|
Nonlinear
Exploration and Multiple Paths
Nonlinear, intertwining plots are some of my
very favorites, but in this article I’ll focus on nonlinear
navigation through a game world. Presenting the player with multiple
paths, rather than a railroad track course, is a great way to keep
him from getting stuck on something. If he can’t progress on one
path, he can go try another path and come back to the first one
later. Ironically, multiple path games are prone to some of the
worst kinds of “getting stuck” there are.
As always, let’s look at examples of games
that did things right (Mario64 and, Castlevania: Symphony of the
Night), a game that did things wrong (WarioLand 3).
Mario64
First let’s look the structure of Mario64’s
world, then the constraints on how a player must progress through
it.
Mario’s Castle is in effect “the map”
commonly used in platform games. It’s the gateway to all 15 levels
of the game, yet it’s a level in itself (though a very friendly,
non-threatening one). Each of the 15 levels has 6
missions (plus a seventh
secret mission to find 100 coins). The multiple missions per level
are a great way of reusing the level and making it feel full of
things to do.
Each mission rewards the player with a
“star.” 15 levels times 7 stars, plus 15 stars in Mario’s
Castle totals 120 stars. As I discuss in my article on hiding
secrets in platform games, Mario64 has one of two goals, depending
on the player. One player might wish to simply reach the end of the
game, beat the final boss, and “win.” A “gamer,” on the
other hand, will be satisfied with nothing less than collecting all
120 stars.
Mario64’s nonlinear structure is simple and
effective: the classic “string of pearls” design. Initially,
Mario’s Castle acts as a gateway to only a few levels. The player
has many choices (but not too many) and may play these few levels in
any order. Within a level, there’s also nonlinearity despite the
illusion to the contrary. Each level begins with mission 1 of 6
unlocked, and beating that mission opens up mission 2, and so forth.
Though ostensibly linear, for the vast majority of missions, the
player may wander around a level and find the star associated with
any mission whether it has been unlocked or not. So basically, the
player can solve the 6 missions in a level in any order.
From many options, there are eventually only
few. At first a level has 6 missions, but when 5 are solved,
there’s only one left, obviously. Only one path. At first, there
are several levels to play, but when all but one of them is
completed, there’s only one level left. Again, insultingly
obvious, but the point is progression of many choices funneling down
to one.
Before this funneling down to just one choice
occurs, the player has earned enough stars to open a door leading to
another set of levels. Suddenly, many new choices open up as the
player reaches the second “pearl” so to speak. Once again, these
many choices will become few once mostly solved, and another door
opening a final set of levels will be revealed. The third
“pearl.”
 |
|
Mario64 was the
prototype for the modern platform game's mission structure. |
Let’s examine the effects of this structure.
First, it’s great for the kind of player who is out to “win”
the game rather than discover all its secrets. This player is never
bothered with missions too hard for him since he can progress to the
next set of levels without completing every single mission in the
first set. Since the player can solve the missions of each set of
levels in virtually any order, the game designer does not even need
to know exactly which ones this type of player will find too
difficult. The player can skip whichever missions he can’t solve.
A forgiving design, both to the player and the designer.
This structure is also great for the player who
is after all 120 stars. Mario64’s structure may seem equivalent to
a linear structure in which all 120 stars must be acquired in order.
Far from it. There’s the little detail of gaining a few items
along the way (metal cap, winged cap, invisible cap) that make
possible the solving of previously impossible missions, but that’s
not even the point. Playing through a predefined order of 120
missions is a daunting and scary task. What if the player gets stuck
on mission #37? He has to keep playing #37 forever, never able to
see the rest of the game at all? This system makes the player want
to give up and stop playing if he ever encounters a significant
snag.
Mario64’s system, though, allows the player
to simply skip that mission and come back later. Maybe later he’ll
be stuck on it still. That’s ok. He can skip it again and try
again even later. The player has so many things to do that being
stuck on any one of them doesn’t make the game boring—he can
always move on to other things.
A final note about Mario64’s structure is
that getting stuck on any single mission does not violate my
Localized Puzzle Theorem: “The player enjoys knowing that
somewhere in this particular, small chunk of the game is something
he’s supposed to do. He enjoys knowing when he’s done it. That
same task buried in the vastness of a huge game-world is no longer a
fun task to solve, mainly because 99% of solving it become finding
where the heck the task is, and only 1% is actually solving it.”
Thankfully, Mario64 doesn’t suffer from this
problem, even though some other games allowing nonlinear navigation
do. In Mario64, when a player can’t solve mission 3 in world 5,
the task does not get lost in the vastness of the game-world. He can
very easily check his progress to see that he hasn’t solved
mission 3 of level 5, and he can also very easily go directly there
to solve it.
Castlevania: Symphony of the Night
(PlayStation)
This game, one of the very best on the
PlayStation, employs a totally different type of nonlinear
exploration than Mario64. While Mario64 is a collection of 15 levels
connected through Mario’s Castle, Castlevania: SOTN is just one
really, really large 2D, sidescrolling world. There are no
“levels” at all.
At first, the player is only able to access a
very small part of the world. There are no artificial boundaries
where a “level” ends, but there are a bunch of places the player
just can’t get to yet. There’s a really high cliff over there,
impassible spikes over there, water down there…if only you could
swim. As the player explores, he uncovers one-by-one what eventually
becomes a huge arsenal of special moves and powers. Now that he can
turn into a wolf he can make that long jump and access new parts of
the world that were previously inaccessible. Now that he can turn
into a bat he can fly to the top of a room and reach that ledge. Now
that he has the scuba gear he can swim through the underwater
catacombs, and so on and so on. (It’s the same formula as Metroid
(NES) and Super Metroid (SNES).)
 |
|
This is the map of
entire world of Castlevania: SOTN. Remind you of Metroid?
Hardcore gamers will appreciate the 210.1% of map completion
shown. |
This gives the feeling of very organic,
expanding world as the game progresses. As more and more of the
world is uncovered, traveling around becomes more difficult, which
is exactly why the game has several strategically placed
teleportation rooms. The player may teleport from any teleporter
he’s discovered to any teleporter
he’s discovered. Notice that this doesn’t help the player
get to new places at all, just to cross the old, already-traversed
places faster. Good design.
This type of non-linear exploration really
forces the player to keep on his toes and memorize the world. It’s
easy enough to remember where the water is once you get the scuba
gear, and that you should probably go there…but what about the
power of the bat? Do you really remember every ledge that was too
high to reach? This becomes the “skill” in this type of game.
Every time the player reaches some obstacle that he’s not able to
pass, he has to realize that it can’t be passed yet, speculate as
to what would allow him to pass it, then remember both the obstacle
and its location later, once he finally gets whichever item or power
is needed. And here we have run into the problem of my Localized
Puzzle Theorem….
The player now has the power of the bat. What
now? Well, somewhere in the vastness of the game-world is a ledge he
couldn’t reach before and he’s going to have to search and
search over territory he’s already been over to find it. A
“string of pearls” design is absolutely deadly in this kind of
game. Imagine if you needed to find the double jump in order to
reach the part of the world that contained the bat, the wolf, and
the scuba gear. Imagine that all your options funneled down to this
one task—finding the double jump ability—and that new tasks are
only opened up after you find it. What a nightmare! You’ll have to
roam through a huge world trying to find the one little, tiny thing
you missed, which leads to that double jump ability. You have no
idea where to look. Furthermore, you probably have no idea that
it’s even the double jump ability you need! The point is that this
type of game requires the designer to keep as many paths open to the
player as possible at all times. Anytime the player’s options
funnel down to only one, there is the chance of violating the
Localized Puzzle Theorem.
So how did Castlevania: Symphony of the Night
fare? Early in the game, I did encounter such a bottleneck. I had to
figure out that if I waited in the clock room, the blocks would
rearrange themselves such that I could climb through a hatch in the
ceiling. I did not realize this and was doomed to wander endlessly
having no idea what I was supposed to do. By chance, I finally
figured out what to do, and after that I never got stuck again.
After that, I felt like I always had at least a few paths open to
me. Perhaps it was due to my superior skills of memorizing every
obstacle and predicting which future power would be needed to pass
it. Haha.
In any case, this is a game of superior design
quality and I hope I have not done it too much injustice by viewing
it through the narrow design lens focusing only on its nonlinear
structure.
Nonlinearity Gone Wrong: WarioLand 3
(GameBoy)
WarioLand 3 consists of about 20 levels, each
with 4 missions. Unlike Mario64, these missions cannot be completed
in any old order. In fact, very much unlike Mario64, the player can
never solve all 4 missions in a row. There game is nonlinear in the
extreme.
For example, solving a mission on level 1 will
open up level 2. There’s no indication of whether another mission
on level 1 can be solved or not. In fact, no other missions can be
solved yet. The player will have to eventually give up on level 1 in
frustration and move on to level 2. Solving a mission there will
open up another level, but again, the player doesn’t know whether
he should move on or not. The player also gains more powers as the
game progresses, allowing him to solve missions that were previously
unsolvable. Now that the player can stop blocks below him, for
example, where should he go? Does it allow him to solve a mission on
level 3 now? On level 12? Who knows?
WarioLand 3 does try to help the player out,
barely. After a new power is gained, the map screen shows various
events (a tornado, a blinking light, etc.) on certain level
entrances. This signifies that the new power can be used to solve
missions on those levels. This is of little consolation to the
GameBoy player who plays a few minutes here and there, a very
reasonable play pattern on a portable system. Is the player really
going to remember that after getting power x that levels 4 and 5
blinked?
WarioLand 3’s nonlinearity ends up being so
confusing and unstructured that it constantly suffers from the
problems of the Localized Puzzle Theorem. Giving the player multiple
simultaneous paths is great thing to do, as long as you also take
steps to keep the player from getting lost in the sea of
possibilities.
Talk
back! Discuss this article in the forums.
|
