"I've given up making estimates."
Styg, developer-extraordinaire of Underrail

Every now and then, someone asks, “What’s taking them so long? Why is the game still in development?” The questions are understandable. You hear that a game has been in development for more than a decade and you think that a decade is a very long time. Only the infamously mismanaged projects like Duke Nukem Forever take that long and nothing good ever comes out of it.

Since we’ve just released 2 new locations – 21 out of 22 locations are now available – and the light at the end of the tunnel is shining impossibly bright, now would be a good time to tell you a tale.

***

It starts deceptively simple.

You’ve been imagining your own RPG for a while. One day you decide to take it a step further and start toying with a character system in Excel. Then you start jotting down notes on the setting; maybe you even write a few quests as a mental exercise. Why, you’re half way there already! All you need is an engine to scotch-tape it together and you’re done!

Where there’s a will, there’s a way. Months later you have an engine, one or two likeminded, overly enthusiastic and always inexperienced (if they were experienced, they wouldn’t be there) people, the very first, hastily created screen to reinforce the notion that you’re in the business of making RPGs and the business is a-booming.

You make plans. Big plans. I want my dream RPG to have a crafting system! And an alchemy system! And a reputation system! And three factions with mutually exclusive questlines! No, make it four! Four? You have to think BIG! Make it seven!

Shortly afterwards, the honeymoon phase of imagining an RPG ends and the hard work of making one begins. That’s where most new undertakings wither and die. That’s where you discover how insanely complex RPGs are and that it would take a couple of years just to put together a functioning combat system.

You’re stubborn or stupid (or both) and you push forward. After all, when the combat system is done, you’ll almost be there. Not really. Well, if you want to make a combat game and you don’t mind cutting a few corners, maybe. Make some maps, sprinkle them with enemies, and your masterpiece is complete!

But what if you want to make a game with a complex story and branching questlines? What if you want to a make a game where Choices & Consequences aren’t an afterthought but the main feature? Then you’d best get ready for the long haul because the combat system is only the first step on a very long road.

***

We started working on the game in March 2004. I don’t know why I picked this particular date as the official start date, because ‘started working’ is a very vague and misleading term when it’s just a couple of guys across the internet, taking the first steps toward learning the trade of making indie RPGs without any budget whatsoever and making mistakes to learn from.

In 2005 we realized that our homemade 2D engine won’t take us all the way to the finish line and switched to Torque Gaming Engine (TGE). I’d say we chose it for its robust scripting system, but back then there wasn’t much to choose from. So we switched to the only available 3D engine that happened to have a very robust scripting system. Best $100 I’ve ever spent.

At that point we still had nothing but ideas, some models, crappy textures, a couple of quests, and a character that could march up and down the map – the living proof that we were making an RPG. One look at the poor bastard and you knew: this here is an RPG in the making.

Like any business Fellowship of an RPG, we needed a strategy:

1. Announce an RPG
2. ???
3. Profit!!!

How to get to point 3 wasn’t immediately clear, so we decided to make an RPG in 4 easy steps:

1. Character system
2. Item system & Inventory
3. Combat system
4. The rest

Basically, to keep the enthusiasm going, we needed a sense of progress, aka something playable. To have even a tiny map where two characters can fight to the death over and over again, you need a rudimentary combat system in place, and to have a combat system in place, you need a character system and an inventory system with equippable items. Sounds easy? It’s anything but.

It took us 3 years and 5 months to get there (keep in mind that we’re talking about working part-time, a few hours after work plus the weekends). On August 30, 2007 we posted the very first combat video, which was the first real achievement that every wannabe developer dreams of – his game in action.

The next step was a combat demo. You’d think it’s a very small step from the video but it’s not. A video needs basic functionality to give an overall presentation and show the main features of the combat system. A playable combat demo (the arena) needs:

• A fully realized combat system with all the attacks, bells and whistles.
  
• A character generator.
  
• More or less decent AI
  
• A hundred or so items with balanced stats, metal upgrades, descriptions, etc.
  
• A way to loot corpses and transfer the ill-gotten gains into your weight-based inventory
  
• A store where you can sell and buy things
  
• A dialogue system for basic instructions and scripting (move from one fight to the next, remove old corpses, heal the player character, close the gates to prevent the player from running all over the place, etc)
  
• Save/Load functionality
  
• A decent level, like an arena, for example
  
• Ending slides
  
• Animated spectators cheering for you when you land a good critical
  
• A menu, help file, TONS of small things.

We released it – our very own playable demo! – on Dec 25, 2009. That’s 2 years and 3 months later for those who’re counting. We started beta testing it in September 2009 if I recall correctly, so 4 months of it was testing and tweaking. Still working part-time as at that point we didn’t make a dime yet.

The next step was a fully playable demo showing not just the combat system but everything: the setting, starting town with about 30 quests, multiple solutions, choices & consequences, dialogue trees, crafting, journal, etc. It’s pointless listing all the things it required for their name is legion.

Once again (and not for the last time), it redefined our understanding of the complexity of the game we were trying to make. Once again, it took years.

The demo was released in March 2012, after being tested by about 40 people for 8 months. The release highlighted a lot of issues and mistakes, so we spent the next 8 months processing feedback, changing and redesigning things that didn’t work well.

When you team is small, you can’t delegate these things to someone else. You have to stop working on the other areas and go back and start making changes to what you thought was already done, making you feel that it’s a never-ending story and you’ll be stuck forever in this form of development hell.

In the end, it took us a year and seven months to do the first 3 locations (the starting town and two nearby locations) and then it took us sixteen months (the beta test and post-release tweaking) to get it to the point where it was actually good. Combined, that’s three years of work.

Overall, it took us 8 years (!) and 7 months to get to the stage where we had 3 playable locations and all the systems (character, combat, dialogue, crafting, alchemy, items, and journal) in place. At this point any idiot would have realized that it’s going to take a LONG time to do the remaining 19 locations, but we aren’t just any idiots so we kept clinging to hope that NOW things are going to go faster.

We started taking pre-orders when we released the second playable demo. We hoped to sell a few thousand copies but sold only 242 copies in the first month, less than half in the second month, 25-40 copies a month after. Thus we had no choice but to continue working part-time.

We dubbed Teron and two locations Chapter 1 and started working on Chapter 2: the second city and 5 nearby locations. With all the tools and assets in place, it took us about a year plus a 2 months beta test – a noticeable improvement over the first Chapter.

We released the game on Steam (aka a MUCH wider audience) in Nov 2013 and again spent the next several months processing the feedback, improving the game, and adding things that should have been there in the first place, like a tutorial, tool-tips, feedback mechanisms, etc. We kept paying attention to feedback and suggestions and improving the game. Those who remember our very first release in March 2012 (R1) can testify how far we’ve gone and appreciate the difference between R1 and R9.3 (the last update).

We released 5 more locations in April 2014, another one in May, and started working on Chapter 3: the third city and two nearby locations. We hoped for the best and aimed for the end of the year. On paper it made sense. We had a pretty schedule for every item on the list and all. Not surprisingly (hindsight is always 20/20), it took us a year to do it, including 15 weeks of testing and tweaking.

Now we know three things:

• With everything in place, it takes us a year to do a city with 30+ quests. No matter how much you want to do it faster, you can’t.
  
• It took us roughly 3 years to put together 16 locations. So we can probably make a 15-location game in 3 years, now that we have all the tools, the systems, and (most importantly) experience.
  
• Patience is a virtue.

We were asked if ‘starting big’ was a mistake. In general, I’d say yes and agree that starting small is often repeated, yet often ignored advice. The appeal won’t become obvious until you spend several years and produce a lot of stuff with little to show. Of course, the downside is that a ‘small’ rpg would still take years but won’t offer much. If you set out to make your own RPG, might as well make an RPG you want to make, but be prepared –it will take the better part of a decade, if you’re lucky.

The hardest part for me was getting to the first 3 playable locations, which took 8 years, and realizing that a) we HAVE to spend months to make it better and b) it’s only 3 locations out of 22. It took us another 3 years to get to 21 locations and it’s a very surreal feeling. I knew we’d get there one day, but knowing is one thing, being there is another.

Once again, thank you for supporting us over the years. I don’t exaggerate when I say that we wouldn’t have made it this far without people supporting and believing in us.

PS.

So what does a decade of work get you?

112 hand-scripted quests
Over 520,000 words (average book is 100-150,000 words, the first Harry Potter book has 76,944 words, Fellowship of the Ring - 177,227 words)
Over 100 named characters
Over 750 characters with unique IDs
Over 500 animations
Over 300 items (not counting metal upgrades and variations)
30 different attack types (split between different weapons and items)
1 well developed setting