You may think if this article as observations on Caramel-Flavored Critical Fumbles. It examines a certain flavor of play that you may either love or hate. A supporter might claim:
- One never does anything perfectly always, eventually you'll mess up.
- Critical fumbles mirror critical hits; if you can get critical hits you should be able to get critical misses.
- Critical fumbles add a sense of randomness and/or fun to the game.
While someone else might claim:
- Critical fumbles hurt the PCs more than the opponents.
- Critical fumbles cause more experienced fighters to fumble their weapons or kill themselves more than a level 1 commoner who's never picked up a sword before.
- Critical fumbles aren't in the core rules.
In the spirit of being scientific and logical as opposed to opinionated, we can gather information about the game. To represent the abstract concept of getting better and honing your skills, in D&D you gain experience and thus gain the mechanic known as levels. To specifically represent combat prowess and accuracy with, as you gain your experience, you gain something called a Base Attack Bonus. This is D&D's way of assigning your skill with a number. Logically speaking, someone who is more skilled at fighting will fail/fumble less than someone who is not. In terms of BAB, this means that someone with higher BAB should fail/fumble less than someone with a lesser BAB.
The D&D battle system functions in a way that a character gets so many attempts at hitting their opponent to deal damage in one 6 minute interval, or otherwise known as a turn. This is a given in our mathematical equation. Also, we are given that at 6 BAB, 11 BAB, and 16 BAB, you gain an iterative attack.
The following chart can be assembled then for any given 6 seconds of a combatant fighting at the absolute best of their ability.
Looking back on this, I feel as though I'm misreading part of your post, and if you want to correct me on this, please do so.
Alternatively, you could surmise that since, as you progress in ability, you learn more varied and complex methods of doing something (in this case scoring a hit), there are more things that you can critically screw up. True, a commoner might not be quite as likely to drop the sword, but he isn't attempting a triple Swan thrust with a half dragon invert** either. If you look at it in this manner, the fact that fumbles increase over time becomes far more likely.
*confirm, in this case, being used to describe a roll to negate something.
**I just made that up, so don't bug me about it please.
The D&D battle system functions in a way that a character gets so many attempts at hitting their opponent to deal damage in one 6 minute interval, or otherwise known as a turn. This is a given in our mathematical equation. Also, we are given that at 6 BAB, 11 BAB, and 16 BAB, you gain an iterative attack.
I wouldn't call it an iterative attack. It's simply another attack in the same round, with a penalty of sorts applied.
The following chart can be assembled then for any given 6 seconds of a combatant fighting at the absolute best of their ability.
Higher level enemies:
Your statement that critical fumbles hurt PCs more than enemies is faulty. Let us consider 2 cases, one where the enemy is a weapon wielding creature and one where the enemy uses natural weapons. Assuming your statement of having more attacks leading directly to more critical fumbles and greater harm, we can consider how the CR system generally works. In the first case, the enemy wields a weapon. The CR system states the enemy is almost always higher CR and therefore has a higher BAB than their PC counterpart. By your logic, this enemy is hurt more by the critical fumble rules than the PCs. If we next examine the case of a enemy using natural weapons, we can glean the effects of the critical fumble rules. Consider that most enemies using natural attacks are creatures using claws, teeth, tail, or some combination of these types of attacks. If you examine most creatures it can be seen that up to a decently high level the average animal has more attacks per round from natural attacks than a full-BAB PC of the same CR. Your reasoning results in the conclusion that the creature is worse off due to the critical fumble rules (I assume critical fumbles can apply to creatures using natural weapons).
Ignoring the fact that enemies most commonly hurt by the critical fumble rules and assuming the statement that it assists them is correct, consider the case of magic. Most casters have a limit on the number of spells they can cast per day via some variable game mechanic. In the case of enemies, the result is that they are usually dead by the end of combat and the PCs are still alive. Since the PCs must be capable of going on to fight further battles, it would be inadvisable for the PC spellcasters to expend all their spells in one encounter. Counter to this, enemies are expected to pull out all the stops to try and survive, expending all their spells and spell-like abilities in the single encounter. Resultantly, magic is ridiculously unbalanced. Martial combatants have no equivalent, even if the new martial maneuvers are included, so critical fumbles would be a good way to do that.
Higher level characters fumble their weapons more:
The probabilistic analysis does show that a character with more attacks per round is likely to fumble their weapon more. If one considers the case of critical hits, it also shows that the character is more likely to critically hit as well. This can be seen by assuming the only way to successfully critical hit is to role 2 natural 20's in a row. Now the chances of doing this once are 1/400, but the probability does go up with the number of attacks (1/400*number of attacks). As a result and assuming the character is crit-able, the probability of scoring a critical hit on any character, no matter the level difference, goes up as well. This is very unbalanced and needs something to offset it.
Natural 1s and 20s -- a solution to critical fumbles:
Rolling natural 1s and 20s normally have the effects of automatically missing or hitting regardless of attack bonus. This seems well balanced, but critical hits of course need a balance as well. Consider the mechanics involved in critical hits. If you roll within the critical range, assuming it would actually hit the enemy and the enemy can be critical-ed, you roll again to see if you hit. If you hit the second time as well then you just confirmed a critical hit. What could possibly be analogous to this for critical fumbles? Since the only guaranteed critical hit occurs when 2 natural 20s are rolled in a row, the only guaranteed critical fumble should occur with equal probability. Since there is equal probability of rolling any given number on a die, 2 natural 1s should suffice.
The only detriment would be against enemies that are not susceptible to critical hits. The probability of scoring a critical hit would drop to zero, but the probability of critically fumbling would be the same. Considering this conundrum from the enemies perspective, without critical fumbles he has a chance of critically hitting you, but you don't have any chance of critically hitting him. With critical fumbles he now has a chance of horribly failing as well. Since you also have this chance, they balance and you're back to the normal problem of being unable to critical certain enemies. This solution seems reasonable for levels below which most people have moderate to high fortification items/armor, and is actually a leniently balanced mechanic since there is no chance of critically fumbling other than on 2 natural 1s where as there exists a chance of critically hitting based more than just 2 natural 20s.
Note that this holds true for auto-kill rules as well, whether you use 20-20-20, crit-20, or whatever. When a PC benefits from the rule, okay, they saved some daily resources, but that's about it. When the enemy benefits from it, though, the party suffers greatly. It's the disposable nature of monsters that makes the difference, not any degree of math.
I use an iterative base attack system that makes this fit very well regardless of what attack fumbles, but with the normal D&D system you may want to rule that the character uses his highest base attack to confirm the fumble (modified by anything the character was doing to alter it such as power attack, expertise, etc..) I'm not sure if that would work favorably, but off hand it’s the quickest fix I can see to keep high levels from always fumbling their last attacks.
Furthermore, the consequences are not proportional. A sudden random failure on the part of a bad guy has minimal consequences on the campaign. He was supposed to die anyway. Random PC failures lead to deaths with permanent consequences (level lost, etc.) that last for the rest of the campaign. A string of random PC failures can cause a TPK, ending the campaign.
To me, the "automatic miss" on a natural 1 is plenty sufficient to model the effects of luck in a melee attack. A commoner won't even notice the rule, he can't hit anything on a "2" anyway. More powerful fighters won't be that badly affected, missing one attack in four isn't as bad as missing everything. But even so, it comes up often enough on single attacks that "should" succeed (ray spells vs touch AC, touch attack for grapple or trip checks, etc) that no attack is certain.
That's enough. You don't need people lopping off their own arms, or striking teammates or disarming themselves, etc. People just don't do that in real life and there isn't any reason to have it happen in heroic fantasy either.
If you have a critical fumble rule, a lance charge with 100 horses will have 5 of them doing something much more serious than just missing their target. That doesn't happen. While freak mishaps do happen, they're far less frequent than a d20 can show. (Maybe 1 in 2000, or 20000, but not one in 20). The equivalent of that in D&D is just missing repeatedly, by a string of natural ones. That's usually plenty bad enough (wasted actions have huge consequences).
It's true that you have a greater chance to fumble per round. What bothers me here is the coin flip analogy: every time you flip a coin, you have a 50% chance of flipping a head. It doesn't matter that in a sequence of N flips, the odds of getting at least 1 head in the sequence increases. The coin flips are still independent.
Here, a more advanced fighter gets more opportunities to attack. The odds of rolling a 1 are the same, 5%, each time. It doesn't matter that performing more attacks gives you more opportunities to fail - you're still performing more attacks. That's the benefit for gaining experience. If you fumble one attack, and you have some attacks left, the fact that you fumbled should have no bearing on the success of the next attack.
I personally don't see anything wrong with that. You have a flat 5% chance to fumble every attack. Gaining experience allows you to make more attacks. That seems fair enough to me.
It means that while a low-level fighter might only fumble once every few battles, a high-level fighter is far more likely to fumble perhaps as much as once per encounter. It makes the fumbles a much greater problem for higher-level fighters than for lower-level fights, and it really doesn't make any sense at all. Your standard mook I can see dropping his crossbow and shooting himself in the boot, maybe. Would Legolas ever accidentally shoot Gandalf? Would Aragorn ever slip and cut off his pointer-finger by accident?
High level fighters don't screw up like that. It's theoretically possible, but it almost never ever comes up.
If you roll a 1, then another 1, you roll a percentile dice.
Character level 1-5, you have a 4% chance to critically fumble.
Character level 6-10 you have a 3% chance.
Character level 11-15 you have a 2% chance.
Character level 16-20 you have a 1% chance.
Character level 21+ never critically fumbles.
So now I have proven that your Caramel-Flavored Critical Fumbles has less nutrition than my Vanilla Fumblelessness. I'm not telling you that you shouldn't eat what flavor you like, but don't be offended, nor try to force people to eat your Caramel-Flavored Critical Fumbles, when you know, now, that they aren't good for you, and you choose to use them merely because you like the flavor, not because it has nutritional value. I personally don't care much for Caramel-Flavored Critical Fumbles myself, and wouldn't ever eat it knowing its lack of nutritional value.
Cute analogy, except that it's no good. Nutrition contains a value judgment. If you're intending to be "impartial" in this exercise (although it doesn't really seem like you are), avoid value judgments.
For what it's worth, I support Stargate's recovery system. By making the "confirmation" roll a BAB check vs. a set DC (I recommend 13), fumbles do in fact become impossible at a given point for warriors of great enough skill (like +12 BAB). That fixes your "unfixable" flaw.
I think his thread should address the issue well enough for at least 75% of the groups out there.I have a question, and a method of a 'recovery system' that you might not have thought of. Why not just have the player need to 'confirm'* his fumble against a set DC using his highest BAB? This makes it so that the number needing to be rolled decreases over time, making it harder to 'confirm' the fumble as your level progresses, and since you're using only the highest base attack bonus for this roll, it will always reflect the best you could possibly be.
If you click the link to “otto the bugbear's critical fumble system,” you'll note that that's exactly what I suggested, while he actually took it and kind of added bonuses, I suggested straight BAB vs a DC. This still fails to address the issue of being able to fumble twice in a round, as opposed to not being able to do it at all. I used the numbers represented in that thread to actually make this one.
It's true that you have a greater chance to fumble per round. What bothers me here is the coin flip analogy: every time you flip a coin, you have a 50% chance of flipping a head. It doesn't matter that in a sequence of N flips, the odds of getting at least 1 head in the sequence increases. The coin flips are still independent.
Here, a more advanced fighter gets more opportunities to attack. The odds of rolling a 1 are the same, 5%, each time. It doesn't matter that performing more attacks gives you more opportunities to fail - you're still performing more attacks. That's the benefit for gaining experience. If you fumble one attack, and you have some attacks left, the fact that you fumbled should have no bearing on the success of the next attack.
That's exactly where the problem comes in. You have a flat chance per attack to fumble (or threaten a fumble). As a fighter increases in levels and becomes more skillful, he is rewarded by having more opportunities to attack per time given.
I have heard the explanation of combat as being an abstraction where you're attacking, you're making more than 1 attack in those 6 seconds, but at a low BAB you're only skilled enough to make one attack that might score significant damage. It's not like you attack in 1 second, and then sit there for another 5 seconds, As you gain more attacks, that represents your ability to get more attacks in.
You don't need to agree with the abstraction, but it does show, that rolls of the dice are meant to actually resolve combat, and it uses math. When the math is in the favor or Mr. level 1 commoner, over Mr level 20 fighter, something needs to be done to fix the math.
Really the point is logically: You have two people. One has been fighting and studying fighting for years and knows the ins and outs of everything. The other just picked up his weapon for the first time. They fight each other for 1 minute's time. You would think that the clumsy newbie would be fumbling around his weapon quite a bit during that minute while the experienced fighter probably wouldn't even do it at all, due to all the practice he's done.
Now you put this is D&D terms. You have two people. One is a level 20 fighter. The other is a level 1 commoner. They fight each other for 1 minute's time (10 rounds). The fighter makes 40 attacks, and has a 5% chance with each to fumble, while the level 1 commoner makes 10 attacks, and has a 5% chance with each to fumble. The fighter, statistically, will kill himself, break his weapon, trip and fall, cut himself, drop his weapon or whatever kind of fumble more often than the commoner.
That's not logical. [i]At all[/i]. The commoner is less skilled than the fighter so they should at least have the same chance of failure, if not the fighter have less.
This doesn't make sense because not only does the fighter get a better chance to hit as his level increases, but he also gets more attacks. He is so skilled he can make more attacks, all at better accuracy than the commoner. He shouldn't be penalized for using the mechanic that is supposed to represent his increase in skill.
It's simple. If you allow critical fumbles on a natural 1, you are introducing a mechanic that is statistically worse for those who are more skilled at fighting.
Your argument is that with anything if you attack more, you increase your chances. Yes you do, but the mechanics introduce more attacks as you get better, so the fumble mechanic needs to take that into account.
You're looking at it with tunnel vision. A commoner fighting to the BEST of his ability gets 1 attack. The fighter attacking with the BEST of his ability gets 4 attacks. The fighter is not "simply attacking more times" he's attacking AT THE BEST of his ability.
When you have a coin flipping contest, to see who's the luckiest, you tell everyone to flip the coin 200 times. You don't say "ok flip it as many times as you can, and whoever has more heads loses," because then everyone will just flip it 0 times, because they can't get any heads. If you set a quota - a minimum number of flips that need to be flipped, then everyone will stop at that minimum. If everyone goes as fast as they can, the slowest, least skilled at flipping will be rewarded by having the highest probable chance of winning - which is completely counter-intuitive.
Higher level enemies:
Your statement that critical fumbles hurt PCs more than enemies is faulty. Let us consider 2 cases, one where the enemy is a weapon wielding creature and one where the enemy uses natural weapons. Assuming your statement of having more attacks leading directly to more critical fumbles and greater harm, we can consider how the CR system generally works. In the first case, the enemy wields a weapon. The CR system states the enemy is almost always higher CR and therefore has a higher BAB than their PC counterpart. By your logic, this enemy is hurt more by the critical fumble rules than the PCs. If we next examine the case of a enemy using natural weapons, we can glean the effects of the critical fumble rules. Consider that most enemies using natural attacks are creatures using claws, teeth, tail, or some combination of these types of attacks. If you examine most creatures it can be seen that up to a decently high level the average animal has more attacks per round from natural attacks than a full-BAB PC of the same CR. Your reasoning results in the conclusion that the creature is worse off due to the critical fumble rules (I assume critical fumbles can apply to creatures using natural weapons).
That is a good example, but it's only true when comparing natural weapons to manufactured weapons - not the big picture. Anything that has a random large effect hurts the PCs more. Why? Because if an enemy critical fumbles and causes [insert negative effect here] and he is easier to kill, ok, well you were only going to kill it in the first place, that just made it easier. If a player critical fumbles, he it could cost him his money, treasure, or even life!
This is direct offshoot of the concept that "harder critical hits" and similar things, hurts the players more than helps them, because they have to live with the repercussions of being on the negative side of the stick, while all NPCs and monsters are, in fact, expendable, while the characters are not.
Higher level characters fumble their weapons more:
The probabilistic analysis does show that a character with more attacks per round is likely to fumble their weapon more. If one considers the case of critical hits, it also shows that the character is more likely to critically hit as well. This can be seen by assuming the only way to successfully critical hit is to role 2 natural 20's in a row. Now the chances of doing this once are 1/400, but the probability does go up with the number of attacks (1/400*number of attacks). As a result and assuming the character is crit-able, the probability of scoring a critical hit on any character, no matter the level difference, goes up as well. This is very unbalanced and needs something to offset it.
I pointed out that it's illogical and statistically worse, and thusly unbalanced than critical fumbles affect characters as they increase their skill. If you don't agree that characters are becoming more skilled as you level, then that's a pretty tough claim to back up. But as characters gain levels they SHOULD be getting better at critting, if not at least staying the same. They are getting better.
If you're going to try to put "balance" on the table, then you might as well consider things where a wizard will never ever ever ever fumble his fireball, meaning out of all those gestures that he has to do he ALWAYS does it flawlessly, unless forced to make a concentration check. Casters are already seen as more powerful than melee, and melee is subpar, with possible exception of the first few levels. How would it be "very unbalanced and need to be offset" of melee actually being able to crit more often as they level? In fact, there are many cases, that a crit is simply just overkill, and didn't even matter that you did double to quadruple damage, because the original hit would have killed it anyway.
--------------
Again, I apologize for the poor structure, if I had organized it a bit better it might be clearer to see the whole big picture.
A natural 1 is an automatic miss and a potential critical fumble. You roll to confirm it.
A result below 5 means that you handle your weapon so poorly that in case of non-magical weapons they break and you risk hitting yourself. You roll a third d20. A result below 5 means you hit yourself while a natural 1 means a critcal hit on yourself.
This system is fun at low levels (you can break your weapon quite easily and replacing it in the middle of nowhere can prove quite a challenge), while at high levels(when most PCs have magical equipment) the chance of something going wrong is not so big(you have to roll below 5 on 3 d20 in a row). So far we've had only one victim of this system: an about 10th level half-dragon orc barbarian who scored a crit on himself with a greataxe (the damage was something like 3d12+60).
If only the first attack can be a fumble (you rolled a natural 1) and you have to confirm the miss you get a scenario where the risk of fumble is 5% * (chance to miss your first attack). As a player raises in lvl the dynamic of CC and hit/miss in D&D generally changes. As a melee character is more and more likely to hit with his/her first attack the (chance to miss your first attack) goes down. This means that as the players lvl the amount of fumbles go down.
Since a fumble at our table means that you provoke an attack of opportunity it usually doesn't kill the fumbling character either. :angelhide
* Basically special AC threat monsters increase the chance of fumbling and since some CR have more of those some lvls are slightly more prone to fumbles depending on how the DM uses his or her MM's.
How does this help? Add your highest BaB to the d100 roll to get the result. That way a lvl 10 Fighter never has a chance to insta-kill himself, while a 10th level wizard has a 5% chance of doing just that. If you get a 100+ result you manage to avoid the situation entirely and resolve the attack as if you could get a 1 on your attack roll.
One remedy for this is requiring a straight DC 10 Recovery check using his BAB+Dex. Failure confirms the critical fumble. For the sake of illustration, we will compare 4 characters: Ftr1, Ftr6, Ftr11, and Ftr17; all having Dex 12. For first approximation, assume the rule "only one critical fumble allowed per round" but allowing it to be any attack roll. The second approximation computes the total number of fumbles expected over 10,000 turns.
Ftr1
BAB +1
Recovery +2
He has a 5% chance of rolling nat 1, with a 35% chance of failing his save. So we expect he will fumble critically1.75% of the time.
With multiple fumbles allowed, we expect 175 fumbles for every 10,000 turns (in this case, one turn is one attack)
Ftr6
BAB +6/+1
Recovery +7
He has a 10% chance of rolling at least one nat 1, with a 10% chance of failing his save. So we expect he will fumble critically 1% of the time.
With multiple fumbles allowed, we expect 95.5 fumbles for every 10,000 turns (here one turn is TWO attacks).
Ftr11
BAB +11/+6/+1
Recovery +12
He has a 15% chance of rolling nat 1, with a 5% chance of failing his save (only on a second nat 1). So we expect he will fumble critically 0.75% of the time.
With multiple fumbles allowed, we expect slightly more than 68 fumbles for every 10,000 turns (here one turn is THREE attacks).
Ftr17
BAB +17/+11/+6/+1
Recovery +18
He has a 20% chance of rolling nat 1, with a 5% chance of failing his save (only on a second nat 1). So we expect he will fumble critically 1% of the time (an increase over 16th level for sure, however...).
With multiple fumbles allowed, we expect slightly more than 22 fumbles for every 10,000 turns (here one turn is FOUR attacks).
Cheers,
StoneStokes
I personally don't see anything wrong with that. You have a flat 5% chance to fumble every attack. Gaining experience allows you to make more attacks. That seems fair enough to me.
The problem starts at the point where rolling a 1 means more than simply not hitting - I see it as very wrong when an experienced fighter has in the end a chance three times higher to disarm, harm or do otherwise bad things to himself than the much-quoted level-1-commoner in the same time span. Hell, I regard it as quite faulty if said warrior has the same chance as the commoner!
I have a question, and a method of a 'recovery system' that you might not have thought of. Why not just have the player need to 'confirm'* his fumble against a set DC using his highest BAB? This makes it so that the number needing to be rolled decreases over time, making it harder to 'confirm' the fumble as your level progresses, and since you're using only the highest base attack bonus for this roll, it will always reflect the best you could possibly be.
The problem I see here is that a single DC is not wide enough - if we have a flat DC of, say, 25, lowlevel characters would have no chance of not confirming a fumble while highlevel ones would have no fumbles whatsoever. This merely creates a zone in the midlevels where a true chance of avoiding fumbles exists. It may be a way of implementing a fumble rule, but it feels... bad to me.
Higher level enemies:
Your statement that critical fumbles hurt PCs more than enemies is faulty. Let us consider 2 cases, one where the enemy is a weapon wielding creature and one where the enemy uses natural weapons. Assuming your statement of having more attacks leading directly to more critical fumbles and greater harm, we can consider how the CR system generally works. In the first case, the enemy wields a weapon. The CR system states the enemy is almost always higher CR and therefore has a higher BAB than their PC counterpart. By your logic, this enemy is hurt more by the critical fumble rules than the PCs. If we next examine the case of a enemy using natural weapons, we can glean the effects of the critical fumble rules. Consider that most enemies using natural attacks are creatures using claws, teeth, tail, or some combination of these types of attacks. If you examine most creatures it can be seen that up to a decently high level the average animal has more attacks per round from natural attacks than a full-BAB PC of the same CR. Your reasoning results in the conclusion that the creature is worse off due to the critical fumble rules (I assume critical fumbles can apply to creatures using natural weapons).
The problem here is that while a monster usually fights exactly a single fight and is then replaced by a new one that does not have to deal with the penalties the old one may have incurred due to critical misses, characters fight a campaign full of battles. Meaning that, yeah, in the context of the game, characters do many more attacks than any single monsters. It's the same with save-or-die attacks: One hitting a monster means a fight ends the way it probably would have ended anyway; one hitting a character means "end of story" when no cheap resurrection is available.
Ignoring the fact that enemies most commonly hurt by the critical fumble rules and assuming the statement that it assists them is correct, consider the case of magic. Most casters have a limit on the number of spells they can cast per day via some variable game mechanic. In the case of enemies, the result is that they are usually dead by the end of combat and the PCs are still alive. Since the PCs must be capable of going on to fight further battles, it would be inadvisable for the PC spellcasters to expend all their spells in one encounter. Counter to this, enemies are expected to pull out all the stops to try and survive, expending all their spells and spell-like abilities in the single encounter. Resultantly, magic is ridiculously unbalanced. Martial combatants have no equivalent, even if the new martial maneuvers are included, so critical fumbles would be a good way to do that.
You are forgetting that only few characters have no ways to expend resources. The most common currency of martial-guys is HP, but there are others - barbarian rages, paladin smites and a rogue's wand charges. So what remains? The fighter. Compare a cleric or wizard to this poor guy and tell me again the fighters need to be downsized.
As a result and assuming the character is crit-able, the probability of scoring a critical hit on any character, no matter the level difference, goes up as well. This is very unbalanced and needs something to offset it.
Why does it do so? I always thought going up in level means "getting better", not "getting some abilities that have to be absolutely balanced with drawbacks so you stay the same".
Your statement that critical fumbles hurt PCs more than enemies is faulty.
You are missing the meaning behind this. The meaning is this: The game is stacked in favor of the PCs. If you break everything down to it's pure statistic outcomes, and remove every ounce of randomness from the game, the PCs should win.
Now, that wouldn't be any fun, and so an element of randomness is thrown in. D20s for attacks, saves, skills checks, etc, and damage rolls. This throws some excitement into the game. However, every amount of randomness that is included gives a greater and greater chance of a statistical outlier occurring: the PCs losing.
This is because the PCs are the ones that are supposed to go through multiple encounters and survive. Basically, if an enemy suddenly crit-fails and impales himself on his own sword, the outcome hasn't really changed; the PCs were supposed to win anyways, they've just done so a little quicker and without expending as many resources.
If a PC suddenly crit fails and impales himself on his own sword, suddenly the intended outcome has been drastically altered. This is why crit failures hurt the PCs more than the monsters. This is true of any extremely random event, like save or die spells or vorpal weapons.
The D&D battle system functions in a way that a character gets so many attempts at hitting their opponent to deal damage in one 6 minute interval, or otherwise known as a turn.
I think you mean 6 second.
-EE
If they succeed in the check, the result is usually something humorous but not mechanically different from "you missed"; in fact, for highly skilled or magically enhanced (e.g. True Strike) warriors, it might still result in a hit against a lightly armored foe ("1" is treated as "-10"). For example, perhaps your blade passes harmlessly under the target's armpit, but skewers the target behind him (oops... hope it was an enemy!). Perhaps instead the foe's last attack your grip on the hilt, but your quick reflexes allowed you to recover the weapon before he could take advantage of your slip.
If they fail in the check -- and degree of failure is a guideline -- then I get to throw in more severe results... or not, as I choose. At a minimum, the attacker loses all remaining attacks for the round.
However, "Vanilla with Chocolate Chips" does not address the issue that more attacks = greater chance to fumble in a round. This is true, but roughly half of all "1s" in my game result in the attacker stopping -- only 1 fumble is possible, because no more attacks occur. And since the opponents usually have more attacks (natural weapon routines, or higher BAB), my players seem to like it. Some would prefer a nice Dark Chocolate ("oh, you fumbled? roll critical damage against yourself!"), some like Vanilla but put up with the small chocolate chips, and all are content that they got to have some ice cream.
Mathematically and mechanically though, I believe I agree that "No critical fumbles" is the best way to go. You miss 5% of the time, regardless of your skill; more skill gives more attacks, which means there is a greater chance of a hit in your routine, despite that 5% chance of a "1".
- A character can only check to fumble once in a round - all other Nat 1's thereafter are a normal miss. It does not matter if a critical fumble actually occurred or not.
- Confirm fumble with your current attack bonus vs DC 20. Do this as many times as you have attacks per round until you either confirm the fumble (fail to rol 20+) or fail to confirm (rolled 20+). It does not matter if you are making a full attack or a single attack - check up to as many times as you have attacks as if you were making a full attack.
- On a confirmed fumble, use whatever results you wish to use (random chart, etc).
- A critical fumble that has been successfully checked against (rolled 20+) is a normal miss.
This method works exactly the same as confirming a critical hit, and the AC is assumed to be 20. Using current as opposed to highest attack bonus simulates that even a high level character may fumble as badly as a 1st level commoner if the fumble occurs on the lowest attack bonus.
This is the method that works for me and doesn't punish higher level characters in that they have more chances to devalue a critical fumble into a normal miss, and yet it remains a simple system (at least for me).
Your players are present in every single encounter in the game (one would assume) and thus over their character's lifetime have a MUCH higher chance to score multiple "critical fumbles".
The lifespan of any given baddie is relatively short. The PC's blow through encounter after encounter, and eventually their Crit Fumble number will come up. That bugbear that took you a round to kill though, has a relatively small chance of fumbling.
In This way, it hurts the PCs more, not because of some mathematical imbalance, but because of the sheer difference in dice-rolled-over-lifetime.
For example, consider a system where a natural 1 is a "critical miss threat." If you confirm the critical miss your attack misses and you lose all remaining attacks in a round (possibly along with other unfortunate results, if the DM so desires). If you don't confirm the critical miss the attack still misses, but otherwise your round continues as normal. If the chance to confirm a critical hit is static, then yes, things do get worse for the attacker. What is reasonable metric to use? The average proportion of a full attack retained despite critical misses fits this criterion (though others might as well). For example, if a character with rolls a 1, assume he always critically misses. This means the proportion of his full-attack he can expect to retain on average is .95, since 95% of the time he retains 100% of his full attacks, but 5% of the time he loses all of them. But if the same character takes 2 attacks, he retains on average just .92625 of his full attack. The proportion retained for n attacks if every 1 is automatically a critical miss is:
(1/#attacks)*Sum(.95^i,i goes from 1 to #attacks)
But what if the chance of critically missing changes? Well, if pc represents the chance of a critical miss we can replace .95 with (1-pc). Since BAB represents a character's overall skill at attacking and changes relatively infrequently, I make pc a function of BAB. Because this chance is dynamic, in an actual game setting it must be easy to calculate, and so I represent pc in the following form:
pc=.05*(j-k*bab)/100 where j,k are real numbers (to be fitted with integers later)
The form above assumes every 1 is a critical miss threat, and that the player will be rolling a d100 to confirm, which, it turns out, provides adequate granularity for the system to scale from BAB 0 to 20.
A further question is what I would like the fundamental chance of critically missing to be. For simplicity, I assume that, over the entire game, the average proportion of a round remaining is the same as a single attack at BAB 0. Thus, this target value is (1-pc(bab=0)) = 1-.05*j/100.
The goal is to find, for a particular target proportion, the value of k that minimizes the square errors. For each BAB and its corresponding # of normal attacks (i.e. 2 at BAB 6, 3 at 11, and so on) I calculate the proportion in terms of k, and then take the sum of (proportion-target)^2 over all 21 base attack bonuses, and minimize this function with respect to k. By doing this procedure for every j from 0 to 100 I generate a list of the k's that optimize for 101 different target values, and look for those k's that are most nearly integer values. Of particular interest are the (j,k) pairs (29, .997111), (58, 1.98518), and (88, 2.99756). These have target proportions, respectively, of
.9855, .971, and .956.
Which pair makes the most sense to use depends wholly on how often you want critical misses to appear in the game. Since the change in the root mean square error is small for small changes in j rounding these values to the nearest 10 makes for almost no statistical difference.
If I were to use this method in my own game, desiring a real risk of losing your whole turn, I would use (j,k)=(90,3). This has a target proportion of .955, and over BAB from 0 to 20 an average proportion of .954102. Thus, over the whole course of the game, no matter one's BAB, one can expect to lose very nearly 4.5% of any given full attack because of critical misses.
Other ways of calculating these probabilities make sense as well. For example, what if I wanted instead to not just count attacks lost due to critical misses (that, as used above, cause one to lose all further attacks in a round) but also automisses that don't affect future attack rolls? This would have a more involved function for the proportion retained, but the rest of the procedure is similar. I may work it out at some later point.
For a different system, where a natural 1 is treated as just an automatic miss (with some probability p) or as an attack that doesn't automatically fail (i.e. it's "just a 1", with probability 1-p) a similar analysis can be done. A quantity I might want to preserve over the standard BAB/#attack range is the probability that 1 or more automatic misses occur in any given full attack. If the probability of any single critical miss threat being upgraded to "just a 1" is (j+k*bab)/100 then a similar analysis gives the following (j,k) pairs: (32, 2.98058), (54, 2.02025), (77,1.01218). These give results that aren't quite as consistent across base attack bonus as might be desired, but they are all vastly improved compared to a static chance, which grows quickly for a large number of attacks.
It is worth noting that since this method attempts only to scale the various probabilities with the extra attacks gained through BAB, extra attacks from TWF, haste, etc. suffer the same problems as described in the original post. In the case of TWF I don't have much sympathy, since they can, at best, make a single attack with one of their weapons only as well as someone else without the feat. In this case, I think it makes sense that more attacks leads to a greater chance of problems in a round. Any system of scaled critical misses still helps them, however. In any case, a magical source of a single extra attack could have the effect mitigated by treating the character's BAB as 5 higher than normal when a critical miss threat comes up, representing the magical effect granting an increased ability to avoid critical misses. Finally, since the system is scaled only for levels 1-20, there should be some limits if attached to epic play. Examples might be a minimum chance of critically missing that is always present, or simply not allowing epic BAB increases to apply.
The problem I see here is that a single DC is not wide enough - if we have a flat DC of, say, 25, lowlevel characters would have no chance of not confirming a fumble while highlevel ones would have no fumbles whatsoever. This merely creates a zone in the midlevels where a true chance of avoiding fumbles exists. It may be a way of implementing a fumble rule, but it feels... bad to me.
But isn't that exactly what you advocate? A system in which a character goes from a high percent chance to fumble all the way to it being a practical impossibility?
For the record, high level characters could still fumble by rolling double one's, as the second one still means automiss.
I have heard the explanation of combat as being an abstraction where you're attacking, you're making more than 1 attack in those 6 seconds, but at a low BAB you're only skilled enough to make one attack that might score significant damage. It's not like you attack in 1 second, and then sit there for another 5 seconds, As you gain more attacks, that represents your ability to get more attacks in.
I'd like to take this opportunity to point out that, while this works in a lot of cases (and prevents the "I stand still for 5 seconds and swing my sword once" issue), the model breaks down when dealing with ranged weaponry. Unless, of course, you force archers to carry 4-6 times as many arrows as they actually get to roll attacks for.
So, I'm not too sure about your vanilla fumbleless system. But I've been playing with Crit fumbles since day 1. It’s a part of D&D to me...
- If you roll a 1, you are threatened with committing a fumble.
- Roll a d20 to confirm. If you roll equal to or greater than your HD, then you fumble.
- Combining these two rolls, the chance of confirming a fumble is 1/20x(21-HD)/20, or (21-HD)/400.
That confirmation roll of d20 vs your HD ensures that higher level characters have proportionally less chance to fumble, and epic characters won't fumble at all.
Using this rule, a 1HD character will still have a 1/20 (5%) chance to fumble per attack. A 2HD character will have 0.05x19/20 or 4.75% chance to fumble. A 20th level character will have a 1/400 chance to fumble, per attack. An epic level character (21HD or greater) has a zero or negative chance to fumble, so no fumbling is possible.
Where a fighter can make multiple attacks per round, a 1st-level fighter has a 1/20 (5%) chance of fumbling, and a 16-20th level fighter making 4 attacks will have a 1/100 (1%) chance of fumbling (0.0025x4).
That seems fair, no? There's still a small chance to fumble, but it becomes more remote at high levels. You can substitute character level for HD, but I personally think HD makes more sense, because a multiple-HD race at 1st level should have better attacking skills and therefore have less chance to fumble.
-A
In my game, if you roll a one, you roll again, and if the second roll would have missed, you messed something up. What exactly that was is up to me.
If there's fighter with four attacks, and one of his first three is a critical miss, I generally take away the rest of his attacks, saying he messed up the timing on his attacks, or got his sword caught unexpectedly in a chink in the opponent's armor, or something of the sort. Aside from the lost attacks, there is no penalty. If it's a wizard with a crossbow, however, he might accidentally shoot his allies. Bowstrings often snap on natural ones as well (a bad thing to happen, but after combat, replacing the bowstring costs less than a silver), and occasionally a critical miss can give the opponent a free disarm attempt or attack of opportunity. By not using a table or a rule, I can customize the failure to the situation. My players have learned to avoid pillars, because they have a tendency to get axes stuck in them.
Basically, I try to handle fumbles by ad hoc-ing what exactly the problem is, and if I can't think of a good way to explain it (How the heck does a wolf critically miss on a bite attack?) I ignore it entirely. As long as the players trust me as a DM not to be cruel and petty (which they generally do) critical misses are simply ways to add fluff to a generally crunchy battle.
Ftr11
So we expect he will fumble critically 0.75% of the time.
With multiple fumbles allowed, we expect slightly more than 68 fumbles for every 10,000 turns (here one turn is THREE attacks).
Ftr17
So we expect he will fumble critically 1% of the time (an increase over 16th level for sure, however...).
With multiple fumbles allowed, we expect slightly more than 22 fumbles for every 10,000 turns (here one turn is FOUR attacks).
Help me out here. Statistics has never been my strong suit. The 17th level fighter has a slightly higher percentage of confirming a crit than the 11th and he has one more attack per round than the 11th level fighter. How does that work out to less fumbles in 10,000 turns if you let him fumble multiple times per turn? I'm not saying you're wrong, I'm just curious as to how that happens.
-----
And now, for my own 2 CP on the issue:
In the past, we've always used the method of confirming a fumble by making a reflex save DC 10. Success means you "auto miss" and lose the rest of your attack sequence, failure means something else happens (usually your weapon slips and goes flying out of your hands or your bowstring snaps.) We've never had issues with that, and no one has yet complained about higher level characters fumbling more often, although I do agree the odds favour that result. Of course, we rarely end up damaging our allies and we never end up hurting ourselves, so for the most part it doesn't bother players that much.
Having read this thread, I think I'll change the system to use BAB instead of Reflex Save. I might up the DC to 15, too. I agree that a character's chance of confirming the fumble ought to go down as he increases in level. However, I also think the confirmation should use the same modifiers as the original attack roll (like crits), which means that if you fumble on your fourth attack of the round, you have a greater chance of confirming than if you had done on your first attack of the round. That makes sense to me. At higher levels you learn how to effectively strike more often in a 6 second interval, but your later attacks are worse than your first (as represented by lower BAB), so the odds of fumbling if you do slip up ought to be that much worse. Since this confirmation uses all the same modifiers as the attack roll, this also has the benefit of making it statistically impossible to fumble when True Strike is in effect. Again, this makes sense to me. Granted you can still miss with True Strike in effect, but then you can do that with RAW anyway, so I don't feel too bad about it.
I realize that none of this addresses the OP's concern about "using fumbles hurts the PCs." Maybe it does, but we find that it adds a certain amount of fun to the game, even if it makes the PCs that much more likely to lose. At the end of the day, that's all that matters, yes? As long as everyone is having fun. And, all statistics aside, the fumbling system we use has not, as yet, caused any PCs to die or caused an encounter to be lost (by itself, anyway.) There have been a couple of situations where an early fumble convinced the party to retreat, but those were all situations where the party knew they were getting in over their heads anyway. The fumble just convinced them to think twice about where the line was between bravery and stupidity.
Ftr11
BAB +11/+6/+1
Recovery +12
He has a 15% chance of rolling nat 1, with a 5% chance of failing his save (only on a second nat 1). So we expect he will fumble critically 0.75% of the time.
With multiple fumbles allowed, we expect slightly more than 68 fumbles for every 10,000 turns (here one turn is THREE attacks).
Ftr17
BAB +17/+11/+6/+1
Recovery +18
He has a 20% chance of rolling nat 1, with a 5% chance of failing his save (only on a second nat 1). So we expect he will fumble critically 1% of the time (an increase over 16th level for sure, however...).
With multiple fumbles allowed, we expect slightly more than 22 fumbles for every 10,000 turns (here one turn is FOUR attacks).
Cheers,
StoneStokes
I have a question: How does the fighter who has the .75% chance of critically fumbling fumble 68 times per 10k, while the fighter with a 1% chance fumble 22 times?
I have a question: How does the fighter who has the .75% chance of critically fumbling fumble 68 times per 10k, while the fighter with a 1% chance fumble 22 times?
I don't know... Because rolling a d20 4 times, you don't have a 20% chance to roll a 1, you have ~18.5% chance to roll at least one natural 1. Statistics of multiple events happening in succession works a bit differently than simply one event happening.
I noticed a lot of people propose threatening a critical fumble or what I call a "critical fumble recovery system." This post operates under the principle that you have a 5% chance to fumble once in a round at 0 BAB, and as you raise your BAB it either stays the same or decreases. As an alternative side note, every time you add another attack your chance to fumble jumps, while BAB scaling type recovery systems give a small increase per level. This means the "sweet spots" of recovery systems will always be, when it's based on BAB, and not the number of attacks, the level before you gain another attack.
Note that while the recovery systems may scale from poor to "getting there," without implementing multiple rules (such as you only can fumble once a round and you have a reasonable recovery system) most of the issues don't even come close to fixing the problem.
Indeed, like I said, if your systems work for you, good for you. But without addressing the issues that a level 20 truly does fumble more often than a level 1 commoner in the same amount of time fighting, it's just a disguise for critical fumbles as a way to make melee kill themselves a little less than implementing them without some sort of recovery system. It lessens, but does not remove the problem.
If the first attack was a critical threat, then any other extra attack that succeeded for the rest of the round was treated as a critical hit. Missing was raised to hitting for half damage, and only a natural 1 would miss.
Likewise, if the first attack was a threat to fumble, then any additional attacks that miss counted as a fumble and negated the rest of the character's available attacks for the round. Hits counted only as half damage, and only a natural 20 would count as full damage. If the first attack was a normal hit or miss, then the rest of the attack rolls were treated at face value, ignoring any special significance of natural 1s or 20s.
I still play this variant, and it works well. Seasoned fighters do not incur extra fumbles because they can only ever fumble once per round. However, the same seasoned fighter, if his first attack is a critical threat, becomes a whirlwind of carnage for the round.
http://boards1.wizards.com/showthread.php?t=466430
The first post is a clinical description to my summary up there. The system actually results in potentially fewer combat rolls of the die each round.
I'd also add that most crit fumble rules don't just hurt high-BAB types more, they also hurt two-weapon fighters more. In my opinion, TWF is already so inferior to using a greatsword for most classes that I don't think it needs any more penalties.
I'd also agree that melee types are already inferior to spellcasters, and making them consistently stab themselves in the face while the wizard never once accidentally magic missiles himself makes the situation worse.
Also, some people don't fully grasp the statistical details in some of their solutions. First, just using an "only one fumble per round" still leaves high level fighters with a greater chance per round to fumble. And if your fumble result includes denying them the rest of their attacks, they are being doubly punished.
Systems that use BAB vs a set DC are still mostly punishing stronger fighters more than weak ones. If the 20th level wizard is for some reason trying to hit something with his staff, it hardly matters if he hits himself with it instead. He won't hurt himself much when he fumbles. If he loses his iterative attacks, it doesn't really matter since it wasn't going to have much effect on the enemy. If your 20th level fighter with an acidic keen greatsword hits himself, he's going to be taking a ton of damage. When he loses his iterative attacks, it is a significant advantage to the enemy. So even though the wizard's got a greater chance to fumble than the fighter, it hardly matters since the wizard's primary purpose isn't fighting, he almost never does it, and when he does, it's already understood that it's going to be largely ineffectual. The fighter does nothing but attack, so even though his chance is less, it's going to come up all the time, and when he gets punished for attacking it will have a very negative effect on the battle.
I don't know... Because rolling a d20 4 times, you don't have a 20% chance to roll a 1, you have ~18.5% chance to roll at least one natural 1. Statistics of multiple events happening in succession works a bit differently than simply one event happening.
Can you elaborate some? It main thing I want to know is why the fighter who has a higher chance of fumbling fumbles less the one with the lower chance.
If you have a critical fumble rule, a lance charge with 100 horses will have 5 of them doing something much more serious than just missing their target. That doesn't happen.
There was a rather entertaining analysis of the Rolemaster (or possibly Runequest, it was a while ago) critical failure tables in a gaming magazine a while ago. In a massed battle situation with several hundred average soldiers fighting each other, both sides would suffer more casualties from critical failures than the enemy, including a couple who would somehow cut their own heads off.
My own beef with fumbles is that they ramp things up in favour of classes that don't need attack rolls. The casters can throw out area or targeted spells with no unusual risk of failure.
...
Also, some people don't fully grasp the statistical details in some of their solutions. First, just using an "only one fumble per round" still leaves high level fighters with a greater chance per round to fumble. And if your fumble result includes denying them the rest of their attacks, they are being doubly punished.
...
I'm simply curious how this can be (since my crit fumble variant fixer does this.)
If you can only fumble once per round, and the fumble denies the rest of the attacks... A} how do fumbles statistically occur more often for higher level fighters and b} considering that a fighter, using such a system, can get in multiple successful attempts with their extra attacks before potentially fumbling, how denying them their remaining attacks is overly harsh?
Can you elaborate some? It main thing I want to know is why the fighter who has a higher chance of fumbling fumbles less the one with the lower chance.
Explaining that is the simple concept of explaining statistics.
For example a 0 BAB commoner has 1 attack in one round, doing his best; we'll call this Weak Person. A 6 BAB fighter has 2 attacks in one round, doing his best; we'll call this Strong Person.
Weak Person attacks and rolls 1d20. There is a 5% chance that he will roll a natural 1.
Strong Person attacks and rolls 2d20. There is a 5% chance that he will roll a natural 1 on the first die, and a 5% chance also on rolling a natural 1 independent of the first die, on the second. There is also a chance that both are natural ones, which the Weak Person never had a chance to do.
For the non-statistically attuned person we'll make a simple chart of all the possible outcomes:
How easily a critical fumble can be obtained should depend greatly on how severe the penalty is for missing. The greater the danger by missing that occurs the harder it should be to miss. So if you critical miss and it causes you to hit yourself or ally I would consider that a pretty hefty penalty, but if the miss only presents a chance you may fall, drop your weapon, etc. then the chance can be increased. Remembering that the reason we play to have some excitement and fun, not defeat the players, make new character every game session etc.
When attempting to come up with systems that provided chances to cut an allies head off to just slip and fall I found that the game slowed too much. To provide a monster an AoO sounds good on the surface but what I have found is that at low levels this extra attack by the enemy can be catastrophic to the group. If suddenly the fighter is too weak to stand in the front row everyone is in jeopardy. Breaking every weapon is too frequent in my games as well.
My System:
I use a straight forward fast setup. If a character rolls a natural 1 they threaten to critical fumble. Next they roll to hit the targets Touch AC top confirm. If confirmed the weapon they are wielding takes damage as if they hit their own weapon (with all the str, magic, etc. added in) and the weapon takes damage subtracting hardness (a sword smashes into the floor, an arrow hit the bow and shatters, etc.) I the confirmation is another natural 1 then the damage is as if the weapon critical hit itself. After the weapon takes enough damage it breaks (may take multiple misses). This also allows the weapon to be repaired between adventures.
I do realize that some weapons (like daggers) are nearly impossible to break unless the wielder is very strong, but it does provide suspense without battle turning events on the first critical fumble. Now a player who doesn't take care of there gear will eventually have it break. I have been using this system for about 1.5 years and had only one magic weapon break and couple small weapons. But every time a player rolls a 1 everyone moves forward in their chairs with excitement and a bit of fear.
I'm simply curious how this can be (since my crit fumble variant fixer does this.)
If you can only fumble once per round, and the fumble denies the rest of the attacks... A} how do fumbles statistically occur more often for higher level fighters and b} considering that a fighter, using such a system, can get in multiple successful attempts with their extra attacks before potentially fumbling, how denying them their remaining attacks is overly harsh?
A) You are right that per attack, the number is exactly the same for all characters. But since high level fighters get more attacks per round, they also get more fumbles per round. To use simple numbers, if you just say that a 1 is a fumble, then someone who only has one attack per round will only have a 5% chance of fumbling per round. Someone with 4 attacks per round will have the same 5% chance of fumbling on their first attack of the round, but then another 5% on each other attack, for a total chance of about 18.55% of fumbling per round.
B) Well, this is more just my opinion than a mathematical thing. What I was saying is that in a typical game of D&D, the fighter's main damage comes from the fact that he gets multiple attacks per round. That's how his damage scales up. This is opposed to a class like the wizard, whose damage scales up purely by getting more powerful spells, not more spells per round. If you deny the fighter one or more of his extra attacks 18% of the time along with also making him break his weapon or hit himself, yet never deny the wizard any of his damage from spells or make him accidentally burn his spellbook or lightning bolt himself in the face, then you are piling lots of unfair penalties at once onto the poor fighter.
A) You are right that per attack, the number is exactly the same for all characters. But since high level fighters get more attacks per round, they also get more fumbles per round. To use simple numbers, if you just say that a 1 is a fumble, then someone who only has one attack per round will only have a 5% chance of fumbling per round. Someone with 4 attacks per round will have the same 5% chance of fumbling on their first attack of the round, but then another 5% on each other attack, for a total chance of about 18.55% of fumbling per round.
No no... look back... you stated that using a system whereby fumbles are limited to one per round that the higher level fighter fumbles more often. That's what I'm confused about. The chance to fumble on the first roll is 5%, then on their second roll it is still 5% for the roll, but the chances of rolling higher than 1 on the first roll, and then a natural 1 on the second is slightly less than 5%. So for the round, the chances of fumbling on each attack is 5% standing, but the chances of fumbling within a given round is more like 5~4.3% since a single fumble would negate the possibility to fumble again in the same round.
I hope my math is sound... the idea is this:
Chance to fumble first attack 5% (same as everyone.)
Chance to succeed first attack and fumble second attack 4.75%.
Chance to succeed first two attacks and fumble third attack 4.51%.
Chance to succeed first three attacks and fumble fourth attack 4.29%
The thing is probability is the same for every roll. Increasing number of rolls doesn't change probability. However, statistically speaking, limiting a player to one fumble per round actually lowers the number of fumbles per round a multi-attack fighter will experience by ever so little simply by capping the number of fumbles possible.
Am I wrong? I may need it explained. I've never taken a statistics course.
In my game, whenever a character or creature rolls a natural 1 it gives all opponents that threaten that character or creature an Attack of Opportunity for that attack roll.
This is essentially what I do, as well, Juggernaut, except that I allow an Attack of Opportunity for only the target of the fumbled attack. Limiting it in this way seems more balanced, and it's quicker.
I think the problem with almost all of the critical fumble systems is their veering off into varied results ("you drop your weapon in a random square"; "you slip and fall"; "you lose a move action" etc.). It's very difficult to balance (to quantify) the different lethalness of the different results.
Critical hits mean one thing: you increase damage. Therefore, if critical fumbles are to be included (beyond the extant auto miss), they should essentially mean one thing, as well. I think the marriage of critical fumbles to attacks of opportunity is a good, simple fit.
The thing is probability is the same for every roll. Increasing number of rolls doesn't change probability. However, statistically speaking, limiting a player to one fumble per round actually lowers the number of fumbles per round a multi-attack fighter will experience by ever so little simply by capping the number of fumbles possible.
Even if you limit the number of fumbles you can threaten to one, the more attacks you have the higher chance of rolling at least one one increases with each additional die rolled. What this means is that while you remove the ability to fumble twice, you're still upping the fumble chance by adding more attacks, which are rewards and shouldn't be penalties for being more skilled. This means that while he doesn't fumble phenomenally more than a level 1 commoner, he still does fumble more often per round than a level 1 commoner, even if it's a slightly reduced number.
You may go scroll up to see a simple table of the outcomes, you can see, explicitly that 2 rolled 20s have a higher chance of a 1 than one.
The chance to fumble on the first roll is 5%, then on their second roll it is still 5% for the roll, but the chances of rolling higher than 1 on the first roll, and then a natural 1 on the second is slightly less than 5%.
The chance of the specific event where the first attack succeeds and the second fumbles is slightly less than 5%, but the odds of a 1 coming up on the second die roll is exactly the same as that of it coming up on the first - 5%. The odds of witnessing a natural 1 on any attack is exactly 5%, regardless of how many attacks have been rolled before, and obviously, the more times you roll the more chances you have of witnessing the feared result. Therefore, the more attacks one has per round, the greater the odds of fumbling.
Your calculations are the odds of witnessing a fumble on a specific attack roll during a full attack action, which is not the same thing as the odds of witnessing a fumble on any attack roll during a full attack action.
I hope my math is sound... the idea is this:
Chance to fumble first attack 5% (same as everyone.)
Chance to succeed first attack and fumble second attack 4.75%.
Chance to succeed first two attacks and fumble third attack 4.51%.
Chance to succeed first three attacks and fumble fourth attack 4.29%
The thing is probability is the same for every roll. Increasing number of rolls doesn't change probability. However, statistically speaking, limiting a player to one fumble per round actually lowers the number of fumbles per round a multi-attack fighter will experience by ever so little simply by capping the number of fumbles possible.
Am I wrong? I may need it explained. I've never taken a statistics course.
You are wrong. You have calculated the odds of fumbling on the first attack in a round, the odds of fumbling on the second, on the third and on the fourth. Individually. These are four different events.
The odds of witnessing a fumble at any time is the union of the above four events. Since rolling a die multiple times produces independent results (the result of the second roll is not influenced by the first roll), the probability of seeing a fumble at any time is the sum of all individual possibilities. If you add up all four of your numbers, you'll find the sum is 18.55%
Also, note that if a particular character only has three attacks in a full attack action, then obviously the odds of him succeeding three times and fumbling on the fourth is 0%, since he doesn't have a fourth attack. If you add 5 + 4.75 + 4.51 + 0, you get the odds of fumbling with three attacks, which is less than the odds of fumbling with four attacks, as expected.
No no... look back... you stated that using a system whereby fumbles are limited to one per round that the higher level fighter fumbles more often. That's what I'm confused about. The chance to fumble on the first roll is 5%, then on their second roll it is still 5% for the roll, but the chances of rolling higher than 1 on the first roll, and then a natural 1 on the second is slightly less than 5%. So for the round, the chances of fumbling on each attack is 5% standing, but the chances of fumbling within a given round is more like 5~4.3% since a single fumble would negate the possibility to fumble again in the same round.
I hope my math is sound... the idea is this:
Chance to fumble first attack 5% (same as everyone.)
Chance to succeed first attack and fumble second attack 4.75%.
Chance to succeed first two attacks and fumble third attack 4.51%.
Chance to succeed first three attacks and fumble fourth attack 4.29%
The thing is probability is the same for every roll. Increasing number of rolls doesn't change probability. However, statistically speaking, limiting a player to one fumble per round actually lowers the number of fumbles per round a multi-attack fighter will experience by ever so little simply by capping the number of fumbles possible.
Am I wrong? I may need it explained. I've never taken a statistics course.
Hmm, I'm not sure what you're asking. The numbers you have there are correct for the chances of a 1 coming up, but you add them altogether for a total chance of 18.55% of at least one "1" coming up. The fact that you're going to stop rolling as soon as you get that 1 doesn't actually lower the probability of getting the 1.
Think of it like this: I say I'm going to flip a coin ten billion times, what are the chances that it will come up heads at some point? Pretty high. Now if I tell you I'm going to stop flipping as soon as it does come up heads, what's the probability? Still exactly the same. If it comes up as heads on the very first flip, that doesn't mean it's changed the probability of the whole thing to 50%. The probability's still 99% or whatever, because if it hadn't come up heads, I would just keep flipping it.
I agree with OP’s points, especially the basic notion that higher level characters, will, in fact fumble more often due to more attacks. However we have house ruled (maybe this is published somewhere as well) that on a natural 1, you need to make a DC 14 Reflex save to maintain your weapon. To remedy the multiple attacks thing=higher chance of rolling yet another 1 on the save, we eliminate the 1 equating in a fail on this particular save. So, as a result, higher level characters with low Reflex saves do drop their weapons more often than low level characters, because they are swinging more wildly, and low level chars who manage to gain boatloads of attacks (Read: Optimization Boards) will be the most frequent fumblers. This method also reduces the number of fumbles, and generally makes the higher level characters less prone to butter finger follies.
For the sake of argument, though, let's assume that the answer was "yes." What would be the best way to set up critical fumbles? We want a system that somewhat mirrors critical hits, but we also want to make sure that as you get more attacks (and theoretically better at melee combat), your chance of fumbling goes down.
Critical hits - whenever you get into critical range on any of your hits (usually a natural 20), you roll to hit and see if you crit.
If we try the same approach for fumbles, we will end up with the problem mentioned earlier. The problem is that we are essentially using a logical OR operator on each combat hit. The logical OR operator makes it so that as you get more hits the chances of it happening are increased (which is what you want for critical hits). For critical misses, we want the opposite, and should use the opposite operator, the logical AND operator. You should be checking whether they missed on all of their attacks rather than if they missed on any one of their attacks.
Of course, if we were to require that a character critically miss (roll a natural 1) on all of their attack rolls, critical misses become extremely unlikely pretty quickly (.025% for 2 attack rolls). The better solution is to check critical fumbles in the opposite order that you check critical hits. For critical hits, if you score in the threat range, you reroll to hit to see if you actually crit. For critical misses, if you miss on all of your attacks, you then reroll to see if you land in the threat range ( a natural 1).
The next step is to figure out what happens when you critically miss. For critically hitting, you deal more damage than a regular hit. For missing, though, you can't deal any less damage than the 0 you are already doing for missing. Since I am reluctant to have critical misses heal the opponent, damage must be done to my team. There are plenty of ways to have this accomplished. I think hitting yourself or your teammate is kinda lame. You could drop your weapon when you critically miss, but I think it would just make more sense to incur an attack of opportunity when you fumble. Swinging too hard and leaving yourself off balance was a significant problem in the middle ages.
If I were to implement something like critical misses into my campaign, it would probably work something like this.
Critical hits mean one thing: you increase damage. Therefore, if critical fumbles are to be included (beyond the extant auto miss), they should essentially mean one thing, as well. I think the marriage of critical fumbles to attacks of opportunity is a good, simple fit.
Instead of natural 1's causing the AoOs, having attack rolls that fail by more than 10 cause AOOs may be more elegant allowing for competent combatants with many attacks not suffer many fumbles while making mooks and fools who power attack for everything all the blasted time suffer greatly.
If we try the same approach for fumbles, we will end up with the problem mentioned earlier. The problem is that we are essentially using a logical OR operator on each combat hit. The logical OR operator makes it so that as you get more hits the chances of it happening are increased (which is what you want for critical hits). For critical misses, we want the opposite, and should use the opposite operator, the logical AND operator. You should be checking whether they missed on all of their attacks rather than if they missed on any one of their attacks.
I like this idea. Another system for fumbles which occurs to me now is this: You can only critically miss if you roll a natural 1 on your first attack (which will be at your highest BAB by default.) The confirmation roll uses all the same modifiers as that first attack roll and has a set DC of 15 (or so.) A second nat.1 is as auto fumble, failing to hit the DC is a fumble, anything else is a normal miss. Whether or not the attack sequence can be continued is up to the individual DM (although my preference is for "not.") Rolling a natural 1 on any subsequent attack is an ordinary miss.
I tend to play games that favour cinematic action over realistic combat. I don't think the d20 system is terribly realistic to begin with, so I don't use the d20 system when I want to tell realistic stories. This new fumble system I suggest here seems to solve the issue of better fighters fumbling less (since the confirmation is based on BAB and you only get one shot at fumbling per round.) It also makes good cinematic sense, to my mind, since you will either frack up spectacularly at the start of the round or else you will follow through more or less as normal.
This may have been said, if so, I apologize.
I agree with OP’s points, especially the basic notion that higher level characters, will, in fact fumble more often due to more attacks. However we have house ruled (maybe this is published somewhere as well) that on a natural 1, you need to make a DC 14 Reflex save to maintain your weapon. To remedy the multiple attacks thing=higher chance of rolling yet another 1 on the save, we eliminate the 1 equating in a fail on this particular save.
The problem with that is what I illustrated in the second part of the post. The fact that the flat DC 14 Reflex save does not alleviate the problem, rather it simply lessens it. The problem of fumbling more often even for a character with a good reflex save, is still higher than a 1st level commoner.
I will use a 1st level commoner and a fighter from level 1-20 to illustrate how your reflex save works:
Commoner has a 30% chance to make save.
- 1 Attack (Level 1-2) - Poor Reflex Save - Need a 0% chance or higher chance to make save to negate fumble
Fighter has 30% chance to make save. This reduces overall chance. It works for this BAB set.
_ - 1 Attack (Level 3-5) - Poor Reflex Save - Need a 0% chance or higher chance to make save to negate fumble.
Fighter has 35% chance to make save. This reduces overall chance. It works for this BAB set.
_ - 2 Attacks (Level 6-8) - Poor Reflex Save - Need a 50% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 2 attacks.
Fighter has a 40% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
_ - 2 Attacks (Level 9-10) - Poor Reflex Save - Need a 50% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 2 attacks.
Fighter has a 45% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
_ - 3 Attacks (Level 11) - Poor Reflex Save - Need a 65% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 3 attacks.
Fighter has a 45% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
_ - 3 Attacks (Level 12-14) - Poor Reflex Save - Need a 65% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 3 attacks.
Fighter has a 50% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
_ - 3 Attacks (Level 15) - Poor Reflex Save - Need a 65% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 3 attacks.
Fighter has a 55% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
_ - 4 Attacks (Level 16-17) - Poor Reflex Save - Need a 75% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 4 attacks.
Fighter has a 55% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
_ - 4 Attacks (Level 18-20) - Poor Reflex Save - Need a 75% chance or higher chance to make save to negate fumble to counteract the increased chance of rolling a one for having 4 attacks.
Fighter has a 60% chance to make save, this is not high enough, and the Fighter having a BAB in this range will statistically fumble more often than a 1st level commoner.
Overall? this means that with your system, your average fighter once he gets good enough to swing his sword twice on a fullround attack, he's still fumbling more than the commoner. Yes he's fumbling less than an auto fumble, but he's still fumbling more often than the commoner.
If the reason you put it in there was to remove the problem then it's not doing it, it's just lessening it. It's like putting on a spare tire when you get a flat. Yeah, it works better than driving on a flat, but it's no real tire.
These events give some flavour and excitement to battles. I know that the PCs are heroes and will always triumph against overwhelming odds but in all great books or films doesn’t the hero occasionally drop his weapon, have to use a different tactic because of a mishap. This adds to the drama. Small events such as twisting your ankle while attempting to attack would mean you could not run after monsters for the rest of the battle. However if that is what you do every battle suddenly you have to change tactic for example getting a long range weapon out instead to attack those far from you. This allows players to be creative with their battle plans and sometimes have to change them on the fly.
For example an Elf ranger in a seriously unlucky set of rolls broke his bowstring and twisted his ankle. While the dwarven fighter obviously sharing the bad luck fell over while attacking. This battle was won, wounds healed and weapons mended. but the players were taking about it for weeks afterwards because the other PCs had to save the dwarf and elf and the fact they survived through a dangerous situation by thinking on their feet.
However critical fumbles do not usually happen as often as that. You can create your own critical fumble tables with D%. (i use the v 3 fumble table) Give events with more possibility more of a range in the dice roll and if you choose you could have no rolls that equate instant death. Events such as dropping a weapon, falling over or even dazing yourself for a round can change the flow of battle. If you have ever find your PCs treat most battles with the same tactics, formation and even attacks each round this adds a little unpredictability.
In this system if we treat each natural 1 as a miss with a roll on the fumble table then all levels have the same chance of messing up badly. However if for every 5 levels a modified sheet is made, for example replacing really bad rolls with an increased range on the normal hit that simply misses you can alter the chance of messing up badly without changing the change of messing up. Reflecting that even a trained individual may have bad luck but may be able to recover from it better than someone of a lower level.
(How the heck does a wolf critically miss on a bite attack?)Simple. The wolf bites it's tongue.
In all of our games we have used the rule that rolling a 1 means that you left an obvious opening in your defence, and thus you incur an AoO, and your turn ends. Offhand, I cannot remember if it's the RAW rule or not, but it seems to work pretty well in our case. If you're using a bow or some such, the string snaps and you have to spend a standard action to repair it.
How the
do you repair a bow string in a standard action? Now I can see a bowstring coming loose from its mounting (this has happened with some bows I have owned) but if a string breaks and you don't have a replacement your (In my game it takes five rounds to replace a bowstring if you break one.)That was off topic but I needed it. Anywho, Critical fumbles are a house rule so it is a flavor and my flavor is cinnamon a d% chart with a 25% chance of no ill effect and the rest of the effects ending your turn so that you only get one chance at a fumble unless you didn't suffer from a fumble. And the reason I want them in the game is because if nothing ever goes wrong then the PCs just get victory after victory and sometimes I like to feel like I won even if I didn't because whoever died died cause they rolled a 1.
PS My crit fumble system doesn't allow instant death or broken weapons {except non-magical bows and who gives a
about non-magical bows} and no-one who has ever died in any of my campaigns have died in the same combat that they crit fumbled. Not saying it won't ever happen, but after playing 3 years it hasn't happened yet.Plus it's cinematic.
High level fighters can deal with it - quick draw, improved unarmed strike and other options let you recover quicker than your average Joe Shmoe.
Here's a table: irrelevant results are ignored so fumbles will end up slightly less then 5% per attack:
Roll 1d10
1-2 Weapon break chance: melee weapon - resolve as if the opponent made a sunder attack. Bow/xbow - string snaps. Black powder gun 50% jams, 50% blows up (take 1d3 damage), modern guns always jam. Thrown weapon - always breaks. Touch attacks, natural attacks, rays - ignore.
3-4 Ricochet Missile/thrown weapon, ray ricochets off target square (roll scatter and roll to hit using BA +0 vs first target in its path). All other weapons ignore.
5-6 Weapon stuck/entangled. Axes, picks or similar get stuck in enemy shield (if they have one) or into an adjacent wall/column or the ground. Standard action to unstick, shields with weapons stuck in them must be dropped or wielder must use a standard action to remove the weapon (can be a useful fumble!). Swords ignore. Missile weapons, rays, natural attacks ignore.
7-8 Missile weapons/rays firing into melee only: resolve hit against a random combatants (even someone the shooter has no LOS to since the combatants are moving.
9 Drop weapon chance: Melee weapons only, resolve as if opponent had made disarm attempt.
10 Nonproficient characters only: resolve hit against yourself.
Let me know what you think. Watch Bruce Lee's out-takes with the nunchuka - he fumbled!
There should also be a fumble table for spellcaster concentration checks....
How the
It's been floating around a fair amount of time before I started playing with this group. I cannot remember who started the rule though... I think the theory was that if you have a good skill with a bow then you can restring it really fast or something like that. Heck if I know the real reason though. (It’s quite possible it was done to shut one guy that used to be in the group up.) It does add a bit of interest to combat without having to add to the dice rolled though.
1. On a natural 1, make a confirmation roll.
2. Roll 1d20+BAB+Dex. You need to equal or exceed the attacking creature’s CR. If you fail, you roll on the critical miss table. If you roll another natural 1, you crit yourself.
Crit miss table (d%)
01-10 hit yourself
11-25 provoke attack of opportunity
26-50 lose remaining attacks (if any)
51-75 drop your weapon
76-90 -4 penalty on remaining attacks (if any)
91-00 hit opponent with +2 Str, but drop your weapon
-Drake-
Fumbles Houserule
When you make an attack roll and get a natural 1 (the d20 shows 1), you miss regardless of your target’s Armor Class, and you have scored a threat.
The miss might be a critical miss (or "fumble").
To find out if it’s a critical miss, you immediately make a Reflex saving throw against DC 10.
On a successful saving throw, your miss is just a regular miss.
A failed saving throw indicates your original miss is a critical miss.
A critical miss means that you suffer from one of the following effects:
- You take a -2 penalty to AC for one round (i.e. to the same initiative count in the next round).
- You take a -2 penalty to AC for one round (i.e. to the same initiative count in the next round).
- You take a -2 penalty to all attack rolls for one round (e.g. AoOs, or if you have multiple attacks for having a high BAB).
- You take 1d4 points of nonlethal damage, e.g. for twisting your ankle or wrist (the specifics are the DM's discretion).
You cannot do a critical miss only once per round.
While the DC may be raised (e.g. if you fail the saving throw by more than 5), a more experienced character will score less fumbles than a 1st level commoner would, if only for having a higher Ref BSB.
There should also be a fumble table for spellcaster concentration checks....
This is the part that I've always had a problem with. Casters already dominate the high-level game; critical fumbles just punish the guy who is ALREADY weaker because of his class! It's not enough that casters blow away any non-caster class at high level -- now the non-casters also have to lose actions, take extra damage, or damage their weapons while they're doing their best to contribute. No thank you.
Cute analogy, except that it's no good. Nutrition contains a value judgment. If you're intending to be "impartial" in this exercise (although it doesn't really seem like you are), avoid value judgments.
For what it's worth, I support Stargate's recovery system. By making the "confirmation" roll a BAB check vs. a set DC (I recommend 13), fumbles do in fact become impossible at a given point for warriors of great enough skill (like +12 BAB). That fixes your "unfixable" flaw.
I like the idea of a recovery mechanic tied to the critical fumble (like the confirmation mechanic is tied to the critical hit).
Proposed Fumble Mechanic
You roll a d20 and if you roll above your BAB you fumble, if you roll equal or under you just miss. Once you have BAB 20 or greater you become immune to fumbling an attack roll.
Com1 = 1 attack per round, 5% chance to fumble, 0% chance of recovery
Ftr20 = 4 attacks per round, 5% chance to fumble (per attack), 100% chance of recovery...
This would allow the higher skilled (read: BAB) characters to avoid being the brunt of the usual fumble rules.
d100
1-14% Miss target
15-25% Hit Self for 50% Damage
26-36% Hit Self for 100% Damage
37-47% Hit Self For 200% Damage
48-55% Possible (Roll to check with -2 to hit) Hit Target for 50% Damage but Provoke a AoO
56-61% Possible (Roll to check with a -6 to hit) Hit Target for 200% Damage but Provoke a AoO
62-100% Lose next Standard Action
Reasons? You miss your target . No need for a reason on that. Hit self for Half damage? You hit your hand or foot or something.
Hit target and you slip or slide provoking an AoO but also have a chance of hitting the guy with your wild swing/miss.
Lose a turn? You miss so bad you almost lose your weapon. Spend a Standard action to collect the weapon/Focus on the battle again.
These even apply to Touch attacks (Like Touch Attack Spells). I.e. you slapped your own hand with the spell, making it cast or something else.
This is the part that I've always had a problem with. Casters already dominate the high-level game; critical fumbles just punish the guy who is ALREADY weaker because of his class! It's not enough that casters blow away any non-caster class at high level -- now the non-casters also have to lose actions, take extra damage, or damage their weapons while they're doing their best to contribute. No thank you.
I did apply the fumbles to rays and touch attacks, but I guess that's not enough since so many spells require no attack rolls. So here's an idea for spellcasting fumbles:
Whenever a spellcaster rolls a 1 on a concentration check roll 1d10, if the roll is lower than the effective spell level (including metamagic feats) a fumble occurs (so higher level spells are far more dangerous). You could also have something similar for Use magic device roll fumbles.
Resolve fumbles by rolling 2d6
Fumbles
2-3 - Energy switch - the spell creates the exact opposite energy type to intended (fire becomes cold, positive becomes negative, DM decides if there's no clear "opposite")
4-5 - Backfire! If the spell is an attack, then treat the spellcaster as "ground zero" of the attack. If the spell was a "buff" then either A) It has the exact opposite effect (Mage armor reduces AC, Bear's Strength lowers Str etc...) or B) the Buff is applied to the spellcaster's nearest attacker (DM decides or flip a coin). Summoning spells summon an enemy. Scrying spells result in a loss of sight/hearing or whatever the spell was based on. Other spells inflict 1d4 per spell level damage on the caster.
6-7 Lose the spell and one random spell from your highest available slot (So a 5th level wizard loses the spell he's casting and another 3rd level slot, or 2nd level if his 3rd level spells are used up)
8-9 Exhaustion: The magical energies sap the spellcaster's strength - take 1d6 nonlethal damage per spell level. Even if not knocked out, treat the spellcaster as "fatigued"
10-11 Unstable: The spellcaster loses the spell and must make a Concentration check to cast any further spells (even if a concentration check is not normally required). If the spellcaster would need to make a check anyway, then such checks have a penalty equal to the level of the spell which "fumbled"
12 Major disruption! I can't think of any major disruption effects. It should depend on the spell level - a level 9 spell could open a portal to the plane least "friendly" to the spellcaster... Lower level spells could just roll twice on the table with negative effects stacking. Any ideas?
I don't think healing spells should fumble though.
To everyone who uses the DC 10 Reflex method, it nags me a little that the Fighter doesn't have this as a good save. As in, the little sneaky Rogues and Bards are going to be better at preventing fumbles than the Fighter. That seems odd.
I agree.
I find it odd though, that this article focuses on the statistics of critical fumbles, and the effects that has on the flavor of them.
It seems a lot of people have been using this article as an excuse to chime in on how they implement critical fumbles without even mentioning relevance to the original post. While I have no moderation responsibilities or influence to control the direction this thread takes, I would rather see posts that pertain to the original article, rather than using it as a cheap way to spread your personal houserule that doesn't even address the concerns expressed in the original post.
It's been floating around a fair amount of time before I started playing with this group. I cannot remember who started the rule though... I think the theory was that if you have a good skill with a bow then you can restring it really fast or something like that. Heck if I know the real reason though. (It’s quite possible it was done to shut one guy that used to be in the group up.) It does add a bit of interest to combat without having to add to the dice rolled though.
Well the way I posit that someone would have good skill with that aspect of a bow would be to say they have the rapid reload feat or several—say 5--ranks in craft (bowsmithing).
If you roll a natural 1, you miss, period. Roll the attack again.
If your second roll would have hit your opponent, you simply missed. Best case scenario. (Treated just as if you missed as normal.)
If your second roll would have missed, you do something stupid, but generally not harmful to you or your allies. (We call this a critical miss.)
If your second roll is another natural one, something rather bad is about to happen. (We call this a critical failure)
For mathematical comparisons, a lvl 1, 10, and 20 Fighter (STR of 10 and no miscellaneous weapon bonuses to hit) with a club attempt to hit an enemy with an AC of 25.
Per attack, the lvl 1 fighter therefore has an overall .25% (1/400) chance of a critical failure, a 4.5% (9/200) chance of a critical miss, a 90.25% (1/400 + 9/10) chance of missing, a 4.75% (19/400) chance of hitting, and a .25% (1/400) chance of critting.
Per attack, the lvl 10 fighter has an overall .25% (1/400) chance of a critical failure, a 3.25% (13/400) chance of a critical miss, a 66.5% (3/200 + 13/20) chance of missing, a 28.5% (1/4 + 7/200) chance of hitting, and a 1.5% (3/200) chance of critting.
Per attack, the lvl 20 fighter has an overall .25% (1/400) chance of a critical failure, a 1% (1/100) chance of a critical miss, a 19% (1/25 + 3/20) chance of missing, a 95% (3/4 + 1/5) chance of hitting, and a 4% (1/25) chance of critting.
So against the same opponent, you always have that .25% chance of critically failing, but your BAB (As well as other to-hit bonuses) will greatly reduce your chances of critically missing and missing in general, and increase your chances of hitting and hitting with criticals.
Critical Misses are a system that works against the players. Any enemy is likely to be in only one encounter. And that will be the encounter it dies most of the time. That is a limited chance for an enemy to roll a critical miss. Players, being in every encounter have a much greater chance to roll critical misses.
But Critical Hits are also a system that works against the players. For the exact same reason. An enemy is likely to be in only one encounter. That's a limited chance to receive a critical hit. Players, being in every encounter, have countless chances to receive a critical hit. Also, the stronger the enemy, the harder it is to confirm a critical. So, most confirmed crits will be against the weaker foes. BBEG's are more likely to just be auto hits.
So, you balance a system that hurts the players (crit hits) with a system that also hurts the players (crit misses)? And this is without touching the greater miss chance at higher levels.
I don't know if this will help, but here's the system my group uses.
If you roll a natural 1, you miss, period. Roll the attack again.
If your second roll would have hit your opponent, you simply missed. Best case scenario. (Treated just as if you missed as normal.)
If your second roll would have missed, you do something stupid, but generally not harmful to you or your allies. (We call this a critical miss.)
If your second roll is another natural one, something rather bad is about to happen. (We call this a critical failure)
I like this method much better than the rolling of saving throws, etc. However, there is only a 1/400th chance of a critical failure - something that will happen very rarely indeed. It will tend to happen to the best fighters more often because they do more hitting.
The advantage of this system is that the good fighter will do better than the non-melee classes. The chances of rolling a 1 will be the same for everyone, but the chances of a critical miss will be less for the character who hits better.
My variation on this - and this is entirely theoretical - is to reroll the attack, and if it misses by 10, apply the critical miss penalty. The easiest penalty to apply is the attack of opportunity.
Vonk rolls a 1, then a 17 - "You swing wildly, and your sword bounces off the Orc's shield".
Rorg rolls a 1, then a 5 - "You cut at his head, miss and overbalance. While trying to recover, the Ogre whacks you with his club."
Perhaps a really bad miss would result in a dropped or broken weapon. Very unlikely to happen to the 17 strength level 4 fighter with a +2 longsword, but quite likely for the sorcerer with a dagger.
The reasons I would prefer to avoid doing saving throws are:
- It's a different thing to look up. When doing combat, I know what's needed to hit something. Switching to another throw is messy.
- A fighter should do well in combat if he hits with high bonuses. He shouldn't do badly because he has a bad saving throw.
The reason for having critical misses is that it is more fun. If they aren't fun, don't do it. The objection that they penalise fighters more is valid, but I think can be addressed by making them less likely for a character with high hit bonuses. That they make the party in general more vulnerable is certainly true. On those grounds, I'd rule out the more deadly consequences.
One of the players, a Dwarf Barbarian, plays his character so well it's amusing in the first place. But his luck with rolls makes him the king of comedy in the game. He has, due to playing his character as stupid as he should be and the roll of the die, done the following:
Dipped his beard in the contents of a latrine.
Been mistaken for a female Dwarf prostitute by a Half-Orc commoner, and treated as such.
Ran off without his party down the hall in a dungeon after an enemy, lost him, got angry, smashed a vase and then had to hide from everyone who heard said vase smash against the floor.
Ran ahead of the party cause they were talking too much and fell down a trap full of skeletons all by himself.
Ran from a party member he angered, then ran back with a much angrier monster in tow.
Drank bad ale with a smile only to regurgitate it later in battle.
And a ton of other things I can't quite recall right now that were caused by his impulsiveness and odd luck. (He has only a few ranks in listen, yet never rolls lower than an 18 on a listen check. Spot checks, not so much.)
The group knows this is the player and the character and we laugh quite well when another one of the Dwarf's crazy situation arises. It's no different with critical misses and critical failures. It's funny, normally. Somebody rolls a one and suddenly everyone else is on the edge of their seat, looking to see that next die roll. People start laughing when the second 1 shows up, even before the DM has a chance to weave the result. It's entertainment and it's fun.
Granted Fighters and Barbarians swing more, but the mage has to roll touch attacks with some of his spells, and from experience I can tell you a critical failure on a scorching ray is no less funny than with a longsword, and usually much more devastating. Plus, the result of a critical failure of a melee weapon usually result in hitting a teammate, which sucks for the party, but isn't a direct blow against the Fighter or Barbarian.
Some people seem to think the Critical Hit system needs a Critical miss system to add some kind of balance. A.K.A: If you can crit hit, you should be able to crit miss.
I agree with your whole post. I find the logic that it needs to be "balanced" by critical misses is a flawed argument. As you explain, players not only can critical hit, but so can their enemies. It's balanced because both sides can crit.
I agree with your whole post. I find the logic that it needs to be "balanced" by critical misses is a flawed argument. As you explain, players not only can critical hit, but so can their enemies. It's balanced because both sides can crit.
By that logic, critical failures would also be balanced because both sides can do it. The PCs get more opportunities to critically fail, but they also get more opportunities to critically hit. Still balanced.
The only two valid complaints I've ever heard about critical failure systems, in this thread or any other, are the idea about fighters failing more often than commoners and the point that stabbing yourself or your friend by accident spoils the fun. I agree that both of those issues are problematic, but they can both be resolved fairly easily. There have been plenty of suggestions already in this thread alone, some better than others.
At the end of the day it all comes down to personal choice, like the OP said. Some people like the flavour provided by crit failures, others don't. You can do math until the cows come home, list example after example; you're not going to stop the people who like fumbles from liking fumbles, nor will you stop the people who hate fumbles from hating them. The only thing to be done is to sit down with your group and decide whether or not you want to use fumbles in your own game and more power to you whichever way you choose.
By that logic, critical failures would also be balanced because both sides can do it. The PCs get more opportunities to critically fail, but they also get more opportunities to critically hit. Still balanced.
Almost. While you are technically correct, you're taking my quote out of context. The emergence of that truth doesn't make my argument invalid.
The argument is that "the ability to crit more is too powerful, and therefor must be offset by critical fumbles to balance it. It's already balanced, thus the logical deduction is incorrect. It would be correct to say:
The ability to crit is balanced because both enemies and PCs can do it, so if you give the ability to fumble to both enemy and PCs, it retains the balance.
The only fault in that is that it really does hurt the PCs, because of the concept that once a player is subject to it, it changes things in the monster's favor, but if the monster is subject to it, it only speeds up the process of what was going to happen anyway - the PCs winning.
The casual logic in coming to the conclusion that it needs a counter balancing mechanic is faulty.
The only two valid complaints I've ever heard about critical failure systems, in this thread or any other, are the idea about fighters failing more often than commoners and the point that stabbing yourself or your friend by accident spoils the fun. I agree that both of those issues are problematic, but they can both be resolved fairly easily. There have been plenty of suggestions already in this thread alone, some better than others.
At the end of the day it all comes down to personal choice, like the OP said. Some people like the flavour provided by crit failures, others don't. You can do math until the cows come home, list example after example; you're not going to stop the people who like fumbles from liking fumbles, nor will you stop the people who hate fumbles from hating them. The only thing to be done is to sit down with your group and decide whether or not you want to use fumbles in your own game and more power to you whichever way you choose.
If you haven't noticed, I am the OP. It is a personal choice. Some people like the immersion and excitement that falling flat on your ass after 20 levels of fighting brings to the table. I cannot take that away from them. But it's easier to understand why someone makes a choice or how to relate to that person if you have a logical reason for your choice.
To contrast it. Pointbuy is "balanced." I loathe pointbuy myself, but I can relate with people and understand why they like pointbuy regardless of my preference to rolling stats.
This really is an exercise to show people why many of us who don't like critical fumbles have a valid reason to dislike them. We are not evil vile people. Well... not unless you call logic evil and vile ;). But what it really comes down to, is preference. It's a lot of work to address the issues of critical fumbles and actually not make the fighter get worse while doing so. But it also shows an outline on how to go about it, if you wish to address the issues with it, while still keeping it - which is not a quick easy fix.
If you value both the flavor of critical fumbles and logical issues with them, this is a great starting point to creating a system that allows you to implement them without breaking the suspension of disbelief that a fighter that trains hard should become better, not worse, at wielding his weapon.
In contrast again, I will point at my preference for rolling again. I value rolling, much more than pointbuy. This is the "non-logical" point of view. At least at first glance
Show
(Put aside the fact that different classes are imbalanced, and that has more of an effect on balance than stats, and that different classes gain more out of stats than others)
Now there is a difference between low level orc and BBEG, but honestly, wait to see how much your players love to see the BBEG swing his Balor's Sword and overbalance and fall so they all rush him and hack away while he's down, then take their AoOs as he stands and tell me it hurts the players more.
Also, for spell effects my players (who had the largest say in the crit fumble system we use, I never even brought it up, but they did) we developed a crit save system. On a nat 20 against a fort or will save you are immune to that spell for the rest of the encounter, on a reflex nat 20 you take no damage. I think I am going to implement a crit failure on concentration checks for spellcasters too, but I'll put that to group vote.
Not to mention the fact that saying the enemy usually appears once, isn't really true. You as the DM are “the enemy” and you appear every game and, on average you Crit fail as much as the party.
You're missing the point; your comment assumes a false premise. You're acting as if PCs and monsters are 'balanced' to begin with; they are not. A monster is meant to die, and the PCs are meant to live. The players are supposed to succeed.
Thus, any effect that randomly hits both monsters and PCs will be more detrimental to the PCs (not balanced, as you claim), because monsters are supposed to lose. As has been mentioned several times, any random effect that's bad for monsters doesn't change the outcome of the event; it only changes the circumstances of that outcome.
By contrast, a random effect that's bad for PCs very well may change the outcome of the battle -- even if it stands no chance of killing a PC, the party will have to spend more resources, thus shortening the adventuring 'day' or putting them at risk of death in later combat for want of those resources.
Not to mention the fact that saying the enemy usually appears once, isn't really true. You as the DM are “the enemy” and you appear every game and, on average you Crit fail as much as the party.
Anyway, the point of the saying "the enemy only appears once" is that each monster -- each DM-controlled character -- arrives in the game with all its resources available. Full HP, all its SLAs available, whatever.
By contrast, the PCs are often entering battle after having spent HP, spell slots, daily uses, and so on. The DM does control all monsters, but each monster does not suffer from attrition. PCs do that.
Ok, crits and fumbles cause more harm to the players than it benefits them. Anyone who still doubts can begin by reading the sidebar on page 26 of the DMG.
Now, I have some problems with the different fumble effects that people have suggested. Many of the consequences involve the victim dropping his weapon, falling prone, breaking his weapon, etc. The game already has mechanics for this. Disarm and the Improved Disarm feat. Trip attacks and the Improved Trip feat. Sunder and the Improved Sunder feat. If the DM wants players to drop their weapons, he should create villains that use that attack.
Then there's the clumsy fighter approach. Hitting himself with a sword? Hitting an ally? Even a 1st level fighter is supposed to be a highly trained warrior. Not some guy who picked up a sword for the first time. Hit the enemy, not your ally is probably one of the first lessons they learn.
One of the ideas posted in this thread is that a fumble makes all enemies who threaten the fumbler get to make a free attack of opportunity. This fumble in particular hurts fighters more than any other class. At least, fighters who take the cleave / great cleave or the whirlwind attack feats. Great cleave fighters want to move near as many enemies as possible, to get as many attacks as possible. Even with an additional system to reduce the fighters chance of fumbling at higher levels, the extra attacks still mean more chances to fumble. And when he does fumble, he's most likely to be surrounded by multiple enemies who now each have a chance to hit, or possibly crit.
A fighter trying to use Great Cleave or Whirlwind attack to full effect is already putting himself in harms way. Do you really need to add to his danger?
Now there is a difference between low level orc and BBEG, but honestly, wait to see how much your players love to see the BBEG swing his Balor's Sword and overbalance and fall so they all rush him and hack away while he's down, then take their AoOs as he stands and tell me it hurts the players more.
It hurts the players more. A BBEG should have a higher AC than a standard encounter for the parties level. Higher AC = lower chance of confirming a crit. And the more work a DM puts into making high level fighters fumble less than low level commoners, the less likely the BBEG is to fumble as well. So the BBEG is more likely to crit, and less likely to fumble than the party. And if the BBEG does fumble?
As a player, why would I love this? The final battle that I've been looking forward to for months, and the BBEG falls over? The term is anticlimactic. A battle ends early because a party member crits, that’s a story to tell. A battle ends early because the villain cuts himself to death on his own weapon? That's a game to forget.
I agree with Sphyre. Most fumble rules I've seen (including those in this thread) make no sense, penalize the players more than the enemies, and those which are a bit more advanced are often too complicated.
The main reason why I refuse to use them, however, is because they punish melee characters. Only them, aside from touch spells. Now, what's so great about nerfing the weakest classes in the game? I don't get it.
As I stated in my first post, the first reason is that players and DMs alike feel a need to "balance" the critical hit system with a critical miss system. If a player can do very well on a combat roll, they should be able to do very badly as well. It feels fairer. Really, it is not. But it feels that way.
It's a type of house rule that keeps popping up. A variant is the auto kill threat. If you roll a 20 when confirming a critical hit, you threaten an auto kill. Roll again, and if the next roll is a confirmed hit, the creature is killed outright. I've seen the idea pop up in multiple games. It seems fun. But it always gets shot down when the group realizes that there is something wrong when a 1st level kobold can kill an epic level PC. I'm certain that some groups do use a rule like this in their games.
People keep coming up with critical fumbles for the same reason that they eat junk food. It's not good for you, but it tastes good. Hey, look. I finally tied my posts into the actual theme of the post, The Flavor of Critical Fumbles
Fumbles are handled in the same manner as a critical threat. Roll a second attack with the same bonuses and penalties.
If you do not confirm, you simply miss.
A confirmed miss results in being 'Dazed' for 1 round.
A confirmed miss by 5 or more results in 'Dazed' and Flat-Footed until your next initiative.
In addition to the penalties above, if confirmed with a natural 1, the weapon is ineffective (broken, lost, etc..) until the appropriate action is taken.
As DM, I determine the flavor of the event on the spot. 'Dazed' is a misnomer, but the intended effect is the same. The character is unable to make any additional attacks until he 'regains his balance'.
I understand the math that makes characters periodically more likely to fumble at high levels, especially fighters, but has anyone ever considered that the player should elect not to use all his attacks? By the time characters are taking -10 or -12 (Flurry) on his third attack they often miss any target worth hitting. Acting in haste should increase the rate of fumbles.
People keep coming up with critical fumbles for the same reason that they eat junk food. It's not good for you, but it tastes good. Hey, look. I finally tied my posts into the actual theme of the post, The Flavor of Critical Fumbles
:D You made me smile. I agree with your entire post, except for the fact that I don't think it tastes good... but that ties in more with the 'flavor' of critical fumbles. Everyone has different tastes, but it's also nice to know the nutritional value of it too, isn't it? There's more than one reason I'm against them. They take away my fun
- even in the current game that I play in with a DM who DMs critical fumbles. Whenever I roll a 1, I stop having fun, even if I make my DC 15 Dex check, with my non-dexterous character.:D You made me smile. I agree with your entire post, except for the fact that I don't think it tastes good... but that ties in more with the 'flavor' of critical fumbles. Everyone has different tastes, but it's also nice to know the nutritional value of it too, isn't it? There's more than one reason I'm against them. They take away my fun
I don't like the taste either, but it seems that many do. I've played with crit fumbles, and without. I prefer without.
I've tried to keep fumbles simple. Like many here, I used a second roll to confirm the severity of the error:
As DM, I determine the flavor of the event on the spot. 'Dazed' is a misnomer, but the intended effect is the same. The character is unable to make any additional attacks until he 'regains his balance'.
I understand the math that makes characters periodically more likely to fumble at high levels, especially fighters, but has anyone ever considered that the player should elect not to use all his attacks? By the time characters are taking -10 or -12 (Flurry) on his third attack they often miss any target worth hitting. Acting in haste should increase the rate of fumbles.
Well, there are exceptions. Sometimes we find DM's that do seem to be out to punish the melee types.
Using a similar logic that attacking "in haste" makes the fighter worse off, you could decide that Rangers who attack in anger should also have a worse time. So, for example, whenever a ranger attacks a favored enemy, he should have penalties on his attack roll and AC. He's just too angry to strike or defend himself properly. Does this sound like it makes the ranger more fun to play?
As for the hasty bit, he's not hacking away randomly. As your BAB goes up, your character not only becomes more precise when attacking, but he learns to attack more quickly. He is improving. Getting better. More skilled. In no way does it make sense for him to become more clumsy as he goes up in level.
The best advantage of having a high BAB is that you gain more attacks. But it can only be used when the player gets to use a full round action to attack. Anything more than a 5 foot step and the melee type only gets a single attack. Depending on the type of enemy, and the DM's play style, full round attacks may be common or rare. Why should a fighter NOT take his extra attacks when he gets the chance? Those extra attacks are supposed to represent the benefits of his training.
A Natural 20 on a skill check, save, or attack roll is always a success. (Whether rolled by a PC or NPC) Additionally a natural 20 on an attack roll is an automatic critical hit (no confirmation roll), weapons with magically expanded threat ranges are crits if they could hit the AC on a 18, 19 or whatever the weapon threatens.
Conversely a natural 1 on any of the above is always a failure, but not necessarily a "fumble" per say. It has no negative effect to PC or NPC except for the failure to accomplish what you wanted to do.
His reasoning for this is simple, sometimes, even the insanely powered Barbarian is going to be foiled by his environment while trying to hit an opponent. Sometimes the beguiler with a bluff of something near +90 is going to fudge a word or get a little nervous and not be able to be convincing.
Personally, I think this way makes for less rolls and a more active (and somewhat more suspenseful style of play).
1) A fighter with a one-use dagger of petrification turned himself to stone with a series of 1s.
2) A barbarian hacked his leg off while trying to open a locked chest with a series of 1s, followed by an amazing crit - on himself, of course.
3) A fighter dropped his primary weapon into a bottomless pit.
4) A high-level two-weapon fighter fumbled so much that he might as well have been blind-folded and on roller skates.
5) Archers all ended up carrying back-up bows because their bowstrings tended to break every 10 rounds.
6) Best (or worst) of all, a BBEG coup-de-graced himself at the beginning of a campaign-capping fight (BoringInfoGuy's worst nightmare
).The first three of these did cause a lot of conversation and general amusement among the players, but you know who didn't enjoy it? The players of these characters who hadn't done anything wrong except to get some lousy rolls.
I just wanted to mention this for the individuals who are posting things like "my system could, theoretically, kill a PC but it never has." If it can and you play long enough it eventually will, and that's probably not a good thing.
-Shaftiel
I got the idea from the Quest game books, but it has been balanced. That way, not only are higher levels far less prone to fumbling, but dual wielders have balance restored to them. And I use the provokes an attack of opportunity variant. This is somewhat rooted in real life, you can over swing a great sword or two handed axe fairly easily, a sword normally, and you really have to mess up to fumble a rapier, or short sword. Also, any weapon 2 sized below you has no fumble chance, unless you are doing something really stupid at the time, in which case it is 1(twirling a dagger to show off comes to mind).
Alternatively, if a PC eats a fumble and carves himself open, he faces a greater chance of dying in the battle -- something which wasn't supposed to be the idea.
How are you so sure? What if the DM wanted that PC to die? What if, just maybe, the game was not meant to be just a cakewalk storyline for the PCs? What if the game was meant to be challenging? I think that's actually the point everyone missed. It's not SUPPOSED to be fair. Fairness is fine for most things, notably statistical analysis. Also a wonderful stat is that statistics are usually not a good idea to rely on too heavily. If the world perfectly followed those rules then every time I rolled a d20 99 times I would know what I was going to roll next.
This was a game, not only for mystery and castle intrigue, but also for combat. Combat was meant to be a challenge, an obstacle, something you had to pay enough attention to and take into account that things might go wrong and how you might recover from that. Combat isn't all about the dice, if the GM thinks things have gone too horribly wrong there are ways he can intervene. For that matter, D&D is not all about the dice. It's about socializing, logical thinking, loopholing, and overcoming or working around obstacles. Fairness is a good idea within the PCs. Monsters are usually advantaged against the PCs somehow with either greater personal strength or strength in numbers. Lets look at this from the monsters side. In a typical adventure, or even a harder one, the monsters still need all the help they can get to overwhelm the PCs. Escaping the box of statistical analysis we realize that the monsters already have lots of advantages the PCs don't, such as no need to save spell, as mentioned earlier. They can have fun with feats such as Improved Sunder, as destroying the party's items is not a problem, money or magik items is unnecessary, and such feats and abilities can be used annoyingly. The rust monster for example. He ruins all of the metal items the party has, because then it gains an advantage, and the beast itself loses nothing. If the party rusted their enemies armor, they could not wear it or sell it later after the battle.
The game of D&D is not fair, nor was it meant to be. Your vanilla flavored critical fumbles might be healthier, but vanilla is boring and life is short anyways. We are but humans, living not even a century. We need excitement in our little lives or there is no point. Sure exercising makes you live longer, but if it's torture, why bother? Short and happy or long and hard? I would prefer being entertained and have my caramel covered critical fumbles over here. I will agree with you that your system does not work, but that does not mean none of them will. Think outside your tiny box Sphyre, work either with the number of attacks you have or your BAB, not the number of attacks granted by BAB. I would suggest a system based solely on BAB, as it is your skill with a weapon.
A simple system would be that if you roll a natural 1, you roll an attack against yourself with flatfooted AC, with an attack bonus of +10 minus your BAB and any magical enhancement of the weapon. Optionally you could include that one could then critical hit themselves, or upon rolling a second natural 1 make an attack on a randomly selected square as if you were blind (50% miss chance), and possibly that you gain an additional -5 on the attack roll for natural weapons representing how much harder it is to claw yourself.
Statistically speaking, this is still not fair. Statistically speaking, I don't think we care. If you had a game, and all you had to do was press a button over and over, it would be boring. Challenges, however fair, are important. If the game was perfectly fair it might be boring. It would be too easy. Now if the game favored the monsters, and PCs were the outclassed underdogs fighting out of their league, that would be an interesting game!
1: Roll against AC to see if you hit yourself
2-5: Roll using the Thrown Weapons chart to see what direction your weapon goes in.
6-10: Drop your weapon at your feet.
11-15: Lose grip on weapon and lose next action
16-20: Treat as a regular miss
She rolled another natural 1, so she had to make a roll modified by her BAB to see if she hit herself. She rolled a natural 20 and then confirmed the critical. In the end, 32 points of damage were handed out by an errant roundhouse kick. Point I am trying to make is that Critical Fumbles are an integral part of the game and definitely make things more interesting.
Also, a roll of 1 does not represent the actor's ineptitude. It represents circumstances beyond his control making his attempt impossible or nigh-impossible. A fumbling warrior (+0) misses on most of his swings, and so doesn't notice the few that would never have worked. A skilled warrior (BAB +11) might discount his third attack missing, but when the first misses against a weak opponent, it's not because he screwed up. It's because something else screwed him up. Making him take damage for this is just adding injury to insult.
I know that's the wrong way around.
When a character swings a weapon one time, that character has the same chance of critical failure as does a Ftr40 of critical failure. This is because a critical fail in my eyes is not a simple ineptitude, rather, a horrible mishap that has befallen the victim. A Ftr17 has the same chance of missing a swing and that swing lodging itself into a wall or crack in the floor as a low leveled fighter if they both take one swing.
In my group the ruling is that a character may only critically fail once per round due to the fact that it is ending their action. Swinging a sword 4 times in a period of roughly 4 seconds is far more dangerous then one swing, thus increasing said chance. If a high leveled character could never critically fail because it is "unrealistic" that would be like saying that a great magician would never drop a ball when juggling 15 items simply because hes better.
A higher leveled character naturally takes risks that a low leveled one would never encounter, because a low level PC would not risk his life to the extent of a 20th leveled PC simply because the lower of the two has far less to gain. Higher levels take higher risks because they have higher chances of success and much higher rewards.
Personally I see this challenge is that a character such as a Ftr16 so that she can make a possible total of 4 attacks is taking the chance of missing 4 times. Saying that the increased chance of critical failure is unfair is the equivalent of stating that its unfair that a Ftr16 has an increased chance of missing.
When a character swings a weapon one time, that character has the same chance of critical failure as does a Ftr40 of critical failure. This is because a critical fail in my eyes is not a simple ineptitude, rather, a horrible mishap that has befallen the victim. A Ftr17 has the same chance of missing a swing and that swing lodging itself into a wall or crack in the floor as a low leveled fighter if they both take one swing.
In my group the ruling is that a character may only critically fail once per round due to the fact that it is ending their action. Swinging a sword 4 times in a period of roughly 4 seconds is far more dangerous then one swing, thus increasing said chance. If a high leveled character could never critically fail because it is "unrealistic" that would be like saying that a great magician would never drop a ball when juggling 15 items simply because hes better.
A higher leveled character naturally takes risks that a low leveled one would never encounter, because a low level PC would not risk his life to the extent of a 20th leveled PC simply because the lower of the two has far less to gain. Higher levels take higher risks because they have higher chances of success and much higher rewards.
When a melee character gains levels, his attacks only get more powerful when he gets a more powerful. The innate mechanism for him getting more damage is more attacks. You actually have to not only gain 5 levels to gain more damage, but it's actually 5 BAB; your medium BAB classes still need to reach 5 more BAB, in which takes 7 levels.
On the reverse side, you have spell casters who scale with each level. Only increasing their potency.
"A high level character takes more risks" is built into the fact that he faces much more challenging foes. Anyone fighting a beholder, or ancient dragon will be able to tell you it was no easy feat, but the kobolds he fought, he eventually learns that although crafty, they are rather easy to slay.
You don't need a built in mechanism of "taking more risks" to punish the fighter more, when the fighter is already leaps and bounds one of the most underpowered classes in the game. Implementing critical fumbles simply takes what we have with the fighter and spits on him, and throws him into the trash in an alley for the wolves to eat.
It all depends on your style of play. Are you playing to "have fun" or are you out to "win"?
My group likes using Critical Fumbles because we think it's fun. Then again, we think it's fun when people get swallowed by dragons, fall into deep holes, and various other means of gruesome death. We have 6 people in the group, but usually only 4 are playing since 2 invariably have died and have to make new characters. Only one player seems to have a problem, but he's the munchkiny "beat-the-dm" player.
It all depends on your style of play. Are you playing to "have fun" or are you out to "win"?
Personally, I'm in it to have fun. And since it's a role-playing game, being able to get attached to a character is nice. Critical fumbles, by increasing randomness, increase the chances my character will be suddenly plucked away from me. Death has to be a threat in any adventure, but personally I find it more fun to have the character die because the enemy was better or cleverer, not because my raging barbarian charged and critted himself. Still, each to their own.
It's called "missing."
A critical hit generally does x2 damage. Further, it doesn't actually do that much - a load of damage (bonus dice from weapon, class abilities, and maneuvers) isn't multiplied.
So a critical hit is, for most characters, worth less than two regular hits. You crit once and miss once in an encounter, you've done less total damage than just hitting every time.
Example 1: Gorg the Barbarian is dueling the evil warlord in the throne room. He rolls a 1 on his 1st attack and has a critical fumble threat. He then rolls again, using his total attack bonus against the warlords AC, just like a crit in reverse. He 'misses' that and that and as a penalty loses the rest of his attacks for that round as the warlord gives a lightning fast parry, briefly staggering Gorg.
Essentially, it should work exactly like a critical hit with 'hitting' on the second roll avoiding something worse than just missing. Another important thing is that it should not be overly punative. That is, it should never result in direct damage to the player. Nor should it last more than one round AC, attack penalties, AoO by the target, maybe even action loss and a chance to become prone and the like seem fair game.
Example 2: An Elf Ranger is attacked by a Goblin in the forest. The Goblin rolls a natural one and rolls a 3 on the confirm roll. He trips over a root while trying to attack and stumbles, giving him a -2 AC penalty. The Ranger strikes back, but also rolls 1 but beats the goblins AC on the confimation roll. He still misses, but he dosent leave himself open to counter attack either.
I try to do fumbling contextually, so there's no specific rule for what exactly happens. An archer may shoot his own foot, or may accidentally target the wrong person, or may simply drop his bow. A fighter might expose himself to an AoO, may trip and end up prone, or may simply lose a turn as he recovers.
I will generally impose penalties that make sense both in-game and out-of-game. In a less serious encounter, I may impose harsh or humorous penalties, but in a highly threatening encounter, I will typically impose only minor penalties. As far as in-game context, I will consider the nature of the action to help determine the penalty. An attack on a speedy and wily opponent is likely to result in a dropped weapon, whereas an attack on an opponent who is low to the ground may result in an overswing and a fall to the ground.
The purpose of this rule, at least to me, is to allow the game to breathe a little bit, and to occasionally humble overpowered characters. This rule achieves that goal quite nicely.
another time, I was the Dm, and my freind rolled 2 on a search check. He walked up to the barrel lifted up the lid and dropped it on his arm, breaking it. Now the barbarian went back to the village and the cleric thought it was so funny, he healed it for free. isn't that kind funny in the least?
everyone made fun of the barbarian for the rest of the night. even he thought it was hilarious!
Now, just because you screw up with whatever your doing, doesn't mean you killed yourself,it means that something hilarious is going to happen.