1. Bring in both the casual and the hardcore players
Clash of Clans originally had a much more cartoony and casual look than its current form -- in fact, the game went through numerous visual alterations before the final look was settled on.
"We had this notion that maybe the hardcore players would actually dislike this, and think it was too childish," explains Louhento. "So we had to find a sweet spot, such that it wouldn't alienate the casual players - nothing dark and black and evil and realistic - but on the other hand, it couldn't be too blubby."
This wasn't a quick chop and change, admits Louhento. "It took us a while - it wasn't an easy task!" he says.
What the team eventually settled on was a mixture of realism, and a "super-deformed, Japanese style," adds the project lead. "I'm a huge fan of Pixar, and their characters are appealing to a younger audience, but at the same time, they're cool for adults too. We're also big fans of Capcom characters - strong character art that's really polished."
Within the team of 5-6 people, more than 10 different character concepts were brought forward and abandoned before they found exactly what they were looking for. Discovering that perfect mixture of both casual and realism in the visuals was key to pulling in a wide range of players.
It's not just the visuals, of course. Casual and hardcore players want different types of gameplay, and attempting to mix these together can be tricky.
For the hardcore, Clash of Clans offers online battling elements . "I think there's something about the competitive nature of the game," says Louhento. "We have leaderboards, we have that kind of edge where people think 'Oh I want to be there, so I'll need to upgrade this.' The progression, I think, is more visual."
And for those players who aren't so into attacking each other, there's enough to keep them entertained on the side.
"It's the social elements too," he reasons. "You can chat to clan mates, and donate troops - that feeling of belonging together is really powerful. Once people get into a clan, they are really invested and willing to play for a long time."
2. Have a clear goal from early on
"For the first two months, we did a company-wide demo, and everybody played inside Supercell," says Louhento. "We coded all the basic functionality, all the character behavior - I think we only had the Barbarian character back then though."
He adds, "It was multiplayer, running on a server. And it felt really good."
This was a bit of a blessing for the team -- to have a game that felt good to play from the get-go -- but it really all came down to knowing exactly what their goal was from the very beginning.
"It was an easy project in a way," he admits. "Obviously we put hours and hours into it, but the goal was pretty clear from early on. There were pieces missing, but the structure was there."
And Louhento puts a lot of this down to the company's tablet-first strategy. "We make these games for tablets - we can see how things are scrolling, we think about the framerate, and we've done a lot of work on the gesture controls."
Louhento believes that you can tell when a mobile game hasn't really been developed with usability in mind. "Take The Simpsons: Tapped Out, for example," he says. "It's a great game, but they didn't really put in the effort to think 'Is this button big enough? Is the usability good? Is this really optimal?'"
"We all played hundreds and thousands of hours of Clash of Clans, just to try to iron out everything. If it didn't feel right, let's do it again - let's really make the scrolling and tapping work best," he adds.
And this isn't an new idea either. The public has been choosing the products with the best usability options for many years now, reasons Louhento.
"When Google and Alta Vista came out, they were basically the same thing, but one was a bit more minimalistic, a bit more clear, it performed quicker," he notes.
He continues, "I think that's a big part of it - usability. I remember when we were first looking at iPad games, and there were horrible framerates. We said, 'How can anyone create this UI? Has anyone actually played this game?'
There was plenty of gorgeous artwork and clearly huge amounts of effort put in to make games look great, he says, "but the interfaces and controls weren't done for tablets - you could easily see that it was a port. You could see the developers who had said 'Oh, let's port our PC game for iPad'. And it just felt sucky, so we wanted to make a completely different approach."
3) Don't overdo the tutorial
You may or may not be able to tell, but the tutorial for Clash of Clans was added at the very last minute. Indeed, just weeks before the game was launched onto the App Store, no work had been started on the tutorial at all.
Says Louhento, while it is of course important to teach your players had to handle the game properly, there's far too much emphasis put on tutorials elsewhere in the industry.
"We're not big fans of long, overdone tutorials," he says. "I know Zynga has a lot of what they called the 'Onboarding Stage', and they spend a lot of time and effort with it - there's a whole Onboarding team, and they specialize in looking at the metrics, and deciding things like, 'Let's make them click here, let's have a bigger button here, let's remove the 'x' button in this window' and all that sort of nasty stuff."
"We really don't want to do this nasty stuff," he continues. "If people like what they see and feel comfortable with the environment, then great."
Of course, as new elements are added to the game that need explaining, then the original tutorial will be updated to incorporate this. And, as Louhento adds, "If we see that the retention in the first phase could be a bit better, we play with it a bit - maybe that button is in a weird place, tweaks like that."
But in general, the project lead believes that there's way too much time and money splashed on teaching players every nook and cranny of social games.
4) Healthy competition is great for your company
Clash of Clans and Hay Day were developed by two different teams within Supercell, and the two groups would constantly have friendly digs at each other about how far they were through development, or the features they'd managed to implment most recently.
"We had this healthy competition with the Hay Day team," Louhento explains. "We'd say 'our controls feel better than yours', and every Friday we'd say things to the other team like 'we managed to build this in a week'."
"Then next week the Hay Day team would say 'by the way, we just did this big chunk of code in a week.' So there was some great, healthy competition. Funny competition!" he adds.
This was a huge part of what made development on Clash of Clans not only of a high standard, but also enjoyable.
"I've been making games for 20 years," Louhento says, "and I've never seen this kind of progress - and also joy from making games."
5) Don't overwhelm your players
Both Clash of Clans and Hay Day are updated every few weeks with new content -- new items, new in-app purchases, new characters and the like.
But updates such as these can be tricky. New players can potentially become overwhelmed by hordes of content, while veteran players may feel like new content messes with the equilibrium of the game.
Supercell' s solution is to only make new content available to loyal players, and bring new players in more gradually before throwing everything into their boat.
"A player coming into Clash of Clans won't see certain functionality for the first two weeks," explains Louhento, "and then once they are familiar with the controls, then you can complicate the game a bit, by adding pieces to it."
"We try not to mess around with the very beginning of the game, because we know it works," he adds.
Adding new content must conform with the game's balance between casual and hardcore too. "We're aware of the fact that we can't make it too complex, yet it has to have that hidden depth," Louhento notes. "From the outside it looks relatively simple, but it has that hidden depth."
And when it comes to keeping a complex game like clash of clans balanced, Supercell has a system in place that makes sure any new content doesn't screw around with the inner workings.
An automated testing simulation runs thousands of battles one after the other, throwing in randomly sized armies with different soldier types each time, and then correlates the data to see whether they're an obvious area in which the game can become skewed.
"There might be tricks that players figure out of course," admits Louhento, but for the most part he says that this simulation stage catches all of the bad balancing that could potentially ruin the game.