Frozen Synapse 2 preview

When I meet Paul Kilduff-Taylor, the co-founder of Mode 7 games, he’s just come from telling a panel audience how much he screwed up his last game. Frozen Cortex, he explained, was a tactical combat game dressed up as a future sport simulator – and it turns out there aren’t many people in the Venn diagram where those fandoms intersect. This was a “conceptual failure”, he announced – a badly-chosen “core-fantasy.”

His new game shouldn’t have that problem. Its core fantasy is essentially as follows: go anywhere, steal anything, and screw over anyone in a living cyberpunk city. 

Let’s wind back a few years. Frozen Synapse – Cortex’s precursor – was a futuristic turn-based tactics game in which players simultaneously programmed orders into brainless but heavily-armed clones. Although each actual turn lasted only five seconds, the game let you assign speculative orders to enemy units and simulate the outcome of any possible encounter – creating long, paranoid stretches of planning and counter-planning punctuated by bursts of savage consequence.

But for all these choices, there was no sense of freedom. Its prescient frame story about humans being sidelined by high frequency governing algorithms emphasised the fact that the player was only ever there to execute other people’s ideas. Spinning webs of glowing waypoints across the map, playing out dozens of scenarios in a microsecond of fictional time, you yourself began to feel like an algorithm, deployed for one purpose only.

Frozen Synapse 2 is a little different. I’m looking down at a procedurally-generated city in schematic. AI players are hiring troops, launching missions, and making deals with each other. Paul, laptop open before us, is promising that I can dispatch my drone squads to any building on the grid, and that each one will be populated and generated according to its function (whether bank, hotel, or militarised safehouse). Everything from money transfers to hostage-takings should play out “on the map”, and therefore be open to disruption in real time. Frozen Synapse’s tight little confrontations have become playing pieces in a much larger game.

“This is about what would happen if you could decide how to change the future of a city,” says Paul. “You’ve been sent to investigate a force which is making incursions into the city. But within it there are different factions which control different elements of the infrastructure, and some corporate factions who are all vying for control, and they all relate differently to this incursion force. Some of them want to destroy it, some of them want to ally with it, some of them want to capture it and study. So the idea is that you will talk to them, you’ll get to know their leaders, you’ll figure out what they want.” Then you’ll decide which ones you trust with the fate of the city – and try to guide it towards the future you prefer.

Blue Sunlight, the religious faction from the first game, is here, occupying a chain of temples (Paul checks inside one and puzzles over its occupants – a priest and a flamethrower soldier). The activity log says there’s a police force which has just hired a bunch of soldiers. There’s even a logistics company, the Parcel Tossers (very definitely a placeholder name) whose deliveries will physically happen in game and whose friendship opens up “interesting” options. “I want to capture some of the idea of how a city is a messy compromise of things,” says Paul. Faction leaders can be negotiated with using Civilization-style dialogue boxes, and their groups should spend money and maintain their forces according to the same rules as the player. Some give out mission contracts, but you can also raid them spontaneously – or they can raid you.   

A health warning is required here: FS2 is very early in development, and many of these features are still up in the air. Often, when I ask Paul exactly how something will work, he admits he’s still thinking it through, and sometimes starts doing so loud. So I’ll try to be as clear as he is about what is implemented and what is not.

To show me a typical mission he loads up a “spontaneous assault” on a bank whose buildings are a mix of procedural and authored content. The AI demonstrates new stealth routines, at first patrolling obliviously and then rushing to defensive positions when his soldiers – “vatforms” in FS parlance – are spotted. Unlike in the first game, the player chooses their own squad, using the money they earn to licence vatform “imprints”, and can customise and improve their stats. The vatforms themselves can now take conditional orders, which makes my head hurt when I imagine what it will mean in multiplayer.

There are other new tools. Paul’s team found that FS has a problem with building assaults: “You’d just spawn outside and get shot through the windows.” So now there are smoke grenade units whose bursts can be moved through but not shot through, temporarily transforming the geometry of the maps. There is also what Paul calls the “Frozen-Synapse-mobile”, an upgradeable vehicle which drops off your vatforms and deploys temporary cover.

But the real changes are of course to the context in which these fights play out. Paul wants players to encounter AI squads while out on missions, either deliberately or by chance; you might be trying to beat them to a goal, capture their VIP as a hostage, or assist them to gain their favour. Every building will have “an owner and a purpose”, and many may be full of civilians – perhaps even crowds of them. “I’m really keen that the city doesn’t just feel like a big collection of boxes,” Paul tells me. “The things you do have an impact on the people who live there; I don’t want it to be just a ridiculous playground.”

That is also why the player almost certainly won’t be contesting territory. Ian Hardingham, Paul’s business partner and the game’s main programmer, calls this “painting the map”, and both are intent on avoiding it. “It’s not about conquering. It’s about interacting with the existing state.” Instead, factions are competing in terms of money, technology, force strength, and, crucially, access to the incursion events, whose increasing difficulty will form the “backbone” of the game. Like MMO players disputing who gets the loot from a raid, factions will probably have a formal system for deciding who gets to fight the invaders – and therefore who gets to “farm” them for their technology.

Of course, you can go to war.. But doing sowill ideally have serious consequences, because both Paul and Ian would really rather you reached compromises with the factions than destroyed them, “trying to maintain a power balance with unpredictable people”. Their cited inspirations are Crusader Kings 2 and Alpha Centauri – games of fraught, sometimes hilarious co-existence with AI.

Like Alpha’s kooky tree-huggers and singularitarians, these factions are “goal-driven” entities with diverse beliefs (though hopefully, Paul says, less “caricatured”; two extra writers have been hired to flesh them all out). The intention is that there will be multiple endings depending on what combination of them comes out on top, and that combinations will be key – that the player will need to decide what kind of coalition they are willing to take into the endgame.

Through all this, Mode 7 is committed to “systemic” design. By this they mean that the player’s experience should arise emergently from a set of interconnected processes which unfold naturally according to their own rules, rather than being painted in beforehand by a designer. Paul puts it this way: “The idea of interacting with a unified, complex system is the thing which is exciting and romantic about this kind of game. When you have things that are a bit more scripted, or things that are very kind of overly tailored to the player’s experience, I think you definitely lose that sense of wonder. We want to see if people find that as exciting as we do.”

I’m not the only person who hears all this and thinks of Subversiona procedurally-generated urban heist simulator which ended up as British indie gaming’s great lost project. How will FS2 avoid that fate? “We have core gameplay, in the game right now,” says Paul, firmly. “I’m not having a go at Introversion – they’re fantastic, and obviously hugely successful and talented, and very, very nice people. But the reason Subversion didn’t work is because Chris [Delay] could never find the game. We have the game.”

Soon players will have it too. As with Introversion’s Prison Architect, FS2 will come out first as a closed beta, slated for November this year. Many of its more ambitious features won’t be implemented yet and it probably won’t have an endgame, but it will be playable, and there will probably be no Parcel Tossers. Multiplayer – for many people the real joy of FS1 – will also be expanded.

In one sense, FS2 is aiming in an opposite direction to its predecessor. FS1’s levels followed a classic videogame pattern of decreasing complexity over time: after enough turns, everything in them was either captured or dead. But here, if Mode 7’s plans work out, the AI players will act as renewable complexity generators, breathing life back into the system right up until it stops. So if FS1 was a story of humans being mapped, predicted, and contained – bit parts in someone else’s program – FS2 offers the tantalising prospect of non-humans unchained, both as enemies and friends. For me, that is the really exciting futuristic fantasy: not of freedom, but of peers in the machine.