Ever heard of the "Messiah"? The one game to rule them all, find them, bring them and bind them all in the darkness?

If you ever listen through E3's press conferences (or any other such events), you've probably heard about it already. You know, that one... It's been promised to us gamers a thousand times over, but we're still waiting for it:

The non-linear game where our decisions will truly matter and have consequences.

Just let me make something clear: I do not own any word. I just want to make clear what exactly I'm talking about and I believe I have to define the terms I use for that.

Story-driven

First off, to be story-driven, a game has to feature a story. Note that many wonderful games don't feature stories. Those are not the types of games I want to talk about here.

Featuring a story is not enough though... It's a matter of focus (hence the word "driven"). It is sometimes genuinely hard to tell if a game is story-driven, but it is often enough to ask yourself the following question:

Is the gameplay helping me immerse into the story or is the story helping me care about the gameplay?

Hack'n'slash is a whole genre where story (when well-done) is most often only a background, pure fodder for the gameplay. Try Titan Quest, Torchlight, Champions of Norrath, Kingdoms of Amalur: Reckoning and you'll see. Maybe you'll notice the story, maybe you won't even... But you sure won't keep playing for the story. It works the other way around.

Note that lots of genre are much less story-driven than hack'n'slash, as a rule. Try real-time strategy, first-person shooter, shoot'em'up, sport simulation, platform, puzzle, etc... chances are you'll find way more gameplay-driven than story-driven games in all those genres.

Want an example of an undoubtedly story-driven genre? And I mean genre, not game. Here are two: point'n'click and visual novel. In fact, those genres are sostory-driven, the most common complaint among detractors is "there is no gameplay!"

Interactive

I don't mean interactive gameplay-wise. I mean story-wise.So, no, the fact that you can reach a game over screen if you don't play well enough for your avatar to reach the next cinematic doesn't count. It's gameplay interactivity, not story interactivity.

Thing is, most games are not interactive, story-wise. Even (especially?) very story-driven games.

To tell if a game is interactive story-wise, here are the questions:

1. Does the game present choices for the player to make?
2. Do those choices bear any consequences?
   Slightly different dialog lines do not count as consequence if all alternatives end up at the same point afterwards.
3. Are there often good/correct options?
   I do not mean the morality of the option as in "Save the villagers" vs. "Kill the villagers".
   I mean good/correct option for the player as in "Continue the game" vs. "Game over" or "Better score" vs. "Lower score". Those choices do not feel like real choices (and aren't, IMO). They feel like tests (and are, IMO).
4. Do the choices alter the story itself?
   Again, if the choices only alter aesthetics, gameplay or dialog, it's not all that interactive. To answer yes here, the plot of the game has to be impacted.
5. Do the choices alter the story significantly?
   There be dragons. Who am I to tell anyone what is significant? No one. It's not my point, fortunately. Here's what I mean by using that word: there are two level of stories in most games. There's the overaching "main" plot, and then there's the rest of the game universe. What I call a significant is any impact on the main plot. For instance, if you're trying to save the world, a choice that makes your favorite character die, but doesn't affect your probability of saving the world is not significant. Not to say that it's not important and emotionally important... it just doesn't alter the game's story significantly.
6. Do the choices alter the story long before the end?
   It is relatively easy to put alternative endings at the very conclusion of a game. "Oh, I have like four ideas how this story could end... let's make the player pick." It's nice. It's a shame that many games won't even make that (slight) effort. But who are we kidding? If you wait until the very end to actually impact the story, it means the story isn't truly interactive. Only the ending is, basically. And that doesn't preclude it from being awful, by the way.

Most story-driven games are somewhere between level 0 and 3. Games that make it to 4 and 5 tend to become legends, hailed as true non-linear adventures where your choices really matter. But really, it just means they're exceptionnally interactive (compared to most other games).

Games of level 6 are rare. So rare, in fact, that none comes to my mind as I'm writing this line. None. How sad.

Non-linear

There be way more dragons. And a black hole. Seriously.

I know I'm going to sound like a judgmental know-it-all saying this, but I feel I have to give context to my definition of non-linear...

I have played hundreds of non-linear games.

I have played hundreds of story-driven games.

But I have never played a non-linear story-driven game.

That's right. I'm saying I have never played what I'm talking about. Not that it doesn't exist, mind you. Just that if it does, I have never heard about it. 'cause you can bet I'd play it if I did.

So what do I base my definition of non-linear games, anyway, since I've never played one?

Two things.

First: all the non-linear gameplay-driven (or story-free) games I've played.

Second: table-top RPG, which I'm well aware can never be fully translated into a video game, but I believe there's enough leeway for a solar system between the closest we could get to it and where games are at.

Enough preparation and disclaimers. Here's what must be true about a story-driven game, for it to be non-linear in my book:

• There must always be more than one way to advance story.
  Not all ways have to be available in all playthroughs, of course. If you made a berserker, it's logical that stealth and diplomatic options should be technically locked out for you. But then, if you replay the game and make a bard, they should be available and not the frontal assault.
  Note that I am not talking about gameplay. If you're given a mission to deal with foes, you have many ways to deal with them, but all end up with the same "congratulations, you did it" sequence after the mission, then it's gameplay non-linearity, but the story is still linear. The Deus Ex franchise is the best example I know of for that: great non-linear gameplay and despressingly linear story.
• No event must be built inevitable outside of the story's premise.
  That's where even the most non-linear story-driven games out there (I'm looking at you, Bethesda) fail miserably. It is okay to make an NPC extremely strong. But if the NPC is completely immortal, then you do not meet this criterion. It is okay to equip an antagonist with an extremely powerful shield. But if there is no way to take down that shield. Not even if the player cheats... then you do not meet this criterion. Honestly? No game I know meets this criterion. All story-driven games I've played are so choke-full of built-in inevitable stuff, it makes me laugh to hear anyone discuss their linearity. Guys, there's an event I cannot prevent even if I cheat, like a monster I can't kill even in god mode or a cinematic where I'm stuck watching an NPC die and I can't even try to help. What would you call that? I call that linear.

Part 1?

I wanted to talk about some "rules" I made up and would like to try to follow when I design my own games (a few years from now, *fingers crossed). But the definition part ended up SO HUGE, I decided to postpone the rules part for a later time.

Just a teaser for now: A door is a door is a door is a door.

I've never been a fan of difficulty in video games. I tend to play in Normal or Easy and don't stick to video games that resist me too much.

For the longest time, I thought I was a sore loser, an all-around impatient player(to the point of being bad, no doubt)or just someone who draws too clear a line between games and work to appreciate hardships in a game. I still think all of those things are true to some extent.

But then I played Super Meat Boy and enjoyed it immensely. I was not repelled by the game's in-your-face difficulty. I died thousands of times and kept trying for more.

That led me to believe that there are two kinds of difficulty: the Right and the Punishing.

I compared tons of games I played, some for countless hours, others for mere minutes, and I came up with the following definition.

The Definition

The difficulty in a video game is Right (by opposition to Punishing) if most of the times the player loses:

1. He/she feels like he/she died through faults of his/her own.
2. He/she has a clue as to what must be done to improve and/or overcome the pitfall.

At face value, both conditions seem completely subjective. In the exact same situation, Player A will feel cheated, while Player B will see something he/she did wrong and get an idea for next time. Of course, no matter the game, there will always be some players who'll deem it a cakewalk and some who'll never be able to survive 2 minutes past the tutorial phase. Let's forget about the extremes and even the individuals for a second, to focus on the majority.

How do you make sure most people will feel like it is their fault if they lose? How do you give players clues as how to improve?

The rules

There ARE things the game designers should pay attention to, if they aim to craft an experience that will not turn most gamers off through unwarranted punishments. Here are some (not all) of those things in no particular order:

Rule 1: Fine tune the damn controls!

Never is frustration greater for the player than when he had the right idea but could not convey it in time to its avatar... because the controls failed him/her.

Finicky controls and complex commands are only fine if the related actions are not performed under stress.

Rule 2: Err on the side of the player

Say the controls are as fine-tuned as they can possibly get, yet there is still doubt as to what the player really intended to do. For instance, when the player targets with an analog stick, there will be times when it is impossible to determine exactly what the player was aiming for.

Let us say the game computes 48% chance of something sensible, like the next platform on the way to the exit or the enemy right in front of the avatar, and 52% chance of something utterly silly, like a fall to the avatar's death or the torch just next to the enemy's head...

The game should err on the sensible side. Chances are, that's what the player really wanted to do. If not, most players won't hold it against the game.

Rule 3: Feed back

There might be a perfectly sound chain of consequences between the player's actions and its avatar's untimely death, but that logic is void if it plays in the shadows, behind the player's back.

There has to be some way for the player to tell what's happening, especially if what's happening results in a game over. It doesn't have to be obtrusive like a blinking red screen or verbose like a detailed numerical account of damage dealt. It can be a tiny icon with a helpful tooltip when hovered, a color code, a specific animation, anything that will be seen and eventually identified by the player so that he/she will see what leads from A to B to C to Game Over.

Rule 4: Don't rub it in

Losing sucks.

The game or the player or both had set a goal, a score to break...and it failed. The avatar died, haphazardly crumbling like the mass of polygons it is, while a voice actor's throat was slit in the background... The player's rank has decreased, the "Indestructible" achievement is forever lost, the section has to be replayed yet again, the player's friends will make fun of him/her for days, another match-making session has started and will last at least fifteen minutes and the player feels the bitter taste of failure.

That is enough. There is no need for further punishment beyond what is required structurally. Seriously, there is never any need for horrendously long death sequences, already seen yet unskippable cut-scenes, shortages of checkpoints that induce waste of hours of gameplay or insane death penalties that serve no purpose other than deter players from daring plans.

One of the things that make Super Meat Boy such a gem is the way it rewards the player for dying: the more you die, the more awesome the replay video you're treated to. It doesn't make losing fun (it never is), but it makes it bearable.

Rule 5: Consistency over Realism

Games do not inherently need to be realistic. They are entertainement and, as such, they provide an escape from reality.

If we put aside so-called "serious games" that really are simulation softwares, no game is ever truly realistic. Sure, you might find here and there guns and cars that behave frighteningly close to how the real thing would... but what about the setting? Right.

Does it matter that your elven avatar has magical powers? That the Destroyer of Nations was killed for the 939729532nd time today? That you're surrounded by undead clones hungry for your brains? None of those things are realistic. But they're fun for millions of players, so who cares?

The problems arise with inconsistency.

It is perfectly fine to wield an incredibly powerful italian hunk who can hold on to a ledge by two fingers for 2 weeks in a row, fend off a hundred trained and armored soldiers barely breaking a sweat and fall down 200 feet without a scratch. But why on Earth can he not move furniture around?!?

In the space between what should logically be, in cohesion with the rest of the game universe, and what is actually implemented, there lie hurdles that will surely feel unjustified.

Ultimately, every game has to set its own rules. If game designers do not want their players to bang their heads on the wall yelling "WHYYYYYY?!?!", they would do well to actually decide on those rules, write them down and follow them themselves.

Rule 6: Do not half-tutor

There was a time not so long ago when tutorials were rare. Most games expected the player to read the manual or to figure stuff out on their own. Some hardcore gamers say that age was golden and that games should never take the player by the hand. I am anything but a hardcore player myself and I love to be eased into a game through a good tutorial.

I could probably write a pamphlet twice as long as this one on what makes a tutorial bad or good... for now, let's focus on just one thing: half tutorials. By that, I mean tutorials that introduce the player to only part of the mechanics in place in the game (usually the most easy to explain, which tend to be the most obvious and easily figured out on one's own).

Game designers. Please. Do not ever do that. If you're going to show the ropes, show all the ropes. Do not explain parts of what is needed to play and then throw the player to the wolves with the illusion that he/she is ready to fend for him/herself. That will result in blood, tears and betrayal charges vehemently screamed at the innocent screen.