Trial and Error Based Gameplay

[Koh]

Yes.  THAT trope.  We've to talk about this.

Think about any boss ever.  In order to beat them, you have to learn their strategies, weaknesses (which entails weak points), and any thing else to successfully achieve victory.  But can this be accomplished without trial and error?  One could argue that it doesn't have to be as long as there's room for human error (life bars/multiple hits can be taken, etc.), but then it questions the design, does it not?  But one could also argue that it's inherently trial and error.

What constitutes trial and error besides being forced to do the same thing over and over again until you create the perfect run?  Is there a such thing as "good" trial and error, or is it all bad and frustrating?

[psychospacecow]

One that comes to mind is Pete in Dream Drop Distance. He has an attack that more or less attacks the entire stage and can kill you instantly. You have to jump into it if I remember correctly.

[Arcrom]

I wonder when there is another factor in equation, RNG. In some turn based strategy games, even if you know what kind of attacks the boss can unleash there still a chance, even if is 1%, to cause a miss hit.
 
One example is Xcom Enemy within where there was a soldier with 95% chance to hit an enemy and he simply missed and vice e versa (10 % chance to hit and he targeted the enemy).
 
One of the Hearthstone challenges also follows this pattern, the boss plays the same cards in the same order, but your cards only provides random effects that even knowing his strategy, RNG can simply help or defeat you.

[psychospacecow]

This is true as well. I play a knife juggler into muster for battle. My opponent gets 3 damage to face.
My opponent does it, he wipes my board.

[Arcrom]

An even more sad history is when you play arcane missiles and they decides to visit my opponent face when there are minions on his board.
 
This thread also remembered when I faced the Lord of the Dragovians in Dragon Quest 7 and his first attack paralyzed my entire team. Didn’t matter my combo setup, his attacks or items/gear, I just watched the gang go to the floor after some turns. (The turns that you remain paralyzed is random)

[psychospacecow]

A poster boy for this entire discussion would be Yu-Gi-Oh! Forbidden Memories. You always have 5 cards in hand. You can select cards to fuse them. If they don't fuse, the first card just dies. If they do fuse, you might not like what you get. The final boss is actually a string of about 5-8 boss fights in a row without save points. Every single one of them can summon the strongest monsters in the game. Every match that you win gives you a random card. If you can't summon a stronger monster, you either draw into Dark Hole, Raigeki or lose. The game snowballs faster than a ball of shovels rolling down a mountain in Alaska. Everything is a coin flip. Even the deck you start out with is random.

The fastest run for Forbidden Memories was 3 hours 29 minutes and 27 seconds. Most runs I've seen run for at least a day.
http://www.speedrun.com/Yu-Gi-Oh_Forbidden_Memories
http://www.twitch.tv/theurnator/c/4304064

[Koh]

RNG factoring into it creates Luck Based Mission, if I'm not mistaken, which can be equally as frustrating and annoying.

I Wanna Be the guy and anything that plays like that are probably the most extreme examples of Trial and Error, where there's literally no other way to win unless you fail a bunch of times first to learn what not to do.  I can assure you that kind of frustrating gameplay will not be present in our games.

[recme]

fire emblem awakening: lunatic +. all i gotta say.

[puggsoy]

Stuff like Super Meat Boy and Dustforce seem to fall into this category (I Wanna Be The Guy as well, although I haven't played that). While I can see what you mean by frustrating, I actually do enjoy these kind of games sometimes. The satisfaction of finally fine-tuning your attempts and getting to the end in one go is really great, even if it took 100 or 1000 deaths first. It's a matter of taste.

It's much like trying to break a record in something, be it throwing shot-put a certain number of metres or fitting 100 grapes in your mouth. You're not gonna get it the first time, you're gonna keep failing and keep trying until you finally get it. The satisfaction and skill you've acquired from doing it is why you're doing it. That's the attraction of these sort of games, for me at least.

That's also why high-scores are a thing. You practice and practice until you're #1, or just beat your best friend's score. This requires you to fail and learn from your mistakes, tweaking your technique until you get it right. It also causes you to try out different things, which might be more efficient or faster ways of doing the same thing, even though they're not the most obvious or straightforward ways.

Again, it's a matter of taste, and whether you're willing to persist for the reward. Some people like it, some don't, and either way is totally fine.

[psychospacecow]

fire emblem awakening: lunatic +. all i gotta say.

What do you mean the boss of level 10 is stronger than the final boss on normal?

[Koh]

-le snip-

True, the severity is all according to the player's interest.  

Another thing about it would be if it's really skill based at all, or if it's just a fluffed up version of Quick Time Events.  Knowing a boss can do one of several tactics at any time, and using visual cues to prepare a dodge or counter makes use of the player's skill.  Fighting a boss repeatedly to memorize and regurgitate the same order of attacks, such that you know exactly where and when to stand with your one-hit wonder character is no different than a Quick Time Event, where you just have to press a single button at the proper moment, or else, you go back and do it again.  i.e.  The Impossible Game, that's really just easy as heck once you've memorized the level design, but gives the illusion of difficulty otherwise, 'cause you die in one shot and go all the way back.

Speaking of lolsy Quick Time Events: https://www.youtube.com/watch?v=5gWz46t50rU

[puggsoy]

Again, with stuff like that, it's preference. Super Meat Boy's bosses are literally that, especially Meat Golem; he attacks in a certain sequence, you just have to dodge it all. I suppose it's a souped-up, more complex version of Simon. But instead of just memorising, you also need to develop the skills of handling the character, and mastering specific techniques (e.g. wall-jumping, Dustforce especially has lots of various manoeuvres for improving speed and smoothness).

Even though memorising and timing seems really simple, it's still a solid gameplay mechanic that, when executed well, I personally enjoy.

[Sketchasaurus]

Learning at its very core is trial and error.

Skills are built by making mistakes and trying other strategies, expecting one to be entirely literate in your game's own design "language" is plainly put: bad design.
You need to facilitate the learning of the nuances of your game THROUGH the gameplay of the game; if you expect them to understand how the game "should" be played, people are going to do one or some of the following things:
A ) Be frustrated with interacting with the game "where do I go, what do I do? How do I do this?"
B ) Feel cheated out of their time "I wasn't prepared for this"


There's a huge principle that validates this trope that has been at the very core of game design from the beginning: Bushnell's Law:
"All the best games are easy to learn and difficult to master. They should reward the first quarter and the hundredth."

When designing a game, you can't assume that they know how they're supposed to interact with your game, you design with the lowest common denominator in mind, not those you'd imagine to be proficient from the get-go. Sure, there's a learning curve, and that road to proficiency is paved in trial and error.

What you need to consider is a balance between risk and reward; you pose a challenge to the player, and in the chance that they fail, you provide an opportunity for them to try another option or try again, after their success, you reward that success in-game, but that reward is also something that a player receives from themselves for their efforts: the feeling of accomplishment.
Memorization and pattern recognition is entwined in game design: The behavior of the ball and enemy paddle in pong, the movement of the ships in space invaders, the attack patterns of the enemy ships in Galaga, the ghost movement in Pac man, the jump arcs of the springs in Donkey Kong, even Mario's Jump arc is a predictable pattern that you learn from trial and error in Donkey Kong.

A boss's attack pattern is not the same as a QTE, to even consider that point is laughable. Enemy movement and boss movement ALMOST ALWAYS has a pattern behind it; it's less rewarding(and exponentially more frustrating) if enemy/boss movements are entirely random: THAT is unfair design. QTE's are more to build a sort of "involved suspense" during moments of a cutscene or other "feeling" moment, they only require the basic instinct of button pressing when prompted. Memorization stages require you to use what you have learned about the interaction with the game in order to test your proficiency in that interaction, or to sharpen those skills, that's what the "challenge" of a game is.

[Koh]

A boss's attack pattern is not the same as a QTE, to even consider that point is laughable. Enemy movement and boss movement ALMOST ALWAYS has a pattern behind it; it's less rewarding(and exponentially more frustrating) if enemy/boss movements are entirely random: THAT is unfair design. QTE's are more to build a sort of "involved suspense" during moments of a cutscene or other "feeling" moment, they only require the basic instinct of button pressing when prompted. Memorization stages require you to use what you have learned about the interaction with the game in order to test your proficiency in that interaction, or to sharpen those skills, that's what the "challenge" of a game is.

You misunderstood my use of the word random. The attacks themselves have patterns, but their ordering does not have to. The boss can do one of the following
A: Dash across the screen, causing damage to the player if the player is in its path.
B: Jump over the player and do a ground pound attack, which damages the player if they're caught under it.
C: Fire a zigzag lazer that burns the player on contact.

Each of them would have some sort of visual cue that they're about to happen. But they don't have to have any specific order to them. If they're always in the same order every time, it just becomes a memorize and regurgitate game, like Puggsoy mentioned, Simon, or like Quick Time Events, where it's the same button every time, only the next time you now know the button order. However, if they're randomized, it puts the player's ability to perceive from the visual cues, and their ability to adapt to a random yet controlled event, which requires more skill than simply replaying it over and over to memorize when and where the boss will be, and create the perfect run.



That reminds me, this is a good read that goes hand in hand with this thread, specifically the Live and Learn section.  For the lazy, a copy paste version is below.

Live and learn

As a rule, adventure games should be able to be played from beginning to end without "dying" or saving the game if the player is very careful and very observant.  It is bad design to put puzzles and situations into a game that require a player to die in order to learn what not to do next time.  This is not to say that all death situations should be designed out.  Danger is inherent in drama, but danger should be survivable if the player is clever.
As an exercise, take one complete path through a story game and then tell it to someone else, as if it were a standard story.  If you find places where the main character could not have known a piece of information that was used (the character who learned it died in a previous game), then there is a hole in the plot.

He talks about other aspects of game design there, but this touches on what amounts to the general consensus of bad trial and error.

[Sketchasaurus]

Consider for a moment what particular genre this person is speaking about for a moment; namely being "Adventure Games" within the context of The Monkey Island and similar franchises. Despite his praising of being "clever" or "careful and very observant" such games are laden with trial and error, not in the sense of death then try again, but a series of "try object x on area y, recieve positive/negative feedback, try object z on area y, ad infinitum" The genre of this game is in-effect, on life support and is only hanging on due to those paying homage to their nostalgia.

This article mainly amounts to one man's pretentious rambling about the design of games-now-gathering-dust; and it's idiotic to assume that every aspect of game design within a genre will translate wholly to any of the other genres. The article is being used more as a "confirmation bias" to which "general consensus" cannot be attributed, largely because the demographic, and topic, of the article in question is "Adventure games"
To assume that a singular source is sufficient to arrive to a general consensus is purely close-minded and idiotic.

When I used the word "random" it was in reference more to throwing all caution to the wind and having no conditions or reasoning behind the attack/movement choice. If a boss has 3 different attacks, they behave when different conditions are met, sometimes relative position to the player is often the main condition for attack/movement choices, and it has the Illusion of randomness. random attack patterns would be just that, entirely random, it could cause the boss to accidentally continue moving towards the player or even not attack, that is what I mean by random attack patterns. And actually, randomly moving around the room with different attacks, regardless of where the player is or what they're doing is just poor player/boss interaction.

But, you can manipulate these conditions and fight bosses in the same fashion after figuring out (via trial and error) what player moves prompt what kind of attack is acted out on behalf of the boss. The "random" actions only seem so, but that's due to the fact that every run through is never the exact same, not even memorization-based playthroughs.

What you're referring to is telegraphing attacks, which is a facet of pattern recognition, but also is part of a pre-defined pattern in itself. And the player quickly learns what these symbols mean through interacting with them, sometimes at the cost of HP or lives or whatever arbitrary system is used.

Consider the Red Koopas in Super Mario Bros: They telegraph their movement patterns through their palette; the game introduces the Green Koopas, which walk off of ledges, but you soon learn by observation that the red ones don't walk off of ledges, but the game doesn't outright state anywhere "Red Koopas won't walk off ledges" Instead, you have to observe and gather information. Part of this "trial" is observing the new information presented, an "error" would be assuming it to walk off the ledge and ignoring it (while standing on the same ledge), only to have it turn back around and inflict damage.

Or, take the Buzzy beetles for instance:
Imagine you've never read the manual or heard anything about them before encountering them in the game.
You have the Fire Flower, so far it's been the best weapon in your arsenal while dealing with baddies.
You see a new enemy: dark, smooth-shelled creatures scuttling along
you quickly, almost without thinking, try shooting a fireball at it while you're still sprinting along
But before you can stop, the fireball has no effect, and you accidentally run into it, losing the fire power.
You then try jumping on it, to find that it behaves like the Koopas do when you jump on it

you learn: Buzzy Beetles are fire-proof, but I can still jump on them.

You may have lost your fire flower, but you've still got a chance to try something else without starting the whole level over.

The game doesn't yell at you for messing up, it doesn't outright punish you for trying something that doesn't work, heck, maybe you jump on it in time to still have your fire flower, doesn't matter; people learn by trying things out, not by having things told to them (heck, you might learn from this entire block of text until you try any of this stuff out, and that's totally fine)

Memorization isn't inherently bad, it can be done well and it can be done poorly; the same is true of trial and error. As a simple reality of learning, the principle of trial and error is neither good nor bad.