| HB's Multiplayer Design Treatise | W A V E L E N G T H |
|
| by HellBringer | February 28, 1998 |
|
|
This article covers: | |
|
|
|
|
|
|
|
|
|
Welcome to Article #5 - HB's Treatise on general Multiplayer map design. This article is a direct result of many of the design-related questions I've been asked in email; as well as observing weak points in maps that have been reviewed over at Radium (www.planethalflife.com/radium). I'm not trying to play favorites by mentioning only them; they are simply the most thorough review site for Half-Life maps that I know of. If you know of other map review sites; PLEASE feel free to let me know about them! Also, I mention my latest DM map, "HBDM1" several times here. Once again, this is because I made the map and know all of its details; as well as how the design process went. It is not because I'm trying to hype my stuff over other people's.
Foreward: Genesis
In the beginning all was darkness, and void. And then the Professional designers cried out "Let there be GAMES!". And there were. And most of the time, they were good! And the morning and the evening of this period were the first many years. Then, the great designers cleaved the space, and separated the sky from the ground. They spread objects out - and these directions lent the objects volume; and thus the 3rd dimension was employed! And the designers saw the sales figures, and knew that it was good! And the morning and the evening of this period were the mid-90's.
It is during this tumultuous time that a few trepidatious souls clamored awkwardly (in their struggle for knowledge) up the high pedestals that they themselves had helped construct for the great masters of design. Upon reaching the heights, a brave few begged of the designers for code, specs, and other small droppings of information. These folk descended from the heights enlightened, and holding several bits of information. There was much rejoicing; and vast temples were created to enshrine and spread the word that these things taught. Great tools and fanatical followings arose; and the designers watched from on high, astonished at how much momentum their creations could carry when the users and common-folk got involved.
Nowadays, things are different. No longer must we trek up the treacherous slopes; for we have the masters at Valve coming down to us from on high, doling out information freely. However, all of the technical help that they can provide to the general public STILL does not make for a graceful creation. While I am not going to part any waters, I will state these 10 basic multi-player design commandments, and then explain each in depth:
First Commandment: Quality control
While the rest of the commandments somewhat follow the design process, this commandment was placed above all; as it is THE MOST IMPORTANT one. Why, you ask? Quite simply it's a matter of reputation. If you put out shoddy maps, people will tend to ignore your later works - or at least not give them a fair, un-biased chance. Also, there is quite simply NO POINT in releasing maps of poor quality. If they're Single-Player maps, word will spread and people will either not download them or not finish playing them. If they're Deathmatch maps, servers won't run them. There are lots of bad maps out there already. If you make a map and its not all that great (especially if its your first or second map); don't despair - simply start working on a new, better one! If at some point in the future you figure out a way to make it INTO a good map, you can; if you have already released it, its too late to do that. There's NOTHING wrong with not releasing; and going ahead and releasing poor quality work has NO benefits. You still gain experience from making the map - and if you want feedback; toss it to a few friends to play around with.
Commandment 2: Planning ahead
There seems to be a lot of maps these days that are quite loosely put together. One of the major reasons for poorly laid-out maps is a lack of planning. It is all too easy to get an idea for a "gimmick", special item, trick/trap, or basic theme; and jump into making the whole map at once. This is NOT the way to make good maps! If you do this, it is inevitable that at some point you will run out of ideas; and have an only partially-completed or connected map on your hands. Then you'll have to "toss in" spare rooms and hallways to make everything complete. This leads to maps that are inconsistent and don't "hold themselves together". You ALWAYS want to plan out your maps before-hand in as complete a manner as necessary. Draw on paper, sketch in MS PAINT; do whatever works for YOU that helps you visualize and plan how it's all going to work together. Don't EVER trust your memory, either!! I can personally say that I've lost a great many ideas by having them shortly before falling asleep, and being lazy and not jotting them down. Keep a sketch pad handy, and if you think of something level-related, write a couple of sentences about it... Or draw a little schematic. Do SOMETHING that will trigger those same neurons in your brain later! As for your main level sketch, it doesn't have to be detailed - but it SHOULD give you an idea of how all of the rooms are going to be roughly shaped and connect to each other. For example, here's my rough sketch of "HBDM1":
And here's the Top-down (XY) view of "HBDM1" in its final form (minus the tramway):
Note how they are different; but similar. Its alright for designs to change as they evolve (that "rough sketch" actually changed quite a bit two times before settling on that version) - but having a starting point and an idea of how you want things to be shaped gives you both a clearer idea of what to do in WC, as well as a jump start on getting everything themed and working.
At this point, I'd like to put in one cautionary note; handed down from one of the "masters" of the art: Levelord. His advice regarding paper planning is that paper is 2 dimensional; and maps that are overly-planned in 2 dimensions end up being 2d in execution (i.e. flat, uninteresting, and without multiple vertical levels/floors). This tends to be quite true; so you must strike a balance.
And this is also the perfect time to point out that most under-utilized of all map's aspects: the 3rd (vertical) dimension. BE AWARE of this dimension; and if your map idea can take advantage of it, use it as best you can! Have areas cross over one another, have floor-height changes via ramps and stairs... anything to vary up the vertical sense and orientation, and keep the map from feeling "flat". This is one of the tougher things to do when it comes to overall design; so don't be discouraged - just be aware that its there, and try to make use of it if you can.
Lastly, its best to let designs "simmer" a little. Sketch things out, think about commandment #3, and get some sleep. I know this is tough; especially when you hit on a GREAT idea (see my earlier comment about writing things down) - but it has been my experience that if you take a couple of days to develop the idea BEFORE leaping into WC, you will end up with a more mature design that is more complete; and therefore easier and faster to make. (Note: prototyping a room idea; or testing a couple of hollowed out rooms to check the flow of a map idea during this time is okay - I'm referring to digging in and actually making the "real" map as not necessarily being the best thing)
Third Commandment: Theme
This one goes hand in hand with Commandment #2; and should never be ignored. ANY decent map will have a "feel" and a theme to it. It helps hold the map together and make the user feel as if they are actually SOMEWHERE when playing; instead of running around a bunch of brushes that have been pushed together. You should have an idea for your map of what you want it to look like. Take that and figure out how you're going to keep that "look" as the player moves between different areas. What kind of textures will you use for each room/area? What kind of lighting? How will you place prefabs and small details to preserve this theme? These are all questions you should think about for each major room or area of your map. You don't need super-specific answers for these at this point (it would be good enough to say "Grayish or metallic dirty or worn textures, with some kind of metal or stone tile floor" as an answer to the texture question... you don't have to know that you'll be using texture "-0E44gray9". For the object question, think things like: "Should I use shelves, crates, boxes, computers, or pipes here?"). As you get more experienced and more familiar with the textures/lights you have available for a particular game, these things will become more automatic and natural. But they still bear some thought; and if you are a relatively new mapper, you should spend a fair bit of time thinking about this.
Commandment 4: Flow and Connectivity
This goes hand in hand with commandment #2 as well - a good designer will know how to integrate these into their initial design; but this takes some time and experience. I get a lot of questions about what "connectivity" (a term applied to maps mostly by the Radium guys) actually is. Simply put, connectivity is how well all parts of the areas connect with one another. Its closely tied to game flow; and both are of HIGH importance in any map! Don't be afraid to chop out or re-design areas of your map if they don't connect well. The PRIME thing to remember is that people want to have fun playing; and they want to run around and shoot weapons and kill people - but they want to have fair chances of killing AND being killed! Maps that allow one-sided play may be fun for whoever's in first place - but everyone else will be frustrated and hate it. This is especially true when making a map with traps or a "gimmick". Its FAR too easy to think "I'll make this tough/neat little thing, and if the player figures it all out, they will be able to dominate; yeah!". This is VERY VERY bad!! Because first off, it favors one-sided play; and secondly, anyone who sees or hears how to accomplish this trick will know how to do it. Coupled with all the chatter across the 'net, your "secret" won't stay that way for long, and soon the map will be nothing more of a test to see who can activate your trick/gimmick first. If you design your maps to favor even play amongst all people; then you will allow the PLAYERS to determine who comes in first by their skill level - which is what they want.
The EASIEST (and often BEST) way to improve connectivity and game-flow when making a good deathmatch map, is to ALWAYS provide multiple exits from every room. Two or three is enough; but if you can squeeze in an extra exit someplace without sacrificing speed or theme, do it! This makes it easier to shake-off someone who's firing on you, as well as keep one person from dominating any one area. It ALSO makes your job easier when it comes time for item placement; as you can count on people being able to get to various weapons a little bit easier. Still, keep in mind that the more direct routes to areas will be the most-often used ones. And NEVER EVER put dead-ends into your map unless they are a small niche in a wall or under a stairwell, where people are likely to just duck in and duck out (to pickup an ammo clip or a grenade or something). Even powerful weapons or items are likely to be ignored or passed-over if they are placed in dead-end areas, because they will favor camping to either protect the item or stop others from getting it. Either way, it ruins the fun and effectively takes the weapon out of circulation. See the sketches above for "HBDM1", and notice how every room has 2 or more exits? (barring the alcoves down in the tram-way, but those are deliberate)... Coupled with the item/ammo placement, this was built in to encourage fast and smooth gameplay; and to enhance accessibility and connectivity between all parts of the map.
Another thing about "flow" - what other words come to your mind when you think of what "flow" might mean? For me, I think "fluid, smooth, quick (but not frantic), and slick". Brain storm on your own and see what you can come up with - because this is what the gameplay should really do in a well-designed map... flowww. Deaths and respawns should be at a fairly regular rate (if the map isn't under or over-populated), no one should last for a really really long time, and no one person should constantly die every 2-5 seconds. Also, weapons should circulate between players. People should be able to travel around the whole map fairly well; and get to different items fairly quickly (so that no matter where they respawn from, they are pretty much guaranteed to pick up a weapon before being killed or hurt too much). All of this takes time and experience to develop a good "feel" for it - but there's no substitute for thinking things through until you believe something will work well. Then try things out, go back and re-work them, and repeat the process until things flowww. Once again, don't be afraid to change your map if it needs it! Its YOUR map - sure its not easy or fun to lose work and effort that's been put into the map so far; but YOU control the map, don't let IT control YOU!
Fifth Commandment: Item Placement
Okay, so you've got a decent design, all the areas of the map are accessible - but you're just starting to hit the really difficult part of mapping. In fact, item placement should be one of the LAST things you do (even after detailing and texture alignment); and should continue to be adjusted throughout the playtesting phase. Unfortunately, there are few hard and fast rules about specific item placement; but there ARE some guidelines to follow:
Commandment Six: Speed
This commandment is so sacred and vital, that I am constructing an entire document about it. I will post a chart here, however, of acceptable r_speeds; and say that without a doubt, you should ALWAYS keep speed in mind when designing a level- and remember that if you have a tricked-out system, only a small percentage of people will be able to run things as quickly as you! A lot of people are still running with CPUs in the 200-250mhz range; so don't be overly complex when you don't need to! Also, in a multiplayer game people are going to be concentrating on FIGHTING - not gawking at beautiful scenery. In fact, if your scenery slows the game down or causes the player to get caught on walls or other decorations; they will condemn the map, rather than praise it for its complex or detailed areas!
The following is a list of things that contribute to slow/jerky play:
Remember that anything that changes/moves is something extra that the computer has to calculate for; and in 'net games, that means increased lag as well as possibly lower frame-rates! Also, remember from my lighting tutorial (Tutorial #6) that the game engine "sees through" entities like doors; so if you have two complex areas separated only by a door or short corridor, you can expect there to be very high poly counts! Walls, 90-degree turns and elevation changes (ramps or stairs) are the BEST ways to block VIS. And again, these need to be taken into consideration NOT just from a standpoint of "How is it when I initially jump into the level?", but also from an angle of "How is this going to play with lots of people in it - after they've scorched the walls with lots of decals, there are dead bodies and items laying around adding to entity counts, and lots of combat is going on?"
Lastly, here's that R_Speeds chart (Remember that high entity-polygon counts may lower these recommendations!):
|
|
|
|
|
|
800 should be the upper limits of any reasonable SP map. No monsters and little action should occur in an area with counts this high. |
|
|
|
Areas of slight to moderate action with not-too-complex monsters are acceptable here, 500 is pretty comfortable in SP mode if you aren't going to be having any big firefights. | 500 is acceptable; but 600 is pushing it. You should never go above 600 if you expect normal machines to play this map - especially if they might use a modem connection. |
|
|
This is good for areas where some decent combat or a few large creatures may appear. | This is the ideal range for areas where some combat is likely to occur; remember that a few players exchanging fire can REALLY increase the entity poly counts quickly, though - so still exercise caution. |
|
|
Great for boss areas or large firefights/complex entity areas. | Good for spaces where lots of combat is expected. It'll take a fairly large firefight to slow things down if your count is this low. |
Seventh Commandment: Being thorough
This one is so obvious; but so often overlooked. Before you begin testing or start really delving into item-placement, you should run around in the map and check on everything. Do things flow okay? Are any areas really tough to get to that shouldn't be?? Do room-exits take precise movement to go through?? Remember that in deathmatch, speed is key - people won't want to be getting caught up on doorways or around corners. Places where people might get caught up on wall-details or corners should be "smoothed over" with clip brushes (I'll put up a tutorial on this). Lastly, run through the map 2 or 3 times VERY SLOWLY, and examine every part of the architecture. Make sure all of your textures are aligned and that there are no gaps or errors... This needs to be done with a fine-toothed comb; and can make or break a map; as ugly maps even with great gameplay may be too much of an eyesore for people to enjoy playing- and thus they won't be put up on servers.
In this category goes LIGHTING. Lighting is CRUCIAL! First off, people need to be able to see in any place that they might have to do battle. Secondly, people DON'T like too many dark spots for campers/snipers - especially if its a semi-protected area that is easily defended! Third, Lights should NOT be colored brightly/harshly!! A light value of "255 0 0" for solid red light in a room will be sickening in the game! Use colors that are "blended", and for goodness sakes, DON'T mix gross colors like "Lime Green" with "Gumball Purple". Poor lighting will RUIN an otherwise decent map. Remember back to commandment 3, establishing a theme??? Lighting is part of the equation, too; and is extremely important in setting the atmosphere of a location. Yellowish/Whitish pale lights should be the "base" for your lighting scheme in almost all cases; and think of colored lights as a way to enhance the atmosphere or "tinge" an area a certain shade. Colored lights should not be THE lighting for the area; and make your lighting match the area appropriately- office spaces are likely to have whitish or yellowish lights from incandescents and flourescents. Same with warehouse areas. Military or research areas may have slightly colored lights in some areas to lend a "sci-fi" feeling to them, but remember that if people would be expected to be in an area (if it was a real-life location), the lighting should be good enough to let them work without giving them eyestrain!! Alien locations can have odd-ball colors; as long as they aren't too "in-your-face". Having warning lights or "detail lights" that are colored is perfectly fine; once again, as long as they aren't TOO saturated or over-used. One thing that helps this, is to use texture lighting instead of entity/point-source lighting - see Tutorial #6 for more information on this.
Commandment Eight: Testing
Far too many bad designs could be avoided if the author just STOPPED for a little while, once they were close to completion; and played the map themselves. Also of great importance, is having a couple of friends try out the map. If you make the map, you will be PAINFULLY aware of everything about the map; and thus will NOT be a good judge of how things work. You need people to be dropped into the map and have NO idea what it is or where they are in the map; and see how they do. If you can get someone to set up a server and have a couple of friends deathmatch (If you can, participate in SOME, but not all of the deatmatching), you are MUCH better off. It will be an INVALUABLE tool for gauging weapon/item placement and game-flow. Make SURE you twist your friends' arms until they give you DETAILED feedback about everything they like and didn't like about the map. Remember that you may have some things you REALLY don't want to change, but if the complaint is serious or is repeated by more than one person; then you should do whatever it takes to correct the deficiency - your map will be MUCH better off in the long run because of it.
Things to inquire about for feedback include: Item placement, unused/unpopular areas, lag/low frame-rates, weapon dominance, camping/sniping, difficult or circuitous routes (some people don't take the "hints" you may drop in your map, and thus might not ever use the best methods to get where they want to go; its your job to make sure that the hints are obvious enough so that people can take full advantage of your designs), good/bad "gimmicks", lighting, and ANYTHING else you can think of - even down to texture choice or transitions between texture types. You never HAVE to impliment people's suggestions, but you probably should most of the time; unless its something that you WILL NOT bend on, or it is just a personal taste that one person has and is not echoed by all of the testers.
One important note: If there is something that everyone is complaining about; but is something you think is crucial to the design, or something that you're unwilling to bend on (like a gameplay issue or something you feel that they aren't "getting" how to do right); you STILL need to go back and re-engineer whatever it is that they're unhappy with. You should NEVER say "well my testers just don't get it"... because it is likely that the PUBLIC won't get it either... and your map will flop. Remember, your map has to be playable by people seeing it for the first time; and with NO CLUE as to what's in the map or what you had in mind when you made it. Putting in extra time at this step and even re-doing large chunks of your map will be worth it in the end. For all of these reasons, it should now be clear to you why getting an un-biased, independant report from others is so important.
Ninth Commandment: Evaluation
You would think that this would have been covered in Commandments 7 and 8, right? Well, I split this up into 3 commandments for a very specific reason: You should NEVER just do some cleaning, some testing, and then push the map out the door! ALWAYS repeat the process several times, and make EXTRA sure that any issues are taken care of - once you release the map, its too late to fix anything! So follow commandments 7 and 8, and repeat them a few times. THEN, sit down with your map and give it ANOTHER once-over. Ask yourself if the map is truly worth releasing. Do you think servers might run it? Will people have FUN with the map? Don't worry if its the best map of all time; just make sure that it is complete, clean, and plays well. Then you must decide on one of four courses of action:
Commandment Ten: Post-release
So you went ahead and released your map to the public, now what? Some people jump right into another design. In my opinion, this is WRONG! You should ALWAYS take a short break if you can, get away from your level editor for a couple of days (look, there's that girlfriend / boyfriend / wife / husband / kid / pet / job / school that you forgot you had!!). Let the aftermath "simmer" a bit, just like when you first designed the map. Often times, this reflection will let you not ONLY think of better things for next time; but sometimes also let you get over your ego on a few things that you may have been unwilling to change. Its too easy to get defensive or personally attached to some creation, and NOT do the "right thing" when playtesters tell you to change it; this time lets you settle a bit and realize not to make any of that mistake again. Also, it gives sites a chance to review your map; and you should DEFINITELY read ALL reviews of your map. Take what they say to heart - even take notes if you can! There will undoubtedly be things that are pointed out that you may not have been aware of, or wasn't touched on by your testers. Best to learn them now, rather than repeat any mistakes in your next design!! If you're lucky enough to have the map put on some servers, play on them, and ask people for feedback while playing! Your future designs will only be better because of it.
Conclusion
Realize that solid design, good thinking, feedback, and attention to detail are the keys to making great maps. Everything you do will help you gain experience and become a better mapper, as long as you pay attention; and don't forget the things you learn or the opinions of others! Best of luck to you; and remember:
As always, if you have any questions or problems, feel free to email me!
|
| Half-Life, and the Half-Life logo are trademarks of Valve Software and Sierra Online, used with permission. All screenshots and drawings of Half-Life are (c) copyright Valve Software, 1998. All rights reserved. All original content (text and art) are (c) copyright Wavelength, 1998, and may not be reproduced without permission. |