Most of the time we play games to “beat” or to “conquer” them. It’s a battle between what ever traps and contraptions the game designer put in and player’s skills. Playing field isn’t necessarily level at all, as the designer decides what tools and options player has at their disposal. As a result, there’s certain perceived fairness that player expect. Designer could easily make a level that is impossible to beat, but that wouldn’t be fun to play for long. Players often equate success in game for having fun and failure is something that takes that fun away. Fail enough and game starts feeling tedious, boring or frustrating and player might choose to play something entirely different instead.
But in the same way as success is often integral part of the game play experience, so is failure (I’ll be concentrating mostly on failures that end the game as opposed to ones that just slow you down or set you back). Most of the time the player is facing some sort of obstacles that they need to overcome. Sometimes obstacles can be beaten with pure skill, sometimes there’s element of luck involved. Chess is a good example of game where pure skill dominates: all pieces are visible, starting position is fixed, there is no element of luck involved (except decision which colour to is played by which player). Poker is an example of a game where luck has bigger factor as cards are dealt from a shuffled deck. Skillful player is still able to dominate against less skillful player, especially in the long run. In both games challenge is in beating the opponent, who is trying to do the same thing to you. Failure means losing the game.
Infinifactory is again a game of perfect information: no hidden pieces, no random element of luck. Here the goal is to build 10 objects in the same configuration as the model one and transport them in correct location. If player can’t complete this, they’ll fail. But since there is no time-limit, player is free to keep trying different ways to build the object until they either manage the challenge or give up and move to another game. So there is an element of failure, but it’s completely up to the player to decide when it triggers.
Opposite of this are games like Missile Command and Space Invaders where failure is just a matter of time and player is merely trying to survive as long as possible and score as many points as possible while doing so. In theory game like this could go on forever, but in practice player will sooner or later make enough mistakes causing the game to end.
There are games without fail condition, like How to Die Like Graceful Beast by Aleks Samoylov, where game is more about experiencing story or world created by the author and less about conquering it. Here player’s expectations and previous experiences shape the game and set their goals. More and more experimental games where goals and fail conditions are different from expected norms are being made, as the games in general mature and are used in experiments and moving boundaries of the media.
But how can we make failure fun, if player is conditioned to avoid failure and try to conquer the game? Especially in a game where player is expected to fail a lot of times before being able to finish it? Or in a game where content is generated procedurally and the player is going to see similar situations over and over again?
One very critical thing is that player should be back in action as soon as possible after a failure. Showing a long, 30 second cut scene every time they die, before they can restore previous save and try another method of conquering the problem will eventually drive them away from the game. If you absolutely have to show a long cut scene, make it skippable. The faster the player can get back in action, the more likely they’re continue playing the game I would think. Super Meat Boy is good example of such a game. The player is going to fail lots of times, but every time they’re just thrown back to the beginning of the level and can continue playing immediately. No long cut scenes, no annoying game over screens, just pure action. Super Meat Boy gets another critical aspect right: controls and putting playing in charge of their success or failure. In the game there aren’t random rolls of die or other arbitrary decisions that would kill the player character and send them back to beginning. Player has full control of their destiny and that makes failure much more tolerable.
Common theme, especially in many roguelikes, is that every new failure will give player a little bit information about the game: what is dangerous and what doesn’t often work. When you get randomly hit by piercer in Nethack and die, you might get a little quip about how soft hat didn’t protect your head. Player might be able to deduce from here that hard hat is probably a good idea to wear. Or when you try to take cover behind some barrels in Doom and they explode because of enemy fire, player quickly learns that it’s not a good idea and will likely avoid that in the next time. New information could also be about the game world and provided in a form of short text. Similar idea that many games use for loading screens. So every time player fails, they’ll learn a little bit more about the world and can use that information to survive longer or have more fun in there.
While Space Invaders have simple interactions for leading to failure (get hit by laser and lose your last life), Nethack has much more complex system: player generally loses when their hitpoints drop to zero (if they don’t have amulet of life saving), but they can also drown or petrify. I found it always fun to discover new ways to die as some of them are quite surprising, but still make sense on a retrospect. Rich interactions with environment that make sense in the context of playing world keep game fresh and add verisimilitude. It then makes sense that there are multiple ways to fail and discovering these ways might be an interesting challenge on its own. In any case it keeps boredom setting in too quickly. Seeing same way of dying over and over again adds to repetition and game starts to feel like a chore.
Often failure works as a negative feedback for game progress. Player character loses all their gear, experience points and other knick knacks they have managed to collect. They might have to start from the beginning of the level or even from the beginning of the whole game. But it doesn’t have to be like that. Instead of penalizing the player harshly, a game could nudge them towards completion just so slightly. This should be done in a way that the game isn’t beatable just by dying as many times as possible. Drox Operative for example has a system where after game over you can start a new game with your current ship. Even if you’re having trouble in the beginning, after couple of failed games you probably have managed to build a decent ship that makes easy sectors easier to conquer. Coupled with difficulty setting that scales along your ship, this makes game enjoyable to play and I don’t mind losing a sector.
Another option to nudge player towards completion of the game would be unlocking content not only when they progress, but also when they fail. Imagine a roguelike game where in the beginning you only had basic types of monsters, items, rooms and such available. As you play, more of those are slowly unlocked and introduced to you (sounds like a neat way to build a tutorial too). After dying, the game would take you to beginning of the game in standard roguelike fashion and you would have to start all over again, but with a new world of course, that used all those things that you managed to unlock during previous plays. Game could even detect if you didn’t make further than before and randomly unlock some helpful things from list of locked features. These could include places where player character can rest, heal and gain better gear (inn, hospital and weapon store maybe). They could also include companion characters, so that instead of puppy, player character could start the game with full grown dog, that would help them during their quest.
I always loved playing BattleTech. Big part of the draw was the setting and background stories, but I also love how the game rewards creative play and players. While rules are fair and playing field is level (or as level as the scenario designer made it), players have huge range of options at their disposal. Many decisions came with risk-reward – combination. If ‘mech wants to move fast, running is often an option. But turning on pavement while running can lead to tripping and tumbling down. In one memorable occasion, a pilot failed his check while turning and ‘mech tumbled down. It then proceed to skid over the pavement and crash into a building and then out from the other end. Pilot surprisingly survived without much damage, but the ‘mech did took quite a beating. I didn’t mind much though, because the event was just so spectacular and I have told the story over and over again after that.
Another hilarious event in BattleTech was during a space battle between 4 warships and bunch of aerospace fighters. Warships were coasting in a distance and firing broadsides at each other, while aerospace fighters (who were formed into wings) were duking out in space between the ships. This meant that roughly half of any warship’s weapons were unused during any given turn. Until one wing got to one the wrong side of a battleship (this can easily happen with vectored movement if player isn’t paying close attention) and the battleship in question greeted it with a full broadside. Needless to say that even if it’s really hard to hit to a fast moving aerospace fighter wing with naval scale weapons, they did pulverize the wing completely. Again, the player didn’t mind that their wing got wiped out as it was because of their own mistake and looked so spectacular.
Do you think I managed to cover most of the important aspects here or did I miss something crucial? Is you favourite making-failure-fun – mechanism missing? Drop me a line in comments, as I love to hear other people’s take on this.