Gw Temp


Article - 'When Linearity Attacks!' by Brickroad

An item about Game Design posted on Jan 14, 2004


Brickroad discusses linearity in your game and how best to utilise it


What is linearity?

Before I go anywhere with this, I will define linearity and nonlinearity for the scope of this article.

Linearity is a measure of the amount of options you have while playing a game. A linear game has a "railroad track" to follow; every time you play the game you'll do the same things in roughly the same order. A non-linear game provides a "spider web" to follow; every time you play the game you'll be faced with a series of decisions about where to go and what to do.

Only events linked directly to gameplay are used while measuring linearity. A hidden secret or bonus in an otherwise linear game doesn't necessarily make it less linear; it just means that it happens to be a linear game with a secret or bonus in it.

A perfectly linear game is best thought of as a string; the player starts at one end and moves to the other, with no opportunity to move off the string. Sonic the Hedgehog is a great example of this; there are six Zones with three Acts in each one, all of which must be finished in order. Then the final boss. (Note: there are secrets to collect in the form of Chaos Emeralds to find, and multiple paths to take within the actual stages, but the Emeralds don't affect your game and the paths always lead to the same place.) Sonic the Hedgehog is, therefore, perfectly linear or, for purposes of this article, a string.

A roughly linear game can be envisioned as a tree. The player starts at the base and moves towards the top. Along the way though are branches the player can explore. Inevitably, however, the player is going to have to come back to the trunk so he can advance along in the game. A good example is Super Metroid. There are a good number of items you need, and a fair amount of bosses you must beat, and a rough order in which they must be accomplished. The more branches you take, the better equipped you are when you come back to the trunk. Therefore Super Metroid (and most commercial RPGs) are trees.

A non-linear game is, like I said, a spider web. The player starts in the middle, and can end anywhere along the edge. There are numerous paths, and you often can go back and take a different path if you change your mind. Super Mario 64 is a prime example of a non-linear game. The goal of the game is simple: get 70 stars. Right out of the gate, there are about twenty stars the player can get. The more stars you get, the more you're able to reach. There is an enormous number of combinations that will work and still allow the player to win. Mario 64 is, therefore, a web.

A perfectly non-linear game is one where the player is presented with every available option from the beginning. It's envisioned as a vast open plain. Ultima: Quest of the Avatar is like this. It doesn't matter where your starting point is, since everywhere on the plain is the same. Ultima actually has a quasi-random way of determining your start location. You still have a set amount of tasks to complete, but they can be completed in virtually any order. Some of the stories in SaGa Frontier also qualify as plains.

Why strings and plains don't work.

Of the four examples above, what strikes you as most fun to play? If you answered trees or webs, you're in the vast, vast majority. Most people just find strings suffocatingly confining and very boring after one playthrough. They saw everything the first time they played it, after all... why play it again? As a result, strings usually lean on hellacious difficulty to remain interesting.

I used Sonic 1 as an example earlier, for good reason: the game has a very steep difficulty level, especially after you get into the fourth zone. Most players will hit a plateau eventually - they can get to Zone 4 Act 3, for example, but no further - they always run out of lives there and have to start over. With time they'll be able to progress, just to hit another wall in Zone 6 somewhere. Once they finally do beat the last boss, chances are they'll be able to do it again and again - but they won't, because the game has gotten stale and dull.

Plains have the exact opposite problem. Have you ever played a game which very quickly overwhelmed you? You feel like you're drowning in a gigantic game world with no idea what to do or where to go. Only by wandering aimlessly will you eventually stumble upon a game event which, once completed, leaves you right back where you started - as lost as ever.

Players will give up on strings because they quickly get same-y and frustrating. Players will give up on plains because they don't have any clues to follow. A game that is designed to fall into either of these categories is being aimed at a very narrow audience. This isn't to say they're bad games, just that they won't appeal to most people.

Two kinds of "optional."

Secrets, bonuses, hidden stuff... call it what you will. Options are the key to controlling a game's linearity. There are, however, two kinds of optional; one directly relates to linearity, and the other has nothing to do to it. Ironically, it is the false options that affect how linear a game is, and the true options that do not.

A true option is defined as an event the player can either trigger or neglect, without altering the flow of the game. Ruby and Emerald WEAPONS in FF7 were truly optional - you don't have to beat them to finish the game, and doing so does not give you any rewards. (Yes, I realize the game does give you something for beating both, but the reward is less than what is likely required to accomplish the task in the first place, so they're not really rewards at all.) In other words, if the player can beat either WEAPON he is already skilled or equipped well enough to finish the game, and once they are beaten the game is not any easier or more difficult to finish. Neither WEAPON affects the game in any meaningful way. Thus, they do not make the game more or less linear.

False options often come in sets. The player will likely have to complete one or more of these events, but if they don't do any of them, they'll never win the game. The second world of FF6 is a super duper example. You'll very quickly find yourself with three characters and an airship, and that's the end of the required game. However, it is impossible to defeat Kefka in this state. You must go and complete some of the optional tasks to even stand a chance. And if you think you can sidestep the situation by just levelling your characters up to L99, you've still failed because powerleveling was one of your false options to begin with! Consider the following combinations of totally optional events, any of which will allow you to beat Kefka:

- Go and get the other eleven PC characters, complete no other optional quests.
- Get just a few characters, and teach them all powerful magic.
- Don't get any characters, but some strong weapons and level the three you have very high.
- Complete only the quests necessary to get the "broken" relics (Offering, Gem Box, Economizer, Moogle Charm...).

As you can see, a set of false options has a profound impact on linearity. Each of the four scenarios above will allow you to beat Kefka, but each one comes with its own set of pros and cons. Each has a different set of strategies you must use. Each offers a different challenge for the same player on subsequent playthroughs. And of course, this is neglecting the obvious possibility of "get every single optional thing in the game"... which of course makes trouncing Kefka really easy.

Non-linear for the sake of non-linear.

Nonlinearity is often touted as a "feature". It's often assumed (incorrectly) that the more non-linear a game is, the better it is. I've already shown how this is a dangerous fallacy - leaving the player in a huge world without telling them what to do can lead to frustration, boredom, and the player giving up.

There ought to be a reason parts of the game is non-linear. If there are two paths that will both take me to the end of the level, and one is more difficult than another, why is it more difficult? There's a part in the Race to the Finish bonus stage in SSBM that splits into three paths. One path has high walls lined with spikes, one has pits of lava broken with platforms, and one is a series of moving platforms suspended above a pit. In this case the reason for the non-linear element is obvious - there are twenty-five characters, and some are better suited for certain paths than others.

Now let's talk about Donkey Kong Country 2. The game gives you two players to use - they are identical in every way save two: one character holds objects above her head and can fly and/or float, the other holds objects in front and can cartwheel. The problem here is that the character who can fly is always more advantageous than the character who cannot. The game presents many situations where it is helpful or even required to fly, but no situations where it is helpful or required to cartwheel (cartwheeling will kill a row of enemies, but you can just as easily fly over them). The player can pick the cartwheeling character, sure, but there's never any reason to do so if the flying one is available. The game has two characters just for the sake of having two.

You see this a lot in RPGs. Quests with no meaningful resolution, no reward, and no real reason for existing. In FF9 you can race against Hippaul at only one point in the game, and you don't get anything that affects the game in any meaningful way for doing so. Why do it? SaGa Frontier opens up almost all the locations for all the characters, even though many places don't have any events or quests in them for some of them. Why bother going there? These forms of illusionary nonlinearity are really only lost time.

Giving the player something extra to do just for the sake of having something extra to do is missing the entire point of making the game non-linear to begin with.

A list of stuff to consider for YOUR RPG!

I've discussed a lot of the ins and outs about what makes a game non-linear and why some stuff works and some stuff doesn't. Now we'll try to apply this to your RPG!

- Nonlinearity is cool on a manageable scale. Giving the player something to do other than following the game's core story is usually a good idea. However, sacrificing game flow or breaking up the plot to do it is not. Putting a goofy fun minigame in the middle of an average dungeon the player is in to look for the magic crystal flower of whatever is fun; putting the same minigame in the archvillain's hidden lair while the players are trying to rescue an ally is distracting. (Hey man, we would have been here sooner but we stopped to make sure all the fluffy critters down in the hidden room on floor five were well fed.)

- Here's a paradox for you - the more non-linear your game is, the easier you must make it to stay on track. If your player has only one option of what to do next, they're not likely to forget what it is. If they have four options, they might get sidetracked. If they have ten options, they might leave the flow of the game for hours-long stretches at a time! Be aware that your player might want to abandon their sidequest at any time and get on with the core game - make it easy for them to slip right back into it.

- You know, there's a reason all those false options I talked about earlier happen near the end of most commercial games. Accomplishing just a few of the false options put you right on par with the last boss; accomplishing all of them put you out of the ballpark. The more optional stuff you have during the game, the stronger your player has potential to be. Now you're in trouble - the player who did everything will be overpowered and might get bored with your easy snooze-fest. The player who did little might be underpowered and get frustrated with how difficult things are. Of course, this gets into the realm of game balance, which isn't within the scope of this article... but it's worth mentioning.

- If there's a branch, there ought to be fruit. Keep that in mind. Don't offer an option if there's no reason to take it! Players don't like going out of their way to be greeted with a dead-end. Oftentimes they'll think they missed something (even if they didn't) and spend more time exploring an optional area than they ought to, only adding to their frustration!

- The less linear the game is, the more difficult it is to keep everything in synch. If your game is designed in such a way that it is possible to reach Boss Monster X at level 5 with just one character, Boss Monster X ought to be smackdownable at level 5 with just one character. Keep track of how all the nonlinearly elements interact with each other, and understand that your players will want to try combinations you never dreamed of!

- There's no shame in keeping it simple. If you can't think of anything extra, bonus, optional, or secret to add... don't bother. Chances are the player won't know anything is missing as long as the linear stuff turns out good.

And that's the article.

I hope someone, somewhere finds this helpful in some way, shape, or form.