Monday, 10 November 2014

A Cord of One Strand Is Not Easily Broken: a party-splitting mechanic

Yes, another in the long list of posts that are sort-of-games that I won't get round to doing anything with.

As far as I remember, this particular idea was born from two premises:

  • the kind of grandstanding, heroic scenes beloved of all media in the history of writing, whereby one person nobly steps up to face a challenge while their companions are occupied;
  • the fact that in most games, splitting the party makes the difference between an obstacle (usually a fight) that saps some of your resources, and being ground into a fine paste.

In fact, quite often lone characters are depicted as overcoming challenges far exceeding what the whole party might typically take on. Such scenes include "I'll hold them off as long as I can", freeing captured teammates, lone charges against a horde to strike fear into the enemy, covert missions where a lone intruder has a better chance of success, or being unable to unleash their full capabilities for fear of injuring allies or bystanders.

This usually isn't possible in games, for a few reasons. One is that the power of enemies is generally static, and so a challenge aimed at five people will be far more difficult for one. Some things are physically very difficult: fifty people can lift a car over their heads, one person can't. In some cases there simply isn't time to do a large quantity of work, even if it's very simple.

Another is that one of the major determinants of power is number of actions: a lone character can take far fewer actions. I touched on this stuff briefly while working through Monitors rules. You could call this an Action Bottleneck, and it's a common issue facing NPCs and powerful monsters. The Angry DM discussed this years ago and covered all the essentials. To a lesser extent, it's also an issue for GMs whenever a player can't make a session, or a PC can't participate in a fight.

As soon as party size begins to vary, factorials come into play: a character capable of defeating one enemy per round confronting ten enemies will be attacked 10+9+8+7+6+5+4+3+2+1 times (55), while two characters would each be attacked 15 times, and ten characters would be attacked once each (or not at all if they act first). You'll notice that this is not a linear scale, and reducing the number of player characters significantly cuts their odds of success.

As well as straightforward actions, a smaller group has fewer options for supporting or synergystic actions, or for dealing with two problems simultaneously, like fighting off enemies while repairing a power conduit, or driving through a forest at speed while researching aliens online. They need to choose between making an attack, taking defensive options, using a buffing spell, altering the environment, healing and so on. The last is a key point - a group can usually allow one person to heal while others handle fighting and similar. The lone character can only generally respond to one aspect of their situation and condition, even though there may be multiple things they could really do with tackling.

Also, a lone PC generally gets to act only once during the initiative round, so the whole team of enemies gets to take their actions with no chance for the PCs to respond appropriately - this is particularly important for things like healing a badly-wounded character.

The next problem is that fewer PCs make it easier for adversaries to focus. This is the reverse of the classic boss monster problem. A PC who ends up stun-locked is very little fun to play (as most attempts at the final battle of Icewind Dale II will tell you) and likely to wind up dead quickly.

On the other hand, some challenges are much easier to handle alone. Stealth missions basically always fall into this category. Most groups are fairly disparate, and will feature at least one person who is bad at any given challenge. In additive or synergystic challenges, this isn't a huge problem; in those with negative marking, it can be. There's often a difficult choice for parties, in that it makes sense for the people good at these tasks to go off and do it alone, but without help they probably can't handle the other kinds of challenges they might encounter.

It occurred to me that one way to handle this would be to implement some kind of Conservation of Ninjutsu rules.

Incidentally, I do realise that I've very quickly gone and assumed "challenge" means "fight", but in most RPGs this is the main form of challenge faced by PCs, and the one that scales worst. Social encounters rarely depend on the number of PCs present, although more PCs may mean a wider selection of social skills. Physical challenges are often easier for one skilled person. The time you usually need ninjutsu is when you're doing action-adventure stuff. It's not exclusive, though. High-octane and utterly unrealistic 'hacking' challenges are a sadly well-establish trend in film. Brutal, punishing social encounters are entirely possible.

Tentative rules musing

This seems like it would work best with a dicepool system, and probably with one where scoring extra successes would enable you to do more stuff, or where dice can be split between different actions.

The other key part of the equation is a Ki Pool (it's shorter than Ninjutsu Pool). This pool has a fixed size and is split between PCs participating in a scene. Note that this means if the party is separated, both sections can use the full Ki Pool shared between members of that sub-party.


Characters get one die per point of Attribute ranging from 1-5 (it's simple). Assuming the game is aiming for high-octane action, these will be things like Acrobatics, Misdirection, Laying Low, Demolitions and so on. Dice are rolled to perform actions, and can be split across multiple actions at the cost of reducing your odds. You can only split dice if it's remotely sensible that you'd be able to attempt the set of actions simultaneously or in quick succession, allowing for a heroic and mildly preposterous interpretation of "sensible".

The ki pool contains one die per character in the party, and represents being awesome. The dice in the ki pool are split as evenly as possible between all participants in a scene who are still functional - unconscious, incapacitated or otherwise non-participatory PCs don't count. Ki dice can't be used on boring actions, which means you may need to find ways to reimagine boring actions as being awesome.

To keep the ki pool relevant with smaller parties compared to the Attribute range, a party of two or three characters should use a pool of six ki dice. For obvious reasons, a party of one is not a sensible choice for this system.

While PCs are together, excess successes can be passed to another player providing that any kind of plausible excuse for the assistance can be produced. For example, one PC might pull another into cover as a guard passes, block an incoming attack, catch a falling vase, throw in a quip to distract a suspicious observer, and so on. If their activities don't relate in any way, this probably isn't appropriate.

If PCs split up, the ki pool remains the same size, but is divvied up separatedly by each group. In a party of six characters who split into groups of three, two and one, the ki pool will be available to each group, and characters will receive two, three and six ki dice respectively. "Splitting up" is open to interpretation; for example, if one character wants to defend the main door while the rest try to perform a ritual, you might rule they have split up because they can't really support each other.

The result is that a lone character gains a major boost to their effectiveness, allowing them to compete more or less on a level playing field with the whole group. They aren't quite as effective overall, but there are some advantages to being alone. Because they can split dice between actions, they suffer less from the usual Action Bottleneck: having a significantly larger dicepool essentially makes it likely that they can take about twice as many actions as usual. This mechanic should also encourage the "moment of awesome" effect that stories of this type often include: a character is mechanically capable of doing things under pressure that they simply can't manage under normal circumstances. Similarly, heroic defence of the fallen is baked into the system, because the characters become individually stronger as comrades drop.


  1. It's an interesting rule but I'd need to see it in play to really comment.

    So instead I will comment on a complete tangent of one of your lines. I've found that most social situations are the inverse of the rule of More PCs = Easier Obstacles, often leading to some funny situations.

    While one failed intimidate, three failed diplomacy, and a successful bluff *dice roll* can be justified as one guy being an oaf, three telling him off, and a fourth making up some lie about the scary one, could win an NPC onside, players themselves are rarely that coordinated when they're actually roleplaying it out.

    Therefore one will fail their intimidate when they stand over the witness, one tries to brush him off so he can seduce the information out of her, the third tries to connect over common ground (not touching upon either the seducer or the fellow standing over her and glaring) and the fourth tries to herd all the other PCs away from her to give her some space, while the fourth comes up with a big story about the scary guy, that is then interrupted by the seducer, and so on and so forth.

    1. That's very true! In real life people tend to be moderately okay at turn-taking in conversation, and if you're trying to get a favour or something, often you end up with physical distance between the Face Man plus Askee, and the rest of the group. We understand that going in mob-handed doesn't necessarily help, and tend to let one conversation happen at a time, maybe with different people taking over if one person is clearly floundering.

      But yeah, in games, you quite often have three or four people taking completely different tacks in front of the NPC and not really acknowledging what each other are doing. I think it partly comes down to the way conversations are often half IC and half OOC summaries, but it's a definite thing. And if your example came up in real life, the witness would most likely freak out, clam up and get the heck away from those people.