So, here’s where it’s at with Scout’s Journey development as of now. I apologize in advance for the wall of text. I hope it is worth reading. A better format may be forthcoming.
Game development has been continuous from 2012 until today and is still going strong, with a recent milestone reached in that the plot is completely written. Money is an issue, but that has always been the case and is no blocker. My health is not optimal, and something is being done about that. Also no real blocker, but bothersome and has definitely impacted the pace of development in the last 3 years. However, progress is made all the time.
A major problem is just sheer loneliness; few people in the indie game scene are doing what I am doing, although reports keep popping up of people finishing games after near-decade-long development etc., so others are out there. But day-to-day exchange with similar projects is rare because they almost don’t exist. People from a modding or general indie background often flat out don’t understand what the project is: A standalone story-based first person 3D game with combat, stealth and exploration mechanics. Such a thing is incredibly rare in the indie game world.
- Game started out as a collection of ideas during work on Remake Quake, a Quake mod, from 2007 to 2012. Lots of these ideas didn’t fit the Quake mold.
- SJ became its own project in the summer of 2012, running on a variety of idtech-based engines (RMQengine, Warsow engine, FTE, Darkplaces etc). Tests were also done on Unity relatively early.
- Main asset development is in Blender.
- A dedicated codebase (gamecode) existed for the game while it ran on the idtech engines, and was developed relatively far, especially the RPG aspects (interactive GUI, novel gameplay mechanics, custom all-original NPC dialogue parsing etc were all functional).
- Lots of assets were developed, including two full levels of a scale much bigger than Quake (later divided into sublevels for performance), two more gigantic locations that were leftovers from RMQ, a full weapon lineup, props etc.
- Levels were converted from BSP to mesh with meshcollision relatively early.
- Terrain modelling methods were tested and refined.
- The game went to combination lightmaps/realtime lights pretty early.
- Character modelling has begun but was put on hold when it turned out that the story side of the game was a mess. (Also, that is hard.)
The idtech engines were abandoned along the way simply because I deemed them not reliable/ rugged enough for this type of project. I still believe that was the right decision. Unreal is the goal, but from the end of 2013 and the spring of 2014 onward, most work was done on the storyline, with art and music on the side. I simply needed to bring up the rear, instead of blindly marching on with levels and assets, although those have not been abandoned.
It is simply the case that level design needs to proceed in accordance with the story. If the plot says “Scout goes from here to there”, the levels need to be laid out in the same manner. And if the script says, “Scourge Martyrs attack”, then concept art needs to be made and gameplay mechanics and weapons need to be developed for that enemy type.
This is why the script needs to be in place BEFORE levels and monsters etc. are finalized. And lots of game projects don’t understand this in time. I believe Rhianna Pratchett has mentioned this – you cannot graft a story onto a mostly finished game, it has to be done story first.
A game design document was compiled from roughly 2012 to 2015, ever changing especially in the core gameplay department. Broadly speaking, the game turned from a first person shooter into more of a stealth and exploration game with combat largely optional.
The quake-style encounter design was completely removed and a dynamic runtime-based patrol and AI group system was substituted. In plain English: The levels are populated with patrols and checkpoints etc. by the program during runtime, and enemies can dynamically call for reinforcements or do a tactical retreat. This suits the stealthy gameplay much better. Fundamentals of this were actually implemented in QuakeC already, with the AI controller being nicknamed “Mother” after the computer from the Alien movies. Large influences were the ALife system from Stalker and some work done in RMQ on randomized monster placement.
Writing is incredibly hard to master. (Most people have probably heard of the 10,000 hour rule.) Here’s a chronological breakdown:
- End of 2012, a loose plot emerged from the sequence of locations.
- During 2013, a cast of supporting characters developed.
- Factions and their relations developed in the game world.
- 2014-15, I started to get a grip on the entire writing thing.
- Lots of feedback was received from “beta readers”.
- Consequently, a lot of rewriting happened.
- 2016 – 17 was a time of constant gradual improvements.
- Two (!) artificial languages were developed.
- Some holes in the plot were incredibly hard to fill.
- The protag, Scout, was a difficult character to write.
Only recently did everything line up. The format of the script changed several times, because there really is no standard in game writing. This is uncharted territory. The beta readers and writing buddies were incredibly helpful here.
Would the real Scout please stand up?
As for Scout herself, she originally was supposed to be a female version of Ranger, the Quake protagonist. Then the entire Quake mod thing fell apart. Suddenly, Scout became a person.
She wanted to talk. She started to take on a life of her own. So did other characters. Scout does some outrageous things in the story, such as coming back to life (twice), meeting the gods, having a near death experience, travel between worlds and outside of them, encounter her childhood self, and set off a number of conflicts among the various factions, turning friends to enemies and vice versa.
In the beginning, she was overly passive and lacked agency. Getting her behind the wheel, and having her be an active protagonist and a moving force, was a piece of work. Getting in her head and understanding what makes her special was incredibly difficult. She is different from the usual game protagonist. Even her ingame interface is different from other games. Even the controls differ (WASD is still there, but using tools and weapons is done differently).
Scout carries the entire game though, mechanically and plot wise, so she needed to be rock solid. I’m not sure there is any other game protagonist quite like her.
Over the course of writing, it became clear that not only Scout transcends video game (and human) standards, but the game itself transcends the shooter and stealth genres. Antagonists can be stabbed and slashed, obstacles can be overcome, but Scout’s major personal problem can neither be dodged nor shot to pieces. The end of the script also obliterates the difference between interactive and noninteractive content. It doesn’t even matter anymore.
I’m deeply wondering about how we can really tell stories in games. What the narrative structure can be. How interactive and cinematic toolkits can be combined in this medium.
Practically speaking, it is hard to overstate the importance of the recent milestone. A few things have to be plugged from the synopsis into the main script yet, and some dialogue has to be written out, that kind of stuff. But we’re really on top of things here. We have climbed the mountain. I still have a hard time believing that. It was such a slog. I keep expecting the next sheer cliff face, but I’m not seeing any. I feel like I’m going back to art and asset production soon-ish.
Some other stuff has been accomplished besides writing, namely music work and some mechanics polish as well as research and experimentation. For instance, I have a concept for loot boxes that work, giving you the excitement of opening surprise boxes but without the bullshit monetisation schemes attached. I’m very closely watching things like that.
I’ve been thinking about making videos for a while. I realize walls of text are hard to read. Bear with me.
Help is always welcome. I’m having to cope with a lot of isolation. I barely have people to talk this stuff over with. If you want to read some of the script, look at some of the mechanics, need some qc code, want to talk level design, donate old hardware, man, I would be glad to oblige you.