Dev Diary 17 - Complex Dice Tests
We lied! Today was going to be about the meta-campaign mechanics, but we did a really cool system overhaul instead and we simply have to tell you about it.
Basically, we overhauled a part of our dice system in a pretty major way. It’s involved some fairly dramatic changes to how rolls are done, though all the other systems we’ve created plug pretty seamlessly into it (and as we’ve integrated it, it’s actually let us effectively cut systems now that they’re covered automatically by the new system). It’s one of the deftest bits of game design we’ve ever done and we gotta brag about it.
Development of the Previous System
One of the things that has been a problem for Torchship for a long while is that it wanted to be a dice pool system (roll X dice, looking for Y amount of Z+ results), which doesn’t just have binary pass-failure outcomes. We wanted players to feel competent in their fields, but we also needed there to be interesting difficulties and complications so that stories aren’t just a stateful progression of experts effortlessly performing the tasks they’re experts in.
This is surprisingly difficult!
Nailing down exactly how it would work has gone through about a half-a-dozen iterations, all of which always felt like hacky temporary solutions. The version we came up with before this, which the game has been using for about a year, involved two thresholds on each roll; a “Difficulty” to do the thing, and a “Complexity” tacked onto it that you had to reach to do it without any extra problems.
This served the purpose, but its various incarnations slowed the game down a lot more than it should have, and put too much stress on the GM to work out what these two targets would be and how complications would emerge from it. It was a clunky solution which required a lot of experience to use properly, functioning just well enough to build systems around without ever being stellar on its own.
While working on Must be Tuesday: Revived Edition, which uses a similar dice pool system, my wonderful editor Lexie came up with a really clever system while we were working out the dice odds. In that game, you have a “Skill” target from 6+ to 3+ with a variable dice pool and a number of Successes needed.
Our partial successes there comes from a concept of ‘Scrapes’; dice which are 4+, but don’t meet your Skill target. If you reach the number of Successes you need when you add your Scrapes to your rolls meeting your Skill, you get a partial success! Brilliant, isn’t it? That means everyone has a chance to get by on even hard checks using their worst skills, but it’s never easy.
When we poked at Torchship stuff after testing that system out, we found ourselves wondering if something similar wouldn’t fit here as well. It wouldn’t translate 1-1; Must be Tuesday is about teenagers fighting monsters in a horror/comedy setup, where nobody is doing anything really complicated, and even the people who are the best at things are still only as good at it as, you know, teenagers. It’s not a good tone fit, but it inspired the system we used.
Complexity Certs & Complications
The solution we came up with, which we are so proud of we bumped a whole dev diary for it, is the idea of Complexity Certs.
Basically, we’ve ditched the previous Complexity target from before. Your dice Test just has a single, easily determined Difficulty. In ideal circumstances, you roll a number of dice determined by the tool you’re using, needing to get results over your Cert target. Get as many of those as the Difficulty, you succeed, otherwise you fail. Simple binary outcome to a simple problem.
But you’re playing cosmonauts. You know, you boldly go places you probably shouldn’t. You don’t face simple problems.
When the GM calls for a roll, they can tack on Complexity Certs in accordance to the situation you’re facing. Essentially, they’re saying this roll is a test not just of the ‘Primary Cert’ that determines if you pass or fail, but it’s also a test of some extra skills that have come up because of the number of moving parts involved in the situation.
So while you still only have one Difficulty, you need to meet that difficulty using multiple dice targets to succeed without qualifiers. If you just meet the difficulty on your Primary Cert, but not the Complexity Certs, then the GM can hit you with a Complication that can emerge naturally from the Cert in question. Conversely, you could end up in a situation where you have a better value on your Complexity Cert than the primary, so you could fail, but avert other disasters.
Or you could fail at both, and now you have two problems!
This system elegantly compresses a bunch of things the system needed to do into one quick judgement call by the GM in the moment. We don’t need to have specific penalties for working remotely through a robot, working in a spacesuit, or doing things in low gravity; the GM can just add the Drone Operator, EVA, or Cosmonaut Certs to the Test as Complexity Certs. There’s no limit to the number of Complexity Certs that can get added either, so you can sum up really complex situations with a single roll.
It also made the game’s group test mechanics much simpler and more impactful. Helping can be a complex game design challenge; you want people to be able to give each other a hand, but you need to make sure people can’t simply do it on every single roll to avoid slowdown and the trivialization of gameplay challenges. The way Help works now is allowing you to lend a friend one of your Certs to take on a Complexity Cert, basically monitoring a potential problem for them while they focus on the main task.
As you get XP for Helping or being Helped on Checks where somebody is rolling with a higher Cert than you, you might want to point out potential problems with people’s plans that relate to your expertise as they come up so you can be the one to solve them. It also means that the presence of a Complexity Cert acts as a prompt for characters to step in and help one another out, and rewards a properly multi-disciplinary crew working together to tackle complex problems.
You know. Like… like a Star Trek.
The example we use in the game rules is as follows.
Let’s say you are at a shooting range with your laser pistol, and you want to shoot a target. That’s a straightforward Sharpshooter Cert test. You either hit the target or you don’t. Easy!
But let’s say you’re doing the same thing but in a combat situation where you might get hit in return. The GM can (and is encouraged to) add the Soldier Cert as a Complexity Cert to the roll; Soldier is the Cert that covers tactics, movements, and the use of cover, so if your dice meet the difficulty using your Sharpshooter target, but don’t from your Soldier target, then you probably hit the target but exposed yourself to danger in the process.
Suddenly, we can see the difference between an Olympic target shooter and an infantryman.
Or let’s say you’re a guard posted in a reactor room; if you are doing some shooting there, the GM could throw in Damage Controller as a Complexity Cert to represent the chances of you breaking something vital in the antimatter reactor by throwing lasers everywhere. Suddenly, you have a really good reason to cross-train your guards in engineering skills, at least enough that they know not to shoot the matter/antimatter exchangers.
Or maybe you’re trying to incapacitate an unfamiliar alien creature without killing it; the GM could add Life Scientist. What if you’re doing it in a spacesuit? Add EVA. Knocking out a piece of machinery? Add Technician. Aiming a remote turret instead of doing it yourself? Drone Operator.
Which means you could, conceivably, be in a spacesuit operating a tablet controlling a gun drone non-lethally shooting a strange device on a strange alien in a combat situation inside an engine room… and it all happens with one roll and no need for infinitely stacking penalties.
The biggest knock-on change this has caused is a need for finer gradation between Certs so that the differences come up more often and are less severe. For that reason, we moved the game to d10 pools from d6s; yes, this was an enormously annoying change to make through our draft, and we’re still working out how to rebalance advancement through it. It also means we have to do yet another pass through the Traits, which we were midway through… oh well!
(We have a cool new lever that’s come out of, actually; we can have Traits just make Complexity Certs just not count in appropriate circumstances. Freefaller characters get to ignore 0g penalties, for example, which includes adding Cosmonaut as a Complexity Cert to a lot of rolls).
I’ve submitted Torchship to Metatopia again this year, and I’m really looking forward to running it on the other side of a year of rewrites and de-heartbreakerification. I’m confident it’ll go much better this time around.
Anyway, next Dev Diary will be about the Zinovians, and then we’ll do the meta-campaign mechanics. Unless something even cooler comes up.
snowghoul
dededos
dailybridgerton
discoholicmusic
turnnoffyourmind
