Common Sense vs Level Design
I often say that common sense is one of the most useful assets a designer has when it comes to designing levels (or indeed, anything), because out of everything that has helped me design my maps, common sense has proved the most reliable.
When I refer to common sense, I mean a very basic understanding relying on some obvious principles. In real life, common sense dictates that you wear a coat when it rains, lock doors before you leave your house and don't walk onto a road without checking for oncoming traffic first. In architecture and building design, common sense advises you to leave room for doors to open, ensure that ceilings are high enough for even the tallest person, and that the building has some form of entrance or exit (apparently buildings tend to need entrances). In level design, common sense gives you subtle advice, such as not to drop the player into instant-death lava as soon as they spawn, or not to lock them in a simple, boring room for three hours, or indeed, forever.
What intrigues me however is how often common sense is not used in the design process - I have seen many maps that appeared to be so promising, or alternatively undeniably awful, but when it came to the crunch really turned their fortunes around due to how well common sense was applied.
In this article I'll be covering some of those common sense principles, plus some other practices which have helped me design my maps when I've got really stuck.
Level design allows for two very easy paths - a very simple level, and a very elaborate, complex one. It's very easy to steer to one extreme without even realising it. Maybe the designer just has thousands of clever ideas and wants to add them all, or maybe they have none, or don't know where and how to apply them. Either way, the player will get respectively a) bored and aggravated, or b) lost, confused and aggravated. You can work out if a level is too simple or complex by simply getting someone to play it over a period of around 15 minutes (it really depends on the game).
If the map is too complex, then the hardest part of the process is accepting the fact and understanding you will have to delete some of your work. Of course, as the designer, you'll know the map inside out. You know the routes from A to B, which one is the quickest and the most useful. You can get to A to B via C if you want to. However, the players won't. You must realise that in order to simplify the map appropriately, you will have to delete or redesign some possibly large sections of the level. You might need to remove several hundred brushes and start again, even if you designed the area to perfection. It's a terrible waste, but sometimes the only cure.
When you don't have play testers to examine the map for you, then you have to work on common sense. Look at the map from above (like a plan view). Identify the 'main route' through the map - the 'spine' which players rely upon. You should notice that the map effectively has a major route from which minor routes emanate, often backing back onto the major route, or indeed another minor route. Once you've done that, examine the minor routes - and remove the ones which are unnecessary. Perhaps you have two routes right next to each other, both providing the same opportunities to the player and both leading to the same place. Chances are, they're only confusing the player - remove one.
Identifying these routes can be a nightmare. You may believe every single one has a unique use and you can't decide which one to remove - that's a common problem. Yet something has to be removed. In this case, it's often useful to identify how often the route will be used. If it's not directly connected to the major route of the map, remove it. If it is, but near an uninteresting or useless part of the map, remove it.
Example: 'Torn' (Half-Life: Counter-Strike)
Counter-Strike's 'Torn' is considered by many as a very attractive map - admittedly, it is. However, it is also quite a maze consisting of lots of small routes to and from major hubs of the map. While these add to the atmosphere, they do make it harder to learn (the tall looming buildings don't help the player orient themselves either!)
If even just a few of these routes were removed, and the remaining major connections made more obvious, I believe the map would be far more fun to learn and play.
In my experience, the most aggravating feature in a complex map is a long, thin, twisting hallway with no clear entrance/exit, especially if it branches into several more similar routes. Ideally, such a hallway should have a clear route from beginning to end, and take only a few seconds to traverse. Personally I like to be able to see the exit from the entrance, but that's a matter of taste. If your map requires long, laborious hallways, and you're sure they are absolutely necessary, then you must go ahead - just take care.
A problem with complex maps is that players find it hard to orient themselves - they might know where they are and where they want to be, but how are they to know which direction they're meant to be going? In some maps players subconsciously rely on sunlight or the sky, but some maps don't allow for that. Sometimes a clear path is needed, or at least a change in the environment that signals they're going to head in the correct direction, such as a descent.
Example: 'Avanti' (Half-Life: Team Fortress Classic)
In the TFC map 'Avanti', it is always clear which way you are meant to be going - uphill or downhill. The map starts with defenders on higher ground than the attackers, and hence each team knows which direction the enemy will come from.
One obvious way to give players a sense of direction is the use of signs - provided they are simple. A cool approach to try in large outdoor maps is to have one tall building/landmark visible over and above buildings which players can 'aim' for. Essentially, all a player needs is a landmark (just like a tourist would). If you're going to have tall, looming buildings everywhere, arrange them so that they accentuate routes and don't just act as obstacles.
- Only add routes that are needed - try not to duplicate routes unless absolutely necessary. Be concise.
- Always ask, 'why would someone want to go to this part of the map?'
- Provide the player with landmarks by which they can orient themselves and work out where they are/should be going.
- If the map is too complex, remove routes.
- Consider giving each major area of the map a specific purpose or theme.
One of the most important processes that I've used in my Counter-Strike maps stemmed straight from common-sense, and is also a method I do mention to other level designers when I feel they haven't quite caught on... timing.
Timing is most important in maps with team-oriented goals. In such games, maps commonly have territory for either team plus some neutral ground where you'd expect both teams to meet (this is especially true in Counter-Strike). Ideally you usually want both teams to meet in the neutral ground where neither has a clear advantage over the other side. One way to do this is to see how long it takes to reach the neutral ground when starting from each side, taking the shortest and most obvious route(s), by timing it - to the second. The numbers for each side should be as close to identical as possible - if they're not, then some adjustment is needed (alternatively maybe the game play requires one team to get there before the other - so use the same method to make sure they do get there in time!).
Example: 'Dust' (Half-Life: Counter-Strike)
In CS's 'Dust', the time (in seconds) taken to run from each spawn site to the central hallway and the underpass were checked and the spawn sites altered so that each team would meet at the major choke point of the map at the same time.
In a later version of map the spawn points of one team were moved forward without the same amount of checking... needless to say, they were later returned.
In deathmatch-style maps, the same method cannot be used due to the different game dynamics. However, generally it is advisable that a player spawning in a DM scenario need spend no more than a few seconds before they reach a weapon or power-up.
In single player maps, timing would refer to the time the player might have to react/act on a visual or audio cue, such as the floor beneath them rumbling. Unless the player isn't meant to react in time to revert their action, there should be ample time for them to move or perform some action in order to save themselves. Remember that dying from something unexpected is likely to annoy the player, especially if there was no warning or no chance for them to save themselves before-hand.
'Gimmicks' are often hailed as excellent ways to make multi player maps more interesting, and usually they are - you just have to consider the floating platforms and jump-pads from 'Q3DM17' or the air strikes in 'Crossfire' and 'Double Cross' for some fine examples. These are gimmicks which change the game flow, and once the player understands the mechanics, are still enjoyable no matter how often they are played.
However, some gimmicks are perhaps better labelled as 'surprises' - something happening unexpectedly to the player, or perhaps even triggered by another player which results in a situation the player cannot escape. An example might be a room which instantly kills everyone inside when a switch is pulled. While these are often fun the first few times, eventually they do get tired. Such instant-death traps are never popular to the victim, especially if they're unavoidable. Despite this, many maps (mostly from inexperienced designers) have mechanisms exactly like this.
Imagine if a popular map like for example, Counter-Strike's 'Aztec' were to have a button which would - without warning - electrify the water, instantly killing anyone in it. Players would eventually just avoid the water.
Now imagine if the button instead caused an air raid bombing on the map - again without warning - causing everyone outdoors to be killed instantly. It might be fun the first few times, but eventually it will grow exceedingly tired and the victims will grow similarly annoyed when someone uses it over and over again. The map would soon become unplayable without going outside and risking dying from a bomb - hence players would become as annoyed with the map as they are the people exploiting it.
Such traps are best left to areas where, by succeeding, the player is well-rewarded. However the player should be able to avoid the trap entirely and still have an enjoyable game. If a player could become a victim of a mechanism triggered by another player, a chance for them to escape will always be appreciated (think of Crossfire, Doublecross and Rock2). The best way to check is to play test.
Similarly, if a player is given such power, they should have reason to worry - for example, maybe they have to remain in a particularly vulnerable position in order to trigger the effect. Or perhaps should another player see that someone is about to trigger a trap, they should have ample opportunity to escape or stop them from triggering it at all.
In fast-paced games, players are inherently impatient. They don't like to have to wait around for things - it makes them feel vulnerable and detracts them from the pace. Doors opening slowly or getting in the way are particularly annoying. Make sure such delays are either a) minimal, or b) cleverly placed to make the environment more exciting (such as a slow transport to a high-powered weapon).
- Measure the time it takes for a player to spawn and reach the 'centre' of the map.
- Make sure each team has the time they need (if any) to set up defences, etc.
- Try not to keep the player waiting unnecessarily (unless it's in their favour).
- 'Surprises' don't last long in multi player maps...
Without lights, game maps would either be a) very bright, b) very dark and/or c) very uninteresting. Lighting not only makes maps look more believable, but provides more believable situations for the player, and more opportunities for them to interact with the environment, even if it is just hiding in shadow.
A common pitfall - especially in multi player maps - is the bad use of lighting, thereby hindering the players ability to negotiate the environment. One can imagine how difficult it is to negotiate a 2D game by trial and error alone, so a 3D game is in all likelihood going to be impossible to explore without at least some understanding of the surrounding architecture. That understanding is best provided by making it visible.
If we were to strip 3D games down to the basics, they are essentially mazes. At the very least, a player should be able to see where they are going and/or where they can go. Depending on the importance of a route, the route should be made more or less obvious. It's the sensible thing to do.
Example: 'Storm' (Half-Life: Counter-Strike)
In my eyes, a major flaw that Storm suffers from is the visibility of its routes. Not only are some areas extremely dark when entered, but from a distance, doors set into the architecture can be difficult to see, and even then it is unclear whether you can pass through them. It is incredibly annoying making the effort to reach what appears to be a passage only to discover otherwise.
Note: blocked routes and feedback
Even if the player tries to follow a route that turns out not to be a route at all, they should be forgiven for thinking they could take that route by the game giving some form of feedback as if it is saying, "nice try, but not quite...".
This is similar to the idea of a trap - the player runs up to an open door just to have it close in front of them (at which point they could be ambushed). The response from the game is what should make the player continue playing (even if it is as minor as a 'locked' sound - provided it is used sparingly.)
Contrast is a proven technique in aiding navigation. For example, the exits from the hallway in 'Dust' are clear because, outside, it is quite bright and the yellowness compares to the dull greys indoors. Similarly, patterns can also help draw attention towards areas, just like the yellow and black 'hazard stripes' to be found scattered around Half-Life (and indeed, real life).
In single player maps, contrast is especially useful through silhouettes. I believe it was American McGee who mentioned (several years ago) the enjoyment of placing a Shambler (big jumping creature with long, sharp arms) from Quake in the dark, set onto a light background, leading the player to have to work out from the outline alone what lay before them.
One would assume that by creating dark areas you encourage players to use items such as night-vision or similar, depending on the game. However, you should also consider how often players just notch their brightness/gamma right up on their monitor and/or graphics card and instantly have near-perfect vision in areas most players would consider pitch-black. You also need to think about how players without the gear would need to handle the situation - will they survive? Importantly - will they still want to play?
One fantastic property of darkness however is that it can be used to hide structure - in particular, structure which doesn't even exist or would be too much hassle to create. For example, a pit which when looked down leads to darkness is likely to be avoided by the player since it will look unfriendly - saving you the hassle of having to create anything grusome at the bottom (like spikes which, honestly, are getting old in the tooth). This method of hiding nonexistent geometry through darkness is particularly useful in cavern-style environments where it would be infeasible to map out absolutely every detail - so you don't. You make those areas dark and let the players imagination do it for you. Just ensure that the shadowed areas look like they're naturally in the dark.
- Use light to give players a direction to aim for or, alternatively, avoid.
- Ensure players can immediately see the major exits from an area as they enter it.
- Players don't like walking into walls simply because they couldn't see them.
- If an area is not visible then no one will use it.
- Darkness triggers imagination...
Common sense probably applies more to the geometric design of a map than anything else, especially when the map is intended to forge the real world (as is increasingly the case), and as such this is what is discussed here.
However, should your map be outlandishly odd - maybe it's based in some alien atmosphere or theme - plenty still applies. Even buildings created hundreds of years from now will follow some very basic concepts that we adhere to today and have adhered to for several hundred - if not thousands - of years.
This whole topic could actually spread a far greater area than I'm going to cover here - consider this a basic grounding and some fundamental ideas and suggestions.
It is often thought that appearance alone makes a map - which it clearly doesn't. Similarly, when designing a map visually, as much work should go into the design of the geometry as goes into the textures and shaders which wrap round them. It's a very basic premise, but the balance if often lost - maps either try to get the geometry perfect but use textures that are far from ideal, or use photo-realistic and perfect textures but apply them incorrectly or just badly to the geometry. A balance between the two is needed to achieve the best appearance.
This means that just because a staircase is shaped like a staircase doesn't mean it is a staircase - the shape needs to work with the textures and the surrounding architecture to convince the player it is indeed a staircase.
Just as importantly, once you've decided how your staircase looks - don't deviate from it. Consistency is essential - this applies to everything. It would be bad to have your map suddenly change style or theme for no apparent reason - this only serves to confuse the player and disengage them from the game.
No one needs to be told that the geometry of the map is important - creating the geometry is at the core of the mapping process. Yet it needs to be created properly and with appropriate thought and consideration. This refers to very basic - common sense - stuff, such as ensuring doors don't open into the player when they don't need to, and not creating impossible or ridiculous structures out-of-theme.
This is basically about making sure the geometry of the map is believable and does the jobs it's supposed to be doing. You have to think to yourself, "Now, would I and could I build this in real-life?" and act on your answer. Modern game engines let you create a lot of crazy structures which float and do many ridiculous things, but they don't mimic the behaviour of real buildings and the stresses they're put under. Everyone can glance at a building and decide whether it looks structurally sound or not, and you have to match these opinions in your map. It's all about convincing the player that the place you've created actually exists. If they don't believe that, they'll find it harder to engage in the game.
- Keep design consistent and don't deviate from a set theme and basic dimensions.
- Maps are more believable if they look AND act like their real-world counterparts.
- Consider how the architecture would be designed and constructed in real-life.
- Don't reinvent the wheel - the inspiration you need is all around you.
- Modern engines don't make an unrealistic map look real - you do.
Overall, common sense is perhaps one of the most useful tools a designer has. It is what players are most likely to use when playing the game and hence the kind of interaction you need to prepare for. It's a case of understanding what needs to be present and why it needs to be there.
Of course common sense alone cannot make anyone an expert - but it is a solid foundation from which to build upon.
It is important not to confuse these issues with correctness. Correctness is an attempt to make sure everything in the level could exist in real-life - such as the correct compounds for creating structures and ample support to prevent them from collapsing. It is usually far too extreme and way beyond the focus of general game design to examine such things in detail - although no thought at all is worse than too much. Common sense hence dictates that you adhere to simple, obvious rules without needing to fetch out a calculator and measure stress and metal fatigue...
Similarly, many of the ideas I've expressed can be ignored - after all, games vary wildly and what applies to one may or may not apply to another. Environments differ dramatically but also share basic characteristics. All that is required is some thought and consideration. Imagine the player is dumb - what is the first thing they'll try... and the second? What if they fail at both? Imagine (as a player) how you'd feel after failing, again and again.
Anyone could have written what I have - most of it stemmed from very basic ideas and rules which were then extended with examples. It's all quite obvious when presented, but less apparent otherwise.
Ultimately, this whole article is trying to motivate mappers to think about what they're creating and why they're creating it the way they are. It doesn't take a degree in architecture to create authentic environments, but it does require time and basic understanding. If you don't have those, then you should probably stick to something else.