Tag «talk»

How to make an AAA game in 2 days

Before the Global Game Jam 2016 started I gave a short talk about how to make an AAA game in 2 days (as the GGJ is 48 hours long). Of course I have no idea how to make an AAA game, but I thought that sounds more interesting than "How to polish your game in a day". So yeah, it was just about giving a jam game that small bit polish so it wouldn't look that much like a jam game.

I started by talking about some experiences I made years ago: how 2K contacted me as they had this new game - Assassin's Creed - and they already worked a day on it. Their prototype consisted of the protagonist Günter (or so) walking around. But they didn't know how to make it any better. That's why they consulted me. And here is what I told them.

First I noticed how the movement wasn't very smooth, and I showed them how to use tweens so the player character would look professionally animated, even though it was a single sprite. It was a bit of a pain, as tweens need careful coding. For example, as long as the player is tweening, don't let the user change the direction, and so on. But all the sweat paid off.

(If you use Unity, don’t use iTween. DOTween and LeanTween are quite okay, though!)

Afterwards, I explained how the same applies to the camera: the more movement is there, the better. We all know how cool AAA games use tracking shots for everything. You can do the same! But using Lerp() to make the camera somewhat smoother can be tricky, as sometimes the player can be too fast and not see where they're going. This is why we add some kind of foresight. This talk from the GDC 2015 can be helpful, even when not doing a sidescroller.

Apparently the original AssCreed had some teleportation mechanic, but it looked bland. I advised Lionhead to add some transition effects. Those can also be useful when the player gets hit - just color the screen completely red!

My game Snakoban has another kind of effect for changing levels. It was a bit of work, but everything is better than just changing screens without any transition.


"Never forget to shake the screen," I told the AssCreed developers. "And of course, use particles everywhere." Every new particle in the game is another step to AAA, as they give instant feedback to the user that something is happening or has happened. And they look nice, too, so even as pure decoration they are useful.

Our very own game prototype Power of Love has trails for the player characters. It looks cool, because it adds the illusion of velocity, speed and fast movement.

Power of Love

Even with all the improvements so far, Blizzard's prototype looked kind of flat. So I introduced them to the concept of layers - giving the player a shadow, even a simple one, already creates the illusion of depth. Having a foreground and a background with different scrolling speed ("parallax scrolling") is awesome too.

Although they already worked on the prototype for a day or so, they still missed the most important thing in a game (or any medium, really): emotions! Always take care your game evokes feeling. This is why we improved the story a bit, worked on the colors and chose a cool music that fits. (I think they changed the story later.) Anyway, humor is also cool, but I don't know much about that. It’s up to you, dear reader!

Of course, music is great, but sound effects are needed, too. Just like particles, sounds add a lot to the feedback and the atmosphere. Foot steps, "ouch" sounds, you name it. Sound can even create things that aren't there! Want a forest full of animals? Just play a sound loop with rustling in the leaves and singing birds, and your graphics department can leave early, because they won't have anything to do.

Obi Wan gives a good lession, in this regard: it's only real if it has a sound!

Obi Wan and the sound of laser

For the final touch Crytek added simple light effects, to focus on the important things in the game (the player), and increase the atmosphere. Nobody could believe this was still the same game, just with a bit of bling created within a few hours. And we all know that Assassin's Creed became a big hit!

(You can download the "game" here. It was made with Monkey-X. The tileset is from Silver IV. The dust is from here. The first music track is by my brothers, the second one by Matt Goles. Of course, Assassin's Creed is a trademark of Ubisoft.)

The winners of Ludum Dare, and what to learn from them

This was originally a quick and dirty talk I held in 2014 at our Global Game Jam location in Leipzig, Germany. I wanted to give the audience (mostly students, a lot of them without experience in the art of game jam) an impression on what is possible during a game jam even when you're alone and have to do stuff from scratch; so I showed them the last five Ludum Dare compo winners, which means each of them got the first place in the "Overall" category. The talk would conclude with some best practices, at least in my opinion.

As most of you probably know, Ludum Dare is an online game jam held several times in the year, with the big versions always commencing in April, August and December. Thousands of participants make a game in 48 hours each time, and a lot of them also are part of the quite active community surrounding the jam. Everybody who joins can be sure to get feedback and answers via the blog on ludumdare.com, and via the IRC channel #ludumdare on AfterNET. And don't forget that for three weeks afterwards, every participant can rate all the others' games, so the brutal final ratings for each category (e.g. Graphics, Fun, Mood) are there for the whole world to see. With the high numbers of entries, winning the competition may sound nearly impossible. So let's take a look at some of those who actually did it!


Ludum Dare 24 – Theme: Evolution


Evoland is a Zelda-like with a funny idea implemented well. It has a lot of chests – in each chest you find a feature that hauls the game to the next step of videogame evolution. At first you can only walk right, but you get the ability to walk left soon. After that, you can walk north and south, how cool is that! Scrolling! Colors! Sounds! Weapons! Nicolas Cannasse broke down a pretty standard Zelda clone into its most minimal parts, used those parts to let the player explore them, and made a unique game this way. (Later on he even extended the game and you can now buy it on Steam.)

As a game jam game Evoland is very ambitious and I couldn't believe that it was made in 48 hours. I assume Nicolas had the idea and concept very early in the process, and using a programming language he invented himself might have given him a good headstart.


Ludum Dare 25 – Theme: You Are The Villain

Atomic Creep Spawner

Atomic Creep Spawner!! reminds me a bit of Dungeon Keeper, although you can only do a fraction of what is possible in that game. A pompous knight is raiding your very own dungeon, stealing your money and destroying your orbs, and you have to stop him by spawning a lot of monsters. You create those hordes of zombies and ghosts via simple clicks on the floor, and they find their own way (more or less) to the rampaging knight. A fair bit of AI must have been programmed for this game.

Made by Sébastien Bénard (known as deepnight, one of the most successful Ludum Darers!), Atomic Creep Spawner features great humor and amazing pixel art. Sébastien even found time to include a tutorial at the beginning. What the game lacks in interactivity it makes more than up with polish and love for details.


Ludum Dare 26 – Theme: Minimalism


MONO implements the theme via its graphical style (which looks simplicistic, but is actually very well done), but most certainly not via the gameplay. While at first glance it seems to be a minimalistic game in every sense – you navigate a small sphere through a world of rectangles – the levels soon become more and more diverse, which is what keeps the game interesting. This is a nice trick you can learn from: make a game with only some basic functionality, and if you have the time, add another level with new elements and mechanics inside it – rinse, repeat.

Tim Hantel managed to make a neat dexterity puzzle game with a lot of atmosphere, mostly by adding a fitting soundscape. The gameplay mechanics sometimes make the game a bit frustrating (you die by touching a wall already), but in general the player's death is a forgivable experience: you spawn instantly at the level start again. An important lesson I think.


Ludum Dare 27 – Theme: 10 Seconds


PROBE TEAM could easily be part of the Ludum Dare before – it uses one single color only, enhanced by some cool looking post effects. While I never liked the theme in itself, because I always felt that it would cripple gameplay concepts instead of fertilize them, Andrew Shouldice actually made an interesting type of exploration game out of it. You start little probes which have 10 seconds of fuel each (so be economical with your commands), lead them through some kind of maze and activate triggers to open doors. It feels a bit repetitive, but the moody atmosphere absolutely helps to tolerate and even enjoy it.


Ludum Dare 28 – Theme: You Only Get One

One Take

One Take by Daniël Haazen is a great example for an unusual idea creating a whole new experience. You play as a camera operator taking one continuous take from a movie scene, and all you do is following orders from a film director in the form of short sentences, like "Zoom in on the sheriff" or "Go back to the guy in the alley". The scenes are a bit animated (while most of the action comes from yourself) and actually feel like small movies – already worth an honorable mention in my opinion. The cherry on top are the pixelated newspapers after each level, including reviews about the 'movie' and your performance.



So what can we learn from these great examples of Ludum Dare rapid game development, for our own jam games?

  • Make something simple. All the mentioned games concentrate on a single idea. Be it spawning monsters in a dungeon or moving a small sphere around obstacles – important is to focus the game's concept on one aspect and not trying to add more and more features. Of course, this implies you actually have a nice idea that can be played barebone and that you like.
  • Be inspired by the theme. This one surprised me a bit, probably because more often than not the theme of a jam hinders instead of helps me. But all the winners above are very close to the theme, and that must tell something, probably about inspiration.
  • Think in two dimensions. Although one of the games uses Unity, all of them are 2D games, varying from totally abstract to concrete pixel art. I don't know exactly why that is, but there is probably more than one factor why jam games prefer to be 2D. My guesses are: the game is simpler to make, the graphics look better while being less work, there are a lot of premade frameworks, and you get a nostalgia bonus.
  • Use a framework. The Ludum Dare rules state that you must make your game from scratch, but it is allowed to use libraries, tools and engines that were already made by you or others. So use them! The winners from above utilized Haxe (compiled to Flash), Java with libgdx, Unity and LÖVE. Three of these games were playable in the browser (with plugins), which might also be a small factor adding to their successes. Remember that Ludum Dare's winner are determined by people who have to play thousands of other entries, too – the less problems they have to play your game, the more likely they will play it.
  • Generally you should try to make your game as accessible as possible. Deepnight's winning game Atomic Creep Spawner includes a tutorial, but you don't always have to go this far. Just don't innovate where it isn't needed, and explain things when they come up for the first time, be it via text or picture (better yet, force the user to play it to understand it).
  • The last learning for now is somewhat vague, as the games tackle this very differently. Some of them use a lot of humor, via little comments from the characters for example, others are pretty atmospheric, often thanks to their great use of sound. What we can take away from this is: don't shy away from trying to evoke emotions in the player. It seems the humoristic way is a bit easier than the one with the dark mood and the feels. Speak to the player, or let them explore interesting places. Give them reasons to attach to your game!



Making a game in 48 hours is hard, winning Ludum Dare is even harder. But if you look at the above examples you see it's not impossible! Sure, you need to know your tools in and out – the Ludum Dare Top 10 isn't the place for learning a new programming language. (This could be a bonus learning.) Just keep in mind to have fun and make something worth playing. Everything else comes afterwards.

Embrace Your Limitations – Game Jams Extended

why limit yourself?

In this post I present you the transcript from our talk "Embrace Your Limitations - Game Jams Extended" at the A Maze 2013. Actually, it's just a prettified version of our outline. By the way, we held the talk with a beamer presentation in the background showing some pictures, but as they weren't important for the talk and we're not sure about the license of each picture, you'll just have to imagine them. :-P

Description: "A two day game jam might result in a neat, even innovative little gem - why not expanding it until it becomes a full, polished game? It can be tempting to bloat a game with things it doesn't really need and make it a monster that never will be finished. Thus we want to talk about why you should fall in love with your limitations and how to create cool things with small prerequisites."

1) Who are we?

  • Rat King from Germany - Jana & Friedrich
  • making TRI, since quite some time
  • derived from an Ludum Dare entry, began with a simple idea, so we added a lot of stuff to it
  • became too big - we refocused and embraced our limitations

2) Why do we talk about this?

  • so that you learn from our mistakes
  • this is not about small or big games, but the avoidance of feature creeps
    • because they prevent you from finishing your games!
  • [picture of 1000 Ludum Dare games] cool to be one of them, but even better to have a finished, polished game
    • game jamming as production style
    • helps learning to limit yourself and sticking to your talent/profession
  • limitation is your boss (if you're indie)

3) Constrain your game design

  • gamedev often starts with the thought "I can do ANYTHING"
  • but: limitations help you creating ideas - just think of game jams (restrictions regarding time/theme/technical/etc.)
  • remove features / elements where possible
    • “Perfection is not when there is nothing more to add, but when there is nothing more to remove." (Antoine de Saint-Exupéry) - talking about perfection, but can be applied on design, too (designers often are perfectionists)
    • example 7-Day-Roguelike-Challenge 2012 (Me Against The Mutants): cutting off features before the development is easy; the later the harder
      • in the end we relied on the one cool idea it has (infinity rectangles)
    • throw away features when they're too much or not needed - never be afraid to cut stuff out that doesn’t add fun
    • often leads to clear design and a clear vision
  • the cycle of innovating and testing (= experimenting in a short time frame) only really works when you limit yourself, otherwise it becomes a growing spiral of death

4) Constrain your graphics design

  • good games are memorable and can be recognized with one screenshot
    • limitations help defining your art style so it sticks out
    • games full of prefabs, characters, features are harder to communicate
    • instead of trying to add lots of details - minimalism is win
    • find a visual trademark, for example a recognizable character
  • pixels are cool, but 3D is cool too - limitations are possible with every art style!
  • use a grid
    • necessary in TRI so people can measure distances
    • grids are fundamental, used in design
    • as soon as you are restricted to a grid you can’t add too much stuff anymore - this is good
    • everything becomes deterministic - you can put only one thing at one place (tile) at a time
    • but beware the almighty Minecraft - games get judged by images

5) Constrain your technology

  • KISS - Keep It Simple, Stupid
    • must be in this presentation
    • search for simple solutions, even when the problem looks complex
  • only do stuff the player actually sees
    • you're making a game, not a simulation
    • even simulations are simplified
    • tip: take a step back as the developer, and imagine playing your game as a normal player
  • do not reinvent the wheel ...
    • ... but sometimes you don't really need a full-blown monster-truck wheel
    • if you have/need a simpler solution, go for it
    • avoid third-party-solutions that add too much stuff / features, because it's tempting to use those features
      • limit your project, not your brain
      • let your imagination run wild and the creativity flow, but sometimes it's easy to forget KISS - just remember limitation is your boss now
    • example: for Karoshi! I needed some pathfinding
      • downloaded an A* plugin for Unity with multithreaded, dynamic pathfinding
      • was too general and had some quirks
      • made the development overcomplicated and added a lot of hassle
      • better solution would have been to just roll my own

6) What is needed in order to successfully make a game with limitations?

  • "If I had more time, I would have written a shorter letter."
  • experience in your field
    • realistically evaluate of what you are capable
    • don't try TOO novel stuff, concentrate on one (cool) aspect
    • prevent too many construction sites at once
  • a small team
    • big teams automatically add overhead and the feeling of “you can do more”
  • ideally one cool, perhaps innovative idea
  • no idols
    • be inspired by others, but don't try to copy them
    • you will try to add more stuff if you clone another game, and might even bloat it
    • at least don't copy a whole game, but pick certain features you like

7) Expand your horizon

  • does it really need to be a “real” game?
    • example: Fibrillation only consists of walking around in changing scenes and experiencing a background story
    • even the little bit of voice-overs in the game are too much
  • tip: mix genres, or invent a new one
  • don’t try to fulfill expectations for a genre
    • example: FEZ doesn't need enemies in order to be a good platformer
    • leave stuff out and innovate elsewhere!

8) Cool game jams in YOUR neighbourhood

9) Thanks for listening / reading!

 TL;DR? Limit yourself so you can get things done.