Fortnite Creative
Working with Fortnite Creative
The Setup
When I first decided to use Fortnite Creative, I knew exactly what I wanted to design - at a high-level. I delve deeper into my thoughts on Fortnite in my blog which can be found here. Essentially, I have never really enjoyed the building aspect of Fortnite. My primary design goal is to recreate the feeling I used to get from games like Doom, Duke Nukem 3D, Unreal Tournament, and 007: Golden Eye, but using Fortnite's modern graphics and feature set.
My goal was to create a simple, fast-paced, arms-raced deathmatch.
During my younger days of playing first-person shooter games, I loved getting into a rhythm. There was this feeling that if you figured out the nuances of the level, you could recognize sounds and utilize all of your knowledge to better hunt down your prey. There were moments where it felt like a rhythmic experience, where you could potentially rack up kill after kill if you kept the right momentum. Timing your jumps, turns, and shots felt great. The best levels were large enough so that you could apply your skill, but small enough so that encounters were regular and plentiful. Balance was always essential, alternative choice of weapons and escape routes needed to be present for the best experience. It was okay once in a while to be instantly killed on spawn, but if it occurred too many times, the game became less interesting. On the same token, the best player should always be susceptible to death at any moment. Even the weakest gun should do a decent enough amount of damage so that the tradeoff of frags vs. deaths always remained in a fun ratio. Losing 3:1 still feels fun, while losing 10:1 isn't as exciting. I think the same holds true for winning.
Design goals:
Deathmatch retro-style map
Map should be made for 2-6 players, and best played with 3-4 players.
Design will follow the principle - simple but interesting.
Provide choice to the player. ie; I can go for the Shotgun or the Sniper, but not both at once.
Balance is the most important feature of the map.
A player should always feel both a sense of control and vulnerability.
The map should allow for strategy and progressive learning.
Interactions should be plentiful, but not always chaotic.
Recreate the feeling of Doom, Goldeneye, Unreal Tournament & Duke Nukem 3D - no building mechanics.
A Small Tangent
With my interest and progressive understanding of level and game design, I also have studied board games quite extensively. I find excitement by attempting to step back and decide what makes a particular game successful, fun, and ultimately work or not work.
Whenever I think about balanced gameplay I quite often think about Catan. Catan does a great job of balancing gameplay so players always feel like they have a chance to win. There is nothing worse than losing control in the first 5 minutes of a game and sitting through the next 25 minutes with the equivalent value of a spectator.
Catan solves this problem brilliantly in two ways. The first way is by curving the difficulty. If the game is broken down to a race to 10 points, a simple difficulty curve allows for the first several points to be easily gained, with the remaining points increasing in difficulty to achieve the closer the player gets to the end goal. This prevents players from frequently having a runaway victory.
The other method (which is intertwined with the first) is the use of perception. Consider the scenario where a game of Catan ends with the winner having 10 points, while the losers have 7 and 8 points. The two losers of the match might proclaim it was a close game. Thinking back to the difficulty curve concept, if the 10th point is 10x harder to achieve then the 9 previous points, how close are the players actually? This is a bit of slight of hand, and when executed well and goes unnoticed it's a great way to keep everyone engaged in the gameplay.
I'm pointing this out because I'll be considering these concepts when thinking of ways to keep the deathmatch level balanced and the frag ratio's modest.
/tangent
Design Start
I'm very interested in tech and workflows. Part of my creative process is dabbling with new ideas and testing different workflows. A bit brain dead on exactly what I wanted to do for the map, I thought I'd leverage AI to help kick start my creative process. I decided to leverage both Chat-GPT v3 and Midjourney v5.
First, I started with Chat GPT by asking it to generate a few high level concept level themes:
Next, from this suggestion I landed on a general concept of Haunted Forest.
At this point, I'm not sure how securely I'll stick to the theme, but at least I've got the creative juices flowing.
Lastly, I engaged into further back and forth with my AI friend and ultimately ended up with the following concepts:
OK, so now what?
Over to Midjourney to stimulate my brain, visually.
Prompt 1
"a top down view for level design, the level includes 5 areas of interest connected by different sized choke points --v 5 "
Prompt 2
"top down level design map of a haunted forest, includes 5 areas of interest a bog, a witches den, a broken obelisk, an abandoned cabin, and an alter of spirits"
The final product
The final version, an upscaled variant of the 4th map from Prompt 1
The map I decided on solved a few problems for me. First and foremost it was interesting. With varied terrain, map width, choke points and areas of interest it offers the designer so many creative options to help the player explore. I can already ready envision various height changes, vantage points, surprise turns and escape routes. It is also condensed, meaning in terms of scope and size it fits well with the "simple but, interesting" design goal from my initial plan.
I think we can work with this. Lets get it into prototype.
The Prototype
When I first started this project it was roughly 1 week prior to UEFN being released. UEFN is the Unreal Editor verion of Fornite Creative. Prior to UEFN all editing needed to be done through the gamplay style editor which left some significant challenges in map development. Simple things like accurately placing content, sizing content and navigating the map were less accurate, slower and overall tougher to iterate on compared with UE5.
I did create the first prototype of the level using the original creative editor. Being new to the tool I wanted to establish workflow that would help me to block-out my first concept.
The first thing I did was explore the creative tools for something I could use to help me to block out the level. Given the limitations of the existing editor I was a bit concerned about the effort to swap out the block-outs for final props. Thanks to UEFN this is no longer going to be an issue, swapping block-outs for meshes can use existing UE workflows.
Poking around in the prefab galleries I found the "convergence group" and quickly realized I could manipulate the existing blocks into custom sized blocks. Using the native functionality of the original creative editor I could then duplicate and copy into larger custom sized walls. This would be a reasonable method for blocking out the level.
The "convergence group" galleries also included prefab chunky sets of blocked areas with varying heights, entrances, and cover. I decided that using these prefabs would allow me to get to a playable prototype where I could then evaluate early gameplay.
To make the map playable I adjusted several of the island settings (removed building, added player spawn, adjusted the UI settings to allow for frags, etc.). I then added player spawn points and a few weapon variants to 5-6 locations on the map.
Enlisting the help of my play testers I launched the map for out first deathmatch and you know what?
It was fun... A lot of fun!
Me and my 2 daughters played for a couple of hours racking up frags and finding different ways to sneak up or hide on each other. Lots of laughs were had.
The most significant takeaway was that I was on to something regarding the simplicity. It didn't matter that the map was made up with mashed together blocks and assets, the core gameplay loop needed to be quick, interactive, balanced and fun. The most significant issue I needed to fix however was the sheer size of the map. Running from one end of the map to another was painful. The size of the rooms were extremely large and finding an opponent sometimes proved difficult. At times you could go several minutes without an interaction - not what I'm looking for.
Lets iterate on this a bit more.
Refinement
UEFN is a huge time saver. As mentioned earlier, UEFN came out while I was in the middle of designing this level. I'm not sure exactly how much longer this would have taken me, but I'm thinking days vs. hours for sure. My knowledge of UE5 came in handy here, so using landscape editing and block-out mesh placing techniques I was able to resize and redefine the level in less than a few hours. I was able to reduce the size of the map substantially, playtesting as I went to ensure time between pick-up areas and chokepoints felt right. Have a look through the next couple of images (Images 8, 9 & 10). I I was able to create a reasonable resemblance to the originally planned top down 2D view.
Detailing and Gameplay flow
Since the basic outline and major areas of the map had now been translated into 3D form, I decided it was time to start focusing in on some of the finer details. But... where to start?
Referring back to Image 10, I figured there wasn't much sense spending time adding visual details or immersive experience design if the map flow hadn't yet been addressed and confirmed. I decided there would be much re-work of visuals and design elements if I had to adjust any of the major landscape design or locations.
This portion of design included a lot of playtesting and re-design work. Every change needed to be play tested (even if solo) to confirm proper "feel" and timing. I created the map barriers and ensured players couldn't escape or get stuck between the barriers and landscape. It was also important to ensure the barriers weren't confusing to the user. As a general principle, I didn't want users to be surprised when they ran into a barrier. It should feel expected or natural, as opposed to limiting and surprising.
I used different techniques to create excitement and force tension through speed adjustments.
First pass (whitebox)
In game design I try to follow the rules of 3's and 7's (& sometimes 5's). You can read a small bit about the rule of 7 here. It's the main reason I decided to keep the game to 7 rooms. Referring to Image 10 "Game Beat Design" - in a notebook I mapped out a quick flow chart or "game beat" to better understand the flow of the game and confirm a few things I had noticed in my play tests. After laying it out it became clear that the decision matrix and pacing in sections 1, 2 & 3 needed some work. The pink highlighted lines indicate long travel paths & potential rest periods. This could lead to slow loops early in the game play. Additionally in sections 4 & 5 it is a requirement to utilize path 2b, further introducing interrupted and slow progression.
Whitebox notes
In order to kick off the second iteration of design I created a list of positives and negatives. The intention is to compare the positives with original game design intention while using the negatives as a laundry list of items to resolve in the next design pass.
Overall the level was looking relatively good for this stage. It was playable and I had met a lot of my original design intentions. There were a few glaring negatives I needed to resolve however... I'll summarize of a few of the positives and negatives to give a general sense to the approach.
Opportunities (cont.)
The list of negatives included a few simple opportunities. Consider Images 15a & 15b - The intention is to use the laser to activate the tractor beam which enables the player access to the next room in the level. In the "Bad" design choice I didn't consider using the glass room to show off the activation of the tractor beam. While the player would eventually likely figure out what was altered, it's an unnecessary difficulty that takes away from the game beat and goes against the glass room theme. Conversely a small change in position of the laser receptor makes it obvious what was changed and gives the player a sense of instant reward and purpose. "I'm smart AND I know what I need to do next".
In the same context consider images 16a & 16b - Here we're highlighting the vent traversal. It's not necessary to make this overly complicated for the player. It's simple and fun to jump through the vents, and gives the player a small break. With the original poor design choice it's not clear what we want the player to do. In the better design choice simple wall lighting guides the player.
Further to the initial lighting changes, additional lighting is placed down each hallway so when looking at the ceiling the player can identify which direction the next vent leads - meaning they have more intuitive control and decision making when placing their ceiling portal.
Completion
The final product.
I've uploaded a play-through of the level. Based on the size and amount of puzzles, it would be impossible to highlight all of paths and nuances throughout the gameplay. I've opted to show the shortest route through demonstrating walk-through playthrough from beginning to end.
If you'd like to try out the map for yourself you can find it here.