Let’s face it, everyone hates losing. And no one hates it more than video games players, especially if they are playing solo and no one else is to blame. No one else? Really? Many games create some or all of their content by random, on the fly. And some others use trial-and-error gameplay: in a given situation where several options seem equally good, only one is really going to work while the other will penalize or eliminate the player. So the reason why a player loses may be beyond their reach! Should the use of random and trial-and-error gameplay be forbidden in games? read on.

The metozoic age: a history of random and arbitrary in games

Orginally, major sources of inspiration for video games included:

A set of D&D dice

 pen-and-paper role playing games and board games, which relied heavily on random as evidenced by the crucial role of dice,

A wooden puzzle

Once deconstructed, there’s only one way to put it back together.

 puzzles which required a trial-and-error approach, justifying the concept of arbitrary penalty, and
 skill games such as pinball, where the unavoidable perspective of losing can only be delayed, but also caused by on all sorts of contingencies outside the reach of the player. Note that pinball’s ancestry can be traced back to casino machines.

All these games have a similarity: even if the player makes perfectly good game decisions, on the basis of all the information they have, and even if they play with great skill, they can still lose.

Indeed, early games used random heavily, not only to handle what the player could be confronted with (random enemies appearing using random tactics or power ups appearing randomly) but also how the player could respond: for instance, in RPGs, characters were created by random, fights were simulated using a random function that also emulated dice rolls. Thus, a powerful but unlucky party of characters could be decimated by a seemingly harmless monster.

Not all games followed that pattern. One of the first games to refuse random was the original Pac Man. Each ghost hunts Pac Man following their own trajectory according to deterministic (non-random) algorithms. Thus, the game is designed in such a way that Pac Man can never be stuck in a corridor between two ghosts, an unwinnable situation. It’s probably one of the games that inspired game authors to create fair games, that is, games that will not penalize a player arbitrarily or depending on pure luck.

With the eighties, a new class of games hit the market, wiht constructed levels as opposed to randomly-generated ones. This was the advent of script. In a game, close to everything can be scripted: levels, behaviours, waves of enemies, patterns of attack, etc.

Scripted games opened a whole new way of playing. Instead of relying on luck, concentration and skills, players could learn the game patterns and devise strategies on how to progress in the game. With scripted games, the success of players depend more on their knowledge of the game and their ability to take enlightened decisions on that basis, than on just luck and nimble fingers. Thus, scripted games are generally more equitable (which doesn’t necessarily mean easier, see Battletoads or Rick Dangerous for instance).

Xenon 2, PC, 1989

Here the player can choose between two corridors. Sadly, the right one (selected) means certain death.

That change of focus can be questioned. What kind of player should videogames reward more? The joystick wizz-kid? or one able to learn their lesson by heart? Indeed, script has its limitations. Because the way a scripted game works is entirely determined, players may want to analyse the game to find exploits, counter-intuitive but very efficientt ways to play. Few Pac Man players know, for instance, that there is one safe spot in the maze where ghosts will never catch you. On top of that, randomness can be a very useful way to generate content, as long as it is playable. Scripted games don’t mean the end of trial-and-error gameplay. Adventure games, which were completely scripted, could be completely unforgiving, and could let the player lose for just any reason.

A reasoned use of random and trial-and-error

Modern games have understood these difficulties and only use random and trial-and-error in specific situations.

Colonization, PC, 1994

The player can describe the world they want to explore and let the game generate it.

The game environment can be influenced by random, as long as the resulting situations are fair. For instance, in RPGs, there could be random encounters, but the monsters the player has to fight are not powerful enough to eliminate a well-prepared party. Big fights are part of the storty line, there is an opportunity to save the game before, and they don’t take the player completely by surprise. In "4X" games (explore, expand, exploit, exterminate) such as Civilization or Colonization, the world map can be generated. But the resulting maps are always playable. Games won’t put a player who plays reasonnably well in an impossible situation.

SaGa Frontier 2, Playstation, 2000

When a light bulb appears above the head of a SaGa Frontier character, they gain a new ability. Light bulbs appear completely by random chance.

Some good things could happen if you’re lucky. For instance, a defeated monster could drop a rare item. Or you could score a lucky shot and blast that zombie in the head with one bullet. The reverse, which was not unusual in the games of the eighties (a monster steals your most valuable item and disappears, your trustworthy gun just jammed as zombies are getting closer), just doesn’t happen anymore.

Final Fantasy Tactics, Playstation, 1998

The player can send characters on mission, and know the outcome a few game turns later.

Optional sub-games are allowed to make unlimited use of random or trial-and-error. There nothing wrong to have gambling mini-games like the casino in GTA4. Likewise, when you mix items in Star Ocean, there is no way to tell what the outcome will be. But you don’t have to take that risk. Some games even combine random with delay: you have to wait to see the result of a gamble. That prevents you from exploiting the game. For instance, in Final Fantasy Tactics, you could send your characters in a mission, and find out what happened a bit later in the game.

Such games expect you to "configure" (your party of characters, your equipment, your cities, your armies, your general strategy, your set of pok�mons, you name it) so that you can face situations that can happen by chance with the best possible outcome. You know that the next big boss is vulnerable against fire? equip your fighters with flame weapons! you know that hordes of barbarians could raze your outposts at any time? Protect them with one unit of cavalry and a few archers! The soccer team you will be facing is said to have an offensive style? Buff up your defense!

The return of random and arbitrary games?

If you’re interested in casual or web-based games, you may have noticed a few families of games which use a completely opposed paradigm. In those games, there is just one arbitrary solution, all the others being bound to fail.

The Crimson Room, web game, 2004

The infamous game that reintroduced "pixel clicks".

Games of the "escape the room" type, where the player must, well, escape a room where they are confined, look like point-and-click adventure games. As we all know, mainstream adventure games were swallowed by the action/adventure genre (Resident Evil, Tomb Raider and the like), precisely because action/adventure games are less arbitrary and therefore less frustrating than traditionnal adventure games. "Escape the room" games go in the exact opposite direction: these are point-and-click games with little or no plot, and instead, a focus on frustration, obsfuscated puzzles and "pixel clicks", that is, tiny, unintuitive areas of the screen that must be clicked to progress.

Grow RPG, web game, 2005

The player must drag the icons on both sides to the game field to discover the story. Only one combination will allow them to beat the demon.

The award-winning Grow series, single-handedly created by Japanese game genius On, presents the player with a series of strange items, that must all be used once, but in the order that the player decides. According to that order, some items will grow and interact with others - or not.

At the beginning of the game, there is no way to tell which item goes first. No matter what you do, you will see some interactions between items, which could hint as to what to try next, but which could also be false leads. Of course, there is only one optimal solution, so that’s one chance out of 720 for 6 items and one chance out of 40320 for 8 items to find it. Then, there are many "false" solutions, that is, weird things that wouldn’t happen if you do the game "right", but which are pleasant to see nonetheless.

Hapland, web game, 2005

So many buttons... but only one solution.

A third popular series is Hapland, created by Englishman Robin Allen, and probably inspired by the early 90s Gobliiins games. Ironically, Gobliiins featured the type of gameplay that mainstream games wanted to avoid as they tried to attract casual gamers. Hapland, which goes perhaps further in that direction than Gobliiins, is one of the most popular and critically-acclaimed web games. In Hapland, the player has to open a magical gate, by following a strange combination of actions, some of which requiring an exact timing. In Hapland, any wrong action will make your game impossible to win. The game won’t tell you, of course, but you’re welcome to reset your game anytime. When you begin a game of Hapland, there is no way to find out what to do. But you’ll discover a few things to do, a few actions that link together. After many tries, they could start to make sense. So the whole premise of the game is based on trial-and-error.

These games can use arbitrary gameplay only because they come under a different format and they address an audience who has different expectations. A player that buys a game in a store has, in effect, bought the right to beat the game. They are entitled to a pleasant experience, free of frustration. On the other hand, someone who hears that the crimson room is the most frustrating game ever devised, and still wants to play it for free, only deserves to suffer.

Random and trial-and-error gameplay are therefore not evil per se, but their use should be restricted to some well-identified cases. At any rate, players should never have to lose because of that design choice wihtout having been warned.