Posted on December 30, 2017 by the_hack
(article is not complete!)
I haven’t made any super popular maps yet, but I’ve definitely learned a few things. My best map to date is Dying Sun which is primarily a 1v1 duel map. The layout took a few months and reached 97 alpha iterations before entering beta, most with play testing in-between. Thanks to Joel, a veteran competitive duel player who got me thinking about the finer points of gameplay and level design.
Every map is different and rules are meant to be broken – and these are definitely not rules! Mapping is as much art as science, because there are so many tightly codependent variables at play. These tips are mostly relevant for duel. In duel, balance is important. In deathmatch, nothing is important except fun (IMO!).
You can probably make a great map by accident or intuition, but sometimes having some methods comes in handy. For those getting started, these are the ones I think are valuable. It’s always going to be way messier what is described – the most important thing is to keep going for the best map you can, learning along the way.
Which gametpye are you designing for? Sure, maps get played in multiple game types all the time. Erase, Deck and ASDF seem to be universally enjoyed. However duel, deathmatch, team deathmatch, elimination and domination all have very different objectives. When you have an idea, it will likely be in the context of a single game type. Stick to this game type throughout your design process, at the most picking a secondary, related game type (eg.: duel and TDM, or perhaps DM and elimination or TDM and elimination).
Create a unique value proposition for your map. What “problem” does it solve and what makes it different? Sometimes we just want to be creative and that’s fine, but keep in mind the most successful products in are those that fill a need uniquely.
Or, rather than solve a problem, an idea be a hypothesis, such as for AMPhitheatre: “a coliseum type layout around a frequently recurring U-damage will be fun.” It’s just trying something new.
Whatever it is, it’s easier to create according to a vision. A design process with no masterplan has no completion criteria. I was stuck in such a process when another player said Dying Sun could possibly solve a problem: UT4 maps were often too easy to control, and that it was hard to get back into a duel once getting dominated – even by a similarly skilled player. From that point on it got a lot more focussed.
Wether you are solving a problem or testing a hypothesis, you need to be able to identify the hook. This is how you describe the map in a sentence. For Dying Sun, it’s “a fast, tight map that is hard to control.” It has both a solution (hard to control) and hypothesis (fast, tight) right there in the hook. Identifying the hook isn’t really about advertising your map, it’s more about defining your purpose early in the design process.
Once a hook is established, a general flow would be born from it. The flow is basically how your map supports the hook and should begin as the simplest possible manifestation of the idea, building complexity over time according to what the hook demands.
Along that flow, certain areas and ‘situations’ will separate themselves from the rest of the flow. I find it good tease these out as much as possible. Dying Sun was vertical from the beginning, the two levels connected by two lifts. At a point it was obvious that they were key. Once identified, the map was designed around the situations created by the lifts.
Another example might be the top area of ASDF (aka the quintessential UT4 duel map). Imagine that early in the design, the top area in ASDF was open. Simply adding walls to block sight might have created a classic situation whereby people camp in the area looking for a sneaky shot on players looking to get the high ground. Just tweaking a few angels here would change this classic situation a lot, so I’m guessing it’s the result of having been recognised and emphasised.
Armors beg for situations, but it’s difficult to create a natural feeling situation that works around them. They have to be cool, but controlled. You don’t need to create them deliberately, but look out for them. At some point, it will click and you will know what to do. The 150a situation on Solo is a great example of an armour situation. It’s hard to defend and also attack, requiring a lot of caution and timing on both sides. I agree that’s it’s good to create armor situations that are difficult to fight on top of. Armour situations that demand careful setup and timing are the best.
If you don’t have an armour situations in mind for your map, I wouldn’t sweat it. Look at how the map evolves, and keep thinking about it, and how other changes in the map might support the armour situations. The Dying Sun 150a situation was the last thing to do on the map. I had tried a number of placements and situations but nothing felt right. In the end, what worked was quite simple.
Beyond Hook, flow and situations/areas, you get into smaller details of item placement and exact layout. At the start of the process, they will be quite random, but with effort every detail will support the map’s greater concept.
We’ve talked about the hook and mentioned ‘flow,’ which is a word that gets used a lot. But what even is that?
Flow is kind of an esoteric term, but it primarily refers to the navigation options around a map, and secondarily, to how the weapons and other items are arranged.
There are just a few general things I like to keep in mind, and they change according to circumstance.
Players should have just three or four immediate route options at any time (back, left, right for example), or perhaps two major options and one incidental one. Some limitation in what players can do creates predictability. You want to promote reasonable player engagement, and having too many routes will enable players to evade each other too easily. However you don’t want to make it too linear. Just limit options along each route, so that once visual contact between players is lost, a reasonable guess on their navigation can be made based on items. An intersection with more than 5 possible immediate choices could be possible, but be sure to provide sound cues for at least a couple of choices (eg: jump pad, lift, armour/weapon/ammo).
An interesting phenomena are backchannels in map, that only have two possible destinations. This might be the flak corridor in deck or the back stairs in Erase. If your map has too many such situations, it will feel like a collection of traps. A couple might be fun though.
Line of Sight
Feeding into the risk idea is line of sight or LOS, which needs to be considered and controlled. Excessive LOS is considered by many to be map breaking – but providing opportunities for sneaky snipes and prediction shots adds great suspense and keeps players fully engaged. ASDF is a great example of a map with a good LOS balance. Any time a player is exposed by line of sight they are vulnerable and taking a risk.
A large part of the flow is the travel distance between destinations. They should be comparable and responsible and relative to the payoffs. The risks involved in travelling should also be factored in – a shorter travel time with higher risk might be the equivalent of a longer travel time with lower risk. So perhaps if a particular distance can’t be made longer, perhaps make it harder. For example, any time a player is airborne they are vulnerable and predictable.
Combine both airtime and LOS vulnerability with caution. Make sure your risks are always relative to rewards.
The major armours (100 and 150) should be opposite each other with more than half the map’s distance (in travel time, not literal distance) between them. This is because if a player can easily collect both, it becomes too easy to control the map with a monopoly, which everyone is trying to do.
To counter the potential for a monopoly, it’s mostly advisable to put smaller armours and other items such as vials and healths away from the major armours, though there are great maps that do the opposite. This includes powerful weapons such as shock and the rocket launcher (which other weapons are powerful might depend on your map). This gives an out of control player a chance to get back in the game, by avoiding conflict for the major armours, and building a stack with smaller items. Antiqus, a map by Heresiarch has double healths very close to both major armours, which works to promote engagement over the items, as the risk for each player is lower.
In Dying Sun, a tight map, the out of control player is supported by keeping basically anything that is not a major armour on the second level (with the 50hp close to each) requiring some travel distance. An out of control player could get back on their feet even after a death, by spawning on the second level and quickly collecting some health vials and weapons. The idea isn’t to make the game some kind of socialist experiment, it’s to simply make it hard to control. For a small and fast map, this is essential to avoid runaway games.
Beginners and pros will need to be able to play your map. I like to keep all basic map navigation completely accessible – no special game knowledge should be needed to traverse the map. Above and beyond this, I aim to provide interesting options and shortcuts for more experienced players, based on more advanced movement (“trick jumps”).
The options shouldn’t create too many new ways to get around the map – map flow should be preserved. If the map relies on too much esoteric map knowledge, games between similarly skilled players will be might become unfair because a few obscure options get exploited. Try to offer tricks, but not make them game-breaking.
Trick jumps are fun but shouldn’t be required. ASDF is a good example of a stock map that is dead simple to get around, but provides plenty of options for advanced players. Not many of these options drastically change the flow.
Solo is an interesting map in that it can be navigated completely differently with trick jumps. This exacerbates differences in skill and map knowledge, while also increasing the skill ceiling on this particular map. This isn’t bad! However if all maps were like this, the game might seem overwhelming. As a duel map, Solo is one of the best. For public DMs, it can be a disaster.
Assume your map will be played by duellers who know the map back the front, and can also recognise every sound in the game. While the game is not finished, at the time of writing, a player picking up most items can be heard by another player across the map – some clearly, some less so. This means that based on sound alone, one player will get a snapshot of where the other player is, and make assumptions about where he/she is going (tracking).
Navigation, health and Armour Sounds
Jump pads and lifts are very obvious map elements that give off useful sound cues for tracking, however these will be placed for navigation purposes primarily. Your major armours will likely be in positions for reasons other than sound, although they give off the best sound cues when taken. Health sounds aren’t that audible from outside the vincinity I believe.
As these sound emitting pickups are place on the map for flow reasons, the remaining pickups, vials and ammo, are important sound cues. A trick I have picked up is to place these in areas where there would otherwise be no sound given by a player. This is to assist the players in tracking each other, which forces both of them to move more intelligently. They can also use sound cues to trick each other.
Vials are commonly found in groups of 3-5 and a map might have 10-15 overall. Having some near smaller armours helps support the out of control player and might lead to more balanced games.
The amount of ammo on a map isn’t as important as the number of vials. I like to bunch them in threes, creating a very hard-to-miss pickup sound.
Not everyone has surround sound, so giving clumps of vials and ammo different spacing helps players identify which clumps they are hearing. Eg, the small group of five vials on the top of ASDF, or the long string of three leading up to shock.
All feedback is useful, even when you disagree. Hearing other people’s opinions sometimes strengthens your own idea and helps you refine your idea it in ways you might not have otherwise realised.
Playing the map yourself is great. Try to find players, but if not, bots are fine too. Even better, is watching people play your map. Getting it loaded onto a server once every so often during development, and watching the replays of people playing, will be a revelation. If three people play a game, you have nearly an hour of play to analyse. Watch how players of all levels use the tools you have created.
It’s worth mentioning that you shouldn’t aim to have every version of your map uploaded to a server (server admins are human beings!). Maybe aim to have a version uploaded only with major changes. Every two to four weeks is reasonable. Even better, learn to setup a hub, and host your own man. It’s possible to do such this freely on a home computer, if your internet connection is half decent. Be grateful to whoever helps you test and be thankful for their feedback no matter how salty. Even if you don’t agree, it will help you.
I like to do a lot of testing – not for every single version, but at least for major changes. Sometimes you’re confident something will work and it flatly doesn’t – sometimes things you thought were easy, most people fail most times.
As a “designer,” I tend to go overboard on this, but in general I believe that things should be refined to their essence. In Dying Sun the result of this was it became extremely tight – there are no blocks of walls – every corridor or area is adjacent to another. No platform is longer than it has to be. When the map got to a state and I realised that a single route was too long in travel time (but the rest was ‘OK’), I realised I had to cut 1000 units from the whole map in order to make that route work. I ended up slicing and dicing the map into pieces, and a day later had the same map but 20-30% smaller on each direction, without making a single platform narrower- just by simplifying the layout. The same flow, just faster. This might not be required, but if you have the most simple representation of an idea, you can then consciously expand and complicate it in ways that support the hook, rather than being stuck with what you’ve got. In Dying Sun this meant I was able to add elbow room back into certain areas, and extend corridors where appropriate.
Everything in a map should have a purpose, or an intended use. Dead space (unchanged areas) should be avoided unless absolutely required to balance travel times. Everything should be a ‘situation,’ where if an engagement should occur, it won’t be completely random. Predict and intend as much as you can and build the possible results into masterplan of your map. Easy to say I know…
Alpha I consider everything before the gameplay is set. Beta, I consider finished but in a rough state, such as a shell map built with BSP. I’m pretty sure most of my maps will remain in Beta as this is all I basically want to achieve. This is why I also try to make my maps visually appealing, even though not being fully meshed. I give each iteration a suffix such as -a001 or -b01. Alpha maps are very much more WIP than beta.
Experienced and successful mappers can definitely teach you, but when they discourage you from trying new things, it’s time to tune out. Take good advice and leave the bad – you don’t make something really great copying or remixing the old. (You might make something good, but good is not good enough.)
Most mappers view things from their own particular perspective. It’s very dependant on what games and game modes they prefer. Smart people realise they don’t know that much. Gatekeepers like to create the illusion that they do.
When you do something new, plenty of people will say it won’t work or is stupid. Mostly, these people will never make anything great.
As UT is all about the fun, and the game’s history is rooted in greatness that came from experimentation (UT99 was full of different, experimental maps), and as you make your money in your day job anyway, GO FOR IT. Be creative, but test your ideas.
This video by the Halo team talks about play testing and some issues that involved.
This video is the UT4 level design team talking about their process in a random kind of way.