Part 1 of a series of tutorials I hope to do. I'm archiving them on a blog for reference, but the blog isn't complete enough to post. Nor am I sure if my tutorials are good enough.
This is the first tutorial in what I hope will be a series on concepts of level design. This current tutorial can be used in reference to all source games, but I hope to delve into several specificaly in time. This is not about how to make a map, or how to make a map work. It is about how to make a map feel good when played, and tricks you can use to make the player experience more while working less. Without further ado, let us begin.
Key concepts:
-Simplified Goals
-Level Flow
-Flexibility of play
-Staging and Events
Key concept 1 - Simplified Goals.
No matter what game you are mapping for, goals are the driving force in moving along for the player in an environment. There are two goals that must be established unless you are making a level based around puzzles (Research and Development is a good example of this) The simple goal to finish, and the complex goal to do whatever it takes to drive the story along . One of the many beauties of half-life 2 is that no matter what map you're playing on, these two goals were allways the same. Get away from the combine. Get out of city 17. Get through ravenholm. You can see where I'm going with this. It created an experience where an 11 year old (as I was when I picked up the game) can get to the end of the game without thinking. Especially in single player games, this is important. I myself often pick up a game with the intent only to kill zombies, or combine soldiers, yet somehow find myself making inexorable progress towards the end of the level.
Punctuating simplified progress with events such as defending an area from a zombie horde event or solving a puzzle do not (and should not) normally obstruct this goal unless it can be completed in a reasonable amount of time or the player is made aware well in advance. A good example of this is the event at the end of half life 2 episode 1, where you are told before you arrive at the train station that you will be expected to help in the escape efforts of the citizens. When the player arrives, it is very well explained and everyone of typical intelligence should have no problem grasping the concept and carrying it out. Puzzles should be restricted to rather simple concepts, preferably with a low-brow way of getting it done that may not be as effective as a more complicated method. Puzzles should not, on principal, take more than 5 minutes to solve unless your game is based around them.
Establishing the second goal, that of working the level in to the story, is something that must be concieved and created by the designer or designers. Left 4 dead uses short scripted sequences at the begining of each campaign to establish the already established fact that rescue lies at the end of the final level. The half life series relies scripted events and spoken word from characters you most likely know about. While the goal can have far reaching implications and very complex underlying motivation, it should still be something simple enough so that some one barely paying attention can figure it out. The very concept of the Left 4 Dead campaigns achieves this, while half life 2 only gets as complex as "throw bomb at creature. shoot bomb". It is also worth noting that in multiplayer maps, the player rarely requires any logical justification for their actions. In Team Fortress 2, the ui showing a neutral point and a path that leads to it is clear indication of the task at hand.
Key Concept 2- Level Flow
Any one single player map should take no more than 45 minutes to complete in game. The majority of players tend to lose interest after roughly this amount of time. Give and take is, of course, acceptable, but 45 minutes tends to be an industry standard.
The people at Valve used a very accurate term to relate the idea that people in games get tired of staying focused on fighting for too long. "Combat Fitigue" can make even best made games seem like a chore to play. Crysis suffered from this severely, giving you long periods of uninturrupted combat punctuated by only short periods of rest, with no other activity involved. A player shouldn't have to fight for more than 10 minutes straight under any conditions without rest, unless given an unusually long period before or after a pivotal event. Even a player experiencing a rush of joy or excitement from the experience can only be expected to remain in this state for 15 minutes or less in a single player game.
Combat under a normal situation should have some sort of build up before the main fight. Being thrust headlong into a room with 20 combine overwatch or an entire horde coming at the player leaves the him or her confused and disgruntled. During the most intense fights a player responds best after getting the chance to experience the area they have to work with before they fight with. There's a reason in every left 4 dead campaign that the place you start the finale is almost at the very end of the map (apart from the obvious one of effective use of space). When defending an area, allways allow the player a chance to adjust their situation to whatever they think is best. As you will read later, you allways want the player to feel like failing is their fault, even if they don't enjoy it.
If you were to look at a graph of eventfullness in a level, the line should allways hit the lowest point at several periods in your design. This is made easier in left 4 dead, where the director sees the survivors are getting fitigued and gives them a period with no zombies, horde, or tank. Even multiplayer maps follow this design. In KOTH_Viaduct, there is a deliberately designed space between the spawn room and a building spanning the width of the map that serves as a gateway area (gateways will be covered later, but in essence they are a portal that must be past through to continue on) that is rarely ever actually used for combat. It gives the players on both sides, no matter who is winning, a chance to rest and contemplate their next move.
Key Concept 3 - Flexibility of Play
Having one way to approach any situation will destroy the experience for a player. If they happen to be sub-par at the skills required for that situation, they are sentenced to death. Creating an environment with multiple staging points yeilds two advantages. 1) It allows the player to work to his or her strengths, and -what is most likely the most important gift a level designer can give to a player- 2) It makes the player feel like they have control over what they are doing. Giving players a chance to be clever makes sure they feel more involved, and less like they are simply being forced to obey your ideas.
In single player games, having off-the-beaten-path places that a player can go makes progress feel more variable and the world more real. This is especially so if getting to this place requires clever use of mechanics already present in the game. However, making an area that requires effort or time to get to and then forcing the player to go back hurts more than it helps. Places like these should come with a reward. If a player had to double the length they have to travel to get to the safe house, earning a gaurenteed health kit or pills makes it feel like less of a sacrafice, and more like a grand adventure. The reward doesn't even have to be something truly beneficial to the player, either. A nice view of a combine dropship crashing into a building, or a beautiful vista serves just as well for some players.
When designing an area that isn't necessary, a good rule of thumb is to make it look half as interesting as the correct path. For example, if there are two buildings you can go through in a Left 4 Dead map with only one being the right way, having the correct one lit by an overhead spotlight helps the player know what they are sopposed to be doing. Another idea would be to make the building vibrant, with colors that intentionaly contrast those around. Obtrustive structure works as well, like in the first level of No Mercy in Left 4 dead. The subway tunnel entrance is not only brightly lit by flames, but the only structure around that isn't a 2+ story building.
Key Concept 4 - Staging and Events
Staging and Events work well in conjunction, though they can be used seperately. Staging is simply setting up an area or group of areas so that they encourage the player to focus on key points. As a quick example, a light pointing at a box of ammo in a nightime map makes it pop out to the player so they know it is there to use. An event is anything that occurs in the world actively that causes a reaction in the player. This could be a puzzle, a fight, or a building explosion.
I mentioned staging in key concept 3 briefly, but it deserves some talk in its own right. Part of the problem of making a level for other people is that you have no control over what the player is doing. If you're not careful, you'll find that the player is staring the other way when that plane crash you modeled for two months straight happens. Regardless of wether you are using Trigger_looks or not, staging is a key point of making the player do what you want. The first of the many tools that are effective for staging we'll cover is contrast. The concept of contrast is simple: Make something important different from everything else. If you design a city block with red sided brick houses and wooden doors with black frames that can't be opened, a white door frame on a metal building is going to catch the player's attention. Light is another tool you can use here. It works best with pointing out important things with light in dark areas, but also works (to an extent) using dark areas in light environments. Contrast also works with detail. As a rough example, if you design a city block where most of the houses are a block with a wall texture with one house that uses the wall texture, but has detailed architecture you will have a building that draws attention to itself.
Another method of staging is something I like to call "visual laning". It involves having things happen at the end of a "lane" or an area that compresses the field of view. This could be a street, a hallway, or a tunnel. Laning, when feasable while mapping, is better than contrast alone to force the player to focus on something. In an environment with a narrow field of view, a person has a tendancy to ignore things on the "edges" (by that I mean whatever walls or borders that are creating the lane) because the eye assumes they are not of note. The player then devotes most of their attention to things that occur along the visible lane.
The final method I'll discuss is directional mapping. Directional mapping is a method of adressing the challenge of getting the player to look up, though it also works with down. It involves creating an environment, be it a ramp, ladder, stairs, or some other construct, that forces the player to move to a higher value on the Z axis. The higher the angle, the more you force the player to look up so they know where they are going. This works especially well if you have enemies on the higher angle, so the player can be absolutely sure that they NEED to look up to avoid death, or failure.
Now that you have properly directed the player's attention where you want it, the question of what is going to happen must be adressed. On second thought, I shouldn't say that. You most likely know what's going to happen, but how do you get it to happen in a way the player will find interesting or useful? Because of the high cost of events in computer power and mapping resources, it is a good idea to make the best use of these things as possible.
Now, on to events. There are two types of events: visual, and physical. A visual event is something like watching a helicopter crash, or seeing some combine getting killed by zomies. This type of event relies in the player being unable to effect the event itself, but does not limit their ability to cause it.. Physical events directly effect the player. Being picked out of building rubble, fighting some bad guys, or counterbalancing a seesaw are some physical events.
Visual events are effective for storytelling for two reasons 1) It is cheaper to have visual events with many effects, because you can control what the player sees and optomize what they don't see. 2) Visual events, when done right, cannot be messed up by a player who isn't cheating. The event will occur the way you want it regardless of what is going on, ensuring the player can't inadvertantly ruin the effect for themselves. In order for a visual event to be worth the effort, it has to be both rewarding and informative. Visual events work especially well if the player did something to cause it. It is the perfect way to show the far reaching implications of certain actions on a large scale without risking the player.
Physical events have to be much more carefully monitored. Any time the environment changes a designer must be absolutely sure that a player cannot exploit it, because you can be sure some one will attempt to. This type of event is also expensive. Ensure when making one that it is worth the resources spent, and risks that a player will get in the way. Combat events and puzzles will be covered later in the segments about their specific games.
[QUOTE=melonmonkey;17093172]
[B]Any one map should take no more than 45 minutes to complete. The majority of players tend to lose interest after roughly this amount of time. Give and take is, of course, acceptable, but 45 minutes tends to be an industry standard.[/B][/QUOTE]
You do mean singleplayer, Ingame right?
Nice stuff you got there melonmonkey.:buddy:
Keep up the good work
Yeah. I should probably mention that... But you know, anyone who can keep focus on a single multiplayer map for 45 minutes has far more patience than me.
In hammer, a good map should take 3 weeks plus to complete.
[QUOTE=laptopman;17093378]In hammer, a good map should take 3 weeks plus to complete.[/QUOTE]
Well that just seems like a logic thing <_< but I fixed it. Thanks.
Sorry, you need to Log In to post a reply to this thread.
