Making Multiplayer Maps

This comprehensive guide on multiplayer mapping was kindly provided by Decay.

Let’s Make Some Multiplayer Maps!

Welcome to the world of multiplayer mapping or player-versus-player (PVP) mapping. 

Multiplayer mapping is a more complex beast than people might imagine. Projects of the past often drew on single player maps, either thematically, structurally, or just tossed in deathmatch starts right into the single player areas! Eventually some single-player sets would make a dedicated deathmatch arena off to the side. However, the best crafted maps are ones dedicated to the cause. This guide will provide you with an abstract, thoughtful approach to multiplayer mapping based on the concept of flow. It is not exhaustive, and not centralized around Doom deathmatch, rather it can apply to any game mode, and realistically any map or level making in general, just by taking the core principle of flow.

Before we get into this though, I want to say that there is no one way to make a multiplayer map. This only serves as a guide to help you think about creating them in a more abstract way. There is also no way in determining what will make a map successful. The multiplayer community as of this writing is substantially smaller than it has been historically, and will likely only shrink further. It is very possible, and indeed quite likely, your maps will never be played by the general public outside of you making the effort to drag people to the server, and events held by figures or entities in the community. Success is determined by you, the author. Does the map meet your visions or goals? Do other players comment on it positively? Did it attract other mappers to the cause? Measures of success can be deeply personal, and I do not want to impose any sort of measuring stick.

However, if you do want some sort of measuring stick, here’s my take. Personally, I would consider maps that earn you recognition or acknowledgement a success. A huge success would be any maps that make it into competitive play or compilations. A complete and total success would be a mapset that is continually played without you prompting it. I think this is a fair assessment given the current climate, but this is strictly my opinion and need not be followed, lest you become discouraged. The amount of maps being made for multiplayer has lessened dramatically in the past decade, for a variety of reasons, and I encourage you to make maps for multiplayer if that is what you want to do, regardless of perceived success.

One of the most important features to keep in mind about making a multiplayer map is that it needs to be fun and engaging enough to be replayed many times. Unlike a single-player map which might be replayed a couple times, successful multiplayer maps will be replayed dozens, hundreds, or even thousands of times. Unsuccessful maps will be played once and dumped, and one bad map has enough potential to either kill the server or the mapset. So let’s dive in and begin!

How do I make a “fun” map?

“Fun” is an extremely subjective idea, obviously. What you might envision as working out great might only work in theory, but in reality plays extremely differently. You might make a map with a large central area meant for intense chaos, only to find out people are predominantly playing the side hallways and complaining about how cramped the map is. You have to keep in mind the map is meant for public consumption, meaning you have to temper what you conceive as fun with a more general conception of enjoyment.

Keep in mind that everybody plays at different levels, and it can be extremely difficult to negotiate this. You do not need to be a multiplayer player to make multiplayer maps. However, it is extremely beneficial to play a lot before getting started in multiplayer mapping. There are several reasons for once. First is understanding the general balance of the game, figuring out player speeds which will determine structuration and timings, determining what level of intensity is appropriate which helps determine map size, having a solid understanding of the role of each weapon in multiplayer (as opposed to simply carrying over single-player principles) which influences placement and helps any implementation of custom weapons if intended, and, most importantly, getting an innate sense of what is generally conceived of as enjoyable or fun.

So how does one make a fun map? Fun, like play itself, is unpredictable. The best thing you can do is to simply try to provide the conditions for evoking a psychological reaction that can summed up as “fun”, and that is achieving “flow.”

The Key to Success: Achieving “Flow” in Multiplayer 

No, I don’t mean map flow, though map flow is important to this. Arguably the most crucial factor in making a successful map is enabling players to achieve a psychological “flow”, a term coined by Mihaly Csikszentmihalyi. This applies to any map or level making, be it multiplayer mapping or single-player endeavours, Doom PVP or another game. What you’re looking to do is getting players into the zone, which in turn keeps them wanting to play the map and, more importantly, replay the map to recapture that emotion/feeling/mental state and possibly improve on it. In layman’s terms, this is what makes a map “fun”, in the sense that people enjoy playing it, can grind it out without becoming tired, and want to revisit it.

Flow can be achieved in multiple ways, which I will break into their own sections. Conducive layouts for smooth movement, spawn points that capitalize on the map structuring, weapon balance and item placement, map functions, and appealing logic.

Flow: Layouts and Movement in Multiplayer

Here is the actual map “flow” discussion, the actual layout of the map.

Many aspects of thought often go into the flow before any lines are put down. Is it an arena style map, where players tend to congregate in a central area? Is it an item control map, where items are points of interest that must be checked but also not the central focus? These things can absolutely dictate how the map is built.

Warning: Avoid symmetrical builds like the plague. There are multiple reasons for this. First, players will often be unsure of what side they are on, or where they are precisely in the map. Secondly, they are boring because they are thoroughly uninteresting. Once you know one side, you know the rest of the map. It’s wasted space and opportunity. Thirdly, they offer no rewards to players. Whatever you have on your side, the opponent has on the other. There is no incentive for directed play or strategy, which takes away from developing flow. 

Exception to this rule: in the case of Team based modes, only local symmetry of the team bases should be avoided, as the maps themselves must be naturally symmetrical for balance. IE red side and blue side should be the same, but blue and red bases themselves should not be symmetrical.

One of the great things about Zandronum’s “Last Man Standing” mode is that you can test a map’s structural flow and spawn points easily. You get 200HP and 200 Blue Armour and all the weapons if using the Classic LMS mod. Thus you can see how the map behaves with everybody on equal footing, having all the weapons. You can find dead ends, bad choke points, and often singular paths of travel, indicating travel times in other parts of the map might not be efficient, meaning you should reconsider some of the structure.

Choke points can be good and bad. They can inhibit movement, but also serve as a point of interest. If you’re going to have choke points, consider the spawns and weapons around them. Are they going to cause rage to players? Are projectile-based weapons readily accessible? Will it eventually become a dead-zone nobody will want to visit? What purpose does the choke serve? A balance of hallways and rooms for fighting is required – maps that are solely one or the other often do not work well, though that’s not to say that they absolutely cannot. 

Similarly, dead-ends are not generally a good idea in multiplayer mapping, for their namesake. Doom PVP moves too fast for dead-ends to work most of the time, and if you use dead ends they really need to be incentivizing, with a power-up or power weapon. However, in saying that, if a player can get in a dead-end and out with whatever they went in for, that can contribute to the development of flow. You really need to be deliberate when implementing a dead-end.

Overflow is also something to consider. Too much freedom of movement is not a good thing, as players often complain of being unable to find opponents. In any map, inevitably players will try to find optimal paths for movement, but in a map with too much flow it becomes extremely overbearing and unfun; you end up fighting the map more than each other. Without some built-in guidance, players find no reward or purpose. Predictability generated by the map is a good thing with a proper dosage – not too much, but not too little either.

There are two critical points of directing the flow in a map, not just in movement but also pushing players to move in particular directions. The first is a point of interest. Points of interest are areas where players tend to congregate, usually around a power-up, a high visibility vantage point, or a spot that dominates over newly spawned players.

Side note about power-ups: Power-ups are a good thing, but like what they say, too much of a good thing is bad. You don’t want too many power weapons like the BFG or too many Soulspheres or Megaspheres lying around; having too many important things directly impacts the flow of the maps and may even make it unfun to play. Powerups should be strategically used to improve and dictate the flow and engagement on the map. People will fight over areas with power-ups and generally go hand in hand with being inside a place of interest.

Vantage points or points of interest are very useful in signalling where you want conflict on the map to be stronger or more prevalent, and the more enticing the area is the more likely people will flock to the area to control it. At least one is recommended but it doesn’t hurt having a second place of interest or even a third; the need for more or less is directly correlated to how large you plan to make your map. You plan to make a 32 player map? You’ll need a pretty decent sized map, but majority of the stronger mapsets comfortably support 6-12 players, and it’s highly recommended to not make a super large map unless you have a particular project that gears towards maps that are required to support such numbers.

The second point is spawns, which brings us to the next topic.

Spawn Point Flow: Guiding or directing players

Spawn point placement is arguably the single most important aspect of a map and is a vital part of multiplayer mapping. Just one poorly placed, oriented, or inadequately armed spawn can break a map and condemn it. There are problems if there are too many spawns or too few. Looking in the wrong direction can be the difference between constantly taking damage or having a moment to breathe. Inadequate arming can determine whether or not you will be consistently rolled over.

 You, the author, are determining how to start a player off within the map flow. The orientation/directioning of a spawn is critical as it guides movement, summed up as players determining how to reach points of interest or action in optimal times as dictated by them, not you, how to defend themselves in the face of immediate action. While you can direct the flow of the player by pointing them in a direction you want them to go, remember that, as a principle, free-play, or playfulness, often results in players wanting to play the map as they want to, and if they are too busy fighting your will because it doesn’t quite make sense, that will take away from their overall experience.

When a player spawns, there are many instant thought processes happening, whether that means coordinated action with teams (Deathmatch, Last Man Standing, Capture the Flag, or other team based modes). Defence, offense, rapid movement, doing whatever is necessary to continue winning the game or begin fighting back to gain the lead. Items will factor into this, and no matter how you spread the items in the maps, players will usually move off a spawn in a way that they will perceive will help them win or get a respectable score. Good maps have spawns that are conducive to these goals.

However! There are always exceptions to this kind of rule. All the most popular maps have what one might consider technically a “bad spawn”, such as a pistol start, a tight alcove with a weaker weapon in a high traffic area, or out of the way of everything. However, these “bad spawns” provide a double opportunity for players.

First, it drives up the intensity and gives players a positive feedback when they can get double or multi kills from inconvenient opposition spawns. People like when they can frag easily so it provides extra fun factor when players can get into the flow. Second, for players who can make the most of a bad spawn and capitalize on it, they can feel more accomplished. New or not so good players can get a sense that they are improving if they can get out of a sticky situation from a not-so-great spawn point.

While it is possible some players will gripe about it, it’s not too likely an unbalanced spawn or two will drive players away, depending on how it’s implemented and not too egregious. An empty handed spawn in a pit that is high traffic and guaranteed death offers nothing, as do spawns that face walls. Striking a balance between even spawn and weapon distribution against layout, intensity, and fun factor can be difficult, and this is one of the key factors that needs to be continually tested. Knowing what weapon to place on or by a spawn leads into the next topic.

Multiplayer Weapon Balance: Frags and Flow

Weapon placement is another critical factor in designing a good and fun map. However, it also presents its own dilemmas, and can be one of the trickiest parts of multiplayer mapping. A good map should  enable players to know where all the weapons are and how they are accessed within the first couple of minutes of play, and being able to find a more powerful weapon within 5 seconds of spawning, because that is a long time in Doom multiplayer. This is because the player/game speed allows for a ton of square footage to be covered in that time, meaning opponents have a pretty good chance of finding you quickly even if you spawn on the other side of the map.

A common mistake from predominantly single-player mappers is to treat the Super Shotgun (SSG)  as a super weapon. In fact, it is the primary weapon people prefer to use. Maps with no SSG are extremely unpopular, and same with maps with only one, with a few exceptions such as Pobla8. Having 2 or 3 SSGs is good, along with 1 or 2 of the spawns starting on one of those SSGs. This is one reason why the 32 in 24 series has suffered greatly in playability – maps were often made with a single-player weapon balance mindset, prioritizing the Shotgun (SG) and Chaingun (CG). 

SSG spawns can be an excellent means of fighting back – or a way of continuing to dominate a match. Doom’s spawns are random (the “Spawn Farthest”  DMflags are reviled and should never be used, if only for the fact that you can manipulate it by directing the frag locations), and as such even carefully crafted weapon balancing on spawn points can be completely obliterated, but it is certainly worth making as balanced as possible. Figure out which spawns should be given SSG in relation to traffic and SSG function. If the room is spacious, the SSG is not as useful given its spread, so a Chaingun  or Shotgun might work better on the spawn itself, with a SSG nearby as the player moves closer to others.

SSG only spawns are a downfall because the game now has fewer angles to work with, resulting in one-dimensional play. This is what causes some multiplayer maps to become boring quite quickly, especially as there is less reward for being a more dominant player. Multiplayer maps such as GreenWar Map20 fall victim to this. Although Shotgun spawns are generally considered under powered, being proficient with the Shotgun is considered a valuable skill that can pull you out of jam. It is also more psychologically rewarding to players who can take out the SSG domineering player with one or two well placed Shotgun shots.

Rocket launcher (RL) spawns are dangerous – often players end up killing themselves or running out of ammo quickly. They are not ideal but can certainly spice up game play, as can Plasma Rifle (PR) spawns. BFG spawns should be used sparingly, but can work dependent on how the BFG is placed and its overall usefulness in the map. They are useful for “spamming” – simply holding fire and letting the frags roll in. It is extremely gratifying for players to hear “double kill” “multi-kill” “ultra kill” “m-m-m-monster kill!”, and the more a map can offer this opportunity, the more likely players will want to revisit it, as maps that enable fast and high frag counts hits right to where players want to be psychologically, and appeals greatly to flow.

Weapon distribution helps you create angles in play and different ways of playing. Players will find more replayability in a map as they figure out what weapons work best for what situations, and how to get or bring those weapons to such situations. Creating different angles or dimensions of play via weapon uses coincides in ways they are placed. BFG placement is often in hard to reach places, as it can be very rewarding for the player to get, and in turn can generate different ways of playing the map itself, dependent on the players. Structuring also influences weapon placement, as more difficult structuring will discourage some players from frequently accessing weapons or power ups, and that goes into the next topic.

Multiplayer Map Functions: Striking the balance between technical and accessible

There is a trend in modern multiplayer mapping to tie BFG access with trawling through death traps, or power-ups requiring perfect movement and timing to get. Anything that requires near perfect or uninterrupted movement I call “technical” structures or functions. The number of maps post 2008 relying on this technique has resulted in somewhat cold, sterile maps that are a chore to play rather than fun. Accessible maps would be structured mostly the same, but lifts would be more forgiving, getting the BFG or other power-ups would be seen as integral to game play but not so dangerous to get that it isn’t worthwhile. A similar huge mistake many maps make is having a power-up or power weapon such as the BFG so far away from the activating switch that players are mostly lowering it for other players. This is poor design and results in more frustration than anything else, and is extremely detrimental to enabling flow.

Highly technical structures and functions can result in psychological flow, particularly a feeling of bliss/triumph as difficult to achieve things are accomplished. They are not inherently bad. However, they should not be considered the only option, and often present themselves as a barrier to new or lesser skilled players who do not have the incentive or patience to learn these techniques as ruthless players stomp them time and time again. One of the greatest strengths of older maps is accessibility – anyone can learn them with ease. Accessible maps promote flow more generally because they are smoother to play by nature. Higher levels of forgiveness means mistakes can be made without totally condemning the player, and that encourages continued and repeated play.

Technical maps are often tests of precision skills, and often appeal higher skilled players, who optimize movement and aim. They can be difficult to learn how to play and require much practice. Accessible maps may also take considerable practice to become good on, depending on the age of the map and the skill of other players on the map. Old maps from the 1990s are often simplistic, and are easy to learn and get the hang of, but a player who has spent considerably more time on it (years or decades versus weeks or months) often retains the advantage. It’s worth remembering that the maps of yesteryear are still competitively viable despite being accessible – that tells you that a map’s “competitiveness” is determined by players, not by how technical it might be.

This is relevant when creating new maps, because it is essential to avoid the pitfall of looking at older, popular maps and trying to emulate them. This results in stagnant creativity, where everybody is trying to replicate Dwango5 Map01 or Mount Zero in their own way, which in turn often leads to immediate criticism because although the map is somewhat familiar, the differences are enough to break mental flow. This leads into the next point.

Radical or Formulaic Layouts: Appealing to Doom logic to create flow

There is Doom, and then there is “Not Doom” (TM). There is a certain logic when creating multiplayer maps, and deviating too far from that results in an antipathy towards the map. Something worth striving for is a blend of new ideas and old logic. If you insist on trying radical layouts in a pack, it is worth building up to that with previous maps, essentially breaking in the player to new ideas and new ways of envisioning play. Often the game changing aspects of these are traversing room over room (enabled by 3D floors or in some cases transfer heights as applicable) or radical changes in height variation.

The reason why this matters is because Doom’s multiplayer logic is cornered on a foundation of not having freelook (no looking up or down) nor transfer-heights and 3d floors (or portals, where applicable). This also helps us figure out why remaking maps directly from other games does not work favourably – they need to be “doomified.” Too much height variation or overzealous use of features does not usually resonate with players, requiring them to develop new skill sets. Unfortunately more often than not players do not wish to develop those new skills. A good example is the porting of Quake maps. Without Quake logic and weapons, the maps, more often than not, do not play well in Doom. Dragonfly’s take on “Toxicity” in Progressive Duel 2, retitled “Ardens Oppressio” is a good example of applying Doom logic to a Quake map that changes it enough to be recognizable but also constitutes a new map.

By appealing to established Doom multiplayer player logic, you allow players to get right into the heart of the game without having to abandon their comfort zone. Pyrrhic 1 on 1 is a great example of this, being a recently made mapset that is as oldschool as it can get, while also presenting newer layouts that are attractive to both new and old players. It is easier to make players want to revisit maps if they feel right at home – but you can also make them want to revisit by twisting the logic just enough that they are comfortable but notice a difference they want to address.

Other aspects worth considering

This has been by no means an exhaustive list of what to consider when making multiplayer maps. There are many more factors worth considering, though they would warrant another write up in themselves. Although they have less bearing on achieving flow, they are still critical to map design and purpose. I will briefly cover a few more topics.

When starting to map, ask yourself the following:

Do you require one or more of the following: DECORATE, 3D Floors, Dynamic Lights, New Actors, Classes, Complex Scripts, Extensive ACS support, PK3 or PK7 support, UDMF support, Portals, Open GL, extensive audio file type support, interchangeable texture and flat usage?

If you answered yes to any of those questions, you need Zandronum.

Do you want the map to work on ALL the ports?

Do you want to make a simple 1v1 or ffa map?

Then you should make sure they work in all the ports. Rule of thumb is generally if it works in Odamex, it works in ZD and Zan. If it works in ZD, it works in Zan, but might not work in Odamex. If it works in Zan, it might not work in ZD or Odamex. Odamex and ZD require that WADS are constructed according to the old style, so make sure that the WAD is properly compiled. Zan is more flexible and less prone to crashing with bad compiling.

Getting your wad hosted on each port is different. Each port supports someone setting up their own home server or LAN server. For dedicated hosts, Zandronum supports users setting up their own server via the TSPG service or by asking another server host. Zdaemon and Odamex require that you ask a server host for dedicated server. This is, of course, assuming you aren’t hosting your own set of dedicated servers. Servers will be determining how people access your maps, and it is definitely far more beneficial to be using a dedicated service. However, servers lead us into a different aspect worth discussing.

Flags: Server settings, intended play

Many people here might be confused with the tossing around of terms such as oldschool (OS) and Newschool (NS). Generally it is referring to a series of flags (settings) that determine what your game play experience is like on a technical level.

There are 3 generally accepted forms of flags (mostly from a DM perspective). OS and NS flags are generally highly contested and everyone has their own idea of what is “right”, but these are some general guidelines that influence how the map is built. For example, if you want to implement jumping, you are going to need to follow New School principles.

Important note: Doom multiplayer always utilizes the “weapons stay” dmflag, meaning weapons don’t disappear on pickup; you can pick them up just once per life and have them. Respawning weapons are known as altdeath but by historical experience this is not preferred and would undoubtedly cause some rage. Native respawn rate for all items in doom is 30 second and that’s unacceptable for weapons, so one would have to completely redefine standard items for this to work.

Oldschool (OS)

  • Nightmare skill (native double ammo)
  • Double Ammo
  • No Freelook
  • No Jump
  • No Item Respawn
  • Strict Air control
  • Infinitely Tall Actors
  • Preferred Weapon Order switching

More conservative OS servers might include

  • No Crosshair
  • Silent West Facing Spawns
  • Automatic Weapon Switching

New Oldschool (Rustking style) NOS

  • Ultra Violence skill (to avoid native double ammo)
  • No additional Double Ammo
  • Freelook Enabled
  • No Jump
  • Less strict air control
  • Item Respawn

New School (general dm, not duel) NS

  • Nightmare Skill (for Double Ammo)
  • Double Ammo
  • Freelook enabled
  • Item Respawn
  • Jump enabled
  • High degree of air control

Pick a set of flags (settings) that you want and stick by it. It sucks to play through a wad only to find out that it is very confused about what it wanted to play. See that BFG on a ledge? Well, in NS, I can just jump up there and grab it, breaking any balance it might’ve had if it was OS instead. If you design the maps around intended settings, it will work much better, and make it easier for server hosts to host the wad as you actually intended it.

Detail: Why you should detail a multiplayer map

A map with a strong flow and good item placement is only still a part of the recipe; the other part that helps a map feel more complete is detail. There is a misconception in the multiplayer mapping world that multiplayer or PVP maps do not need detail. This is incorrect. Detail is highly important in making your map not only feel good but also look good and enticing, which contributes to the psychology of flow. Pleasant to look at maps combined with a good flow and item placement will always yield strong replayability. It is the type of detailing that matters the most. Maps that visually inspire wonder and stand out from the dreary old sets of the Dwango days relying on stock Doom 2 textures or Gothic textures always yield positivity.

You may ask yourself, “How much detail does my map need”? It needs enough going on that makes it look alive but at the same time doesn’t hinder gameplay in any way. If you have edges, corners, objects, or pillars that you can easily get caught up on, bumpy floors, a lot of sector light phasing/glowing/flickering going on at one time on one screen, very dark rooms everywhere, then chances are you have over detailed to some degree, or implemented poorly. Floors need to be smooth (Doom engines and port physics all tend to handle small height changes very poorly resulting in loss of control), lighting should be bright enough for seeing other players with ease (intentional dark places aside), light glowing and flickering should be kept subtle (not distracting), enough to be noticed but not enough to impact the gameplay. If your map causes players to have low FPS counts because of how many sectors and effects you have placed on your map, you have over detailed. Although nobody really wants to alter their visions when they make a map, the map benefits greatly from having these types of issues resolved, it can go from having a map with lots of (wasted) potential to being a map with everything just right.

Under no circumstances, which is a big no and I mean NO is to ever draw detail influences from older map packs. You’ll realize that a lot of maps back in the Dwango days were underwhelmingly (not) detailed and in this day and age, you’re better off looking at more modern packs or packs with a higher amount of detail. You can take what you see and remix it to make your own, and give it new life. There is nothing wrong with utilizing inspired themes but direct copying and pasting is boring, so it is highly advised to change it up in any way you see, it is a matter of how innovative you can become with the textures you are using.


Multiplayer mapping is tough. It is multifaceted and each game mode has their own particulars to consider. This only provides basic principles that can be applied across game modes – each one certainly warrants its own guide or tutorial. There is still plenty of other factors to consider, such as play testing, receiving and giving feedback, and factoring in mods.

Of course, you can always just ignore all this and draw lines and see what sticks. Some of the best maps are complete accidents, while some of the most reviled maps are products of too much thought. All this being said, get out there and map! And more importantly, play!

Be the first to comment

Leave a Reply

Your email address will not be published.