War Robots: Frontiers is an evolution of the original War Robots mobile. It's a completely standalone project made from scratch to expand the franchise to the bigger platforms.

The core of the game is the same: 2 teams of 6 players fight over control points. The goal of the match is to hold the points long enough to accumulate required victory score. Each player has a hangar of 5 robots + titan at their disposal. Each mech is unavailable once destroyed. So in the matches, there can be at least 2 high-level strategies - win with control points score or win the war of attrition just by killing all enemy mechs.

Apart from signature gameplay one of the project's USPs is mech builder. Player robots are not set in stone. Player is free to construct their own mech from various parts within a set of limitations.

So, the goal of each team is to accumulate 1000 victory score by holding control points. Score is granted over time to the team holding more control points on the map (CP number is always odd). To capture a control point player's mech needs to enter the CP radius and stay inside uncontested for a period of time. Once captured the control point stays under the team till it's recaptured by the enemy.

MOVEMENT

Player mechs have few mobility options:

• walking - just regular way of moving
• dashing - in clutch situations mechs are able to dash a short distance. Dash spends fuel resource.
• jumping - mechs can launch themselves in ballistic trajectory. Think of Warhammer 40k jump packs. Jumping spends fuel same as dash.
• hovering - by spending additional fuel after jump mech can hover in the air to gain height advantage when needed.

All moves except walking spend a resource called fuel. Fuel regens over time. Think of it as stamina. The idea behind this movement system is to keep gameplay positional as in the original game. Though not as limiting as in the mobile game. Players can jump over pits and can clutch dash to cover when being caught under fire. But those moves are limited by fuel, so can't be used repeatedly. As fuel is a shared resource, each of the moves carries an opportunity cost as well.

SHOOTING

Each mech carries 3-4 weapons. Though all weapons are fired with a single button simultaneously. Partly this is a legacy solution from the mobile game. Though also it's a conscious decision to keep gameplay closer to traditional shooters and not turn it into a MechWarrior with overly complicated controls.

There's a plenty of weapon archetypes - SMGs, Miniguns, Explosives, Sniper weapons. Burst and sustained fire weapons, targeted and indirect, you name it. Weapon system and purpose will be covered later.

ABILITIES

Each mech is equipped with up to 3 abilities. All kinds of shields, healing, damage dealing and so on. Using abilities has no cost other than cooldown. Though some abilities may be limited by charges. They can be used only X times over the match. This is our version of consumables. This brings the game closer to hero shooters. Though mechs are freely constructed and equipped in mech builder, abilities are part of the game RPG system rather than a tool for creating mech identity.

When thinking about what a mech is we had 2 starting conditions:

• It should be perceived as a vehicle. We are a mech shooter, not a reskin of a usual human shooter.
• It should stay casual. No burden of knowledge on mech construction, and no unnecessary penalties.

Vehicle feeling: apart from art components (movement animations) and dynamics balance (feeling of weight, speed), this is also a gameplay mechanics requirement. What usually separates vehicles from living combatants? Our answer is 'being modular'. If the game is not hard-boiled MilSim, human characters are "whole". Our mechs are made of 4 parts - torso, chassis and 2 shoulders. This has some implications for gameplay mechanics and building principles.
No burden of knowledge: Typically different modules have different purposes. Providing mobility, offence, storing fuel/ammo, etc. They should have something to be meaningful. But at the same time, our mechs are vastly different in shapes and geometry. We wanted to have meaning to mech modules but avoid forcing players to learn the composition of each possible mech detail and its purpose. Like in tank games players memorize where are fuel tanks in each tank.
No penalties: Another typical part of vehicle shooters is penalties. Destroyed wheels limit mobility, and damaged weapons limit offence. But the vehicle stays functional even with some structural damage. This creates this feeling of modularity and being a vehicle, not a human. But we wanted to avoid that as well. There's no need to slow gameplay down or force players to play a half-functioning entity.

So together this was quite a conundrum: we wanted to have modularity, but avoid all typical implications and effects that modules bring to the game. Our solution to that was the HP system:
Mech is made of 4 modules, and each module has its own HP. To kill a mech you need to destroy any one of the modules, doesn't matter which particular. Before damaging modules player needs to get through the energy shield protecting the mech.

In my opinion, that's a pretty good solution. It fulfils all the conditions above while also creating some interesting dynamics.
1) This mechanic allows having more weapon dichotomies. In addition to typical burst/sustained, direct/indirect there's also spread/precision. There can be weapons that deal high DPS but distribute it over the whole mech with spread. Or there can be low-ish DPS weapons, but their precision allows the player to focus a module and keep high lethality. So there are weapons good for softening target in general and weapons better for finishing an enemy off.
2) This HP system separates good players from great. By utilizing all mobility options and map elements better players can distribute incoming damage over the mech increasing lifetime.

So overall this system gives us all - creates a mech feeling in gameplay mechanics, provides more design space for content and rewards players for skill. Nifty I'd say.

So, one of main USPs is ability to build your own mechs out of components. There were many ideas what our mech builder would look like. But after all we ended on Front Mission variant.
Mech has a torso, chassis, and 2 shoulders.

• Torso carries a signature ability attached to it. It functions as the core of the mech, and strongly defines playstyle as you can't get rid of the ability in the torso. Also defines mech energy capacity.
• Shoulders usually are for carrying weapons, they have 1-2 weapon slots each. Though the torso can have a weapon slot as well
• Chassis define mobility parameters - speed, fuel limit, fuel regen. Also, define mech weight capacity
• In addition, each mech can equip 1-4 weapons and 1-3 abilities.

We considered more elaborate mech builders. Some were Crossout style, where you build mech out of like lego parts. But in Corssout you build only one car at a time. We have 5 robots in the hangar, so we need to avoid fatigue. Constructing a mech should be streamlined, like using an installation wizard. Select chassis (how much you can carry), select torso (define playstyle) select the rest within the limits one by one.

This construction system operates on two limitations - weight and energy capacity. Each part that can be installed has a requirement of weight, energy, or both. Most of the weight limit goes to mech parts, most of the energy limit goes to abilities, weapons need both in some proportion. This alone creates some build strategies - mechs focused on abilities have less energy to spend on weapons and vice versa. Though it's more elaborate than that.

But that's not a hard rule. There's a variety of content with different requirements to support more mech building strategies. Mech abilities are made in the way to open such an opportunity. There are different types of abilities:

• Regular abilities - moderate power, standard energy requirement
• Consumables - moderate power, low to zero energy requirement, can be only used fixed number of times
• Ultimate - high power, moderate to high energy requirement, need to be charged like in Overwatch. unavailable from start.

So, while there's standard equipment following standard rules - weight for details, energy for abilities, both for weapons - there are workarounds. One player can free some energy by fitting consumable abilities only and therefore equip more energy-hungry weapons (provided they have weight capacity). Or there are old-school kinetic weapons with lower energy requirements so players can fit ultimate abilities into the build with more free energy. But that's closer to talking about our RPG system.

There's a story behind that, but to make it short, there are a lot of analogies with CCGs in our mech builder. A robot is a "deck" of details. Each detail has weight/energy cost like cards have mana cost. And first of all, decks have archetypes: aggro, midrange and control.

• Aggro: plays under the mana curve (cheap and fast) and focuses on dealing face damage fast.
• Control plays above the mana curve (skips early turns, ramps mana, high-cost cards) and focuses on attrition rather than face damage. Waits till the enemy runs out of resources to defend.
• Midrange: a bit of both

In our game:

• Face (nexus or commander HP) is our victory score. One can gain victory score without dealing damage to enemy mechs. Like in card games, you deal damage to face rather than enemy creatures.
• Creatures are robots. You deploy them, you use them to deal damage to the face or other creatures.

Therefore what are our archetypes:

• Aggro - light mechs focused on interacting with control points. Work towards win condition directly. Low weight/energy usage, high mobility, low lifetime.
• Control - heavy mechs focused on eliminating other mechs. Low speed, high survivability, and damage. Due to low speed, they are last to get to control points therefore less interaction with "face".
• Midrange - Medium mechs. Average stats and performance.

GREAT! I've just invented light, medium, and heavy mechs, what an achievement! But there's more to it. First, this framework already gives some guidance for content design so we focus on some options and avoid others. It guides energy/weight costs for mechs of different classes. Second, there's mech builder. Within these archetypes, players have different options. They can spend their energy/weight within the archetype differently. How player builds the mech also correlates with their desired playstyle.

• Gunner - focus on weapons. Spend more cost on weapons, and less on abilities or body. Gunplay focused inside the match.
• Midrange - balanced build between weapons and abilities.
• Caster - invest a lot in abilities, and spend less on weapons. Tactics focused in the match.
• Combo - a subtype of caster that picks a specific set of abilities with high synergy but also high opportunity cost.

So this creates a grid of 3x4 robot classes

I hope class names give some insight, but to provide examples or explain more specific to our game classes:

• Rogue - light robot doing a bit of everything associated with being light.
• Scout - even lower fighting capabilities, but stats and skills designed to take control points FAST. Exploits abilities for that like haste, teleport, and control point lockdown.
• Assassin - a set of weapons and abilities to one-shot enemies. Harder to build within energy/weight limitations.
• Soldier/Brawler - absolutely average mech. Walks, shoots, casts.
• Support - medium mech invested in abilities. Typically this is a support/healer build as heavy and light casters push this out by doing something else better.
• Bladedancer - unspecified medium mech, really depends on abilities set but does something cool.
• Leader - heavy support. Think paladin. Mech designed to lead the charge but not very good at fighting on their own.
• Sniper is being our combo caster due to our mechanics. Sniper rifles are usually burst low DPS high precision/lethality weapons. This is emphasized further with self-buff abilities. Like damage buff, instant reload. So there's a lot of potential for one-two-shot combos.

This grid covers all core archetypes and playstyles as I'm aware. But we are yet to see how inventive our players are. To support this playstyle grid and give players some additional guidance there are pilots

What are our pilots? They are a lot. First of all, they are people. People tell stories, people can be related to. So we wanted human figures in the game to have more opportunities for storytelling, therefore community management and marketing. We wanted players to have something or someone to empathize with. But also pilots are additional ways to customize the build.

We didn't want to make pilots another "stat stick". Mech parts are for stats and we don't want duplicating content. So our pilots provide conditional bonuses. Each pilot has class and personality. Generally, class tells what the pilot does and personality tells how the pilot does that.
For each mech class, there's also a pilot class. The pilot class provides bonuses to relevant mech parts, weapons, and abilities. While personality tells when those bonuses are provided. For example, there's a 'cautious' personality. Cautious pilots get bonuses when enemies do not see them. Or there are reckless pilots, who get bonuses within enemy proximity.

Putting this together, there's a pilot class. Let's say flanker. Flankers get bonuses for SMG weapons and reload speed. There can be a cautious flanker that gets increased reload speed while no enemy sees them. This makes them natural hit-n-runners. But there can be a reckless flanker as well, who gets reload speed bonus when near an enemy. This one can be a more dueling hit-n-run pilot, but maybe it's an assassin with a quicker reload to finish an enemy off if they didn't do it on the first pass. Or there can be a smart flanker. Smart pilots get bonuses to abilities. This one will find use in combo builds.

This system helps the player to build a mech. Pilots they want to use give a hint on what equipment type to use, how to distribute weight/energy capacity, and how to play in the match. But they aren't limiting and promote unusual combinations. Assassin pilots designed for light mechs can find some use on snipers, for example, as both share combo and unseen playstyle. Or maybe even some ultimate combat support that's able to chain-cast.

In addition to mechs in each hangar, there's a Titan or Alpha. Titans serve as a one-time ultimate. First, the player needs to accumulate charge to summon a Titan. They do it by doing something useful - dealing damage, supporting, capturing points. It happens naturally over the game. Titans are closer to characters. They are fixed entities that can't be built, only customized with weapons and some stat modifiers.

Titans were one of the toughest parts to design. We didn't want them to be just "a bigger robot". We wanted them to do something unique, and pivot gameplay from regular patterns. Especially considering that standard patterns do not fit them. Because they are big, they are easy targets. They can't utilize covers and have lower mobility than regular mechs. Therefore much lower dynamics and fewer ways to protect themselves. They just stand and shoot.

We tried a lot with them. One variant was "gun platform". Regular mechs carry 4 weapons, usually 4x of the same. We let our titans carry 4 different weapons and switch between them. Like in traditional shooter. But that didn't work well within our dynamic. The player piloting the titan didn't care about what weapon to use while titan's targets didn't know what to expect. It was "random bullshit go" type of gameplay. Then we tried "space cruiser" model. Inspired by games like "Nexus: the Jupiter Incident" and "Battlefleet Gothic". In space cruiser games player controls huge space battleships. They are slow but carry a lot of weapons and other options. Main gameplay there is energy management. Like how much to give to shields and how much to weapons and to engines. So dynamics and decisions aren't in movement and shooting, but in juggling efficiency of those. We tried to emulate something similar with different stances. Like this is shooting "energy to guns" stance and this is a defensive one with a shield and no shooting. Didn't work as well within our TTK and match speed. Situations either change too fast or too slow to make stances meaningful and not a chore.

So we ended up with betting just on wow-factor. Titans share class with regular robots. In some sense, they are just bigger robots. But they are exceptional at what their class is supposed to do. Nothing too fancy gameplay-wise, but jaw-dropping in implementation. If it's a scout titan maybe they can teleport freely across the map, for example, making them ultimate capture robots. I'm not entirely happy with that, but maybe it's for the best. As it's easy to pick up and jaw-dropping is good for monetizing.

Speaking of monetizing, let's discuss meta a bit. With our mech builder, we felt that CCG is a really close analogy for what we do. We want players to obtain individual details and put them together the way they want. It's deckbuilding.

So our content distribution is close to random boosters. But we made it reverse. Each day shop offers a random set of content and the player can decide to buy it or wait till refresh. With the option to reroll shop.

As you can see on the screen, prices give a hint to buy the whole package instead of individual desired parts. This preview serves a lot of goals. First, it works around gatcha regulation current and possible. Selling content in bulk we give players additional options for their build they wouldn't try otherwise, but those are also a hook, as now they have an incomplete build so this will drive additional purchases eventually. And no booster is perfect. So one may gamble a bit, or a lot, with refreshing offers to get more desired parts in package.

In addition to that, there's crafting similar to Hearthstone. Unwanted details can be dusted and from dust, player can create any detail they want. Also, we employ the full range of standard monetization elements like battle passes, vanity options for players, and premium status.

For the game to be sustainable within this model, we need to provide good variety and scalability to content. There should be enough design space to make new parts. As it's not implemented at the moment, I'm not at liberty to say what exactly is done in the field. It's rather easy for weapons and abilities, as they are "physical". And somewhat harder for robot parts, as they are just stat sticks, and there's a limited number of combinations of stats that make sense. But once current opportunities are exhausted, there are ways. It's possible to introduce passive traits to robot parts to make them unique with the same stats. Or even allow some ways to customize what's "inside" the part. I'm pretty confident that this came has good scalability.

This turned out to be much longer than I wanted it to be. But well, it's a whole live ops game. Hope my reasoning for the decisions described is clear and the cases are interesting enough.