CP_Steel Design Principles - A dyncp post-mortem

fubarFX

The "raw" in "nodraw"
aa
Jun 1, 2009
1,720
1,978
Steel has a weird mode and a fairly confusing flow. It definitely has a severe case of "too confusing for new players". Yet somehow, steel pulls off the gamemode with various design tricks that prevents it from being a complete mess. There is a fairly strict formula to it. This is a guide that looks into steel and its design choices. These are my learnings from designing a map for the dyncp contest.

The main idea driving the whole map is that there are 2 active points at all times. Blu can attack either of 2 points and the map relies on Red being able to defend both at the same time.

Core Layout Rules

Rule #1 The red spawn must connect to all active points

Defenders must have immediate knowledge of where the 2 active points are, having a spawn that connects to all active points fills that need.
Steel could not have had a single spawn that connects to all 5 points, this is why it uses a forward spawn. regardless of which spawn is active however, rule #1 is still fulfilled.
This is part of the main reason why we decided to only have 4 points on cp_hadal, we figured that we could avoid having multiple spawns if we managed to connect all 4 points with a single spawn so we decided to design the map around a single spawn. The defending spawn is the single most important thing in a steel map and everything should be built around it. Word of caution: one of the downside of having 2 spawns is that eventually, defence will be pushed back to a spawn in an area that they had no prior knowledge of, this is a bad design because it is a break in continuity and the defending team might be completely lost and may have a hard time getting back on its feet. steel's spawns are tolerable because they are both very tightly connected to the main point.

Rule #2 Red must be able to quickly switch from defending one point to the other.

This is usually done through the spawn. This is why having a good red spawn is crucial. A good spawn ensures that the defence can quickly switch between the active points at all times and be effective in defending both. An interesting exception to this rule is cp_Sulfur, sulfur has a mechanism that allows you to recover from losing the final point, this greatly reduces the emphasis on defending 2 points at once and removes a sense of urgency associated to the last point. This affords sulfur more liberty with its layout as it does not have to equip the player with the ability to react swiftly, it makes the pacing of the map more manageable even if it is slightly confusing.

Rule #3 All points have a route connecting to the main point

This is pretty straight forward, the final point must be accessible and you should actively shove it in the player's face if he is not in an area where he is being useful. For example, if an attacker goes through B while it's locked, he is guarenteed to end up at E. Likewise, if an attacker goes from A to D while D is locked, he is certain to end up at E eventually. You want to direct the player as much as possible.

Rule #4 All points connect to each other chronologically with a one way path.


Steel is big and confusing, you need one way paths to give a flow to the map and lead the player around. Note how steel's A>B, B>C, C>D are all one way paths. This is a great tool in preventing backtracking and making sure players end up somewhere where they can be useful. The connection between A and D is a little more arbitrary, that one is up to you.

Rule #5 Points get gradually be less interesting.

Steels layouts have progression to them, control points become less and less interesting. This indicates to the player that the focus should gradually shift towards the final point.
Steel's A and B are big, sprawling battlefields. C is slightly cramped and can't accomodate as many players, it shares an attacking path to the main point clearly telling the player that it's something he might be interested in. D is just one of the most boring platforms you can get. This is not a hard rule but it has desirable effects that serve steel well.

Rule #6 increase the focus on the main point by making it gradually more attainable

The main point should be harder to capture at the start to prevent abrupt games. The final point should be meant to act as a tool to distract the defence rather than something that should be capped straight away. There are 2 components to Steel's main point's ease of capture. Accessibility and safety. Accessibility is done by deploying bridges to reward Blu, it allows more classes to be involved on the point. Safety is achieved by adding guardrails to the point making it more secure to navigate. How you make your main point easier to obtain is up to you. With cp_hadal, we wanted the point to always be accessible so we simply made it safer to be on via height advantage and adding room to dodge fire.

Other Observations
Steel has a dynamic chalkboard that communicates the state of the level. while it is helpful, I would not say that it is completely necessary. The information displayed on the chalkboard is mostly redundant with the HUD. The one thing the chalkboard is useful for however, is how it presents some spatial context for how points are laid down and how they connect to each other. This helps the player in forming a mental map of the level. This informed the design decision to not have a dynamic chalkboard in cp_hadal and display a static map of the level instead.
(I'm going to hazard a guess and say that steel's chalkboard was inspired from TC overviews displayed at the begining of each rounds. It was a good inspiration to draw from but I feel like it may have been slightly misguided)

A Note On Dynamic Elements
Dynamic elements are elements that respond to player interactions. However, the player is required to understand those interactions in order to be succesful within the system he is presented with. If you want to look into it, this is a design principle called affordance. Dynamic objects need to communicate not only that they can be interacted with but also how. This is something that Steel is really bad at and can be improved upon.

-Steel's doors, while fairly straightforward, don't communicate that they are interactive. What's more, they don't share the same visual style so the player can't associate a type of door to a type of behaviour. This is addressed in cp_hadal by having all dynamic doors display a lock symbols that clearly indicates the required interaction to open the door.

-In a similar way, There is no clear indication or narrative to cp steel that would explain why the central point is transforming with each captures. This is addressed in cp_hadal by visually connecting the main point to each control points with colored pipes. Connecting the objectives visually helps both in understanding the interactions of capturing points but also in understanding the layout of the map. The same could probably be achieved with wires, or some mechanical contraptions, be creative.

Final Words
Try to rely on map knowledge as little as possible, make something that is intuitive and foolproof. Don't be afraid to be handholding the player throughout. Steel maps are very intimidating at first. Avoiding that should be your main concern.
 
Last edited:

leprecan

L2: Junior Member
Feb 10, 2013
96
89
Although you do touch on the locking doors later on as something not conveyed properly, I think it is worth mentioning earlier in the segment about each capture point rendering E safer.

While D is boring and arguably out of the way for BLU, capturing it shuts down 2 of the easiest routes to E from RED spawn, drastically lengthening RED's rollout times, forcing subpar spam angles and cements the game for BLU.

Despite being the second to last point and therefore not that interesting, it is still very important. Given that the most popular Engineer spot for defending E also relies on support from those two same spawn doors, and can often rely on easy access to spawn locker ammo, it hurts that Sentry spot a lot.

Also, I don't know if it was a mistype, but you wrote that the bridges extending helps Red, when the bridges for the most part only help BLU. The cap time on E is so long that standing on the point isn't an efficient way for RED to defend.
 

YM

LVL100 YM
aa
Dec 5, 2007
7,135
6,056
Rule 5 is 100% bullshit. You've taken one of Steel's poorer features and spun it into a plus that all maps should copy. Making boring points? What?

Don't ever make any part of your maps forgettable.
 

Fruity Snacks

Creator of blackholes & memes. Destroyer of forums
aa
Sep 5, 2010
6,394
5,571
If you mean "More simple and less complex" for #5, then I may agree, but thats a rule for all points, regardless of order. It feels like maybe you've used poor wording on that one?

As YM said, never intentionally make your map (or any part of your map) more forgettable. Each point should be unique, memorable and interesting. Most importantly, each point should (ie needs) be fun! No one wants a dull point to play on, as thats, well, dull. Additionally, if anything, the final point should be the MOST interesting point, and the primary focus of the map, at least visually.
 

fubarFX

The "raw" in "nodraw"
aa
Jun 1, 2009
1,720
1,978
sure, you can make all your points interesting but rule #5 is one that works for steel. I guess my point is, not all your points are going to be of equal strength, if you're looking for a way to order them, that's a good way to go about it. but yeah, I stand by saying D and C are progressively more boring, I just don't think they are as exciting and I think it helps with the player's decision making.

and yeah, thanks to leprecan for pointing out red's spawn getting worst for the final push, that had slipped my mind.
 

Yrr

An Actual Deer
aa
Sep 20, 2015
1,308
2,743
I think that can still be done without sacrificing the enjoyability/memorability of the points tbh

ie gradually making the routes to last more appealing the further you progress

I should want to assault last because its strategic, not because i dont want to play the next point
 

radarhead

Basically? Kind of a Huge Mess
aa
Mar 6, 2013
1,045
625
I 100% disagree. I have literally no idea what steel's D looks like, and that's bad no matter which way you spin it
Steel's D is the dark room with the grating on the floor and the missile silo up above. It also has a water tank that's floating above the ground and clipping into the ceiling. RED spawns close by and enters via a balcony. I recall all of this without looking at a reference.
 

Yrr

An Actual Deer
aa
Sep 20, 2015
1,308
2,743
im pretty sure YM was just making a point and didnt require a description
 

LeSwordfish

semi-trained quasi-professional
aa
Aug 8, 2010
4,102
6,597
I think Clam was just making the "just because you don't remember" counterpoint. For my own part i can recall D fairly clearly.

Making a point less interesting to dissuade players from it seems a little peverse to me - rather like writing a particular part of the game manual in latin, or placing a minor trigger_hurt over the whole area. Surely we're sufficiently skilled to point players in a direction without just making the alternative less fun to play in?
 

Fruity Snacks

Creator of blackholes & memes. Destroyer of forums
aa
Sep 5, 2010
6,394
5,571
You should never discourage someone playing an objective/area/yard by making it 'boring' or 'dull.' You discourage them by redirecting them. Redirect through art and design (but not by boring design). Redirect them to the action. Redirect them away from routes that could take them to places you don't want them to be at.

Just because 'thats how steel does it' doesn't make it right, there's some good things from Steel, which you've highlighted, but this is not one of them.
 

fubarFX

The "raw" in "nodraw"
aa
Jun 1, 2009
1,720
1,978
It might not be the best way to go about it, but as it stands D looks less important because it is. I think that works out just fine. that's the map telling you "capturing points is cool but E is the big money".
And just to be clear, all of the aboves are just observable rules that can be extracted from steel, you can take them or leave them (rules are meant to be broken after all).
 

Yrr

An Actual Deer
aa
Sep 20, 2015
1,308
2,743
i agree the points should look less important as they progress but imo importance shouldnt be confused with interest

less important points should still be interesting since all strategies* should be viable and so players should still be rewarded with an interesting area for pushing all the way to D

*A->E, B->E, C->E, D->E