Friday, January 31, 2020

Fumbles Can Be Great if you Just Make the Perfect Rule

I have done this. I have made the perfect fumble houserule.

And no, fumbles are not "inherently bad" and unfix-able by nature. I did it. I made them good.

We need to stick to our design principles. You know how much I like a well-thought-out houserule. So, if people have come up with houserules for fumbling attack rolls, there must be a reason they did it. What drew them to it? The first question when you are modifying the existing rules is...

What's the Problem?

The usual logic goes: if critical successes are a thing, why not critical failures? If you can have a 5% chance to whoop some ass, why not an equal chance to fail? Doesn't that kind of chaos add a little bit of crazy fun? Simple enough, right?

Test it Out. What Happens?

Uh... well sometimes it's fun. But sometimes your level 15 fighter randomly stabs his own foot despite being a literal master-class swordsman. Approximately once in every 20 attacks.

This is obviously pretty absurd. How the hell can you realistically justify people with such expertise and experience being at any significant risk of failing that badly? Yes, we all know that everybody makes mistakes. But this is ridiculous.

And others will go on to question the inherent merits of the system to begin with. Is the chance for "that kind of chaos" really fun? Does anyone actually love that moment? Maybe the type of player who is really open to laughing at themselves and not taking things too seriously. But for everyone else, it's just needlessly frustrating. The common comparison these days is the tripping mechanic introduced in Super Smash Bros. Brawl. Every time you dash, there's a 1% chance you just fall on your ass. In a competitive fighting game where every move is extremely calculated and precision in execution matters, stuff like this ruins the strategic integrity of the game. Likewise, when I choose to attack an enemy instead of, like, running away or Dodging or something, I shouldn't have to calculate a 5% chance that I'm going to potentially hurt myself and accidentally go below 0 HP or something stupid. The only risk/reward calculation I should have to make is the variability in the die rolls that already exists.

Is There Any Hope for this Rule?

Ordinarily at this point I'd probably say no and give up on trying to make it work. BUT, something else came along that helped strengthen the value of a fumble rule. Similar to this, it's very common for DMs to try finding the right way to adjudicate "called shots" in combat. Players constantly ask to do this ("can I just aim for the medusa's eyes, specifically?") but it's not explicitly addressed in the rules.

The most straightforward solution would be, "sure, but you'll have disadvantage." Sounds reasonable enough. It's exactly the sort of thing we created advantage and disadvantage for to begin with, right? Well... there's a problem with this. Actually, there's two.
1) Disadvantage can't stack.
2) Any amount of disadvantage and advantage present completely cancels out and results in a normal, unaffected roll.
Within their own context, I think these two caveats are ideal. I have never really run into problems with these and I think they are key to make advantage and disadvantage work. However, they are informative for "what types of things should incur disadvantage."

If the only cost to a called shot is disadvantage ("I wanna chop the archer's hand off!"), then that might not be enough. Because if you are already suffering from disadvantage, like from attacking in the darkness, you may as well always opt to do a called shot. "Doesn't stack with the blindness condition anyway, so I may as well try to get a little bit more out of this attack. I have nothing more to risk since I already have disadvantage."And it also doesn't make much logical sense, because aiming your attack more precisely should be distinctly harder when you are blind. And if something grants you advantage, then a called shot would be a much easier option to take. Why not make all your attacks aimed at the hand or the eyes or the throat or whatever? Yeah, it wastes the advantage you got from attacking from hiding or from darkness or from a Bless spell or whatever, but it means you basically just have a normal, unmodified attack. Shit, that's easy money right there.

So if you want to allow called shots, and Players will always ask for it, then you need something else. We still want it to be simple and intuitive, though. I've seen a million ideas for this, you may have too, and I bet a lot of them are good. But my favorite I ever found was this:

"To hit a specific part of something, you must crit. However, you declare what range you want to crit on. The fumble range must also increase the same amount." E.g. you want to chop the nose off? Alright, you decide that your new crit range is not just a natural 20, but any natural 14-20. However, you fumble on a natural roll of 1-7. Whatever you roll in between, just resolve as normal (higher than AC, normal hit. Lower than AC, normal miss). The highest stakes is to declare that crit on 11-20 and fumble on 1-10, making it a 50/50.

I love this. If I'm gunna take a risk when deciding to attack, I want to control that risk at least a little. I've had players with gambling problems before and this is great. It also makes sense. The more precise you are at aiming for something specific, the more reckless you are in landing the attack in general. If you are trying to carefully get the pinky toe, then it's also likely that failing to do so will be a total miss entirely.

But you may have noticed that, for this called shots houserule to work, it requires that you use fumbles. And more importantly, that fumbles mean something significant. They are the sole cost of choosing to try aiming for the throat. They can stack with disadvantage, but they have to be a higher risk than just disadvantage. They are key to this mechanic. So now you have to return to fixing fumbles.

Address the Complaints
  1. Characters who are good at something shouldn't be at significant risk of failing that thing
  2. The higher level you are, the less likely you should be to fumble, since you are just generally all-around better now
  3. 1-in-20 is too frequent, by far. The chaos of fumbling now and then isn't that fun.
Alright, we can then find a way to key fumbles to a character's varying strengths and weaknesses, their overall level (or just progression of power), and narrow down the chance of occurring a bit. I did all three at once, with one change.

TL;DR : Perfect Fumble Houserule

When you get a critical failure (normally, rolling a natural 1 on an attack, and for called shots, rolling whatever critical failure range you decided on), you then are at risk of fumbling. To confirm or overcome the fumble, you 1) roll 1d6 to determine which ability is affected, and 2) make a DC 20 saving throw in that ability. If you fail, incur a penalty based on that ability. Use the following table.

1d6 to determine saving throw type (DC 20)
Strength Saving Throw
Overpowered. Disarmed, weapon 10ft away
Dexterity Saving Throw
Knocked off balance. Fall Prone
Constitution Saving Throw
Strained muscles. -1d4 HP
Intelligence Saving Throw
Leave yourself wide open. -2 AC for 1 turn
Wisdom Saving Throw
Lose coordination. -1d4 on next roll
Charisma Saving Throw
Botched feint. Provoke opportunity attack

What makes this perfect is maybe subtle. A lot of people I know balk at a permanent, fixed DC for some check. It runs counter to how they understand the core mechanic of the game. But it makes sense in this case, because we know that PCs will improve over time. So a DC 20 Strength save may be a bit tricky for a Level 1 fighter, but it's a cakewalk for a level 10 fighter. This addresses complaint #2, and dividing them into each different ability type addresses complaint #1. You'll be much less likely to fail saves in the abilities that you are good at (rogues won't have dexterity fumbles and rangers won't have wisdom fumbles very often) but you are still susceptible to failing at things you've never invested in before.

I've seen these massive charts of fumble results that are based on what type of attack you performed, whether it's a piercing, slashing, or bludgeoning weapon, or if it's a missile, or if it's a fire attack or an ice attack or a force attack or whatever. Reference tables are only good if consulting them requires no more than, like, 2 seconds. I put this small table on my DM screen and that's that. The types of penalties are based on how you fucked up. None of them are that serious, they are roughly balanced, and most will try to do something more interesting than just "you stab your own foot like a dumbass." Even if you fail, you may even be lucky enough that the circumstances nullify any consequence. If no one is near you when your weapon flies 10 feet away, then it won't be too rough to go pick it up next turn. If you hadn't even used your movement yet this turn anyway, then you can stand up from prone before anyone has a chance to take advantage of it. If you fail a Charisma save and provoke an opportunity attack, you might be lucky enough that every enemy near you has already used their reaction this round. But, every now and then, the penalty will be perfect for changing the tide of combat. Getting to make a saving throw at all helps address complaint #3, and luck can make it less awful. But for serving as the cost of a called shot, I think this is a pretty serious risk to take in a situation like that.

And yes, I make all NPCs subject to these rules, as well. I hope this helps you find peace with fumbles and be able to enjoy the benefits they offer in this game, as well as being another example to illustrate what thoughtful design looks like.



  1. Shadow of the Demon Lord solves most of the problems you discuss with advantage/disadvantage in 5e. Worth a check, because boons and banes are a really interesting replacement in my opinion.

    1. I just looked it up and this IS a really interesting replacement. It also just looks like an interesting game in general. Thank you!