Prototypes
If you are set up with your foundations, the next step is proving the concepts actually play well or “how does my game actually play and how do I find the elusive “fun”?
There is no framework for what a prototype represents and needs, as different studios have different approaches to prototyping. For more context see Jason Schrier’s great book - Blood, Sweat, and Pixels: The Triumphant, Turbulent Stories Behind How Video Games Are Made. In the “Dragon Age Inquisition” chapter he notes how prototypes as tools can have different terminology depending on the development studio - i.e Blackbox, Gray Box.
Since “Prototype” can mean different things - “proof of concept” to showcase internally, a prototype for a game convention, a prototype of a feature, location, etc. - how one approaches it depends a lot on time, scale, scope, and studio. This article sets up a framework, followed by an example and the headspace needed to navigate this milestone’s uncertainties. Painted in broad strokes.
Where is the fun?
Fun as a term is packed with Noise (opposite to signal). Players have interpretations of what it means, and developers based on their discipline can see it differently. So start by translating the pillars defined in the deck and detailed in the GDD into gameplay.
Imagination meets creation. If a game or game genre really resonates with the team get these points across to see if what’s imagined makes sense as gameplay. Stick to the core-loop, without expanding the meta (unless the meta is part of the game’s USP or innovation)
Code
If it’s built to scrap or build to last - you’ll need to establish adjustable variables, collisions, basic control schemes, and simple AI functions. I.e: for a genre like turn-based strategy games, add finding the right attack order, effects activation (start or end of turn), stats (vitality, attack power), grid, and character positioning.
Art
If an artistic direction is what sets your game apart try in broad strokes to present at least one part of the environment and a character. Let’s take the example of a time-management mobile game where the controls a rock band. All we needed was a 2D background of a garage and a 3D model of a character that we used to check how the band “plays”. This way we could visualize a guitarist, singer, bass player, and drummer while only having one model. That was enough to get the game rolling. With suitable lighting adjusted in the Unity Engine, we felt comfortable with the art direction.
Camera position
The depth, angle, rotation, and camera positioning are crucial for establishing what the player sees at any given moment. If you have played any action-adventure games like Devil may cry, you realize how critical is camera movement in any given genre. Some have it easier than others.
Core Loop
Don’t try to put everything, rather focus on executing a good chunk of your title’s core loop (main gameplay). What is the main action that drives the rest of the game forward - the jumping, shooting, matching, looting - showcase that, and make it feel good.
Story
The narrative structure - story, world-building, characters, tone, and dialogues - shouldn’t come as the first priority here, as we are in search of the mechanics first and foremost. However, if it’s one of the core pillars of a title, it needs to be flashed out during this phase, do it early.
The purpose of this Milestone is to check in-engine if the type of game you set out to make makes sense.
Thinking of “Marvel Snap” - from simplicity to complexity
Hearthstone Battlegrounds and Marvel Snap both share turn-based card game mechanics, and I’ve poured a fair share amount of 200+ hours into each of them. At the Game Awards and DICE 2022 Snap took home the “Best mobile game” trophy on good merit.
So as an example (deconstruction process), I found it useful to think about how to prototype such a marvel. This is not “reverse engineering”, as no two teams approach prototyping the same. It’s an individual’s observation of one of many paths to go about it. A thought experiment is you will:
Let’s imagine that the visual direction has been established and documentation is on point. The engine and tools for development were chosen. The team is looking to present the fun factor of the Core Loop (no meta at this point).
Classic mechanics of turn-based card battlers - Start and End-Turn with a max of “n” Turns. Because one of the pillars of Snap is fast-passed gameplay, 6 turns just feels like the sweet spot. Arriving at this sweet spot requires the system to be in place and tinkering with time (match length). Snap has a mechanical twist on how turns play. Opposing players play their cards at the same time, and once the End ”Turn” button is pressed, the turn plays out for both. At the end of the game the player with the highest “Total Power“ wins
Energy as a resource - this one is a rule of thumb for card games, as players begin with 1 energy and get +1 each turn (if other mechanics don’t intervene)
Deck & Hand size: There are 12 cards in a deck, the starting hand has 3 cards with a maximum of 7 cards in hand at any time. On average players draw an additional card at the beginning of each turn, unless abilities come into play
Card power and cost - approaching cost is straightforward. Minimum 1 - Maximum 6. Power ties to a character’s innate characteristics. The Hulk has 12 power, while The Infinite has 20, but can only be played if a player skips an entire turn.
Card mechanics: No abilities - i.e Hulk has 12 power, that’s enough to Smash - and On reveal: “If your opponent played a card here this turn, +3 Power“ (Starlord)
Location mechanics- if there’s anything to learn about Valves Artifact, is that it had some good ideas about the number of boards, but it was tedious to manage all that complexity. Having 3 locations with a max of 4 cards provides variety and strategic decision-making (trade-offs) without it becoming a hassle. Similar to cards, the “abilities“ of locations vary.
Prototyping could begin with just one location and its ability - i.e “Destroy the other locations.”… Worldship anyone? Next, create one deck of 12 cards with “x” having “No abilities“ and “y “ with “On Reveal“. Give this deck to 2 different players (or an AI) and see how the games play out. If one location and 6 turns feel good to play, start imagining scale within boundaries, and add one more location without adding new cards. As locations are revealed at the start of the turn and are distributed randomly, this provides the needed chance encounters required in such genres. Observe games with just two locations. My mind goes…if “Total Power“ at the end of turn 6 determines the winner, having 3 locations we can make sure players won’t just place all their eggs in one of the two locations. This adds strategic depth to the game - do I fight for this one location or just misdirect my opponent and win the remaining two?
Note: flip priority and Snap mechanics can be intuited from the GDD or arrived at through experimentation.
Both of us had the spidey emote on, but I didn't get the screenshot in time
The headspace
Prototyping requires a very open-minded approach. You should see the end result, while filling in the gaps, without ever being certain, is this what will ultimately play and look good. Developers are a very brave bunch. Knowing full well that failure always haunts us, success is not guaranteed, we push into the creative unknown, and bring worlds & experiences into the real world.
What was going on in the heads of the team prototyping Kratos’ Leviathan axe in God of War? I don’t know, but God, this series rocks… pun intended! Whatever the headspace the Santa Monica team was in after Cory Bralog came up with the pitch, helped them create one of the best weapons in gaming history. I tried going there:
How does it control? What do button inputs result in the player seeing?
Unlike the Blades of Chaos, the axe can be thrown, but what next?
It retracts after being thrown with a press of a button, like Thor’s hammer!
What are the upgrades available?
How does it fit with the new camera angle?
What does input R1 and R2 on the controller do?
Needs to feel heavy and meaty upon connecting with an enemy - let’s take one attack animation and an enemy “Get Hit” - what needs to happen on impact?
Feels different from the Blades of Chaos - those are dual wielding and much faster and focus on crowd control
As we are in the Norse mythology setting it’s imbued with Ice, not Fire or other elements
It will have Runic Attacks, but how many? I.e. let’s stick with two“special skills” (Runic attacks), as they map out nicely in the chosen control scheme
How does it connect with enemies, the timing of it
I could be well wrong, but my guess is they didn’t choose many items for prototyping but stuck with what mattered most and felt intuitive.
Everyone can stuff their prototypes with showcase features. “Fun” shouldn’t be explained, but experienced. Stick to a solid foundation and build upon it in the next milestones.
Now prototyping in AAA studios is not the same as in an indie studio or one that sets out working on The first tile. The costs involved to make one should guide your decision-making. No studios have the luxury of unlimited experimentation. As Voltaire said “Perfect is the enemy of the good “, so learning when to stop prototyping will do. Focus on the consequences, not on probability (this is probably better than this). There is a long development road ahead of us and the game can change during its Alpha and Beta stages.
Credits: Artwork created by Tanya Timosina
