Home   >   Features

How 18th Parallel is putting mobile Android games onto Indian set-top boxes

Your phone becomes a smart software gamepad
How 18th Parallel is putting mobile Android games onto Indian set-top boxes

Playing the best Android games on your TV, with your smartphone serving as your control pad: that's the mission for 18th Parallel.

Based in Pune, India, the startup's flagship product is Cannonball Arcade, an Android console gaming software platform and managed service for TV operators. At its heart is a carefully curated library of Android games and an adaptive software gamepad platform.

18th Parallel already working for Airtel Digital TV (a national-level satellite TV operator in India) where the service is live, and now five other operators have signed up. The team is working with Vuclip across India and South-East Asia, and iscurrently talking to Western European partners. It has over 50 content partners and a catalogue of some 125 top games.

Fired up

Co-founder and CEO Nishith Shah attended both Gamescom and PG Connects Helsinki this summer, where he explained how their software works.

Tech embedded in the set-top box OS allows existing Android mobile games to be played as console games - even those designed to be played via touch, tilt or swipe. This enables developers to make their games available to Cannonball Arcade with no porting effort.

<em>Pocket Rally</em> is one of 125 games on Cannonball Arcade
Pocket Rally is one of 125 games on Cannonball Arcade

His team runs a full featured service delivery platform in the cloud; operators can offer a subscription service for gaming this way, and the service delivery platform integrates fully into the operator's CRM and billing systems, allowing the operator to offer a tiered subscription service using the standard Android APKs of the games.

Meanwhile, a mobile companion app for iOS and Android smartphones means consumers can skip the purchase of a physical gamepad.

The mobile companion app connects over Bluetooth to the set-top-box and serves as a context-aware smart gamepad. Buttons are large and ergonomically designed to facilitate heads-up gaming. The buttons are custom generated for each game (accelerator and brake when a racing game is launched, shoot button when a shooting game is launched, and so on).

We had the chance for an in-depth interview with Nishith Shah, and here's our three part guide to 18th Parallel's plans for the Cannonball Arcade platform.

The company and its games

The company and its games Tell us a little bit about the formation of 18th Parallel. What's its mission?
Nishith Shah: It was five people originally. Three co-founders, two engineers. All of us were engineers, really. There were five of us together who started off in my apartment – which is as typical as it gets.

The mission was to actually bring fun, affordable gaming to the TV screen. This was way back in November 2011 – before Android TV, before Google TV, before you even had a set-top box with Android on it.

Coming from India, it seemed very obvious. When you look at an Xbox or look at a PlayStation, it’s awesome gaming, but for a market like India, it’s too expensive. When we looked at mobile games, the first thought was, "Wait, it’d be great if that was on TV". Affordable, fun and innovative. It seemed like the obvious thing to do.

So we started off thinking we were going to do our own hardware, our own boxes and look at dedicated game consoles. But very quickly, we saw that Android-powered set-top boxes were coming through TV operators anyway.

So cable TV set-top boxes were going to be running Android, satellite set-top boxes were going to be running Android. And we thought it’d be much smarted to ditch the hardware idea and just ride that wave.

That made a lot of sense. It probably contributed heavily to our survival, because we would surely be out of money and be shut down long ago if we hadn’t done that.

We were fortunate enough to get a seed round the month we started – even though we only had a barebones prototype to prove the concept. This was five angel investors that came together, three of them were my former bosses. It was very validating, personally, and they got us past the first hurdle.

We've now got some money coming into the bank. Whether from a customer, an advance payment, or from an investment, something’s always shown up that’s kept us alive, and we have enough to run for the foreseeable future.

<em>Mechanic Escape </em>on Cannonball Arcade
Mechanic Escape on Cannonball Arcade

But we’re thinking of expanding, as we’re seeing really large opportunities everywhere. So we are in the process of trying to raise a Series A.

What kind of games you can play there right now? What have you signed up?
Mostly casual games, because our market is currently in India, and that’s what we’re seeing is working. We want to try to bring on some more platform games and see how that goes. We're experimenting a lot with our games.

It’s about 125 games in total. What goes live, varies a bit from operator to operator, because each of them has different set-top boxes, and we test every single game on that set-top box. Every once in a while, we’ll find a game that’ll crash or won’t work well and we’ll pull that game off that operator’s list.

Sometimes, we get surprised. There was one game that we really, really loved, but then it did terribly. We tried to give it some new positioning, and we refreshed the collateral, the splash-screen and the icon to make it as attractive as we could, but it didn’t work.

And then there was another one that we thought was really, really obvious. It was a very simple game that we brought on just to fill the catalogue, and it did remarkably well. So we are getting surprised as we go.

Every game we offer is part of a curated list. Every game is tested, and that automatically means there’s a smaller number of games to find in our catalogue. So if your game is on it you’re much, much more likely to get discovered.

We’ve got tens of thousands of boxes with lots of subscribers playing games and we’ve seen that only very few games in our catalogue – a really small number – have, as yet, gone unplayed, compared to the mobile game stores where a very large percentage have never been touched.

What’s your mission at events like Gamescom and Pocket Gamer Connects?
New games. Absolutely. We are continuously trying to bring new games onto our platform. By no means are we done with 125, we want to keep adding new games and we want the catalogue to stay fresh.

<em>Sunny Hill Ride</em> on Cannonball Arcade
Sunny Hill Ride on Cannonball Arcade

What do you think your next big announcement will be? Are you looking for activity in 2018?
We hope to do a big announcement later this year, because one of the customers we’ve signed up is another operator of the size of Airtel, and we are in the late stages of implementation with them, and we are expecting them to announce their Android set-top box soon. There are five more that are signed up and are in the implementation stage.

We’ve signed up Vuclip. Vuclip is launching across India and Southeast Asia. They’re very, very strong in Southeast Asia, that’s a big market for them, so we expect to go live in the region before too long.

We are also doing a few partnerships with set-top box platform providers in Western Europe. There we are signing up for one or two partners, who are well-established companies that power set-top box platforms for some of the larger operators in the region.

These partnerships are yet to be announced but they are signed and are in various stages of development.

The technology behind it

The technology behind it The key ingredient of your platform is that a mobile Android game is ready to go live on it straight away. There’s very little effort on behalf of the developer. Can you talk us through that process?
Nishith Shah: That also came from necessity. We were underfunded and we didn’t have the money to do any minimum guarantees. So we decided the way out was exactly this: 'Let’s create a platform that developers have to do minimal work with and can get away with working on a peer revenue share'.

And if they don’t have to put up anything upfront, then you don’t have to pay them upfront. That was the thought process.

The phone becomes a controller, and it translates it, in real-time, into the inputs that the game needs. So it translates it into fly, build, drive – whatever that game needs. We know which game is running, and there’s an external file that tells you what input it’s going to need. So based on what’s running, it reads the external file and it does the necessary translations.

So, for example, for a racing game, you might want to tilt left to turn left, and tilt right to turn right. But if it’s, let’s say, an endless runner, you might swipe left to turn left, or swipe right to turn right.

In both of those cases, you want the left joystick to do the left and right. So in game one, you say, "okay, generate a tilt". And in game two, it’ll generate a swipe. That’s basically what it’s doing.

<em>Mad Aces</em> on Cannonball Arcade
Mad Aces on Cannonball Arcade

It took us about three or four complete rewrites to get it right. It worked the first time around, but it was really, really laggy. It was good for a demo but it wasn’t great to play with, and it took about a few rewrites to get it usable and playable. Now there’s no latency or lag and it works as good as if it was designed for it.

How much actual work does your team have to do to get a game live?
Not a tremendous amount. We’ve got a game testing team. We are required to test every game input on the target hardware, anyway.

What we do is, these guys play through the whole game, and they are gamers’ gamers in the full sense. They know what keys they want and what action they want. They play through the game, and then they’ll generate one external file that says, "Okay, the ‘A’ button’s going to do this, the ‘X’ button’s going to do that and the ‘Y’ button’s going to do that". And that’s it.

The on-phone gamepad is a game-aware gamepad, so it knows what game is being played. When you start a racing game, they’ll actually show you a little brake pedal. Anything you require for that game – if it’s a shooting game, it’ll show you a "shoot" button, or a "bomb" button, with a joystick to move around. So the controls are custom-built to each game.

We have a library of buttons which are action-aware buttons. Depending on what game is playing, it constructs the gamepad on the fly once it's launched.

There is not a minimum requirement. On most contemporary phones, it’ll work. It’s a fairly lightweight application.

The way it works is it goes out as an official operator companion app, and it works like a soft remote for the most part when you’re watching TV. You can change channels, change the volume, and then when you go into the games menu and launch a game, it turns into a gamepad and ties to the set-top box via Bluetooth.

<em>Pyro Jump</em> on Cannonball Arcade
Pyro Jump on Cannonball Arcade

Every time a new game is added, if it’s using the same buttons that are in the library, it’s fine. If not, we create new buttons, add them to the server, the gamepad downloads them and adds them to its local library.

So what do you think is the future for this kind of gaming? Can gamers expect more of this?
I think casual gaming has room to improve. But who knows? If I were to be a really brash person, I would draw an analogy with what smartphones did to handheld game consoles.

If I were to make the case that there’s room for one device in your pocket and that’s your smartphone, I could just as easily make the same case that there’s room for one device under your TV, and that’s the set-top box.

If your set-top box is going to have the power of a smartphone… the set-top boxes that are being launched now in a place like India, which have quad core CPUs, two GB of RAM, Bluetooth, Wi-Fi, Flash – these are powerful machines. And they’re only getting more powerful. The OS is getting more powerful and the games are getting better.

At what point do consumers say, "Hey, wait, this is good enough!"? Except for the really hardcore, this is good enough. I mean, maybe not this year, maybe not next year, but at some point, right? The technology curve is moving in that direction.

The business model

The business model

What’s the business model for the customer? Do they typically pay a subscription fee?
It's never usually per game, but whether there’s a subscription fee or it’s bundled, though it differs from operator to operator.

In the case of Airtel, there is a subscription fee to the end consumer. In the case of some of the other customers we’ve signed, they have absorbed the cost into their main TV service as a bundled offering. As long as they subscribe to the base offering from that operator, the gaming service is free.

For game developers, what’s the advantage of working with you guys?
There are multiple advantages. One is, it’s a channel that does not cannibalise the Play Store revenue. Sometimes you worry that if you put your game in an operator app store for mobile: will it cannibalise my Play Store revenue? In our case, it’s two completely different worlds, so it does not cannibalise it. It’s only incremental revenues.

The other thing is, discoverability is great on our platform because we are at 125 games by choice, in the sense that we actually have licenses for thousands of games but we have deliberately not made live games that don’t work well on TV or that are duplicates.

<em>Protoxide Death Race</em> on Cannonball Arcade
Protoxide Death Race on Cannonball Arcade

So sometimes you get one game that’s ahead, and then you get three clones of that game. So we’ll just launch the one and not the many clones. That way the catalogue is a lot less cluttered.

What about free-to-play Android games? That can't translate to TV easily.
We do support free-to-play. The jury’s still out on how well it’s going to work on TV, but we’re trying it out and experimenting with it. But you’re right about one thing, certain mechanics within free-to-play games will not work as well.

For example, there’s a very common mechanic that says: you build something, and then you wait for one hour, come back, and you put in some points to speed it up. That works well for mobile, because your phone’s in your pocket; you’ve played for five minutes, and then an hour later you can check back in and see if your castle is built yet or not.

On TV, that’s not going to work. What we’re seeing in behaviour amongst our users is: they don’t play for five minutes at a time, and they don’t play daily, or they don’t play multiple times in a day. They’ll play maybe once a week and it’s usually tied to weekends.

But once they play, they play for a continuous 30, 40 or 60 minutes. We are averaging 22 minutes for a game session right now, which is pretty good, but it’s similar to console behaviour.

What’s your commercial relationship with the developers?
We share 70% of whatever we make. So from any source, whether it’s a subscription revenue or if it’s a royalty in the case of bundling or if it’s in-app purchases – and we’ve not yet started this yet, but we might start doing advertising around it. But any source of revenues, we will share 70%.

You can find out more about 18th Parallel and its plans for the Cannonball Arcade and playing mobile games on your TV at its website