwise.

Cooking In Breath of the Wild

For the bulk of this blog I'll discuss two games that illustrate the power of clever ideas. These are ideas that any competent developer could implement had they the want, not ones with high execution requirements.

First up is cooking in Breath of the Wild. I'll begin with some user reactions. First is a Twitter conversation representative of a lot of the talk about cooking as players first explored the game.

I've awkwardly cropped out the identifying details here as you weirdos can't be trusted.

In a further attempt to lend legitimacy to my opinion I also asked Kirk McKeand, games media writer and BOTW cooking fan, why it's his favorite cooking system.

Before considering these reactions let's examine how cooking in BOTW works. The cooking has two independent layers - a purely formula-driven layer that determines food effects, and a purely arbitrary aesthetic layer that determines graphic and name.

The effect of food — how much health it restores, any additional bonuses like heat resistance or stamina — is determined by a linear formula. (More or less) In practical terms there's no difference between a bird thigh + acorn dish and a prime meat + apple dish. This formula layer has a few fun twists that nod to real cooking — some ingredients like salt have diminishing returns. (Apparently the BOTW devs are sensitive to over-salting. Personally I never put salt on anything. Take that, Gordon Ramsay) But it's very straightforward.

What food is called and what graphic is used is determined by a second unrelated layer. Here the entries are chosen in line with our understanding of cooking with no formula whatsoever: combine two vegetables and you end up with a vegetable dish, combine a fish and a vegetable and you get a fish dish. Many of these are extremely specific; Pumpkin Pie requires four set ingredients. There's no logic behind it other than our real world understanding of how certain dishes are made.

With this in mind let's look at those reactions again. The focus on recipe is something I've seen a lot on social media — people sharing recipes or advocating for a recipe book. But when you understand how the cooking works you see why a recipe book wasn't included: that book would presumably describe the aesthetic layer rather than the functional one. When the people in that tweet thread ask for the recipe what they really want to know (though they don't know they want this) is not how to create that dish, but the cooking formula and the attributes of each ingredient.

Similarly look at the middle of Kirk's quote again: "In Breath of the Wild, you are always experimenting. You throw ingredients together to see what works. Things that make sense usually become the recipe you were expecting: fish and meat becomes surf and turf." This seems to nod to both layers of the cooking systems: the aesthetic layer — that fish and meat becomes the expected surf and turf, and the formula layer — that by experimenting you can determine the the rules of the combining formula and the attributes of ingredients.

From my experience observing player reactions most players fall squarely in the middle of these two systems. They use recipes as anchor points while understanding that there's room for improvisation, very much as in real cooking; they start with a recipe and then season to taste. You find a recipe that you like that restores X hearts, then add an apple or two if you want a little more healing, or add a status-effect ingredient if you want heat resistance or stamina.

Innovation Through Formula

The practical effects of cooking in Breath of the Wild are formulaic. But when people describe it they often speak of exploration, discovery and experimentation. 

In most games cooking or other crafting systems use a recipe / blueprint system. In many games you can't even create something until you officially learn the recipe in-game. And while these systems are often designed to be plausible they are rarely predictable. In Minecraft to make a ladder you build an H-shape out of wood. If you put another stick on the top row does it make a step-ladder? Can you make a metal ladder? Instead of an H why wouldn't you craft a ladder out of an H on its side — surely if you're assembling a ladder you'd assemble it laying on the ground before turning it upright.

In Final Fantasy XIV you can make "Cockatrice Meatballs" (yum!) out of a bunch of stuff that kind of sounds like it might make meatballs. But you can't add extra salt to increase the critical hit percentage they give you. The recipe is plausible (if you believe that "Cockatrice Thigh" makes for a good meatball) but that's all that it is — it's not intuitive or predictable and any number of other recipes would be just as plausible. Put another way, the Cockatrice Meatball recipe is the evolutionary psychology of the cooking word.

These sorts of fixed-recipe systems map to the aesthetic layer in Breath of the Wild while assigning each result a fixed stat line. The brilliance of BOTW cooking is that the arbitrary logic of "a tomato, some olives and some fish creates Salmon Veracruz" affects only the aesthetic of the food, not the function. You don't have to hunt for recipes by trying to read the minds of the designers or wait for NPCs to give you specific formulas. You can experiment, learn the system and create your own recipes. You may never stumble upon the exact formula for fruitcake but you can experimentally derive your own recipe that has the same functional properties.

Without the aesthetic layer you'd be crafting interchangeable buckets of gruel with no categorization, mnemonic or relationship to real cooking. Two apples and a nut would create "Food Item 0xABF0001D" and three apples and a nut would create "Food Item 0xBADF00D". Both systems are required to break out of the normal recipe-bound paradigm while remaining understandable.

Idea vs Execution, Again

Time to circle back around to the framing device of the piece and the discussion of "ideas are cheap, it's the execution that matters." Breath of the Wild cooking was well-received but the execution is not standout. Small touches like the jingle that plays when you cook something and the fact that food items are physical objects that stack in your hands are neat, but those are offset by a UI design that makes assembling dishes a bit clunky. And those neat touches are themselves idea-driven. Nothing about the system screams technical excellence or complexity. The aesthetic layer is most likely defined in a giant Excel spreadsheet somewhere, the same as how it would be defined in other games, and the functional layer is a relatively simple formula and some basic attributes. The strength of the system is squarely in the idea of having food be formula-based while creating an aesthetic layer to add apparent order. "Apparent" is key here because, again, there's no real relationship between the aesthetic and formula aspects. This is a non-obvious idea with straightforward execution. Nearly any game could implement BOTW's cooking system without much trouble — they just didn't. And while there are some good game design reasons to stick to a rigid recipe-based approach it's hard to avoid the conclusion that for many games this is less an active choice and more a default one. Those developers either didn't consider alternate systems or couldn't come up with one that worked.

Future Sight in Xenoblade Chronicles

The second in-depth example of an idea rather than execution-driven success is the future sight mechanic in the Xenoblade Chronicles. (The original Wii / 3DS version, not the sequel or X) In Xenoblade Chronicles the protagonist's ability to see into the future is a major plot point but I'll be talking about a couple mechanical aspects.

Battle Premonitions

During battles in Xenoblade you'll sometimes flash-forward to a devastating enemy attack, giving you a chance to react to it and change the future. Maybe an enemy does a killer 5-hit combo on a character in the flash-forward, but the power of precognition allows you to change the future by using an ability that makes them temporarily invincible

Alfred Hitchcock once famously differentiated surprise and suspense as the difference between a bomb suddenly going off under a table as two characters chat vs the audience watching the oblivious characters knowing a bomb has been planted. The future sight battle mechanic in Xenoblade creates suspense rather than surprise, something rarely found in RPG battles.

Not too long ago I was playing Etrian Odyssee IV and I came across a turtle-style FOE. (A strong monster — FOE stands for "***ing obnoxious enemy") After fighting it for a few rounds it did a tail-wack attack that hit everyone in my party for lethal damage. That's surprise. I had no idea the creature could do that and I had no way to react to it. (This story has a happy ending — the next time I met this creature I turned it into turtle soup) In Xenoblade the tougher an enemy is and the more potentially lethal attacks it makes the more often you see the flash-forwards and the more time you spend with the threat of a ticking bomb looming over you. The tougher the battle the higher the tension, not just because tough battles are naturally tense but because of an explicit mechanically-added element.

An interesting property of this system is that it provides a natural rubber band on difficulty, in that the harder a battle is the more you'll see these warning premonitions and the more chance you have to stave off death, at least temporarily. It's a very clever way of applying an organic asymptotic smoothing to enemy difficulty. It's no coincidence that Lorithia, generally considered the toughest boss, is tough largely because of a sort of deadly lava environmental hazard around her which doesn't trigger flash-forwards.

Nearly every JRPG in the past 30 years has claimed to have a battle-system innovation with a trademarky-sounding name, but while a few of these are interesting (Active Time Battle, Grandia's timeline-based thing, whatever the heck is happening in The Last Remnant) most are fancier in name than execution or are more finicky than good. Both Xenoblade Chronicles X and Xenoblade Chronicles 2 are guilty of this — the "soul voice" system in X feels overcomplicated for the value it adds and 2...well

Item Get!

The final clever idea I want to talk about is what I'll call the "item premonition system" in Xenoblade Chronicles. The system works like this: when you come across an item that can eventually be used to fulfill a quest the game flash-forwards to a scene of you turning in the item to finish the quest, then marks that item in your inventory to let you know that it's quest-relevant. This happens even if you don't yet have the quest.

This is a humble but brilliant particular solution to a largely unsolved class of problem in games: communicating important information to the player that the character has no business knowing. I follow Adrian Chmielarz on Twitter because he calls these sorts of things out consistently.

Many games these days feature collectible items that can eventually be turned in for quest rewards. But what happens when you don't yet have the quest? More broadly speaking in many games there's some information the designers want the players to know for ease-of-use reasons but that the characters shouldn't know. The most common options for dealing with this is ignoring it and letting the player twist in the wind, or calling it out in a way that makes the hand of the designer very apparent. "Hello player, this is Fred Jones, designer of this level. Just want to let you know that you're approaching the mission boundary area, and if you cross it you'll automatically fail for reasons related to our technical implementation and that have no narrative justification. Also just FYI you've found 80% of the items in this area and that berry you just picked up is important because down the road an NPC will ask you for three of them and give you a hat in return. Good talk."

The Xenoblade solution to quest items is simple, very easy to implement and has good in-world justification, not an eye-roll-inducing narrative justification like "losing synchronization" during chase scenes in Assassin's Creed games. I recently saw Naughty Dog employees claiming that the health bar in Uncharted is actually a "luck bar" and...no. I applaud the attempt to come up with an in-world explanation for why Drake can soak up so many bullets but this one just doesn't work any better than "a wizard did it."

The Xenoblade solution isn't a monumental change that pushes the game from a 6 to a 9 at "TheGameDudes.com". But it's a great idea that improves player quality of life without breaking immersion, and elegantly solves a problem left unsolved in most games. If "ideas are cheap" and "everyone has lots of ideas" why are games that solve this basic problem so rare? The reality is that when it comes to these sorts of problems entire game studios often don't have a single good idea and end up just slapping some non-diegetic text onscreen and moving on.

This is the Part Where I Conclude the Essay

The conclusion is the part of the essay where the author summarizes the premise, and that's the part you're reading right now.

There are some game development aphorisms with merit but "ideas are a dime a dozen" is not one of them. AAA game development is largely a technology and production-value-driven medium, but even AAA games benefit from compelling ideas. That Mario can throw a hat at and take over a T-Rex is a great idea that instantly resonated with people. BOTW is a strong technical achievement (as someone who worked on an open world Wii U game let me assure you: it ain't easy!) but the strength is ultimately more in the ideas. On some level the execution in BOTW is a bit weak with repetitive enemy encampments and modest dungeons. The strength of the game is how it relentlessly drives towards certain design aesthetics: the capability of the player over the character and "what you expect is what you get" gameplay. BOTW is not an example of a development team doing what other development teams could not — it's an example of a development team doing what other teams chose not to.

Indie games often live or die by their premise. There are many positive things you can say about Slay the Spire but that it has good production value isn't one of those. There are all sorts of jokey "simulator" games that are middling in all aspects but premise. Cuphead has great execution of its art style, but great execution of a more standard "indie pixel art" aesthetic wouldn't have sold nearly as well.

Of course there are plenty of fine games with no flashy grand ideas, and plenty of "big idea" games that fail. And there's no "idea guy" job where every six months you come up with one cool idea then lean back and wait for everyone to implement it. But ideas very much do matter, and quality idea generation is a skill like any other.