TTRPG Quarterbacking part 2: Solutions, kind of? Mostly vague ideas about solutions.

June 5, 2023

Somehow this one didn’t make the great cohost migration! I’ve converted and back-dated it appropriately, and lightly touched up a header or two, but otherwise it’s intact. - Binary, Sep 22, 2025

So I outlined the problem (as it is) in part 1. This is the part where I think out loud about what to do about it as a designer.

Foreword #

I’m going to put this as the first thing because it can’t be stressed enough and because I don’t really want to hear about how this isn’t an issue at anyone’s specific table or whatever because that’s not the point. Communicate with your players. If someone’s doing it a little too much you can probably just ask them to cut it out and they probably will. Folks are generally pretty reasonable.

We will ignore this going forward because it’s a really glib answer from a DESIGN perspective. But keep it in mind.

The second thing is that at the end of the day you don’t want to like, stifle general table talk unless you’re really, really sure you want to do that. A tabletop game is ultimately going to be an avenue for folks to hang out and sometimes that involves like, giving each other advice. This is more about ways to build and run a game such that naturally it doesn’t come up as much.

Strategic Quarterbacking #

Getting this one out of the way first. In group-based games especially, the biggest source of this is the combination of strict role protection and role requirement. A great example of this is the “holy trinity” of DPS/tank/healer in an MMO. For a noncombat example, think about a heist crew: Shadowrun for instance wants everyone to have specific, specialized roles and you’ll probably want a dedicated face, decker, magic person, getaway rigger, etc. Maybe not all of them for every job but you’ll feel it if one is missing over time.

Ways to mitigate this include:

Similar capabilities #

Make it so everyone can mostly do a similar thing, which is the thing you need to do. In APOCALYPSE FRAME you can basically make any strike team work because they largely all fill the role of “smash face”, the big difference is how much offense they have, how much mobility they have, how much they can get away being exposed along the way, and specific utilities from specific Frame abilities. This isn’t a great solution for every game but for explicitly combat-based systems especially, give this some real consideration: at the end of the day, if the thing you really need to do is move the ball forward, let everyone be able to move the ball forward.

Role overlaps #

Make it easy to fill multiple roles. In Valiant Horizon each class is written to have a major and minor focus between Attacker (extra Harm, breaking defenses, exploitation of positions and conditions), Defender (mitigation, control, and prevention of enemy action), and Support (buffs, healing, and giving allies extra actions). For example, the Berserker can do a lot of AOE Harm and has a fair number of abilities that move and attack (Attacker/Spread and Attacker/Exploit) but also has abilities that give penalties to enemies who don’t attack the Berserker (Defender/Provoke).

For less combat-ish situations like the heist example, challenges can be described in-system in such a way that they can’t only be solved in one fashion. Maybe it’d be better to hack security so help can’t be called, but maybe you can also just physically cut the cyber-phonelines to roughly accomplish the same thing. Focus on outcomes/results rather than means. (You can still make it narratively better to do one or the other - like maybe hacking can bring down autoturrets or reveal things to grab that cutting the lines doesn’t - as long as it’s not a hard barrier. It’s good to make choices count.)

Watch for bottlenecks #

Remove functional bottlenecks - areas where a particular function needs to exist - wherever possible. A big one for combat-focused games is often healing: if only certain classes have access to healing and damage is unavoidable to the extent that healing is expected, then there’s a bottleneck that you need a healer. For this particular instance, you can introduce other ways to avoid that, like damage mitigation, the ability to hamper enemy attacks, spike offense to the degree that you can defeat enemies before they attack, etc. You can also give everyone a way to “fill in” on roles: damage isn’t so one-sided towards attack-focused classes that other people can reasonably do it, non-defensive characters aren’t so comparatively flimsy that you strictly need a tank to draw fire, everyone has an emergency ability to heal themselves, etc.

Tactical Quarterbacking #

This is the meat of what I was getting into in part 1, and what got me thinking about it in the first place. As noted in my first post re: diceless designs, I think the “set” nature of being able to set up combos, etc. can cause this. A major way to combat it is to make situations less “solvable”.

Uncertain Outcomes #

Certainty is a major factor in this, so a major solution is to introduce uncertainty in a few ways. The first obvious move is to…introduce randomization again, of course!

Frankly I think a little randomization is fine, if not generally interesting! But I’d rather it be in the form of “how well did the thing you’re doing work” rather than a pure binary “did it hit or not”. This is the approach Total//Effect games are taking. (Basically fuck to-hit rolls as a base mechanic, they fucking suck. This has nothing to do with quarterbacking, but it’s important to me personally so I’m including it.) Also for games where players have a ton of actions (aka mine, I do this a lot), randomization can slow things down.

Another thought re: uncertain outcomes is to have everyone figure out their actions separately, then reveal at the same time, but this is a pretty specialized kind of mechanic.

Uncertain Situations #

Another idea is to just throw curveballs so players don’t know exactly what to solve for. LUMEN has this built in with its “do something drastic every round as part of the GM Turn” idea: you’re encouraged to constantly be changing the situation. But beyond that, just hiding information until something happens is a classic move. (Sometimes this creates quarterbacking in the way of contingency plans but like, after a certain point if everyone’s that on board, have at.)

This comes with a caveat: as a designer, provide tools and templates and ideas for GMs to do this without constantly having to come up with shit on the fly. Otherwise, it can add more narrative requirements to a role that can already be very overwhelming!

Uncertain Motivations #

Here’s another thing: How well do you trust your quarterback? Are they calling plays that best suit the situation or ones that’ll get them the hall of fame nod?

Games, mostly-co-op board games and video games, sometimes have a betrayer mechanic - thinking of the Battlestar Galactica board game where one player is a cylon so you can’t really trust anyone to truly lead the way on ideas, but you can see this in video games too with SS13, Among Us, etc. The goal is to do the thing, but the goal is also to find whoever the betrayer is.

Now. I don’t think you should do exactly that for a TTRPG! If everyone’s always second guessing everything everyone else does because they’re worried it’ll fuck everyone over then it’ll bog down everything HARD. (Though, if that’s your bag then go wild, sounds like fun to me but I recognize I’m weird.) What you can do as a less extreme measure is introduce personal/secret motivations alongside group motivations, and to instill doubt as to motivations for opinions. Does a warrior’s motivation suggesting the group try and take the demon lord on alone match his protestations about keeping civilians safe by not involving them or indicate that he’s a gloryhound? Is that wizard reticent to attack a shady group because they’re not doing anything wrong or because she needs something out of them? Basically, make it make sense to make your own call sometimes rather than rely on someone else. When you’re not sure you’re 100% on the same page as someone, you’re less likely to accept their recommendations and more likely to do what you feel is best.

Uncertain Communication #

This is one I’m reticent about because I don’t usually like restricting table talk too-too much (see foreword point #2) but: in tense situations, make communication mechanically scarce. Enforce taking actions to communicate things of substance. Give players explicit mechanical abilities to ask for help, or create situations where there are comms blackouts, like when player characters are separated. (I don’t recommend like, running two halves of a game as secret information because that shit’s annoying as fuck over time, and the tension of watching half your party fuck it up for you is palpable.)

You can also make the details each character has uneven, or even just subjective - different characters will notice different details about a situation based on who they are, perfect information is a lie - but that can be a pretty tall ask unless it’s very integrated into the system, and even then you can get weird situations where everyone’s unclear about what kind of information their character knows vs what they know.

On a related side note…

Idea: The Quarterback Is A Gameplay Construct Now #

Alright, hotshot. You think you got what it takes to run this operation? Congrats, you’re now the other characters’ handler. You don’t have an in-action character of your own, you run overwatch from afar and communicate information as best you can. If they hit a communication dead zone, they’re cut off and can’t talk to anyone except anyone local, so better hope you gave applicable advice prior to that.

If it’s a more local situation, you’re now the squad leader. You’ve got one free-action suggestion per round - anything past that needs an action - but also anything intelligent you face knows to go for the guy barking orders first. Best of luck.

Reduce Coordination Complexity #

So one major source of quarterbacking is coordination: you need to do X and they need to do Y so I can do Z, that sort of thing. If the first person doesn’t get the memo and isn’t on board for X then the whole plan could be up shit’s creek.

In general, when there are mechanical combinations, add in fallbacks. Sure, X + Y -> Z is the best outcome, but maybe if you have either X or Y it’s still better than nothing at all.

Reduce stakes out of character #

So honestly, one of the big takeaways I have re: quarterbacking is that it’s mostly when people think there’s going to be a really important outcome from a situation: losing is too painful or winning is too crucial. Honestly though? Unless it’s an especially brutal kind of game, not to spoil anything, but the good guys are gonna win eventually. And if it is that kind of game, generally the idea is to play to find out rather than ensure victory. If people start getting a little too intense about it, just remind them of the goal, whatever it may be. The benefit of TTRPGs is that it’s way easier than a video game to make situations recover from “everything went sideways” because everyone involved has way more narrative control: if things do go badly in a “the good guys will definitely win” kind of game the players and GM can usually come up with a way to bounce back. Encourage people to allow things to go sideways.

Conclusions #

• Try out any and all of this as a dev/GM, but just be careful to make sure it’s actually improving the game. In a lot of cases being too worried about it just slows things down.
• I think about this stuff the normal amount.

Comments