Bennett Foddy interview – the game changer

Bennett Foddy is a one-man gaming phenomenon. Ever since his nigh-impossible and frequently-hilarious running simulator QWOP exploded across the internet, he has juggled his day job as a tutor of moral philosophy at Oxford with the demands of being a cult games developer.

Since then Foddy has produced a succession of bizarre and addictive Flash and iOS games which frequently turn game design on its head. He is in demand as a speaker at game development conferences and is even in the process of producing a title for the PlayStation 3, as part of the Kickstarter-funded Sportsfriends pack.

As fans of Foddy’s games, we wanted to find out what makes the man tick: where his game ideas come from, how he puts them together and what other independent game developers can do to follow in his footsteps.

Before we go any further though, grab a PC or laptop, fire up a browser and head to www.foddy.net. All the games there are free to play, so you can follow us through the feature while trying out his titles for yourself.

THE PHILOSOPHY OF GAME DESIGN

Foddy’s academic work at Jesus College, Oxford focuses on the philosophy of addiction. This may go some way towards explaining why his games manage to be utterly maddening but still keep you coming back for more.

Much of the games’ addictive qualities come from the way he imbues them with a sense of consequence. Foddy has little time for games that don’t allow you to fail, such as the plot-driven blockbusters that are the norm from big publishers.

“The most difficult part about video game design is to make it matter whether you win or lose. Most games now that have a story don’t allow you to fail in ways that would jeopardise that, you have a motivation to play but you don’t have a motivation to play well. I was playing Crysis 2 and I was on autopilot as I don’t care if I lose, nothing bad happens if I lose,” he says.

This approach shows up in the sheer finality of his titles. In QWOP, where you attempt to run the 100m sprint by individually controlling the thighs and calves of your athlete, one false move and your athlete is on his back and it’s game over. You can fail at 10m, you can fail at 99m, it’s still the end and you have to start again from the beginning. In GIRP, where you have to scale a cliff face, if you fall all that’s left is the distance you managed to climb.

QWOP is difficult to master and you can fail at any point, making it more rewarding when you win

Foddy is adamant that being able to frustrate players is integral to designing a fun game. However, it’s important that the frustration remains within the player’s power to overcome. Foddy’s games may be hard, but they are certainly fair; if you fall over in QWOP, drown in GIRP or get bowled out in Little Master Cricket, it’s definitely your fault.

He feels that games that kill you out of the blue are thankfully a thing of the past, when titles were sometimes unfairly difficult in order to artificially inflate the challenge. He points to Amiga title Sword of Sodan as an example of unfair game design.

“You’re walking along, suddenly this thing comes out of the ground and kills you, the only way around it is to play it 100 times to memorise the position of the hazards, and that’s not much fun,” he says.

Sword of Sodan is an example of unfair game design, where you can suddenly get killed out of the blue

Rick Dangerous, with its emphasis on nigh-unavoidable trap deaths, is another example from the Amiga era. Foddy sees the game designer as taking the role of a Dungeon Master; it’s a game played between the developer and the player, where the object is to inflict pain on the player. However, in order for the game to remain fun this pain must be tempered by a sense of fairness. “You’re kind of an evil mastermind,” says Foddy, “it’s like how Doctor Evil always gives Austin Powers a chance at escape.”

PERFECTING CONTROL

Making the player aware that their actions and reactions have consequences is part of drawing them in, but to truly immerse a player you need to think carefully about control systems. Many of Foddy’s games, while superficially simple, play like nothing else. QWOP, the game that made him a cult gaming hero, is a sprint simulator with an insane control system.

It’s hard to know how to simulate such a natural human activity as running on a computer, while still maintaining a challenge. Early titles such as Track & Field or Daley Thompson’s Decathlon relied on hammering buttons or waggling joysticks – the faster the hammering or waggling, the faster your avatar would run.

As fun as Track & Field is, its control system still removes you from the action

That method goes some way towards simulating the physical effort that goes into running, but it’s still removed from the actual action; Track & Field, for example, would also work as a drumming simulator.

In order to get you under the skin of the runner in QWOP, Foddy has made it so that you control his thighs with the Q and W keys, while using O and P to control his calves. This literal interpretation of the process behind running means the game is borderline impossible, but there’s no doubt it keeps you playing.

It’s a similar story with GIRP. In this rock-climbing simulator, each handhold on the cliff is represented by a letter on the keyboard. As long as you hold down the key for that handhold you will continue to hold on to the cliff, but to move up you need to hold the shift key to flex your muscles, then release the current handhold and press the new one before you fall off. The result is that you end up clinging onto the keyboard as if you’re clinging onto the cliff: you have become the climber, in a way that you never truly become Marcus Fenix in Gears of War, able to leap and dive into cover with a single button press.

Foddy’s games let you become the character in a way that modern games, such as Gears of War, don’t

LOWERING THE LATENCY

They may feel difficult to control, but Foddy’s games have some of the most responsive controls there are. If you press a button in QWOP, a part of your runner’s body moves immediately. If you move the mouse even slightly in Little Master Cricket, your cricketer will move his bat – even if it’s straight into your stumps after hitting a six.

The instinctive response of your on-screen avatar is what makes you feel involved, even if the result isn’t what you expected. Foddy likes low-latency controls.

“The sense of embodiment in your player character on screen depends on having a very instantaneous reaction feedback whenever you press the button,” he says. Super Mario Brothers is a good example of a game with low-latency controls.

“Mario, as soon as you press the button, it’s not like he has to wind up and jump. He’s instantly in the air. I think this creates an illusion, your brain expects somebody to have to wind up to jump, so if you press the button and he’s instantly in the air, it’s like you think you must have pressed the button a little while ago,” says Foddy. “It creates a very tight bonding – you feel very strongly in control of Mario when he jumps.”

Foddy believes that Super Mario Bros’ lets you feel “very strongly in control”

This is something that isn’t possible with a game that uses physics simulation, such as most of Foddy’s titles.

“Everything has to accelerate and decelerate,” he says, “the trick as a designer is to try to make it so things react instantaneously when you press a button, even if it’s not the reaction you intended.”

This is at its most stark in Foddy’s two-player title Get On Top, which is best described as a two-player low-gravity sumo simulator with lighter combatants. The two participants’ bodies move using simulated physics, and while a touch of the controls will always make the participants push, pull or jump, the combination of the two participants’ movements and the physics engine means the outcome is sudden and often unexpected.

INTERFACE MINIMALISM

Getting the controls right isn’t the only challenge behind making a game. The best way to draw a player in is to make sure there is as little as possible between them and the mechanics of the game itself – the more minimal the user interface (UI), the better.

“It’s so much better if [the UI] can be made a literal part of the game,” says Foddy.

He feels that this streamlined approach is often missing from big-budget games, where an action might be met by a +100 experience notification or being explicitly told that another character likes you +4 more.

“I think the worst case is something like Fable II, where everything you do a little icon appears with a number. Basically you see far too much behind the scenes when somebody does this, you’re seeing the inner-workings of the game and the whole illusion is lost,” he says.

Progress? Fable II is cluttered with stats

It’s this desire to reduce clutter that drives the literal version of progress in Foddy’s games. GIRP is about rock-climbing and the only indicator of progress is how many metres you make it up the cliff. QWOP is the 100-metre sprint, but with time replaced by the distance you managed to run without falling over.

“You could say that QWOP doesn’t take place mostly on the screen anyway, it’s taking place between you and your keyboard,” says Foddy, “There’s not much to get in the way.”

You can see this approach in Foddy’s two-player titles, too. In PoleRiders and Get On Top, there are just two players and the objective is to score a goal or knock the other player’s head on the floor. The player that succeeds gets a point, and the first to ten is the winner; it’s as simple as that.

Foddy feels that other adversarial games, such as the Street Fighter series or Tekken, are unnecessarily complicated, which means only the very best players understand all the attacking options and so can predict and counter their opponents with intention and skill.

“I think that if you see a health bar you’ve blown it. You are hit, how are you supposed to know how far your health bar will go down? How many little pixels are still left on your health bar? Can you afford to be hit in a certain way? Well, the only way to know is to memorise the precise amount of damage. It’s like a game for robots,” he says.

With no health bar, IK+ is arguably more fun than modern fighting games

Foddy misses games that make you go down in one hit, such as the venerable Amiga martial arts classic, IK+.

MULTIPLAYER GAMES AND THE QUEST FOR RANDOMNESS

Foddy’s multiplayer games are some of the most fun we have ever played. PoleRiders and Get On Top manage to be both hilarious and addictive, and much of this is down to their inherently random nature.

For Foddy, for a multiplayer game to be truly fun it needs to have a random element. In PoleRiders, for example, the fact that you’re carrying an elastic pole and have enough strength to spring the other player across the screen means you’re never sure what the outcome of a particular clash will be.

“PoleRiders is designed to be slightly random … the idea of using physics is that you can be really great at it, but being really great at PoleRiders means that you win nine times out of ten, it doesn’t mean that you win every time,” he says.

Foddy compares the random element of PoleRiders to that of pool. He feels that pool is more fun as a multiplayer game than snooker, as the table has been shrunk enough to introduce an element of randomness: hit the ball hard enough and something will go in. It means you can still enjoy a game of pool against a more skilful friend, whereas go up against a world champion snooker player and there’s a good chance they’ll win the toss and pot everything on the table without you getting a chance to play.

It’s the random nature of PoleRiders that can make a game really fun

This is also a common design flaw in fighting games, albeit one that nowadays is patched out fairly quickly: the infinite combo, where one player can keep landing punches and kicks while the other player’s character is too stunned to react.

“For me the moment that someone did a perfect break it was the end of snooker – the game is no longer interesting,” he says. “From the game designer’s point of view, it’s like the infinite combo. It’s like, hey, look while I do some very hard to perform but ultimately fairly meaningless combination of buttons while you sit there and watch.”

MAKING A MASTERPIECE

Knowing what makes an amazing game and having a great idea is, of course, just the start. Turning it into a playable title is a huge challenge in itself, but it’s easier than it used to be. The reason Foddy didn’t make a game until his late 20s was due to how hard it is to get started on even a very basic level.

“The main thing with games is it’s a barrier to entry up to the point where you can get something moving on the screen, and as soon as you’ve done that you have every element you need to make a functioning game,” he says.

The problem was that when he was first interested in making games you needed to be a serious programmer just to get to that elementary point.

“If you need to have a degree and spend months and months learning machine code to do the simplest things,” says Foddy, “most people are going to get frustrated.”

He uses 1989’s Prince of Persia as an example. The seminal title was written by one man, Jordan Mechner, in Mac II assembly language – essentially the closest you can get to programming in machine code without having to type in ones and zeroes. Have a look at the source and at Mechner’s programming notes to see the kind of complications early programmers had to get their heads around.

Prince of Persia was coded by Jordan Mechner using assembly language

Things got easier in the Amiga era, thanks to programming languages such as the BASIC-derived AMOS but, according to Foddy, “it was still extremely hard, it would take an almost obsessive person to learn to do that by themselves without training”.

As PCs got more powerful, it was game on for Foddy,especially once Flash made it easy to create interactive animations using a simple GUI.

“It took until the point where you could use Flash with drag and drop that it finally got easy enough for me,” he says.

Foddy uses the example of QWOP to show how simple it can be in Flash to get started with a game.

“Well that was really quick, as it’s a really simple game. That’s probably my simplest game, in terms of you just basically set up the rag doll and the simulation and then make sure you can measure how far he’s gone – that’s 99 per cent of the game right there,” he says.

To make QWOP, Foddy used Adobe Flash and something called the World Construction Kit, which integrates the open-source Box2D physics engine with Flash. World Construction Kit allows you to set up a physics-governed world in Flash and create objects within that world that the user can control. For QWOP, Foddy set up a world, then “basically I just dragged him in – I traced a picture of Carl Lewis, I dragged him in, it was really set up very very quickly”.

It’s still not exactly easy to make a Flash game, but Foddy has a valid point; Flash makes it incredibly simple to create, if not a game, then at least an interactive animation.

With essentially no Flash or programming experience and a copy of World Construction Kit, it took us about an hour to create a Flash animation where we could fling bouncy smiley faces around a maze. This gives you a foundation on which to start building your ideas, and everything is easy to tweak; if you want to alter how much one of the balls will bounce, for example, you just need to put a new figure in the Restitution box with no maths required.

It may not win any awards, but in an hour we had the basics of a game

THE JOY OF PHYSICS

Foddy may describe QWOP as easy to make, but not all his games were quite such a breeze. PoleRiders took a long time to get right, chiefly due to what makes the game such fun, the incredible elasticity of the pole.

“It’s very hard to debug that kind of physics game. The poles are like a worst-case scenario for video game physics. The way that video game physics works, is every frame it looks to see if, for example, your pole is jutting into the other person’s body, and if it is they move back a little bit, but it doesn’t always know what to do if you have very thin, very small bodies,” he says. “It doesn’t know which way it should move [the body] back, and you get this cascade of errors, it’s very difficult to tune.”

PoleRiders is a superb two-player game, and while it can be frustrating when one player runs straight through another’s body and goes on to score, it does add to the random element. Foddy admits that if he knew how popular PoleRiders was going to be, he would have debugged it some more. This, he says, is part of the joy of putting games online for free: people are less likely to complain if some bugs remain.

“If you tried to put out a PS3 game that had these kinds of physics glitches, people would be very upset. As it’s a Flash game, you can get away with it,” he says.

Foddy evidently has his work cut out, as a version of PoleRiders is in development for the PS3 as part of the Sportsfriends package.

GETTING A GIRP

PoleRiders may have been tricky to program, but GIRP was even harder. In GIRP, much of the challenge revolves around knowing exactly when to flex your character’s muscles, in order to use his momentum to scale the cliff, and getting these muscles to behave correctly had Foddy stumped.

“I was stalled on GIRP for quite some time, as I couldn’t get the inverse dynamics, the code that controls the movement of his arms, to work,” he says.

The problem is made more complicated due to the fact that the climber’s body in GIRP is not anchored to anything, it’s your flailing legs and swinging body that affect the movement of the arms. Finally a programming colleague, Chris Hecker, who is currently making the cerebral spy game SpyParty, suggested Foddy use inverse kinematics.

“For anywhere you define a target, it works out where the bits of the arm should be to look right,” Foddy explains.

You can see this effect in action in the Air Toss Flash game, where a colleague throws you crumpled-up bits of paper and you have to throw them into the bin. The position of the man’s hand is determined by the position of the mouse pointer, and the rest of the arm rearranges itself to suit. A similar principle is used by factory robots to orientate their various joints to get the tool hand into position.

Air Toss uses inverse kinematics for realistic arm movement

PUBLISHING AND PUBLICISING

Once you’ve made your ground-breaking title, you still have to get people to play it. Foddy makes games in his spare time, and doesn’t spend thousands of pounds advertising his titles. Instead, word about his games has spread organically; QWOP became huge when it was linked to by YouTube “comedic vlogger” Ray William Johnson. Johnson has a huge following, and QWOP quickly spread around the internet and became a phenomenon in its own right, with attendees at the Comic-Con convention turning up dressed as QWOP and falling over a lot.

Getting picked up by high-profile YouTube personalities is both unlikely and, in our opinion and Foddy’s, subscribing to YouTube personalities is “cringeworthy”. However, there are other ways to promote your games. How you approach this depends on the platform you are going to use; promoting free-to-play Flash games on an advertising-supported website requires a different approach to releasing games on a mobile app store, such as for Foddy’s iOS versions of QWOP and Little Master Cricket.

According to Foddy, hosting games on your own website causes network effects that just aren’t present on a mobile or tablet app store. This is due to the ease of promoting your other games on your website, so they benefit from the success of another title that has gone viral.

This happened to Foddy when his first game, 2007’s Too Many Ninjas, spread around the internet. There was a brief spike in traffic to his site, which then tailed off, but when he put a new title up with links to Too Many Ninjas on the same page, the network effect meant that the traffic didn’t tail off to the same degree.

Too Many Ninjas was spread by word of mouth

“If I have a new game I tend to promote it on the other games’ pages, so there’s this aggregation effect,” he says. “When I put up Too Many Ninjas, it got featured on Kotaku and Digg, and so what happened was I had a lot of traffic for a day, maybe four days, then it trailed off to as close to zero as possible. Then, when I put Little Master Cricket on, it didn’t quite trail off to zero, but it trailed off to maybe two-to-five per cent of its original. By the time I added QWOP it wasn’t trailing off to zero any more.”

This means that the amount of money Foddy makes from each new game is boosted by increased numbers of people playing his previous titles. This is an effect that works well on a website, but is less effective on an app store. Unless you have the advertising clout to make your app known among the hundreds of thousands already out there, it’s going to get buried.

“I have recurring income on QWOP for iOS, but that’s mainly because it’s advertised on my website, so people keep coming back to it. Most people I’ve spoken to who publish on iOS find that you publish a game, you get like two weeks of income and it’s literally zero for ever after unless you advertise,” he says.

This is mainly due to the fact that Apple’s App Store, while it is good at promoting new and popular apps in its Featured section, doesn’t give you the same kind of aggregation effect as you get with running your own website. On a website, such as www.foddy.net, you can cultivate the developer as a personality. Just as you may want to buy the latest book from a certain author or latest single from a band, you’ll want to keep an eye on your favourite Flash web developer to see if they have released any new games.

The Apple App Store doesn’t help people find other apps by the same developer easily

On the app store, by contrast, the name of the developer is written in a small font under the large, bold name of the app, and you can only see the rest of the apps written by that developer by clicking the Developer Info button, which isn’t even titled “more by this developer”.

“People don’t go on the app store and go ‘I really like this, let’s see what else this guy has made’,” says Foddy. “The app store is not very good for promoting the idea of an artist, whereas you’re free if you run [a website], you’re free to promote yourself as an auteur.”

Despite this, Foddy is positive about the iOS App Store. For a start, the overheads are fairly minimal. It’s $99 a year to join the developer program, and after that all you need is a Mac to run the SDK and an iPhone or iPad to run the software. This means you can start your own programming studio for around £1,000 upfront along with £64 a year, and get a computer and phone or tablet into the bargain.

He’s also keener on developing for iOS than for Android, at least for the moment. Part of this is that it’s harder to make money from Android than from iOS, as users are generally less inclined to pay for apps up front. Foddy doesn’t want to go down the in-app purchase route, either; part of the joy of his games is that you are given everything up front with nothing to unlock, so your success or otherwise depends entirely on practice and skill. The fragmentation of Android is another factor that raises barriers to entry.

“The development costs are much higher as you’ve got to support all these different devices, whereas on iOS basically it works,” he says. “I wrote Little Master Cricket in 2008 for iOS, first iOS game I made, and it still works on the latest iPhone – that’s not the case for Android. I don’t think it’s viable for a solo developer.”

Foddy also feels that one of the most important things about making a game is knowing when to stop. You can debug until the end of time, but you’re best off accepting that a game is good enough for its target audience and moving onto the next; in Foddy’s words, “I don’t want to come back to work that I’ve already put in the can.”

If you’re putting a game on the web for people to play for free, they will tolerate the occasional bug (such as the slightly random nature of PoleRiders’ collision detection). If you’re expecting people to pay up front, though, they’ll expect a more complete product.

“My iOS version of QWOP is still not perfect, but it’s a lot more perfect than the web version. I felt like if I was going to charge people two dollars for it up front, I feel like I owe a more complete and polished product.” Foddy says. “The nice thing about web publishing is you don’t owe anybody anything, if they don’t want to play your game they don’t have to come and play it.”

Well we certainly do want to play his games, and we hope that you’ve enjoyed one or two as well while reading this article. And with the amount of fun we’ve had playing his titles over the last couple of years, we’re certainly glad Foddy spends more time making new games than ironing out minor bugs in the old ones.