"It's always been a good idea to make a game for a niche," Simon Roth tells me. "I've always found it quite bizarre that people go after this mass market. People quit the industry and go 'I'm going to make tons of money, we're going to make a mobile game with micropayments,' and they don't actually know what it's going to be about. I've known so many people who've done that."
He's talking, in part, about his game Maia - a hard sci-fi god game of moon colonisation which nods firmly to the great era of Maxis and Bullfrog management sims. However, he agrees that in more general terms Kickstarter is also helping the industry return to making games that a relatively small group of people will adore, rather than slightly impressing everyone without bewitching anyone.
"When you're making indie games you always have to make a game that you want to play. I think what Kickstarter has done is underline that. People didn't understand that this was a perfectly valid business model. Now with Kickstarter funding people have realised that you can make these niche titles and that's OK. Take Frozen Synapse, for example. Obviously a very niche title, but to say that it wasn't successful would be crazy.
"I think someone described it as bottom feeding indies, saying that if you cater for the smallest niche, you can actually be quite profitable because on the internet your niche still consists of thousands of people."
Roth picking out Frozen Synapse, Mode Seven's brilliant turn-based tactical squad shooter, is not completely coincidental - he helped out there prior to founding Machine Studios to produce Maia. Even before that, the genial young Englishman has a strong pedigree, interning at Natural Motion when the company worked on GTA before moving on to David Braben's Frontier Developments. Roth also managed to squeeze in a stint with Super Hexagon developer and indie idol Terry Cavanagh prior to the Maia project.
Having worked his concept to the stage of proof, Roth then took to Kickstarter and IndieGoGo to make the final financial push, rounding out his Kickstarter with just over £140,000 from an initial target of £100,000. That cash injection might have been vital, Roth tells me, but the usefulness of the crowd-funding exercise extends far beyond the gathering of capital.
"Rather than just the pre-order model just being important, it's the ability to build your community very quickly."
Roth on the importance of Kickstarter.
"I knew that it was viable - the Kickstarter was literally just a kickstart to our development," he explains. "I'd been working on it for a few months already. The plan was to get it to this Alpha stage and then start selling it, much like Minecraft, if we couldn't get the Kickstarter going. But we got it, so we didn't need to worry.
"Rather than just the pre-order model just being important, it's the ability to build your community very quickly. We had a community that was quite hardcore, that had spotted the game quite quickly - but I've gone from having 100 people interested in the game to about 9000 now. It's a crazy jump. With indie games you're so reliant on word of mouth marketing - I have zero marketing spend. Having people who will evangelise your game for you is pretty important if you want to get the money to build something. It wouldn't be feasible otherwise.
"I genuinely think that there is quite strong community in the comments on my Kickstarter - people would be discussing things, saying 'oh, have you seen this other game?' You see on their avatars who else they've backed - they sort of go around recruiting from other Kickstarters. Exactly 25 per cent of all of my money came from referrals on Kickstarter. I'm sure a lot of that is from people going between projects.
"Most of the people who have backed my project have backed four or five other games, too. On IndieGogo there doesn't appear to be much of that. I had presumed that before, but I now have the data to confirm it. Kickstarter has a massive headstart in that sense. IndieGoGo is getting better, but it's not yet the same."
That community is and always happy to offer advice, occasionally veering into vociferousness, but Roth has been sure to maintain a firm hand on the tiller of development, ensuring that the game retains a coherent creative vision.
"That was actually one of the things that I hated about AAA development," he ponders. "A lot of things got stamped on early because too many people were giving their opinion and they didn't seem feasible. With Kickstarter, I think a lot of people are putting their trust in me to make the right decisions. I do take a lot of feedback from people, and often they're completely right, but when I genuinely don't agree with someone I do say, 'no, sorry.'
"You have to learn to let people down gently, because they are emotionally invested, but at the same time I think these people have the trust in me to make the right decisions. When I have gone back to people and said politely why I think that their idea isn't great, they're quite understanding. Because I will talk to them about it I think people are much more accepting. If it was someone like EA, when they ignore the community, they get very frustrated because there's no communication, no 'why'. I can come back and say why and they're much more accepting of that. So I still have strong control, probably even more than a large studio would have."
The more I chat with Roth, the clearer it becomes that there is a wealth of underlying lore and structure to Maia. Everything is there for a reason, everything is, to an extent, logical and fitted to the '70s influenced core sci-fi framework. He's a man who does his homework, to whom it's clearly important that his world hangs together believably rather than existing merely as a veneer to justify a fun set of game mechanics.
"You have to learn to let people down gently, because they are emotionally invested, but at the same time I think these people have the trust in me to make the right decisions."
On the benefits and perils of an engaged community.
"I think that's the good thing about being able to Kickstart it, being able to take our time with it, to really design it. I was obsessive about getting the sci-fi correct, making it hard sci-fi and making everything correct. I was so obsessed with getting the details correct that I found a star system that we could potentially travel to, that would potentially have planets. I did loads of research, emailed an astronomer friend of mine and was told that I had picked the wrong system, that there was no way there could be planets there. Then the week before last they actually did find planets there, so I got a phone call saying 'Simon, you've predicted a planet!'
"It's in the Tau Ceti system. Astronomers had decided that you couldn't have planets around this star, because the star's metallicity (the proportion of its mass not made up from Hydrogen) was too low, it didn't have the elements required. There was a dust cloud in the area around the sun that I thought obviously had planets in, but everyone said no, we're sure of it. Then three weeks ago they found them.
"It's good to be able to go for these little details, people appreciate that handcrafted-ness. We can do these things because we can carefully think about things, it's much more involved than trying to do it quickest and cheapest, like you'd have to do at a number of other studios."
That attention to detail isn't just limited to the prediction of H-congruent planets in far-off star clusters, however. Every design decision in the game is riveted firmly to an underlying principle which fits snugly into the mise-en-scene Roth is crafting. I ask him about the game's distinctive first-person view, used when the player chooses to directly control one of the robots which populate the game's bases.
Instead of crystal-clear robo-vision inscribed with random feeds of numbers and shimmering vectors, Roth has chosen a visual style which evokes pointillism - a blurred collection of loose blobs which nonetheless result in a perfectly understandable representation of the world around you. It feels oddly low-tech and completely appropriate, and is the result of just as much careful thought as any other of Maia's design decisions.
"There was a whole raft of reasons behind this," Roth tells me, warming to the subject. "Firstly I thought it would be wrong to have a straight possession mode, with this really crisp, rendered view. It was kind of boring and there was no reason for it. I wanted something interesting, to re-imagine what it would look like through the robot's eyes, because obviously they wouldn't need video camera eyes, they would be interpreting the environment themselves.
"There are lots of different things. The shader is actually quite complex, but I looked at how our eyes see - we see the details of the image differently to the colours of the image. So I thought what if I break that up, like if you built your own camera to try and simulate your own eye, how would you do that.
"Then we looked at how we do machine vision, how we detect edges in images and stuff. I was also inspired by an Isaac Asimov story, where a robot who was designed for the moon is accidentally released on earth. He's wandering around and he can't make any visual sense of what's going on around him, because he doesn't have any visual frame of reference. I thought that was interesting to try and visualise what that might look like."
Quite the purist, then - but Roth is also quick to admit that there were pragmatic bonuses too.
"I'm not simulating every individual tooth in someone's head."
On where to draw the line with complex simulation.
"There's also some technical reasons, because it means I don't have to make the texture maps and the shaders look good up close,"he chuckles. "In Dungeon Keeper, you could posses your creatures, and every creature would have its own sort of character through its vision. Originally, when we were going to have more robots, we were going to do that, have every robot having its own signature way that it sees the world."
I get the impression that Roth is the sort of person who could quite happily keep adding new layers of realism and interdependency to his game's mechanics until the cows came home, but he's also very aware of the dangers inherent to that approach. Discussing the clear influence of the notoriously Byzantine Dwarf Fortress, Roth explains that he'll never go as far as that game when it comes to modelling the most minute of possibilities.
"The complexity thing is difficult," Roth says, the subject seemingly weighing heavy on him. "I'm on the DF forums and they were the first people to pick up on the game and get excited about it. The thing is, a lot of the complexity in Dwarf Fortress is very trivial, some of it is quite unnecessary. A lot of that is important to the character of the game, but at the same time I'm taking some of those things from the complexity and trying to feed them back into the gameplay. Things like my eco system simulation being directly linked to the gameplay and the complexity of the lighting and the fluids and things.
"The Dwarf Fortress guys either love that or hate it, but I'm not simulating every individual tooth in someone's head."
That's precisely what Dwarf Fortress does, among a wealth of other seemingly completely trivial information, like allowing a dwarf's eyelids to be injured, resulting in lost sleep. Teeth lost in battles will also litter the floor of Dwarven strongholds, upsetting other dwarves and generally making the place untidy until they're cleaned up. This sort of incredible level of simulation has proven to be both the thing which attracts the uber-core to DF and drives almost everyone else away, resulting in a massively obscure yet ultimately tremendously rewarding game which is only played by a criminally small audience.
As one of those players, and not a very good one, I'm interested in knowing where Roth intends to settle on the spectrum of realism.
"I'm taking that bare minimum of accessibility and going far above that. So I'm trying to remove all the GUI that you see in these management games, where you get stats, lots and lots of stats. As soon as you start getting stats in a lot of games, you start trying to game the system, you optimise your build. We're going to be doing that much more organically.
"So we have complexity, but we show it through much more subtle real world things. So when a character is hurt, they're not going to have damage stats, they'll be bleeding from different parts of their body. If something is damaged, it'll start smoking. You'll need to keep an eye on things by watching them, to see how they're performing, rather than having a page of stats that says: 'this is damaged'.
"That's hopefully more accessible than a lot of city builders and management sims, which can get a bit bogged down," he adds. "And I hate coding GUIs anyway."
An alpha build of Maia will be available to backers for free and for a fee to other later this month, with the full game arriving sometime this summer.