Preventing "god-skills" in more freeform, rules-light RPGs

I’m working on some game mechanics and aesthetics for my RPG.

It is meant to be good for new players, and unique characters. The core mechanic is that you roll a d4 for anything, stepped up if you have anything relevant on your character sheet, stepped up again if you’re at a fictional advantage. Characters have a few sentences about themselves as Truths on their sheet, such as “I am a dwarven warrior from (…)” or “I am a witch.” or “I can use wild, unstable fire magic.”

In playtesting, the spellcasters (and not only) tended to be narratively (if not mathematically) overpowered, and tended to throw the “god-skill” of magic at every problem.

Other times, players were upset when I suggested that something they’d attempt might be outside of their abilities, which were used in a different context before.

So in a game where relatively freeform abilities are the core, how would you prevent them from becoming Godlike Swiss Army Knives of flexibility? I like the problem-solving brought on by limitations, and preparation, as well as the character development of pursuing new spells. How do I draw a clearer line between “You know how to do that” and “You don’t know how to do that”, mechanically?


I’d go for very specific definitions. The more powerful the more it should cost.

1 Like

Perhaps an issue here is the d4 basis. Bonuses are harder to manage then one might think, and a 25% bonus rising to a 50% bonus with situational crafty is just huge.

You might consider that you set magic tests at a 0 in 4 (afterall not everyone can do magic) or even a -1 if the effect is big. Alternatively consider using bigger dice for harder things (with a roll under - e.g. sure you can summon a god beast with you beast summoner power… with a roll under three on a d12, also on 3 - 6 it arrives still, but is mad, at you.)

The players wanting to be all powerful gods is simply something of a play style or table problem. Ultralights and FKR stuff doesn’t really work unless the table trusts eachother and cedes authority to the referee. This tends to be an issue with folks coming from low trust systems where RAW or player narrative control are emphisized. Not really a system issue but one if the downfalls of FKR more generally.


I think there’s a reason that magic in most RPGs usually has some kind of cost that mundane actions don’t have. You have to spend spell slots or spell points or mana, you can only cast any given spell X times per day, etc.

As for players trying to flex their skills to work in more situations, it’s hard to say what’s going on there without more information. Gus seems to think it’s a 5E-playstyle-mismatch issue, but maybe the players want to use the tools they’re given in creative ways (a very “OSR” thing to do) and in your playtest you didn’t let them.

I am also working on a game with freeform skills, but I have reduced the number of variables to one—can you apply a skill to this situation or not? There’s no adjustment for advantage or anything else. Unskilled rolls d20, skilled rolls d20+d12, target 13+.

I think reducing the number of variables in that way—making it a binary skilled/unskilled—helps to make freeform resolution feel less arbitrary. One only has to make a loose ruling on one thing instead of multiple things.

1 Like

Along the lines of magic coming with a cost, you can also think about more powerful/broadly applicable abilities having more consequential failures (failed spells don’t just fizzle, they blow up in your face). I’m running a game of The Pool (which is a different style of game but I think the example is still relevant) where one of the characters has magical abilities that are potentially applicable to almost any conflict or challenge. What keeps him from drawing on it in every challenge then? For one thing, the look and feel of magic is very big and showy: so even a successful use of magic is going to be obvious and potentially disruptive. For another, a failure while using magic opens the possibility for magical complications that are much worse than complications from a mundane failure.


Some limitations are due. Options:

  1. Decide what they can do at character creation with the possibility of learning more rituals during the adventures. Something like adventure points I Don't Remember That Move: Adventure Points

  2. They can do whatever with some general cost/benefit limitations. I like the sorcerer rules from World of Dungeons hack Streets of Marienburg. They go something like:
    Range/how noticeable:
    Touch, intently
    Someone nearby, subtly
    Someone far away, loudly
    All around you, conspiuosly

Casting time/Duration:
instant and it lassts moments
a minute and it lasts 10 minutes
10 minutes and it lasts 1 hour
1 hour ritual and it lasts 1h for each hp spent (I personally don’t like magic that causes self-harm but that’s what’s written)

And the effects include:

  • a mundane task, at range
  • a mundane task performed instantly
  • the purpose of an object not worth more than 10cp

The last one is real fun and creative, like you could get a whole squad drunk since alcohol is cheap, and so much else!

  1. This series of psots would be super-interesting to develop a freeform magic system around. Basically ties range, duration and other stuff to actual tangible objects like: “this spell lasts as much as this candle” or “this spell affects everything my shadow touches” The Nine and Thirty Kingdoms: Conceptual Magic: Delays and Triggers

  2. Smart wizards. Wizards can do anything that a modern smartphone can. That’s it. Cast light, record and reproduce sounds, talk with someone at a distance, etc


#4 is great. “There’s an app for that!”


My current project, Hexingtide (minimalist rules in a 12pg zine) has the same potential pitfall:

Character stats and abilities are open ended descriptors with a +1 to +4 bonus applied to a d10 roll to beat a target number.

If you wanted to min-max munchkin it, a player could spent their points on a very broad pair of +4 bonuses (e.g. Ultimate Goku Fighting Powerz +4 and Everyone Loves Me Because I’m a Cool Dude +4) and try to bludgeon their way through most situations.

Where I try to address that - in the games I’m playtesting and written in the rules - is in two ways:

  1. The table must be on the same page about the setting, genre, and power level of the game. Yeah, in the comics Batman can hang with the rest of the Justice League. That can translate to Hexingtide, as a +3 is a +3 across the board because it’s highlighted by the character regardless of if it’s a Bazooka +3 or a Demonic Curse Passed from Father to Son +3. The diagetic effect of the bonus levels, however, is what needs to be consistent (i.e. we all need to be on the same page about what to expect from a +3 level).
  2. Once everyone’s on the same page about what to expect, it’s incumbent on the GM to recognize potentially overly broad powers during character creation (and when players have enough XP to buy more) and nip it in the bud by defining them more restrictively - and often specifically - or by splitting them up into multiple abilities. For example, I have a player who wanted Werewolf +3, but we quickly realized that’s really broad, so we broke it into Lupine Senses +2 and Werewolf Agility +1. If abilities continue to want to be used super broadly as the game goes on, we revise and refine as needed. We’re all learning together at the table - and have to, if we want to take advantage of lightweight rules like this.

So that’s all pretty specific to my project, but hopefully it’s an idea on how that issue may be resolved in your context?

The players wanting to be all powerful gods is simply something of a play style or table problem. Ultralights and FKR stuff doesn’t really work unless the table trusts eachother and cedes authority to the referee.

Yep, that. Much more to the point than my rambling!

1 Like

One idea would be something like this:

“Once you use a tag, circle it. When using a circled tag increase Strain* by 1”

Or fatigue, stress whatever bad your system has

Or maybe link number of circled tags to XP at the end of the session