Monte Cook On Fumble Mechanics

Fumble mechanics have been part of the tabletop RPG experience for decades. Even where games don't have a fumble mechanic, many players house rule them in. A fumble is the opposite of a critical hit (or critical success) - its most common manifestation is a roll of 1 in a d20-based game (with a roll of 20 being the critical). Veteran game designer Monte Cook has some thoughts on fumble mechanics, and talks about them and how his Numenera RPG (and all of the Cypher System line) use an "intrusion" instead.

​

It can be a divisive issue. If you're like me, you've experimented with fumble mechanics of various kinds over the years. When I was 12, I remember one character accidentally shooting a fellow character in the back of the head and killing him. Monte Cook's thoughts on the matter are that "we don’t want to run games that “punish” players for rolling bad. A GM intrusion isn’t meant to be “punishment”—it’s meant to make things more interesting. But a fumble, for many people, just seems like a moment for everyone to laugh at them, and that’s not always fun."

If you look around, you'll find dozens of fumble house rules for most games. They clearly provide a draw to those who like to tinker with their games. But many games deliberately do not include any such rule.

You can read the rest of Monte's article here. What are your thoughts on fumble mechanics?

 

[pemerton]

In the traditional role of the Game-Master, as world-builder and impartial adjudicator, it is improper to take the capabilities of the party into consideration at any point.

<snip>

"Fair" means that it's set-up without bias. The dragon is what it is, without regard for who the PCs are. The GM doesn't throw in a young white dragon because the party is high level and has tons of fire effects; and the GM doesn't make it an ancient red dragon because it has the best chance of killing the party. The GM disregards the existence of the PCs as meta-game information that cannot possibly matter, and figures out what kind of creature makes the most sense for this place in the world.

This whole issue is something as a tangent. But to me it seems relevant to the tangent that my Moldvay Basic Set included a module called Keep on the Borderlands which has been designed keeping in mind the capabilities of 1st to 3rd level PCs; and my Cook/Marsh Expert Set included a module called The Isle of Dread which has been designed keeping mind the capabilities of PCs above 3rd level.

It seems to me that the process of design of both modules was, in fact, first to think about the level of PCs that the module would be written for; then to think of a setting and a population for that setting that might provide suitable opposition for such PCs; and then to write a module that instantiates those ideas.

I've got no view on whether or not either module is fair; but I think they are both pretty traditional, and yet the method of authorship is not the one that you describe.

The recommended level (or CR) could be descriptive rather than proscriptive. It's not that the Necromancer is raising an army of CR 4 undead, and there's a CR 7 dragon involved, because this adventure is recommended for characters of level 5-8; it's that this adventure is recommended for these levels because these are the kinds of NPCs involved.

<snip>

The Necromancer isn't working with the level-appropriate dragon because it's a game and the whole thing is designed for the party of a certain level; the Necromancer is working with this dragon because lesser dragons couldn't help her and she is beneath the notice of more powerful dragons.

This is confused. Your first post on this topic is about fairness on the part of the GM. The ingame motivation of imaginary beings has no relevance to that.

In the fiction, the reason the necromancer is working with the dragon is because lesser dragons couldn't help her and she is beneath the notice of more powerful dragons. But the reason the fiction has been authored this way, by the GM, is because the whole thing is designed for a party of a certain level.

There is no tension or contradiction here.

These worlds need to be judged on their inherent potential for adventure. If you base it on the players or their characters, then you are violating impartiality.

If you don't want to be fair, then feel free to ignore the rules, or play some other game that doesn't care about fairness. You can't take a biased starting position and then claim to be fair, though.

Impartiality to whom? And fairness to whom?

No duties of impartiality or fairness are owed to possible worlds, or to imaginary people. Those duties may well be owed by the GM (a real person) to his/her players (other real people), but it's not at all clear to me how the GM violates any such duties by designing an adventure with the intention that it be fun for those people to play using their PCs.

 

[pemerton]

No the problem is that since a fumble is a sub-category of failure... by definition a fumble will always entail a failure... until what differentiates a fumble from a regular failure is defined... there's no way to determine whether something is or isn't a regular failure or a fumble...

As for catastrophic failures... Luke gets his hand cut off in a lightsaber duel.. how do you model that?

In D&D you can't. Unless Vader is wielding a very specific magical weapon, there is no mechanic for translating hit point loss into dismemberment. This is a consequence of the abstraction of D&D, which I have already mentioned a couple of time upthread.

In Rolemaster, or Runequest, or Burning Wheel, or plenty of other FRPGs, on the other hand, the combat mechanics are not abstract like D&D's, and they produce concrete results that include specific injuries.

what differentiates a regular failure from a fumble fiction wise and mechanics wise for you?

Nothing. I don't think there is any general concept of "failure" or "fumble".

It all depends on system.

In RM, for instance, attacks produce one of three results: a hit (which may deliver concussion hit loss and/or specific injuries or conditions via the crit tables); a "miss", which may be a dodge or a parry or a shield block (there are rules for determining this, and then for determining weapon/shield breakage if appropriate); or a "fumble", which is rolled on a table and generally leads to the attacker suffering some sort of adverse consequence (such as some damage or other debuff).

In RM, a non-attack skill check is generally resolved on a table which has different results depending on the overall result of the roll + bonus + situational mods. Each skill (or skill category) has a different table, but in general low results (eg -25 or down) result in some sort of adverse consequence for the PC and/or a debuff; mid-results (eg 06 to 75) leave the situation unchanged but typically (not always) permit a reroll, which in the fiction costs time; upper results short of outright success (eg 76 to 110) may allow a reroll with a bonus, or produce a change in the situation less than what the PC was hoping to achieve (eg jump 70% of intended distance); and then there are results that produce outright success.

There is no notion of a "fumble" for these non-attack skill checks, although d100 rolls below 06 are "open-ended low", meaning that the d% is rerolled and the result subtracted from the first roll to determine the die result before bonuses and mods are added.

RM also requires a roll for spell-casting. For non-attack spells, success occurs on a roll of 03 or up, but on a roll of 02 or 01 the spell is not cast as intended and a roll must be made on the spell failure table. The result of that roll can very from there being no additional consequence, through the caster suffering damage and/or debuff due to internalising the magic, through to the caster dying from the failed casting.

Burning Wheel also requires a roll for spell-casting, and if the roll fails requires another roll to determine the consequences of failure. Consequences can include no additional result; or a random magical effect generated; or an unwanted summoning occurring.

If someone at my table failed a BW spell-casting roll, and then the result of the failure roll was not no additional result but rather was unwanted summoning, then I wouldn't be described if another player described that as a fumbling of the spell. But fumbling here isn't a technical term.

 

[Maxperson]

No, they absolutely do not. The 3e DMG has an excellent little article on randomness in the game that sums it up nicely - crits favour the monsters. Full stop. Because it doesn't matter that you crit three times more often. I have unlimited monsters. I only have to crit a small amount of times to change the nature of the game.

Sure, in the white room you have unlimited monsters. In the reality of game play, though, your monsters are highly limited. You don't go anywhere near throwing unlimited monsters at the PCs. If you did, they'd die no matter what.

Since it's a fact that the DM is going to limit the number of monsters PCs encounter, crits are going to favor the PCs since they crit many times more often.

 

[Maxperson]

Good grief. FOUR people before me said EXACLTY the same thing as I did - that there's no reason we need that to be a fumble - the regular rules work perfectly well for that scene - but you choose ME to pick an argument with? Look, when four people tell you the exact same thing, perhaps, just perhaps, you might want to consider that your argument might not be the work of beauty you think it is.

So then the answer is that you can't show how it's not a fumble. You can only say it isn't required which isn't the same thing at all.

When a scene can be modelled without resorting to extraneous mechanics, why would I add extra mechanics to model the scene? What's the point?

Increased enjoyment is the point.

The argument was made that fumbles are common in genre fiction. Han brings down the blast door, Han steps on a twig, apparently Luke stumbles and falls, swordsmen kill allied swordsmen all the time. But, when proof is asked for - actual examples from genre fiction, suddenly it's shown that no, there's no need for fumble mechanics and in fact, fumble mechanics would actually make following genre conventions MORE difficult. Because, outside of some very corner case examples, our heroes don't catastrophically fail in genre fiction.

Irrelevant. Need plays no part in this. You aren't required to need fumbles for something to be modeled as a fumble. Every last one of those examples is also an example of a fumble. All of them. You can model them as fumble or not, depending on how you want to play.

 

[Imaro]

In D&D you can't. Unless Vader is wielding a very specific magical weapon, there is no mechanic for translating hit point loss into dismemberment. This is a consequence of the abstraction of D&D, which I have already mentioned a couple of time upthread.

But it can be accomplished with fumble rules...

In Rolemaster, or Runequest, or Burning Wheel, or plenty of other FRPGs, on the other hand, the combat mechanics are not abstract like D&D's, and they produce concrete results that include specific injuries.

Yes but even those systems are probably (I'm unfamiliar with them... except for the Mongoose version of Runequest so I can't be 100% sure) lacking in certain areas and unable to recreate certain situations that arise in fiction.

Nothing. I don't think there is any general concept of "failure" or "fumble".

If this is true how can you and others claim certain things that arise in the fiction aren't fumbles?

It all depends on system.

Exactly... so depending on the system, if you want to model certain actions that arise in fiction (with the regularity that they occur in said fiction) it may be necessary to introduce a "fumble" mechanic. Now whether this is a good or bad thing is entirely group dependent but to claim that fumbles are unnecessary is to assume you know what game experience any and every group playing a particular game is looking for...

In RM, for instance, attacks produce one of three results: a hit (which may deliver concussion hit loss and/or specific injuries or conditions via the crit tables); a "miss", which may be a dodge or a parry or a shield block (there are rules for determining this, and then for determining weapon/shield breakage if appropriate); or a "fumble", which is rolled on a table and generally leads to the attacker suffering some sort of adverse consequence (such as some damage or other debuff).

In RM, a non-attack skill check is generally resolved on a table which has different results depending on the overall result of the roll + bonus + situational mods. Each skill (or skill category) has a different table, but in general low results (eg -25 or down) result in some sort of adverse consequence for the PC and/or a debuff; mid-results (eg 06 to 75) leave the situation unchanged but typically (not always) permit a reroll, which in the fiction costs time; upper results short of outright success (eg 76 to 110) may allow a reroll with a bonus, or produce a change in the situation less than what the PC was hoping to achieve (eg jump 70% of intended distance); and then there are results that produce outright success.

There is no notion of a "fumble" for these non-attack skill checks, although d100 rolls below 06 are "open-ended low", meaning that the d% is rerolled and the result subtracted from the first roll to determine the die result before bonuses and mods are added.

RM also requires a roll for spell-casting. For non-attack spells, success occurs on a roll of 03 or up, but on a roll of 02 or 01 the spell is not cast as intended and a roll must be made on the spell failure table. The result of that roll can very from there being no additional consequence, through the caster suffering damage and/or debuff due to internalising the magic, through to the caster dying from the failed casting.

But we aren't speaking specifically to RM...

Burning Wheel also requires a roll for spell-casting, and if the roll fails requires another roll to determine the consequences of failure. Consequences can include no additional result; or a random magical effect generated; or an unwanted summoning occurring.

If someone at my table failed a BW spell-casting roll, and then the result of the failure roll was not no additional result but rather was unwanted summoning, then I wouldn't be described if another player described that as a fumbling of the spell. But fumbling here isn't a technical term.

There is still a limit here to the effects that can be produced by these failures... correct? What this does is set the precedent for what a failure constitutes in these particular games... however if a group (for whatever reason) wants a wider variety of results, more severe results, or even to give the GM the power to decide a specific and more relevant result of failure... well then fumble rules could be introduced to take care of that and thus a "need" for fumble rules arises.

What I don't understand is how you and others like [MENTION=22779]Hussar[/MENTION] are dismissing examples of "fumbles" when there is no default game system we are discussing to determine whether they can or can't be modeled without fumble rules? Yes I'm sure somewhere for every example presented you can find a particular system that could model it without recourse to fumble rules... but the point is no system is capable of modelling all possibilities of these failures and thus the "need" as well as what constitutes a "fumble" is entirely group and rules system dependent.

 

[Imaro]

Critical success on Vader's part. Given how often lightsaber's hack off limbs, even of highly skilled practitioners, I'm going to give credit to Vader rather than try to convince myself that Luke bungled his defense.

You can't do this in D&D (which was the system being referenced) which was my point, a new system is necessary to model it(or are we now switching to another system to "prove" it's not a fumble?). Now you could model it a some new type of critical success but it could also be modeled as a fumble... so you saying you would model it in a different way doesn't prove it's not a fumble...

This is not really idiosyncratic to Hussar.

Never claimed it was... in fact I specifically called out it was him as well as a few other posters... Still doesn't prove or disprove the correctness of making said assertion though...

If you're going to have a distinction between failure and fumble terms, fumble needs to mean something distinct from just failing and that pretty much requires an effect that is even worse than failure. Traditional definitions of fumble suggest clumsiness, bungling, or making a mess of something and that too suggests more than simply not succeeding. So is it really controversial that people assume that a fumble somehow harms the PC or his allies?

No it does not pretty much mean it "requires" an effect that is even worse than failure... This is totally dependent on the system being used and what it's limitations are... which is why the arguing that these fictional examples aren't fumbles is so silly. Until we agree upon what the system is and what a fumble constitutes in said system... both anything and nothing are valid as "fumbles".

While I agree with your traditional defintion of fumbles... I in no way see how clumsiness, bungling, or making a mess of something during an action must logically and in all situations lead to severed limbs, the gouging out of friends eyes and decapitations... That's how you and a few others personally choose to define a fumble but taht's all it is, your personal view of what a "fumble" is.

 

[Maxperson]

No it does not pretty much mean it "requires" an effect that is even worse than failure... This is totally dependent on the system being used and what it's limitations are... which is why the arguing that these fictional examples aren't fumbles is so silly. Until we agree upon what the system is and what a fumble constitutes in said system... both anything and nothing are valid as "fumbles".

Correct. There is no such thing as "worse than failure" with a fumble. Fumbles are simply a degree of failure, not something worse.

While I agree with your traditional defintion of fumbles... I in no way see how clumsiness, bungling, or making a mess of something during an action must logically and in all situations lead to severed limbs, the gouging out of friends eyes and decapitations... That's how you and a few others personally choose to define a fumble but taht's all it is, your personal view of what a "fumble" is.

It's fallacious arguing. They have to point out the most extreme and absurd situations in order to feel right about their positions, even though such extremes are not at all required on any fumble chart.

 

[Bedrockgames]

As for catastrophic failures... Luke gets his hand cut off in a lightsaber duel.. how do you model that?

.

That could have been a critical on Vader's part, not a fumble on Luke's.

 

[Celebrim]

Now show how that high level 3E fighter with 4 attacks per round screws up 4 times as often as Bob the 1st level warrior. Fumble rules just screw over exceptional PCs, glad we never used them.

Fumble rules do screw the PC's, but not for the reason you cite. You have in your mind a particular implementation of fumble rules, but that isn't the only way to organize them.

Consider the way criticals are handled in 3e. Normally, you critical only when you throw a 20 on an attack roll. But you don't automatically critical. You only threaten to critical on a 20. Your chance of actually hitting after throwing a 20 depends on your attack bonus. The more likely you are to hit, the more likely you are to critical. Moreover, as you increase in skill and a level, your expected chance of achieving a critical tends to increase as well. You start threatening to critical on a 19 or 20. Some of my player's character threaten to critical on a 12+ and basically miss only on a 2+. A true mook would only critical them 1 time in 400, whereas they critical a true mook about 43% of the time.

The same thing generally applies to fumbles. The more skilled you are, the less likely each individual attack results in a fumble. For example, there are a whole class of fumbles that provoke balance checks, and the consequences of the fumble - tripping, stumbling, or falling down - only happen on a failed skill check. A skilled rogue (for example) is basically immune to fumbles of this class. And there is a whole class of fumbles that provoke endurance checks, where the consequences of the fumble (becoming fatigued, being winded, pulling a muscle) only happen if you fail a constitution check. This basically never happens to the stouter members of the party with their magically enhanced constitutions and Endurance feats. All but the most serious ones they are immune to. Likewise, there are several sorts of fumbles where you only fumble if a second attack roll at your full BAB would also miss. This rarely happens to a skilled fighter.

Despite all this, fumbles do hurt the PC's more than NPC's. The simple fact of the matter is that most NPCs are going to lose anyway (however you rationalize it, that's true). As such, it doesn't really change anything if an NPC fumbles occasionally. The NPC just loses very slightly quicker. But when a PC fumbles, it changes the dynamics. The odds that a PC will experience a game changing run of bad luck that results in an unavoidable defeat increase.

I certainly don't think that fumbles are for everyone. The only reason I use them is that they add variety to D&D's sometimes overly abstract combat, which might otherwise not be marked by anyone imagining events happening at all, but rather only by the rolling of dice and the marking down of numbers in tally columns to see who won. The occasional unexpected narratable event in a combat gives D&D combat more life.

 

[Hussar]

No, I don't agree with you at all. These worlds need to be judged on their inherent potential for adventure. If you base it on the players or their characters, then you are violating impartiality.

If you don't want to be fair, then feel free to ignore the rules, or play some other game that doesn't care about fairness. You can't take a biased starting position and then claim to be fair, though.

You can't have it both ways. Either you're building scenarios specifically for world building purposes or you are building scenarios that would be fun to play through. You can't have both. As soon as you decide "fun to play through" is a criteria, then world building considerations take a back seat. IOW, your 6th level necromancer raising undead is 6th level, not because it makes world building sense, but because you are building a scenario for characters of a specific range of levels. There's a reason the necromancer isn't 3rd level or 12th level and it has zero to do with "impartiality" and everything to do with building a fun scenario.