FORECAST: Simultaneous Turn-Based Tactics

Hi all!

I wanted to talk about 3D Environment Workflows and touch on Level Design, building Requirements and using References.
Ths DevLog touches briefly on all of these points, but I'd love to learn how everyone else goes about this. I didn't want to make the video too long so it's a fairly brief overview on how I go about my own design process - with no in-depth breakdown.
Below I've tried to explain my thought process in a bit more detail and how I tackle level design. Splitting it into 6 key stages:

Gathering/Building Requirements
I think this is something everyone does instinctively, but by taking it to a more active level we can improve the consistency of our work and ensure that our levels convey the feeling we want, to the players.

a. Gameplay Requirements
Forecast is a very unique turn-based tactics game, every round is simultaneous and all spells are directional. Because of this our level design needs to be clear and concise, I'm trying to ensure that the levels created for Forecast take this into account.

I handle this by understanding how many potential conflict zones a map has, how many path choices are available, how wide paths are and how many rounds it takes to reach the opposing team. Changing these factors allows me to change the gameplay of a level, a more open level creates a greater deal of uncertainty - predicting your enemies becomes much more difficult and it forces players to be a lot more careful and thoughtful in their planning phase. On the other side, smaller maps lead to more immediate conflict and promote aggressive playstyles.
I create a storyboard of references in PureRef which I refer to when blocking out the level and when creating materials in Substance Painter. So far my Storyboard always feels too small so, the bigger the better!
( I get references from all over, but Artstation tends to provide some very high quality art with a good understanding of materials )

b. Story Requirements
It's important to remember that while the level needs to be fun to play, it also needs to tie into the story in some way, especially if working on an RPG or single player game. Currently I do this by considering the surrounding environment (location) and materials used (setting), I'm trying to get a better understanding for my hero pieces, ensuring that there are elements inside each level which will serve as story anchors in future.
Personally this is an area I'm trying to improve on, it is less tangible than Gameplay Requirements and unless you have a fully written script or planned campaign it feels difficult to line this up. My current focus has been on interesting levels that fit the game's tone - as we progress in development I would like to take this up a notch and cater levels toward lore events also.

Both the Gameplay and Story Requirements help me decide on a few key factors, these are:
1. The size/scale of the level
2. The location/setting, be it indoor/outdoor, in a tundra or a sinking pirate ship
3. The materials used, from the story requirements and location/setting I can determine which materials will be needed
There are many more considerations such as lighting, colour balance, contrast and points of interest, however as the sole artist it is difficult for me to focus on all of these points, so the 3 listed are the centre of my attention.

A note on a project specific level: Forecast's combat can be incredibly busy and chaotic, spells tend to be flying and exploding everywhere, most of the time. Due to this I try to keep the saturation down, relying on muted and pastel colours in my environments to ensure that the spell effects do not have to contest with the environment for visibility. I want the spells to take centre stage and not get washed out.

Block outs / Grey Boxing & Playtesting
Once I've determined what kind of level I want to make and some of the key attributes, I create a quick block out in Blender, matching the scaling with the engine (UE4 in our case). Importing this into a new UE4 scene and setting it up for playtesting shortly afterwards. We built Forecast's core combat system and a couple of characters with spells first, so that we could playtest the rest of our mechanics and content going forward. As such, it is relatively easy and painless for us to throw something together and test it.

From the playtesting I determine if the block out fulfils my Gameplay Requirements defined previously. For the Portal Dimension this was to create a smaller level (but not so small that you can gib the opposing team during Round 1) with limited movement options, and with varied terrain elevation (I wanted to experiment with this and see how it felt mechanically).
This phase has some back and forth, but if your core gameplay mechanics are in, this becomes much easier and faster. In future I would like to take advantage of our networking and have a core of players who playtest block outs and provide feedback on what layouts they found most fun or interesting to play on. Quite an ambitious objective, but I think it would be awesome!
Once I'm happy with the block out, I go back to Blender and begin separating out the individual pieces and segments of the level from the block out and build them into more detailed, low-poly meshes.

Low-Poly Modelling
There isn't a great deal to say about this phase. Story Requirements focused, considering which Hero Pieces to create to help me convey meaning and story points or build on the game's setting.
It's also important to consider what materials will be used as that will greatly affect the geometry of the assets.
I create High-Poly models (just bevelled and subdivided) to control edge flow when baking my PBR maps in Substance Painter. I do this for pretty much all of the assets in the level unless it's not required. Usually this is just a quick touching up of the corners/edges, but often I sculpt some wear into bricks or natural surfaces. I prefer to handle normal map details in Substance Painter by creating a Height layer and baking it into the normals later, this is largely because it's quite precise and easy to do, and I don't have access to ZBrush nor the time to learn it on top of developing Forecast and my fulltime job. Though I would love to get into it ...

Material/Texture Creation in Substance Painter
This stage also focuses on story requirements, setting the tone with material creation and ensuring the right level of colour balance / contrast to keep hero pieces & points of interest in focus. For the Portal Dimension I tried to be a bit more bold, more reflective surfaces, slightly more saturation and more pattern details. Overall I like the look I achieved, but I feel that it's a bit too much at times, though my personal preference artistically has always been on more subtle tones.
I try to work out the core materials needed for the scene and frontload my time on creating customisable Smart Materials to help me quickly texture the rest of the assets.

UE4 & Setup
After I have created all of the individual assets I set them up in UE4 using the Low-Poly block out in Blender as a reference, everything scales 1:1 from Blender to UE4 for us, so this is a pretty quick process.
Next I work on the rendering of the scene, adding more detailed lighting and highlighting key elements in the scene. I adjust post-processing effects and add reflection spheres as necessary. The Portal Dimension is almost entirely marble so there are pretty much spheres everywhere.
My folder structure is pretty basic, I have a "tile-set" folder for each environment, split into Meshes, Materials and Textures. I find it's easy to find stuff from individual levels if I want to reuse them later this way.

Polish & Final Playtesting
When that's all complete I throw down map boundaries, invisible walls, correct collision on individual meshes and ensure that Unit's path predictably everywhere and can't do anything weird. I also check that there's nothing breaking in the level such as being able to path out, spells colliding with things they shouldn't or material effects overriding each other (for example we had a water plane with a higher priority than our spell effects, particle systems weren't visible if they overlapped the water plane).

Finally I shout at the developer to implement anything leftover, and then I agonise over missing details before moving on to the next task.

I hope this has been informative or at least interesting to some of you! This isn't THE way to structure your design process, it's just what I'm doing currently. I'm constantly looking to improve and would love to hear your feedback, how you tackle it and anything you think might be worth sharing.

My goals going forward are:
1. Spend more time gathering and looking at references
2. Improve my understanding and usage of lighting in UE4, currently my lighting is very basic and it's such a powerful tool in level design so I really want to up my game there.
3. Understand my materials earlier. I often find myself changing my mind on materials when I'm at the texturing phase, it would save time if I settled comfortably during my reference gathering phase.

Thanks for reading!

Quote

What is Forecast?
Forecast is an ambitious fantasy turn-based tactics game packed with thought-provoking combat, experienced through dozens of playstyles. Journey through the afterlife, gather magical beings and grow the forces of your very own pocket dimension to aid you on your quest for revival.

Footage?
Check out our Youtube channel for gameplay!

How does it work?
The core gameplay of Forecast focuses on a planning phase and an action phase. Simultaneous means that both your team AND all opposing teams will carry out planned actions at the same time. All attacks will be in the form of powerful area or directional spells. There is no hit chance, so no missing those 99% shots! If your spell collides with a unit, it's going to feel it.

How can I stay in the loop?
➤ Website: https://www.forecast-game.com/
➤ Twitter: https://twitter.com/ForecastGame
➤ Youtube: https://www.youtube.com/channel/UCqwOzws8qy7LdnUrNtuwQ3g
➤ Reddit: https://www.reddit.com/r/ForecastGame/