Thursday, July 21, 2011

Trial-and-Error: Balancing on a Razor's Edge

Games that involve a lot of dying or restarting are said to have trial-and-error gameplay; the player doesn't know what to expect or what to do until it's too late, and then has to do it again. Trial-and-error has become widely regarded as a bad thing. When someone uses the phrase "trial-and-error" when referring to a game, they usually mean that as a criticism. However, trial-and-error doesn't have to be a bad thing. In fact, trial-and-error can be a very powerful tool for creating an engaging, rewarding gaming experience. There's a very delicate balance involved with making trial-and-error effective, otherwise you're likely to end up with an unsatisfying gaming experience, but it's something that shouldn't be so readily dismissed.

Click to continue reading about the nuances of trial and error.

Trial-and-error gets a bad rap primarily for being frustrating. Everyone's gotten stuck on a level in a video game, forced to replay the same sections over and over, trying to figure out the solution to the problem. Try one thing, die/fail, reload, try again. This mechanic brings the pacing of the game to a crawl as the player becomes unable to make progress. Typically, whatever progress they earn is through tedium and sheer external willpower.

Trial-and-error can ruin the immersion of certain games. Resurrecting after death tends to remind the player that they're playing a game. But in the case of survival-horror, the genre relies on suspense and tension to evoke its horror. If a player keeps dying and going through a sequence repeatedly, the surprises lose their effect and there's no more tension. My recent experience with Siren demonstrated precisely how trial-and-error can ruin a good experience. I was looking forward to what was supposed to be a great survival-horror game, but the mechanics of figuring out what to do in the level ruined the feeling of immersion for me.

Some other games that feature trial-and-error gameplay, typically for the worse, which you may be able to relate to:
Mirror's Edge: This is basically a first-person platformer. You run along skyscraper rooftops, leaping across pits of death, running away from helicopters, climbing up walls and bounding around obstacles. The game commands a fast pace, often with armed enemies chasing you and frantic music, but it's not easy to realize exactly where you're supposed to go when you're speeding through the level. You often end up dying and restarting the section, or coming to a stop and awkwardly looking around trying to figure out what to do. The pits of death are especially trying. 
Hitman: Silent Assassin: Like a lot of stealth games, H:SA often involves a lot of trial-and-error to learn the map layouts, the routine cycles of NPCs, and environmental set-pieces. Each mission gives you a specific objective and there are usually multiple different ways to get your kill, but you end up failing a lot of times until you learn the best possible way to do things. This is often exacerbated by needing to do multiple consecutive things that you probably don't even know about, or the game not letting you do what feels like an obvious solution, and it gets even more frustrating when something goes wrong after patiently waiting for the right moment.
Call of Cthulhu: Dark Corners of the Earth: A broodingly atmospheric action/survival-horror, DCOTE has one major sequence relatively early that is based almost entirely on trial-and-error. You're staying in a hotel and are awoken by local townsfolk trying to break into your room to kill you. You have to latch deadbolts, push things in front of the doors, move to different rooms, all in a very precise fashion. There are like a dozen different things that have to be done in a certain order, within a certain time frame, and there's no way to tell what has to be done until you've died in each subsequent scenario. 
Trial-and-error is not an unfair criticism. There are plenty of occasions and examples where it really is just a pain in the butt. Sometimes, the only thing it adds to the experience is extra frustration. Some developers may think it adds challenge to their game, which may certainly be true, but it's not always a good challenge.The above games aren't bad at all---in fact they're pretty good---but the trial-and-error still makes them annoying at times.

Other games, however, have used elements of trial-and-error to create a more satisfying experience. Some games have absurd difficulties that trigger a more base desire to be challenged, in order to struggle and succeed against overwhelming odds. Usually, these are represented by "Nightmare" or "Insane" difficulty modes. But this kind of trial-and-error based challenge, used in moderation, can be a very compelling mechanic.

In order to feel accomplishment in a game, it's sometimes good to know that failure is an option---that there are consequences for your actions. Winning at a task that you can't lose is often a hollow victory; anyone can do it, and it took no special effort for you to do it. Knowing that you can die or fail at something, and then succeeding at it on your first try can feel very satisfying; if you've died a few times on a task, it might make the satisfaction even greater once you've mastered the technique.

But simply failing at a task is not enough to make an experience feel rewarding. If a player dies because of circumstances beyond their control, or for reasons they aren't able to comprehend, then they gain nothing. In order for trial-and-error to feel rewarding, the player has to learn things about themselves and about the game; there has to be an underlying reasoning for why they've failed, preferably in ways that affect their understanding of the entire game, not just one isolated instance.

In some games, players resort to trial-and-error as a way to learn more about the game. They play a sequence and think to themselves "How can I do this section better," then go back and try to improve their performance. As such, trial-and-error becomes an intrinsic tool, not a forced mechanism. If you've ever played a game where you quick-saved before every conversation or before entering a dynamic battle, just so that you can reload and try it again with a different strategy, then you know what I'm talking about.

In fact, quick-saving is a valuable asset for avoiding some of the tedium of trial-and-error. Quick-saving allows players to save their progress without breaking the flow of the experience, and gives them the power to try different strategies with efficiency. And of course, quick-saving limits the amount of time players spend re-playing parts of a game, since they can save immediately before the specific part that's giving them troubles.

And, of course, good atmosphere and/or story can help make trial-and-error bearable or even enjoyable. There has to be some sort of reason to keep trying things over again, some compelling motivation beyond wanting to beat the game or hoping it gets better. I've put up with annoying deaths and raged at my monitor enough times in plenty of games, and continued playing just because I was so engaged with the atmosphere and immersion. 

To put some of these ideas into context, I'll describe examples from a couple of games that implement effective trial-and-error.
Gothic: Gothic is a game that can be rough for new players because there are a of lot rules that aren't readily apparent when starting out. Many of the player's actions have specific consequences that you wouldn't necessarily predict, and NPCs behave a bit differently than in other games. As a result, the player has to learn the mechanics of the game system (including its fictional world) in order to be successful.  But fortunately, every "error" you make presents itself as a learning opportunity that shapes your later experiences.
Most of the enemies are stronger than you and can kill your easily, so it's up to you to figure out which ones you can handle at any given time. Different enemies require different tactics to take down. A lot of NPCs lie to you and manipulate you when you're just starting out. NPCs react if you draw a weapon or visibly see you stealing from them. All of these things can get you killed if you don't know what you're doing, and it usually takes a death or two in order to figure out how the game works. But once you learn these things, you feel more empowered to tackle the challenge.
Spelunky: A 2-D platformer with randomized levels and challenging gameplay. There are a lot of instant deaths awaiting you in Spelunky, and a lot of things that will trip you up and damage your already limited supply of health. Randomizing the levels helps to alleviate a lot of the tedium, since you're not playing the same level each time, and it also motivates you try things again because you know the level will be different (for better or for worse). 
Like with Gothic, though, every time you commit an error you learn something about how the game works, and you use that knowledge on subsequent trials to do better. You see a suspicious-looking rock and get shot with an arrow; you now know to avoid those. You throw a pot against a wall and a spider jumps at you; you now know to break those with your whip to pre-emptively hit any spiders. 
Portal: Whether you realized it or not, Portal is all about trial-and-error. Most puzzle games are. In Portal, you're given tools to solve a puzzle. You try one thing, it doesn't work. You try another thing, it also doesn't work. But with each subsequent trial and error you move closer to solving the puzzle and figuring out what has to be done.
The consequences for "failing" in Portal are relatively slim---you don't have to replay major sequences and you definitely don't lose any "progress" when you die, since the progress towards solving the puzzle is all in your brain. But it also helps that Portal has an excellent atmosphere that compels you to continue playing, even after you've fallen into that pit of death for the 9th time in a row. 
Not all games and not all kinds of games benefit from trial-and-error. Sometimes it's fun to have a fluid, easy experience that doesn't require the player to test his brain figuring things out on his own. Personally, I can find a lot of satisfaction from trial-and-error, because I tend to think like a scientist. It's fun for me to test alternative options until I find the best solution. As Thomas Edison once said, after finally making a functional lightbulb on his 1000th attempt, "I didn't fail 999 times; I succeeded in finding 999 ways to not make a lightbulb." 

But trial-and-error is an easily misused mechanism in gaming, and it can be difficult to implement just right, for it to be satisfying and not tedious. It can be an effective way to instill challenge into a game experience, while giving the player a sense of feedback about their performance. Having consequences in a game can go a long way towards creating an engaging experience. Hopefully this goes to show that trial-and-error is not all bad, and can in fact add significant depth to a game when implemented properly.

And just for the fun of it (and since there weren't any pictures breaking up that wall-o-text), here are some videos demonstrating absurd trial-and-error. Rather entertaining stuff. The first video is of a hacked Super Mario Bros ROM which you may have already seen by now. WARNING: contains lots of profanity, potentially NSFW. The second video is of I Wanna Be The Guy, which basically satirizes the hardcore difficulties of original NES games. 

No comments:

Post a Comment