Thursday, January 04, 2018

After Death

The CalFree in Chains launch went really well, honestly better than I was expecting. I've been pleased both by the number of people who have played it and their reactions. It's now been over two years since Shadowrun Hong Kong came out, and I was mentally prepared for a graveyard reception; instead, after less than two weeks out, over one thousand players have already subscribed to my campaign. As with all of my projects, I'd come to kind of hate it in the immediate run-up to release, and it's been a relief to see people appreciating its virtues instead of the flaws I've been obsessed about.

I liked the format of my post-mortem on The Caldecott Caper, so let's follow the same breakdown for CFiC. These all probably qualify as minor spoilers.

What Went Well: Technical

With each of these games, I've become more aware of the advantages and shortcomings of the engine, and have invested time and effort up front that's paid off significant dividends in my overall productivity and capabilities.

Music Replacement

My biggest disappointment with Hong Kong by far was losing the classic cyberpunk soundtrack from Dead Man's Switch and Dragonfall. I'd vaguely remembered reading a while ago about some earlier modders who experimented with music replacement, so I dug up some old threads and spent a fair amount of time creating a similar feature for Hong Kong. This was absolutely necessary for all the rest of my work, there's no way I would have made this game without access to a broader range of tunes.

This isn't perfect; in particular, since I don't belong to any of the paid development programs my players have needed to put up with intimidating Windows Defender warnings and false positive trojan notifications. But it sounds like a fair amount of players are willing to go through it, which is wonderful to see. I'm sure that's largely because of pre-existing trust based on my previous mod work, I doubt as many people would have installed this program if it was just from some rando.

Alpha Testing

This was one of the things I explicitly called out in my previous post-mortem. This time around, I had a plan in place from the start for how I wanted to approach testing, including a dedicated alpha-test phase. I was much more aggressive about recruiting testers, and ended up with a good number. I think about two-dozen people expressed interest, of those about a dozen subscribed, of those about a half-dozen gave (incredibly useful) feedback, and a grand total of one actually completed the game.

I really should have expanded the alpha-test window by at least a week. In retrospect, I was asking players to put in one or two hours a day over the course of eleven days, which is a pretty big commitment. Of course, some people will power through a campaign like this in a single sitting, but none of those joined this test. It all worked out all right in the end, though. I only had a handful of major issues crop up, and they were all identified and fixed prior to the public release, which was my main goal with this phase. This has been by far my smoothest launch with the fewest patches.


It's amazing that I managed to hit my date. I started working on this in earnest early in the summer. At first I thought I might go more quickly than with Caldecott, thanks to my increased familiarity with the editor and less need for experimentation. I think it ended up taking longer, though, mostly due to the increased reactivity in CFiC and the addition of alternate peaceful routes to multiple scenes. Still, I had enough of a buffer in there that I managed to nail my desired release date without cutting any planned features. How often does that happen?!

What Went Well: Artistic

My devlog posts were weird to write, because they're like 90% focused on the technical aspect of making the game, and not the creative/artistic/emotional elements that were my entire motivation for creating it. Anyways, here are some new things I tried for CalFree in Chains that I think turned out pretty well.


I covered this in more detail in a previous devlog post, but I'm pleased with how the optional side-quests turned out. I like how varied they are: narratively and mechanically, each feels distinct. They are more forgiving than the side-quests for Caldecott or Antumbra, offering alternate routes to completion or gracefully diminishing returns. And, although it was a late addition, I like the multiple ways that they can tie back into the game's main plot and directly impact (while not defining) the climax of the game.


The dog is 100% inessential, but I really liked the idea and was happy to make her work. It turned out to be a lot harder than I thought, and I learned a lot about how the engine handles prefabs and genders and stuff. There are still occasional bugs caused by the dog, and it would be a lot simpler to remove her entirely or just make her a him and have him follow you like Dante did. But I think it's all worthwhile: I love how she moves and behaves a bit differently, and despite me totally stealing the model and portrait, she still feels very different from Dante.

I think that completely optional content like the dog is paradoxically some of the most important stuff in an RPG. Adopting and interacting with a dog isn't a demand I make of the player, it's a choice that they make, and it can become a significant part of their story.


I felt nervous about matching the impact of Caldecott, which had at least one major plot twist that caught a lot of players off guard. Here, I ended up basically doing opposite twists from Caldecott, which feels really cheap and formulaic, but actually seems to have turned out well. I think CFiC is less reliant on the twists than Caldecott, which also helps.


I intentionally selected a variety of cameos, plot points, and other elements from Antumbra Saga and The Caldecott Caper to reappear here. I don't think they're overdone, and players seem to appreciate them: I view them as little "thank you" notes for the people who've been with me since the beginning, and they double as extra flavor for those who are just joining us now.

I am curious how future players will approach these campaigns. Antumbra Saga and Caldecott were fairly separate, while Caldecott and CFiC are rather closely linked, and I've been gently advising players to play Caldecott first. I think that CFiC should still work on its own, so it will be interesting to see whether that's true, and how players feel if they go backwards through those stories.


I think most characters turned out OK. I've kind of got crew members down to a formula at this point, and that formula seems to be working, as crew-related compliments are some of the most common I'm receiving. Recurring characters like Moonflower seem to have an appropriate amount of story to uncover, and I think it was good to get players involved in her life as well as just listening to her story.

There were a few weak points. In particular, I think Rick was bland. I grabbed him because he felt vibrant in Antumbra Saga, wasn't explored much in that game, and his position as a Ranger gave direct access to an important relevant faction. But since Rick never goes anywhere in CFiC, his live-free-or-die-hard voice was kind of wasted. I should have explored his personal story like I did with other recurring NPCs, or given him an arc of some sort. (Not that every character needs one, necessarily. Maybe it's OK for some characters to be more interesting than others, instead of trying too hard to make each person Unique and Important and Interesting.)

The one-off characters in this campaign, like Sira Ikeda and Weetabrix and Yamazaki, are narrowly drawn, but I think they're distinct and work well enough. I deliberately pick just one or maybe two aspects of such characters to focus on, and I think that works out fine.

Varied Gameplay

I was bummed that Caldecott ended up mostly being "Talk-and-Shoot: The Game", and one of my explicit goals at the start of CFiC was to include more puzzles and other non-standard gameplay. The puzzles are honestly not all that great, but I think that even mediocre puzzles really help to break up the monotony and give players a different sense of achievement. And I think I learned from my experience on Antumbra Saga, too: at least judging from the early waves of comments, a lot fewer players are getting stuck on how to proceed. I forget if I mentioned this on an earlier devlog, but I added in-game hints that are spoken by your companions after you've gotten stuck on a puzzle, which hopefully gives a graceful nudge to players.


Some players have complained that combat can be a slog, which I'll readily admit (especially if playing on easier difficulty levels), but on the whole I think CFiC has the best combat of any of my games. The general consensus seems to be that it's challenging but not ridiculous, which is exactly what I was going for. I made a lot more unique fights here than in Caldecott, with boss-fight mechanics or alternate objectives or environmental obstacles or other twists to shake things up, which I think helps keep it fresh. And, as noted above, I think that the inclusion of more non-combat gameplay helps make the combat more interesting when you return to it.

One of the few significant updates I'm still thinking of adding to CFiC is an "easy mode": poking through the editor, I see that HBS specifically down-tuned the difficulty for easy-mode players, above and beyond the standard engine changes: in particular, they will often remove a couple of enemies from the map. I think that should help make CFiC more accessible for people who are interested in the story and don't want to focus on character builds or tactics.

What Was Mixed

There were a few things that... didn't go exactly as I imagined, but still had some positive impact or unexpected side-effects.

Political Themes

CalFree in Chains is not a subtle game, at all. The Caldecott Caper was also inspired by social and political issues, but it mostly worked through metaphors at one degree removed. Out of the 500+ comments on that campaign, maybe one or two have indicated that they get what it's about. Which is totally fine! I deliberately wanted to encourage a certain kind of thinking, rather than thinking about a specific topic, in the (very possibly misguided) belief that self-directed discovery will be more effective at impacting hearts and minds than on-the-nose diatribes.

CFiC, though, waves its flag from the opening crawl through the final credits. It isn't just political: it tells you how to read itself, describes what it's doing. (One of my favorite bits to create was a sidequest where you create a propaganda trideo in order to sway the opinions and behavior of your fellow Californians.) I expected a backlash, and have been a little surprised that this aspect hasn't received much notice. A few people have specifically commented on it, and seem to appreciate my treatment of those themes, but not many.

Which I guess is good - I didn't want a backlash - but also a little odd, since it was my entire reason for creating the game in the first place. I'm not sure yet how to interpret the low-key response. It may be that my audience is going "Yep yep, of course, we're all on the same page and it isn't really worth talking about." Or maybe people are digesting it. My greatest fear is that it's just become background noise... one major risk of epic save-the-world plots is that they tend to blend together after a while, and themes might be seen as just flavoring rather than essence.


Villians are always my weakest characters. They're usually the last thing I figure out during the planning stages, long after I've finalized the crew and the overall shape of the plot.  I had trouble with Gavan for a very long time, as I wanted to make him distinct from previous villains like Claude Bullion and Kroin. I was delighted once I finally figured out Gavan's voice, which ended up being really distinct and uniquely hateable. I'm still not sure whether or not it was a good idea to include a look into Gavan's personal life... it seems kind of petty in contrast to the scale of the plot, but I think it's also another way to humanize and particularize the harms he causes, so maybe it works, I dunno.

There are fewer villains in CFiC than in Caldecott, which I think is okay... I was concerned that Caldecott had too many. One thing I did miss in CFiC was infighting amongst your enemies, but I think that's inevitable given the story of the Protectorate.

In the future, I'd like to start shifting more towards antagonists instead of villains. I have a lot more fun with antagonists, people like Dorbi and Slagarm who you can butt heads with but who aren't necessarily evil.

Release Date

I selected December 22nd for the completely arbitrary reason that it was the day Bright came out on Netflix, and I thought that it might prompt some people already familiar with Shadowrun to look for related content. In hindsight, there doesn't seem to have been any such link, though a savvier marketer than me might have turned this into a good promotion opportunity.

It did end up dropping before the Christmas and New Year holidays, and I have no idea whether that helped or hurt it. On the one hand, more people are off work and have free time for fun. On the other hand, a lot of people are busy traveling and/or spending time with families, while others have gotten new games as gifts and are less likely to reinstall a two-year-old game.

At the very least, the holiday release doesn't seem to have killed my download numbers, and there's a chance it may have boosted them. ¯\_(ツ)_/¯

What Went Poorly

These aren't necessarily failures, but elements that didn't turn out as well as I hoped or weren't worth the effort I put into them.


I realized way too late that the romance arc(s) were crowding out the main plot and weakening the core of the campaign. I totally should have seen that coming... I can't find the article right now, but people have observed that while you can write a story about a thousand different things, as soon as you add a love interest to that story, it always becomes a love story. I'd been thinking about romance as a stakes-raiser, an optional component that can strengthen a player's emotional attachment to the story. Instead, for the (many!) players who pursue a romance, your personal relationship became the central element of the game, which was the exact opposite of my intention.

I frantically made some last-minute changes to the ending to make this less distressing to players, but in the process further muddled the message of the game. I dunno. I do think that the romance works really well on its own terms - it does some things that I haven't seen before in a video game, and people seem to enjoy it (or at least have a strong response to the arc). In retrospect, I wish that I had saved this for a future, non-Shadowrun project, and just omitted the love interests from this one. It would have been cool and more thematically appropriate to have similarly deep platonic relationship arcs (possibly with casual hookups for funtimes, which in all honesty feels more Shadowrun).

Solution: Think more critically at the start about whether romance works with the story I want to tell. Just because I enjoy writing romance and players enjoy reading them doesn't mean that I should always include them. Or, conversely, just make a game about romance in the first place and build around that instead.

Crew Influence

I had a nagging feeling before that I was making this too complicated, and after playing through the game twice I think I was right. I fell into the classic trap of making something fun for the designer instead of for the player. I loved the idea of a value-neutral companion influence system rather than the positive/negative favor system of The Caldecott Caper, which was encouraging metagaming over roleplaying. But the system I created was opaque and not predictable. I think players have a general sense of "oh, cool, I'm making a difference in my crew's lives", but no specific sense of achievement or deliberate guidance.

I do still think that value-neutral design is the way to go, but melding it with a standard approval-style counter was a bust. I now think that Dragon Age-style "hardening" dilemmas are the way to go: instead of a smoothly graduated progression, just have a handful of big, impactful decisions that players can carefully consider and which will have big, discrete outcomes. Which, now that I think about it, is exactly what the crew loyalty missions are in the official Shadowrun campaigns! I feel a little silly to have spent so much thought and effort on something, only to find out that a superior solution was right in front of me the whole time, but hey, at least I eventually figured it out & won't soon forget!

Solution: Think reeeeeally hard before implementing complex systems. Unless the game is actually about that complex system, I'll probably end up confusing players and putting in too much work for too little (or even negative) benefit.

Content Allocation

This isn't a huge problem, but I think that the earlier missions are generally bigger and more interesting than the later ones. FORGED IN BATTLE, for example, comes very early on and crosses a lot of territory, has you fight a ton of different factions and enemy types, includes puzzles and a boss battle, allows you to recruit a new companion, ties in to the previous campaigns, and advances the central plot. In contrast, later missions like WASTE and PROTECT take place on relatively small maps, have just a couple of enemy types, and will probably be quicker to beat. Similarly, your first few visits to the hub will introduce a bunch of new characters, kick off new side-quests, and have lots of new conversations. On the last couple of runs, though, some characters won't have anything new to say, and apart from the romances there aren't many surprises.

I don't think it's necessarily bad to front-load the best content, as it's most likely to hook players and get them to keep playing. And it may make sense to speed things up as we approach the finish, especially if people are getting bored. But it does feel kind of like I ran out of steam and had fewer ideas near the end.

Solution: This will be hard to quantify, but I'd like to get a better picture of how much content will be in each scene during the design phase. That will let me anticipate if some sections are doing relatively too much or too little, and reallocate content if there's a big discrepancy.


This just came in recently: one of my players has color-blindness, and wasn't able to solve a puzzle that relies on colors. I felt bad about that... he or she was able to get unstuck, but it's something I should have anticipated.

Solution: In the short term, I'll be pushing out an update with some form of work-around for players who aren't able to complete the puzzle, just as soon as I figure out what that work-around is. For future projects, I'll look for an accessibility checklist or some other resource that can help me identify similar issues in advance so I can design for them properly in the first place.

Next Steps

Around the end of this month I'll write one final post where I go over the subscriber numbers and other metrics associated with the campaign. That will probably be interesting to other creators thinking of making stuff for Hong Kong, as well as giving me some ballpark numbers for players who might be interested in following me to fresh platforms. And that will probably be it! I'll keep maintaining these campaigns for the forseeable future, but it's time to stop nattering about them on this blog.

1 comment:

  1. Maybe it's just because I'm a fellow modder (and the one who finished the alpha build) but I really enjoyed alot of the technical challenges you included.

    Too many RPGs have a paragon/renegade see thru mechanic for dealing with your companions, so yours was a refreshing change. Do wish the colors of the occasional pop-ups were a little clearer though. (Red being the 'enemy colored' responses was a bit confusing)

    I loved interacting with the dog, and was probably my favorite part of returning to the hub... So in my mind that trouble was well worth it.

    That and trying to figure out what the sister was trying to spell with the 's-a-l-e' cards and balloons. Every time I came back they were some new combonations...

    Playing through on the alpha, I tried not to play as 'me', but down the path that I as a modder find the weakest in my own works. Very pleasantly suprised by your work (again), and really look forward to following whatever works you come up with in the future... Be they shadowrun campaigns, or something else.

    Now I gotta go sit down and enjoy the campaign as 'me'.