Introduction

The purpose of this document is to provide guidance and insight for designers who are creating or working on a multiplayer level. I will address such topics as Flow, Item Placement, Initial Design, Architecture, and Testing.

Although Capture the Flag and other team games are rarely addressed specifically throughout this document, because they are typically for a minimum of four players (two teams of two), with a higher number more often being the case (e.g. 4 on 4, 6 on 6). That being said, many of these guidelines will apply to those types of games as well. (The major new issue in a cooperative/team game is how the new goals will affect gameplay. For example, if capturing the flag and returning to your base is more important than killing your opponents, then a speed power-up may become more important than a better weapon. For another example, consider a location in the map that might be very difficult to hold in free-for-all play, but would become very easy to control for two teammates.)

There are many accepted design principles that apply to level design in general. These will not be discussed in depth in this document, and include such things as:

However, there are many aspects of the multiplayer experience that can be handled incorrectly if approached from a single-player point of view. This will often result in the production of lower quality, unbalanced, poorly planned levels that will provide a disappointing multiplayer experience. For emphasis: You cannot reliably design good multiplayer levels from a single player point of view.

Since overall level flow and item placement are two of the ways in which multiplayer level design differs most dramatically from single player level design, these two aspects will be mentioned first, then followed by more general design principles. Important: The following design guidelines in this document are general rules. As with all general rules, there are always exceptions and special cases. Sometimes good level designers can ignore some of these guidelines and produce excellent levels... but it's not the way to bet.

Flow

For purposes of this document, flow is defined as a combination of direction of movement, speed of movement, and pace of movement through a level. In a level with an extremely high degree of flow, a player will be able to move at a relatively consistent pace from any area of the level to another with a minimum of dramatic changes in direction and speed. In a level with poor flow, there will be starts and stops, awkward geometry to navigate, edges and corners to get stuck on, and many dead-end areas.

Multiple entrances and exits:

Ideally, any major area in a level should have at least two (and preferably at least three) ways in or out (e.g. a room might have two hallways leading into it, a ledge above it that the player can drop from, and hole in the floor that a player can jump into to get to another area (three ways in, three ways out). To explore the idea of multiple entrances and exits, and the resulting effects on gameplay a bit further, imagine a room like the one below. "W" is a powerful weapon, and "H" is a health kit, and the room has exits/entrances on the north and east walls, and is somewhat flat and unremarkable:

Diagram 1

It's relatively simple in this particular example for a player to "camp" the weapon by standing in the corner and keeping a sharp eye on both doorways, attacking any player who tries to enter, and using the health kit to counterbalance any damage he or she might have received. Keep in mind that this "camper" does not have quite the same advantage that he would in a typical first person shooter (i.e. It's much more difficult to be sneaky in a game where the other player can easily look at your part of the split screen to see where you are), but it's still a tactic that can have a significant effect on gameplay. Consider the change below:

Diagram 2

Another door has been added on the west, the health has been moved to the NW corner, and the weapon has been moved to the south wall. Three doorways across a much wider field of view are more difficult to watch than two, and if the player still tries to camp the weapon, he or she has to move back and forth a bit more to obtain the health. This also leaves the camper open to attack by two or more other players from radically different directions (and it's much more difficult to watch what two or three people are doing on the split screen than it is to watch what one person is doing). For another approach, consider the following:

Diagram 3

The weapon has been moved to a spot between the two doorways where any player who moves through that corner of the room can quickly grab it, and the health is now by itself in the SW corner. Above the health is a shaft from the room above--impossible to travel up, but easy to see down and jump down. This produces some interesting gameplay possibilities and tactics. If a player is camping the weapon, he can be attacked easily through either doorway, and will find it difficult to watch both doorways at once. Suppose he decides to wait in the far corner, picking up health if he needs it, and ambush a player trying to get the weapon? It may work once or twice, but when his opponent catches on, the player will be attacked from a position of relative safety above, or his opponent will simply pay more attention to the player's position via the split screen. Relatively simple changes in room design and item placement can produce much more complex and flexible gaming situations.

Note that teleporters, which instantly whisk a player from place to place, can serve to increase connectivity and flow within a level if the level geometry itself is uncooperative. However, teleporters are also easily abused by being used as a quick fix for substandard level design that shouldn't have seen the light of day in the first place.

Clipping Geometry

"Invisible boxes", "clip brushes", "see-through walls"--different terms for unseen geometry that aids the player in navigating through the level with minimal difficulty. Ideally, this aid should be of the subtle variety--anything that is too intrusive might distract the player from any immersion in the game world that has been created.

For example, if there is a slightly protruding arch in a hallway that players tend to get caught on when moving down the hallway, the designer could place an invisible box along the length of the hallway on both sides with the inner plane of the box flush with the inner edge of the arch.

What if the arch sides didn't protrude slightly, but instead stuck quite far into the hallway? A box that kept the player from getting anywhere near the wall would be an obvious and blatant "fix", but the designer could place a wedge-shaped brush on the near and far edges of both sides of the arch to gently force the player out and around the arch as they passed by.

Note:

The use of clipping geometry will differ somewhat depending on whether the game in question is first person or third person. Something that might work well, and feel relatively unobtrusive from a first person point of view, might be very obvious and clumsy when experienced from a third-person perspective (and vice versa). This is just one area where a great deal of playtesting and feedback is essential.

Dead ends:

In general, dead ends are a bad idea in any multiplayer level for a number of reasons:

  1. Dead ends promote poor flow. If a player has to stop or do a U-turn at the end of a dead end passage, then that area is somewhat awkward and clumsy.
  2. Dead ends are boring and/or frustrating. The player has to travel back through an area that he or she has just seen.
  3. Dead ends can easily result in "no-win" situations for a player. If he or she is trapped in a dead-end, there is no option for a tactical retreat.

However, although the preceding points are generally true, in specific situations dead-ends can be useful (e.g. a powerful weapon or item can be placed in a dangerous dead-end in order to properly balance the value of the item with the risk involved in obtaining it).

Summary:

Item Placement

Poor item placement can turn an otherwise solid multiplayer level into an unbalanced and irritating gaming environment and can interfere dramatically with flow. Excellent item placement can add much-needed spice to an otherwise forgettable level, and accentuate the architecture and environment that has been created.

The items in a multiplayer environment can be divided into four basic types:

  1. Offensive Items (e.g. weapons, ammo)

    There is generally a maximum amount of damage that a player is able to inflict in a given period of time in a given situation. Offensive items increase that amount.

  2. Defensive Items (e.g. health, armor)

    Defensive items increase the amount of damage that a player is able to endure, make that damage have less effect on the player, or allow the player to avoid those effects.

  3. Special/Other Items (e.g. binoculars, mine sweeper, jet pack)

    These are items that can somehow change the balance of the game in a way that isn't purely offensive or defensive (but could strengthen offense or defense for a player, depending on the player's particular situation).

  4. Team Items

    These are items that somehow affect game goals in cooperative play. The best-known example would be the flags in traditional two-team Capture the Flag. Flags are traditionally placed in two opposing bases that often have the same layout, geometry, and item placement (to more easily avoid giving one team a subtle advantage over the other).

    As a side note: As previously mentioned, this document does not focus on capture the flag (and similar games) to any great degree. However, one of the simplest ways to introduce a CTF-like element into a map for fewer players is to have some single power item or power spot on the map that a player gets points for holding or capturing.

Item Quantity and Placement

There is a fine balance to item quantity. There should be enough items to make it relatively easy to get the most basic necessities (e.g. basic weapons, some degree of health/protection), but not so many items that the challenge is eliminated and the player is stumbling over some new item every few steps.

Generally, there should be fewer of the more powerful items in a level. There would be no reason to pick up the weaker items if there was a better item nearby. The more rare and powerful items can also be placed in locations that are more difficult and/or more dangerous to reach.

There is nothing necessarily wrong with placing a powerful item in plain sight in the middle of the level where it is easily reachable by all the players. This placement in itself can add an element of danger as players wait nearby, simply watching the item, and attack other players as they approach. Just realize that much of the action will occur around that powerful item, and that there should be sufficient incentive for players to travel to other parts of the level.

Powerful items can also be used to "balance" the level. In other words, if there is a powerful item or weapon at a certain location within a level, a good designer will be likely to put a similarly powerful item or weapon in another area of the level. This accomplishes three things:

  1. It makes it less effective to try to "camp" either item.
  2. It encourages players to move gameplay around the level.
  3. It makes it harder for a player to continually have both items and more easily control the game.

Finally, some weapons can be placed in such a way that they are not only balanced to some extent, but also encourage more gameflow and movement through a level. One simple example would be to place a sniper rifle in an enclosed area in the depths of a level--in order to make the best use of it, a player would have to get the item, and then travel up to the top of a high tower to get the best vantage point from which to snipe at other players.

Ammunition and Minor Item Placement

The placement of ammunition (if it exists in the game separately from the weapons), and the placement of minor items can be a much more subtle process then the placement of powerful items, and can be approached in different ways. Furthermore, many of the fine points will be very dependent on specific game mechanics. For example, a game with differing levels of health (or healing potion, or whatever generic "more life" item it happens to have) can have a much more complicated and "fine-tuned" item layout than a game with only one type of healing item.

If a game has multiple weapon types and multiple ammo types (or even multiple ammo types for each weapon), this will result in more fine-tuning and more complicated decisions for the designer. A good general rule to remember is that if a player has everything he/she needs in one area, then there's little reason (gameplay-wise) to leave that area and explore the rest of the map.

Item Setting

It can add significant atmosphere and "feel" to a level if the items are placed in appropriate settings, and not just strewn about in relatively equidistant spots. One good (albeit subjective) rule of thumb: Every area of a level should be attractive enough for a player to want to visit it.

Creating a proper item setting is a much more subjective process than some of the ideas that have been mentioned previously, as it deals with artistry and aesthetics rather than easily quantifiable factors such as damage and movement. Items, especially powerful items, are best placed like a gemstone placed in a ring. Impressive and/or detailed geometry, eye-catching lighting, or even props and other items can all be combined to create a memorable setting for items.

Camping Revisited

As was touched on above, it is very easy to create a situation in a multiplayer level wherein a powerful item (or even a not-so-powerful item) is placed in such a way that it is very easy to defend once it's obtained, and a player can "camp-out" at that location and dominate others who attack that position or try to get that item.

For example, a machine-gun with a large supply of ammo and a health kit are placed at the end of a long corridor, behind a pillbox with a small "gunner slot" to shoot through. A player can stay there for a long time racking up victories with relative ease. While this example is an extreme one for illustrative purposes, it is easy to make this mistake in more subtle ways. This mistake becomes less likely if the designer uses the "at least two ways out" guideline, and incorporates some sort of vulnerability into every major item placement.

Item Placement and Player Start Locations

There seem to be two schools of thought on placement of player start locations relative to weapons and items, the first being: "Players should have to work to get good items/weapons. Gameplay becomes boring when players always have access to all the good items immediately upon starting a level." The opposing point of view goes something like: "When it's a difficult process to get good items and weapons, then the player who wins any particular skirmish always has the advantage, since he/she already has all the good stuff, and the defeated player has to restart, recollect items, and possibly fight off a beefed-up opponent while doing so."

There is no clear answer or definitive formula to resolving this issue. Both points have some validity, and it will usually be safest to try to place your player spawn points while keeping both these points in mind. This is an issue that is usually resolved best with a great deal of playtesting.

Ideally, player start locations should be placed with the following additional things in mind:

  1. Player starts should not be in a direct line of sight with each other. If they are, this potentially eliminates a major part of a good multiplayer game: maneuvering and responding based on where your opponent is (or where you think he is), and reacting to his movements with appropriate strategy or tactics.
  2. Player starts should be placed in places that are "off the beaten path" to some extent. It can put a player at an unfair disadvantage if he/she appears in the middle of a central combat area in the level, and can be frustrating if he/she is immediately defeated before gaining any real momentum.
  3. There should always be at least two nearby exits from any player start location. A player spawning into the game in a no-win situation (because a beefed-up opponent has them trapped in a dead end) is simply a result of poor level design.

Secrets

Finally, placing items in "secret" locations is generally a bad idea in multiplayer levels, since there will often be one or more players who don't know how to obtain the item (bad enough), but may be unaware that it even exists (worse).

This sets up a dynamic wherein one player can easily dominate another player or players, only because of the "insider" knowledge that he or she possesses, and results in a blatantly unfair situation which can frustrate and anger players. (Note that I am not referring to items that are simply very difficult to obtain. If everyone knows where it is, it isn't a secret.)

Again, although the preceding is generally true, there are some ways to make secrets work in a limited way in multiplayer games:

  1. The secret shouldn't be a "game-winner". A secret that gives someone an overwhelming advantage in a game = bad idea. A secret that helps a player slightly, or that simply gives some background color, or information of some kind about the game world = good idea.
  2. Secrets that are a relative "one-shot" (i.e. once the secret is discovered, pretty much all the players will know about it) are much less unbalancing.
  3. Secrets that have a random factor can work. These can be fun without being too unbalancing. For example, suppose there's a somewhat out-of-the-way spot where a powerful weapon will appear 5% of the time instead of the regular health that appears the rest of the time. Further suppose that there is no additional ammo for the weapon, and that there is no other weapon of this type in the level. This results in a player randomly finding this weapon on rare occasion and only using it for a very short time (thus being likely to establish no serious advantage). In a situation like this, "insider" information can be fun and can produce some interesting gameplay situations (as players begin to shadow the other player trying to find out where the "odd" weapon came from).

Summary:

Initial Design

The initial design process can be a dramatically different one for different designers. Some individuals greatly enjoy it, because it allows them to visualize the level in broad strokes and come up with various ideas without necessarily needing to address some of the more "tedious" or "exacting" details that will appear near the end of the construction process. Other designers struggle to come up with a new and creative idea, or a broad outline, but excel in providing the fine points of a level's look and feel.

Some level architects plan out their levels in exacting detail on grid paper beforehand, or work from detailed concept sketches, while others simply start from scratch, allowing ideas to evolve as they work. Both approaches have their pros and cons: A high level of preplanning assures that the designer won't wander off down the wrong track and possibly waste a great deal of time and energy, but can also stifle creativity and force a designer into "mental blinders" that reduce his or her potential.

Summary:

General Testing and Game Mechanics

Testing is at least as important in multiplayer levels as it is in single player levels, and some would say that it's more important because the actions of a group of players are more unpredictable than the actions of a single player. While multiplayer levels are simpler in some ways than single player levels, players in a multiplayer setting can try new things (and find new problems) that might not have occurred if they were playing alone. Some design questions become exponentially more complicated when designing levels with a multiplayer focus. A few of the basic points to consider in testing and game mechanics:

Testing Start Locations

If you are one person testing a multiplayer level, it's easy to overlook non-functional or flawed start locations, especially if the start location is not always randomized, and if you do not have any sort of artificial opponents. Always make sure all start locations work consistently and correctly.

Gameplay habits

We all have a tendency to do things in a certain way, and repeat habits. The only way to be sure that the gameplay in a level isn't broken in some major way is to have the level playtested by someone (preferably many players) other than the designer. That being understood, you can at least playtest better as a designer by doing everything you can to break up your habits--if you find yourself always following a particular path in a level, then consciously go another way. Pretend that you haven't memorized every nook and cranny, and try to play like a new player: "Gee, I wonder what's over here..." Try to look at your map with new eyes, and you will often find problems or possibilities that you didn't realize were there.

Gameplay mechanics

Be aware that all games--even all multiplayer combat games--have different (sometimes radically different) gameplay mechanics. A couple of notable examples:

Camera Angle

Lines of sight are as important in multiplayer level design as they are in single player level design. Being able to see an enemy, or be seen by an enemy, is a key factor to victory. When playing from a third-person perspective (again, depending to some extent on camera movement) it's relatively easy to see where players are in relation to one another, and, if you are in a relatively high position, to potentially get a bird's-eye perspective on the entire field of play. In addition, when you are in a low position in a third-person game, it can be quite difficult to see what is above you when compared with a first-person game. When playing a first-person game, your field of view is limited horizontally to approximately 90 degrees, so losing track of your opponent can happen in the literal blink of an eye.

In a first-person game, it's pretty much impossible to see anything that your in-game character wouldn't be able to see (i.e. you see through the character's eyes). In a third-person game, the circumstances involved in having a third-person camera view that doesn't necessarily change consistently with player movement can result in a variety of unusual possibilities at any given moment:

  1. If it's a console game, no players can see the others directly, but everyone still knows where all the players are by looking at split screens.
  2. No players can see the others directly, but players can shoot the other players (e.g. with a weapon like the grenade launcher or mortar that can shoot in an ascending-descending arc).
  3. Players can see each other directly, but players can't shoot each other. (This can happen if the players are positioned in such a way that the camera sees around a corner or over an obstacle for each player.)
  4. One player can see and shoot at another player without it being possible for the other player to see him directly, or hit him with return fire. (This could happen if, for instance, a player had a high vantage point, and was behind an obstacle of some sort. Direct fire from the other player would hit the obstacle, and arcing fire would either overshoot or collide with other geometry.)

Line of sight issues are further complicated by the fact that any player might be able to kneel or drop prone at any time, which could change any of the above situations. It is also possible in some third-person games for a player to change the camera angle without actually moving (by rotating the camera in place). This makes it possible to stand facing in one direction, but keep a 360-degree watch. This would, of course, only provide an advantage in certain situations, and if you have the control skills to make it useful.

Also, consider the effect of ground cover on combat. In a first-person game, the heavy use of ground cover (e.g. bushes, low walls, obstacles) can easily obscure the field of play and add a hide-and-sneak element, emphasizing the importance of accurate prediction of an opponent's tactics. In a third person game with any sort of height to the camera angle, this sort of ground cover is more of a simple obstruction to movement than a serious influence on tactics and strategy. These issues become very important to a level designer when questions about gameplay, balance, and tactics arise.

Auto-Aiming

For another example, consider the subject of auto-aiming. Auto-aiming (when the computer/game system does some of the work of aiming for the player), gives a very different feel to a skirmish, and a player must concentrate more on positioning, movement, and any other ways in which he or she can help the auto-aiming system along, and less on accurate crosshair positioning and shot-timing.

Summary:

Research

To make good multiplayer levels, and to continue to grow in his or her skills, a designer needs to play lots of good multiplayer levels, and, unfortunately, the only way to play lots of good levels is to wade through even more levels that aren't so good. While you're playing a particular level, analyze what is working in that particular level--very simply, what makes that level good and not bad? If you don't know quite specific answers to that question, then you may not be able to create the same great gameplay and fun experience in your levels, and if you do, it may be by accident rather than by design.

Finally, write things down. That may sound obvious and slightly juvenile, but you will not remember important things if you don't. I have a simple text file called "tips" that I just copy and paste tidbits of various kinds into--technical tips, design tips, interesting gaming anecdotes, "here's a great idea for a level" bits, obscure design facts, and so forth. If you have any doubt if you should put something in, go ahead and put it in, then review the file periodically and weed out things that are outdated, have lost their usefulness, or were just never quite as useful as you thought they might be.

Summary:

Links