Triple 20s in a row rule

[Shotaro]

Yeah just wondered if/when it had happened to anyone else? We are playing tomb of horrors and one of the players (who ALWAYS dies) created a level 10 fighter with 100+hp, 30 or so AC and was introducing himself to the party when he got caught in a trap.

20, 20, 20 - the first time he had been targeted and the 1 in 8000 instant kill causes an arrow to hit him square in the eye socket and out through the back of his neck via his spinal cord. He falls to the floor limp and lifeless. The party set about looting him.

[Flickerdart]

That's not really a rule. Or rather, not a printed rule.

[Shotaro]

That's not really a rule. Or rather, not a printed rule.

IIRC it's in the players handbook or DMG for 3.5 - It's a variant rule yes, but every group I have ever played in has made use of it.

[Vemynal]

I like it, its fun. Really allows the unexpected to happen lol

[umbergod]

had it happen i think twice in my 14 years of gaming, both in 3.0-3.5 dnd settings. most memorable one was a setting where the pcs were supposed to die, because it transported us to another plane. i refused. it was i think level 5. DM sent an Advanced Roc after me. triple 20s with a crossbow, instakill.

[Half-orc Bard]

I had something similar happen, I was in a group low level I was a bard. We were fighting spiders, and I shot a crossbow bolt at one of the spiders, i get a 1 then a 1 then a 1. the dm decides that the bolt hits a rock and bounces back and kills our cleric. We killed the spiders and looted the cleric, yay for natural 1s!

[Mando Knight]

It's a terrible rule from a player's perspective: the more attack rolls a player faces, the more likely he is to instantly die horribly.

The chance of getting insta-killed by a level 16 Fighter? It's not 1/8000, it's .05%, each round he gets a full attack. The longer you fight, the more likely that one of the attacks you'll receive will instantly kill you. The DM might lose one or two monsters per campaign this way... but if it happens to a player, then suddenly they lose their character. And the same character will generally be the target of several dozens of attacks over his career.

Let's take the Fighter again. Let's assume the party is built around letting melee mix it up in melee while the casters support them, rather than the casters playing god and all that. If the Fighter (on average) is subject to five attacks per encounter, goes through ten encounters per level, and wants to get from level 1 to 20, he has a 11.2% chance of dying due to a triple-20. If he's subject to more attacks (such as the enemy getting several full attacks themselves), then his chances of dying go up further.

[AMFV]

It's a terrible rule from a player's perspective: the more attack rolls a player faces, the more likely he is to instantly die horribly.

The chance of getting insta-killed by a level 16 Fighter? It's not 1/8000, it's .05%, each round he gets a full attack. The longer you fight, the more likely that one of the attacks you'll receive will instantly kill you. The DM might lose one or two monsters per campaign this way... but if it happens to a player, then suddenly they lose their character. And the same character will generally be the target of several dozens of attacks over his career.

Let's take the Fighter again. Let's assume the party is built around letting melee mix it up in melee while the casters support them, rather than the casters playing god and all that. If the Fighter (on average) is subject to five attacks per encounter, goes through ten encounters per level, and wants to get from level 1 to 20, he has a 11.2% chance of dying due to a triple-20. If he's subject to more attacks (such as the enemy getting several full attacks themselves), then his chances of dying go up further.

Except that probability doesn't work that way. The odds of 3 consecutive twenties are (1/20*1/20/*1/20) which is .0125%. The odds increase dramatically with each consecutive twenty, as you need to not only roll twenty, but roll it three times.

[Kamai]

We had a bard die a horrible death to a crossbow bolt by a variant of that rule (2 nat 20s and a 3rd roll that hit = fort save or die). It was the first session of that bard, and we vowed never to use that rule again.

[JaronK]

The rule should never apply to PCs, since they take so many attacks. It's a terrible rule for tanking characters that are liable to be hit by a lot of attacks. It's hilarious when applied to NPCs of course. If you want it to apply to PCs, at least make it "automatically reduced to -1 hitpoints" so they don't just instantly lose their character.

JaronK

[Drelua]

Except that probability doesn't work that way. The odds of 3 consecutive twenties are (1/20*1/20/*1/20) which is .0125%. The odds increase dramatically with each consecutive twenty, as you need to not only roll twenty, but roll it three times.

He meant for that .0125% chance to be multiplied by 4 for the fighter's 4 attacks, which gives you .05%.

Anyway, I use this rule and it's come up a few times but never against players. Of course, I suspect that my DM would pretend it didn't happen if it did. It's usually his PCs that roll, and he's done this against the main protagonist when someone else was DMing with his ranger, and many other times. Makes for a good laugh, but tends to end encounters too quick. He also has multi-crits, so if you roll a threat on your confirmation, you get to roll yo confirm again for a double crit, and this can go on until you get a quintuple crit, which is also an insta-kill. It used to be that you'd take the multiplier and multiply it, so a double crit with a x2 multiplier would be x4, a triple crit would be x8, etc. I recently talked him into adding multipliers how you're supposed to, so it's be x2, x3, x4 instead of being quite so bad, but needless to say, I usually take a weapon with a high threat range and get that up to 15-20 as soon as possible.

[Cog]

Except that probability doesn't work that way. The odds of 3 consecutive twenties are (1/20*1/20/*1/20) which is .0125%. The odds increase dramatically with each consecutive twenty, as you need to not only roll twenty, but roll it three times.

0.0125% per attack. Even discounting Haste, attacks of opportunity, and the like, iteratives turn that into the 0.05% per round that was quoted.

[Krazzman]

The rule should never apply to PCs, since they take so many attacks. It's a terrible rule for tanking characters that are liable to be hit by a lot of attacks. It's hilarious when applied to NPCs of course. If you want it to apply to PCs, at least make it "automatically reduced to -1 hitpoints" so they don't just instantly lose their character.

JaronK

I approve of this...at least that's the one I use in my Campaign...

Have a nice Day,
Krazzman

[AMFV]

0.0125% per attack. Even discounting Haste, attacks of opportunity, and the like, iteratives turn that into the 0.05% per round that was quoted.

Ah that is correct then. That's still fairly remote however, but a lot less than it should be for an instakill rule.

[legomaster00156]

I've heard of a more reasonable houserule. For every 20 beyond the first, you add an extra damage die.

[sreservoir]

0.0125% per attack. Even discounting Haste, attacks of opportunity, and the like, iteratives turn that into the 0.05% per round that was quoted.

81 attacks gets you to just a bit more than a 1/20 chance to die.

LD50 is 5545 attacks.

if a character has an average of 20 attacks rolled against them per encounter in the encounters it takes them to get from level 1 to 20, it will die half the time under this rule.

[Flickerdart]

Five attacks per encounter seems very low, considering that many low-level feats are claw/claw/bite animals, a number of which have Pounce, and the other most popular opponents for early levels are hordes of monstrous humanoids. Additionally, the figure for gaining a level is 13.33 encounters, not 10. So we're looking at about 10 attacks (assuming the baseline of two opponents, one of whom is felled after a single full attack and the other after two) against the Fighter each encounter, over 266.6 encounters. That's 2666 attacks, or 33.3%. With this rule in effect, every third frontliner will have died of a fluke sometime in his career.

[marcielle]

Our rogue one shot a kraken using this rule. We were level 3 :smallbiggrin:
Storyline went right out the window. We decided to go off and do what our DM wanted us to do anyway, just to throw him a bone.

[Metahuman1]

The Dm introduced that rule just before we had a boss battle once. We were a fourth or fifth level party and going up against a hugh creature of some kind, I forget what exactly.

Anyway, the party's elf wizard come up on her initiative, and attacks with her long bow. Nat 20, then nat 20 again, then nat 20 a third time. First time using that D20 ever. One shot kill, and the funny thing was she won Initiative for the encounter, so the big boss the dm spent like two hours planning that week was done in on call for initiative and three attack rolls form the wizard!

[BoutsofInsanity]

Best way to rule this, for fighters especially, for natural 1's, only the first attack counts as the natural 1 rule. The other attacks from a full attacking fighter only miss. This stops a level 1 fighter from screwing up less then a level 20 fighter. The Natural 20 stuff shouldnt apply to monsters because they roll so much.

Thankyou Big_Teej for the houserule

[Darkomn]

Its happened to me once, its actually a pretty good story.

DM was mad at the group, specifically me for guessing his plots by accident (his own fault I didn't know he would rip of Oblivion), and he read somewhere online that it was a good idea to throw a few UNwinnable fights at the party to show them who is boss. So we are traveling through the mountains and are warned of a "snow beast" the warlock(4E) rolls knowledge to see what kind of monster it could be and he gets a list of a few possibilities, none of them too bad so we continue on. Then as we are traveling through the thick snow (half movement speed, important) we come across this angle singing/speaking some awesome sounding language, DM tells us to roll initiative, I win but delay because it doesn't really look like a bad guy. The ranger delays too, the warlock player is sitting there shivering because apparently his knowledge told him more then he shared with the rest of us but we don't meta game and now its the angle's turn. It flies up with a speed of at least 8 and one hits the warlock, at this the ranger takes a shot and misses with an 18. Well dang its snowing and I can move 30 ft a turn and it can move 80, and attacks are useless what to do.

Well my group plays with the house rule that you can declare any action for your character and if you roll double 20s you can do it. Now you would have to be pretty stupid to use that to do anything except "I cut off its head" most of the time, but I'm a mage and don't even have a dagger but I also don't want to do something lame like I shoot magic missile down its neck and into its lungs. After a little thought I decide to go for something more cinematic, and equally effective rules wise, I fay step(4E) into him and rip out his hear (or other important organ)....
20...
20...
Me: Ha it works!!!
DM: not so fast roll again...
20...
DM: ok guys take a break while I think of ways to punish our Eledrin friend

so 20 minutes later it was dead and I was less one arm but still a great story.

EDIT: UNwinnable
Thanks for the catch Greenish, that's what I get for listing to the little red lines

[Mando Knight]

Except that probability doesn't work that way. The odds of 3 consecutive twenties are (1/20*1/20/*1/20) which is .0125%. The odds increase dramatically with each consecutive twenty, as you need to not only roll twenty, but roll it three times.
(1/20)³ = 1/8000 = 0.000125. Probability may not be my specialty, but I do know how to check my math.

(To determine the probability of instant death due to a triple-20 rolled during an arbitrary number (X) of attacks, take (1-(7999/8000)^X)*100%. BTW, if you somehow need to endure 5550 attacks, you have almost exactly a 50/50 chance of not getting hit with a triple-20. But that's if the HP damage from that many attacks doesn't kill you first.)

[Devronq]

We have used to rule pretty much since 3.0 came out and come into play many many times. Although you have all these formulas explaining how often it can happen it doesn't actually mean that's how often it's going to happen. Ive seen it happen twice in the same session and ive seen it not happen for 6months worth of sessions. Ive been playing more or less with the same group for 14years and we just recently changed the rule to automatically put you at -1 instead, which is exactly what someone else suggested as well. We also completely removed the massive damage rule because its far, far easy to activate and way way to difficult to live through.

[Madcrafter]

I've seen it happen three times in my DnD career. First time, a lv 4 rogue in my campaign encountered a pit fiend trapped in a magic circle. Decided that running up and trying to stab it in the face was the best idea, 20 20 20, dead. Second and third times were both in the same session, a boss monster (half troll fire giant thing) got triple twenties against one of the PCs, but he survived because of a displacement spell (the DM has a giant red die that flashes on a 20. It lit up a lot on that encounter.) That same session, a different player one shotted an regular monster.

It makes the game a lot more fun IMO, although I agree that it probably shouldn't be used against players.

[Mando Knight]

Although you have all these formulas explaining how often it can happen it doesn't actually mean that's how often it's going to happen.
Probability is what probability does: it predicts the likelihood of something occurring, but will not tell you exactly how many times in a given period it will occur.

[faceroll]

Ah that is correct then. That's still fairly remote however, but a lot less than it should be for an instakill rule.

1 in 20 is hardly fairly remote. In science, the cut-off for statistical significance in most fields is 0.05. But playing DnD has shown me that a 5% chance of something happening is really quite large.

[Greenish]

he read somewhere online that it was a good idea to throw a few winnable fights at the partyI tend to agree with that. If the players have to run away or be killed/captured in every battle, it gets old pretty fast.

[Flickerdart]

We also completely removed the massive damage rule because its far, far easy to activate and way way to difficult to live through.
By the time that 50 damage per attack becomes commonplace, a DC15 Fortitude save is no longer a real threat.

[Lord Ruby34]

I believe that natural ones would like to have a word with you.

[noparlpf]

Personally I dislike this rule incredibly. I don't even like the nat 1 auto-fail/nat 20 auto-success rule. 5% of the time even the most experienced warrior slips up and misses the broad side of a barn? I can understand it happening some of the time, but 5% is way too high. Likewise, a 1st level commoner can manage to find the one chink in a god's defenses 5% of the time. What? (Sure, said god probably has DR and it won't actually damage them if it's only a 1st level commoner, but still.)

[Trodon]

My friend's character got killed from a kick to the shin from an 11 year old boy that way...

[Mando Knight]

1 in 20 is hardly fairly remote. In science, the cut-off for statistical significance in most fields is 0.05. But playing DnD has shown me that a 5% chance of something happening is really quite large.

Instagib is 1/8000. It's fairly remote, but not remote enough if you're going to be adventuring.

[Drelua]

Personally I dislike this rule incredibly. I don't even like the nat 1 auto-fail/nat 20 auto-success rule. 5% of the time even the most experienced warrior slips up and misses the broad side of a barn? I can understand it happening some of the time, but 5% is way too high. Likewise, a 1st level commoner can manage to find the one chink in a god's defenses 5% of the time. What? (Sure, said god probably has DR and it won't actually damage them if it's only a 1st level commoner, but still.)

That's exactly why my group plays with a 20 counting as a 30 and a 1 counting as -10, though I've been meaning to tell my DM that auto-success is only supposed to work on attacks and saves, not skills and, well, everything else.

[noparlpf]

That's exactly why my group plays with a 20 counting as a 30 and a 1 counting as -10, though I've been meaning to tell my DM that auto-success is only supposed to work on attacks and saves, not skills and, well, everything else.

Yeah, my group plays with it for skills and the like too. I recently read the line in the PHB that says it doesn't work like that and I've been meaning to bring it up to the DMs.
And hey, I think we tried 20=>30 and 1=>-4 or -10 a few times. It's still not quite what I want though.
I do think there should be a chance of failure (say you sneeze, or a mosquito flies into your ear, or you stumble on a bit of gravel) but that it shouldn't be 5%.

[faceroll]

Instagib is 1/8000. It's fairly remote, but not remote enough if you're going to be adventuring.

On a single attack, but on the attack routine of a creature with 4+ attacks a round? The chances rise dramatically.

[Shotaro]

It's a terrible rule from a player's perspective: the more attack rolls a player faces, the more likely he is to instantly die horribly.

The chance of getting insta-killed by a level 16 Fighter? It's not 1/8000, it's .05%, each round he gets a full attack. The longer you fight, the more likely that one of the attacks you'll receive will instantly kill you. The DM might lose one or two monsters per campaign this way... but if it happens to a player, then suddenly they lose their character. And the same character will generally be the target of several dozens of attacks over his career.

Let's take the Fighter again. Let's assume the party is built around letting melee mix it up in melee while the casters support them, rather than the casters playing god and all that. If the Fighter (on average) is subject to five attacks per encounter, goes through ten encounters per level, and wants to get from level 1 to 20, he has a 11.2% chance of dying due to a triple-20. If he's subject to more attacks (such as the enemy getting several full attacks themselves), then his chances of dying go up further.

Statistically you are correct. I feel I was remiss in my original post as usually I would ask the players to make a fortitude save or die. However since this particular instance occured within the TOMB OF HORRORS! (dun dun dun etc) I thought I would run with it to make the whole thing just that little bit more lethal.

[Shotaro]

On a single attack, but on the attack routine of a creature with 4+ attacks a round? The chances rise dramatically.

Very true in fact we can safely assume that as has been said, 1/8000 per attack soon adds up and when dealing with a creature or character with say, 8 attacks it becomes as high as 0.1% I still gotta say 1/1000 is extremely unlikely but I understand where you are coming from.

[faceroll]

/facepalm.

0.05% is not the same as 0.05. My bad.

[137beth]

By the time that 50 damage per attack becomes commonplace, a DC15 Fortitude save is no longer a real threat.

I hate the 50 damage massive death rule. It turns the game into "who will roll a natural 1 first?"
Actually, I play with the open-ended-rolls variant in ELH, if you roll a natural 1, reroll and subtract 20 (a natural 20 gives you a reroll at +20). It still gets annoying making fortitude saves after every damage roll, especially since there's almost no chance of failing a DC 15 check using this variant.

[Blisstake]

The chance of getting insta-killed by a level 16 Fighter? It's not 1/8000, it's .05%, each round he gets a full attack. The longer you fight, the more likely that one of the attacks you'll receive will instantly kill you. The DM might lose one or two monsters per campaign this way... but if it happens to a player, then suddenly they lose their character. And the same character will generally be the target of several dozens of attacks over his career.

Probability does not work additively. For example, if you flip a coin, with a 50% chance of getting tails, flipping it twice doesn't give a 100% chance of tails coming up at some point. The actual odds of getting triple 20'd in a given round with four attack per round is

(1-(.999875^4))x100%

Which I can't calculate at the moment, but it's less than .05%

[Douglas]

Yeah, my group plays with it for skills and the like too. I recently read the line in the PHB that says it doesn't work like that and I've been meaning to bring it up to the DMs.
And hey, I think we tried 20=>30 and 1=>-4 or -10 a few times. It's still not quite what I want though.
I do think there should be a chance of failure (say you sneeze, or a mosquito flies into your ear, or you stumble on a bit of gravel) but that it shouldn't be 5%.
The problem with auto-success for skill checks is that it introduces things like ordinary level 1 commoners jumping to the moon just by trying 20 times.


Probability does not work additively. For example, if you flip a coin, with a 50% chance of getting tails, flipping it twice doesn't give a 100% chance of tails coming up at some point. The actual odds of getting triple 20'd in a given round with four attack per round is

(1-(.999875^4))x100%

Which I can't calculate at the moment, but it's less than .05%
You are technically correct, but the margin is ridiculously tiny.

[Mando Knight]

Probability does not work additively. For example, if you flip a coin, with a 50% chance of getting tails, flipping it twice doesn't give a 100% chance of tails coming up at some point. The actual odds of getting triple 20'd in a given round with four attack per round is

(1-(.999875^4))x100%

Which I can't calculate at the moment, but it's less than .05%

Actually, it's almost exactly 0.05%. It's 0.0499906% and some change. I provided the formula I used later on, which is basically the same as the one you gave.

[Acanous]

Now, mind you in the beginning of your career, things are not going to *Last* multiple rounds, most things that get multiple attacks do so at a negative, and by the time you get to the point where things get 4 swings on a full attack with any degree of regularity or accuracy, you have the money to be wearing armor of Fortification.

In situations where you're a clothy wizard, you might not invest in this, but a clanking tank most certainly should.
Finally, how many times an encounter is the tank going to be subject to full attacks? Depending on the actions of the rest of the party, he may never be attacked at all, or may have to stand in a doorframe and soak for 10 rounds. I propose an average of one full attack per combat, which is to say, a reasonably skilled party, and a reasonably skilled DM.

So! let's do this and show our work, shall we?

Level 1-5. 13.3*5= 66.5. *0.00125%=0.083% chance of dying to triple 20's. So less than one tenth of one percent, that's pretty reasonable.
Level 6-10. 26.6*5= 133. *0.00125%= 0.16625%. So there's a little less than two tenths of a percent. It happens to two out of every one thousand four hundred adventurers who make it to level 10.
11-15 is where things get sticky. 39.9*5= 199.5. *0.00125= 0.249375. So you're looking at one quarter of one percent. that's scary-high from an adventuring point of view. This is likely the level range wherin the most PCs die to the three twenties rule. By 16-20, you can affoard armor of heavy fortification. (Or one of a good number of other things that make you immune to criticals.)
If you were silly and left yourself open to critical threats in this bracket,
16-20. 53.2*5= 266. *0.00125= 0.3325%. 0.3% is pretty dang high, and you're looking at effectively one tenth of a percent per twenty required.

Of course, this is ALL much, much less of a concern than the 5%!! chance of death to any of a number of save-or-die spells that will be flung around within the 11-15 and 16-20 brackets. Further, enemies in those brackets will be relying less on attack rolls and more upon spells or spell-like-abilities.

Really, the combat coup-de-gras rule of three 20's in a row DOES help the player. While the DM may be expected to make as many as 667 attack rolls against a member of a fairly high skilled party, that party's player, assuming he is of the persuasion to make attack rolls, is likely going to roll MORE attacks than that. Sure, it's probably only going to kill a mook, wheras the DM might kill a PC, but the chance is so small that ANY kill with 3 20's is pretty damn epic.

I have, in my 11 years of gaming, only ever seen this happen in combat twice. I've seen it happen OUT of combat an additional 4 times (Skill challenges, craft checks, and when multiple saves are called for from death-trap-rooms, for example)

[Tibbaerrohwen]

I can only remember it happening once to me. Though we did have a few close calls in the past where we were all praying that something would intervene.

I was playing a rogue in one of my first 3.X campaigns, long before I knew how to effectively build a rogue. I can't melee to save my life and am up against two bruisers, albeit for stealing something necessarily dangerous. Que the triple 1's, and one falls on a candlestick.

Most of the time now, we hope it doesn't happen.

[faceroll]

Now, mind you in the beginning of your career, things are not going to *Last* multiple rounds, most things that get multiple attacks do so at a negative, and by the time you get to the point where things get 4 swings on a full attack with any degree of regularity or accuracy, you have the money to be wearing armor of Fortification.

If you're going with the three 20s is an auto-kill rule, you could have -infinity to hit and still kill your opponent in about 1 out of 8000 attacks.

[candycorn]

0.0125% per attack. Even discounting Haste, attacks of opportunity, and the like, iteratives turn that into the 0.05% per round that was quoted.

Statistics do not work that way. If they did, then that means that, in 8,000 attacks, there is a 100% chance that a triple 20 will be rolled.

There are no guarantees in probability. For you to survive an attack, your odds are 99.9875% For you to survive 4, you cube that percentage, for a final result of: 99.95000937%

Close, but not quite.

In 100 attacks, your odds of surviving them all are 98.7577%
Assuming that you are attacked, on average, 4 times per encounter, 14 encounters per level, and 20 levels? You take 1120 attacks.

Your odds of surviving this? 86.935% That means that you have a 13.065%, more or less, in 1120 attacks, of falling victim to an event with a 0.0125% chance of occurring.

8000 attacks? You still have 36.786% chance of surviving.

In other words, as the attacks mount, the amount each lowers the odds gets smaller and smaller.

25,000 attacks? 4.393% chance to survive.

In other words?
In 1 attack, the chance of death is 0.0125%.
the first 1,119 attacks increased your chance of death by 13%.
the next 6,880 attacks? Increased chance of death by 50.149%
the next 17,000 attacks? Increased chance of death by 32.393%

1 attack - 0.0125% per attack
Next 1,119? 0.000116% per attack
Next 6,880? 0.000073% per attack
Next 17000? 0.000019% per attack

Add on that a player can resurrect? And that further devalues the impact to the PC.

This thing is a rarity. It will likely happen as some point over your gaming life. But it will not likely happen on one character. Maybe 1 in 7 or 8 characters.

[candycorn]

Also, factor that on average, action economy is with the player. On average, the party will make more attacks than the enemy.

Take a level 1 group.
4 PCs vs a CR 1 orc.

Take a level 3 group.
4 PC's vs an Ogre.

A level 5 group.
4 PC's vs 2 Ogres.

There are exceptions to this, but by and large, a party will drop the enemy with less attacks.

Take the level 3 group. An 18 strength fighter's greatsword damage is 2d6+6, average 13. Assuming weapon focus, he's got 65% accuracy, and will likely win initiative. Add a charge, and we'll say average weighted damage is around 9.

Rogue, 20 dex (halfling). weapon finesse, and two weapon fighting. Likely wins initiative. Throws a dagger, after moving towards. Likely to hit, deals 1d3+2d6 damage (average 9, 6 after accuracy weighting)

Wizard could grease for the smart move, but let's go only damage. Magic missile. 2d4+2, average 7.

Ogre, on average, down about 22, of its 29. Ogre will attack once, likely hitting for 16 (we'll say weighted damage 12, 75% accuracy on the attack, representing attacking an AC 14 character)

Cleric will likely flank and aid another, and round 2, another couple attacks will end it. Party makes multiple attacks, and total attacks in the encounter is low. 1 enemy attack made, 2 tops, for an actual average of 0.25 to 0.5 attacks per player for the encounter.

[Trekkin]

Just out of curiousity, could a character be feasibly made to absolutely maximize the number of attacks per round to the exclusion of everything else, just to take advantage of this rule? There must be a way to get a statistically significant chance of this happening.

[candycorn]

Just out of curiousity, could a character be feasibly made to absolutely maximize the number of attacks per round to the exclusion of everything else, just to take advantage of this rule? There must be a way to get a statistically significant chance of this happening.

A character could... You'd start with pounce, and slap as many natural attacks as you can on.

Assuming you can get 12 (doable)? 600 rounds of combat yields a 50% or higher chance.

Assuming you can get 26 (hard, but possible)? About 250 rounds of combat.

If accuracy is no object, let's look at a level Thri-kreen Fighter 15/Barbarian 1. The multiweapon fighting tree, 4 held weapons, 4 elbow blades, 2 boot blades, 2 knee blades, a mouthpick weapon, and a prehensile tail weapon?

4 attacks with each, 14 weapons, total 56 attacks per round. That'd be 120ish rounds for a 50% chance. In any one round, it's have about a 0.7% chance of getting one.

[stainboy]

f you want it to apply to PCs, at least make it "automatically reduced to -1 hitpoints" so they don't just instantly lose their character.


I've used a house rule like that to make to make RAW instant death more palatable. Slay Living stops your heart but the party has a few rounds to give you magical CPR. I'm a fan.

And yeah, if you do something like that it's not a big deal to make every attack a 1/8000 chance of instant death. (It's probably not 1/8000 though, at least not on a physical die. Some people throw dice in a way that makes them likely to roll the same or adjacent faces over and over.)

[candycorn]

On a single attack, but on the attack routine of a creature with 4+ attacks a round? The chances rise dramatically.

0.0125%, or 1 in 8000, for 1 attack.
0.05%, or 1 in 2000, for 4 attacks.
0.5%, or 1 in 200, for 40 attacks.
4.88%, or about 1 in 20, for 400 attacks.
39.35%, about 2 in 5, for 4000 attacks.

So, in other words, if that creature with 4 attacks, did nothing but full attack, every round, for 40 minutes, it would have less than a 50% chance of seeing those odds.

In fact, 5544 attacks is as close to a coin toss as it gets, at 49.995% chance to see a triple 20.

[TroubleBrewing]

Just out of curiousity, could a character be feasibly made to absolutely maximize the number of attacks per round to the exclusion of everything else, just to take advantage of this rule? There must be a way to get a statistically significant chance of this happening.

Well, if you use the Lizzipede (http://brilliantgameologists.com/boards/index.php?PHPSESSID=1fs1qbotsticru00pidd7pslo5&topic=146.0;wap2), which gets 1,067,212 attacks/round, you end up auto-gibbing 133 dudes. And, assuming the "triple 1's and you kill yourself in a comically embarrassing way" variant, you end up auto-gibbing yourself 133 times, I might add.

[etrpgb]

4 attacks, 0.0125 chance of autokills means
1-(1-.0125)^4 = 0.04907... chance to autokill per round

If you get more attacks, i.e. 6... then
1-(1-.0125)^6 = 0.07269...

Math a part, I also agree it is a stupid rule. It is another way to improve life to casters that usually does not have to worry about attacks.

[candycorn]

4 attacks, 0.0125 chance of autokills means
1-(1-.0125)^4 = 0.04907... chance to autokill per round

If you get more attacks, i.e. 6... then
1-(1-.0125)^6 = 0.07269...

Math a part, I also agree it is a stupid rule. It is another way to improve life to casters that usually does not have to worry about attacks.

It actually provides a no save just die to the people that DO worry about attacks. It's rare as all hell, but it's there. This actually benefits attackers, ever so slightly.

Now, double 20 and confirm is another method, which has a much higher chance, depending on accuracy.

[tiercel]

Just out of curiousity, could a character be feasibly made to absolutely maximize the number of attacks per round to the exclusion of everything else, just to take advantage of this rule? There must be a way to get a statistically significant chance of this happening.

Leadership/Thrallherd/Diplomancy springs to mind, if "20s in a row" rule doesn't only work on PCs' rolls.

For that matter, any town/city above a certain size should be essentially immune to many large rampaging single monsters -- a few thousand peasants attacking with slings and normal pebbles at ten range increments can create a 500' radius Zone o' Statistical Death.

(OK, a little more complicated than that, given the Peasant Horde of Pebbly Doom is distributed over some area, but you get the idea.)

For your own personal army, just invest in a large range increment weapon you can buy a lot of (e.g. composite longbow), have your followers just launch nonproficient attacks if necessary at 1100' (or more, if you use some higher-range weapon en masse).

Heck, just the single natural-20 rule means you'll be getting plenty of regular hits, which will put down pretty much anything w/o DR. Some judicial purchases of magical and/or alchemical ammo can alleviate this too, of course.

[Seharvepernfan]

He's not missing the broad side of a barn during practice, he's missing in the heat of battle. Chances are, that barn is armored and moving around and surrounded by the guys friends.

EDIT: ...uh, ok, apparently a ninja clan wants my head.

[etrpgb]

@tiercel
nice thinking! 20 commoners with longbows have more than 93% of chance of killing a <put here a large monster which is not outright immune to arrows>!

[Devronq]

Sorry i know this is slightly off topic but massive damage fort save 15? wow thats week... umm my book says DC.15 plus the damage over 50. ex 100 dam is fort DC.65 which makes it much much stronger and easier to fail.

[Pigkappa]

Yeah, my group plays with it for skills and the like too. I recently read the line in the PHB that says it doesn't work like that and I've been meaning to bring it up to the DMs.
And hey, I think we tried 20=>30 and 1=>-4 or -10 a few times. It's still not quite what I want though.
I do think there should be a chance of failure (say you sneeze, or a mosquito flies into your ear, or you stumble on a bit of gravel) but that it shouldn't be 5%.

The most effective solution I found is this.

If you roll a 20, now roll 1d4. Your result is 20 + 1d4 - 1. If you got a 4 on the 1d4, roll 1d4 again.

So for example: 1d20 ---> 20! I can roll 1d4!
1d4 ---> 4! I can roll again!
1d4 ---> 2! Damnit!
Result = 20 + (4-1) + (2-1) = 24.

The same is possible if you roll a natural 1, with negative results (1d20 ---> 1, 1d4 ---> 3 means 1 - (3-1) = -1). The rule is complicated, but it doesn't make the game much slower since you only need to do this on a natural 1 or 20 and only if you can still succeed with a result lower than 1 or fail with a result greater than 20. And that's quite rare in D&D.

[tiercel]

@etrpgb

You'd like to have a lot more than 20 commoners to put down a big beastie, given than you'd expect around one normal autohit (natural 20) for every twenty schmucks cluelessly flinging arrows around. If you do use the instagib triple-20 rule, your chance of invoking it with only 20 attacks is relatively small (but still higher than leveraging your own personal attacks).

The extreme range helps though -- you may get off several rounds of Random Projectiles before your enemy can close to enough range to reply with anything, which only helps your odds.

-----

As a side note, I gave an unwary party of 8th-9th level PCs fits this way by having a small goblin army attack them with statistics. First, every goblin had a missile weapon, counting on natural 20s to produce at least a few hits every turn (more for low-AC PCs). With the goblins spread out in skirmish formation, no one AoE could easily obliterate enough of the low-hp targets. When the PCs closed, goblins went for statistical grappling (dogpile each PC with a ton o'goblins, then surround the dogpile with more goblins Aiding Another).

The PCs won of course, but they had a newfound respect for bog-standard CR 1/3 goblins in sufficient numbers and with good tactics against superior foes.

[Acanous]

I did something like that with a bard, once. Had all my level 1 followers retrain into Rangers, with Mounted Combat and Mounted Archery. The DM refused to let me play them in the field, instead saying random encounters stopped happening, and all our enemies that didn't die in the resulting massacre now lived in underground fortresses.

I was pretty sad about the terrain/combat change, but it makes sense. DnD enemies that can't respond with a similarly well-equipped army are just going to die.

[Frozen_Feet]

... for everyone arguing about how evil it is to have random chance to kill your characters, you do realize perfectly normal attack rolls and saving throws are much more likely to kill them than triple-20s?

[Flickerdart]

... for everyone arguing about how evil it is to have random chance to kill your characters, you do realize perfectly normal attack rolls and saving throws are much more likely to kill them than triple-20s?
Perfectly normal attacks and saving throws have less chance involved, or at least don't feel as bad - it's one thing when you're facing a powerful wizard BBEG whose save DCs are high because he's that good, and quite another when you're facing goblin lackey #357 and his nonmagical crossbow.

[tiercel]

... for everyone arguing about how evil it is to have random chance to kill your characters, you do realize perfectly normal attack rolls and saving throws are much more likely to kill them than triple-20s?

The difference is that you can mitigate your PC's risk against normal attacks with AC, miss chance, resistances, immunities, DR, SR, save boosts, temp hp, healing, etc.

Against the triple-20 option, your ONLY defense is to prevent an attack roll from ever being made, or miss chance. While it is a smaller risk overall, it's one that's much harder to do anything about in general.

Also the risk is just as great regardless of the caliber of your foe -- even a relatively green adventurer can pretty much scoff at the standard attacks of a common housecat, but is just as vulnerable at being triple-20ed by a housecat's attack as one from a Storm Giant Warblade.

Also, dramatically, it sucks a lot less to have your character go down to Strahd von Zarovich or even single-handedly locked in melee with numberless undead hordes to save your companions, than to get ganked in the eye by Orc #1673's thrown javelin upon entering Area 27 of the dungeon just before your companions puree him and his three buddies. [Edit: partially ninja'd. Still.]

[Shotaro]

I think in terms of storytelling, it matters when how or why it happens, for instance if the fighter had just blindly charged into BBEG#3 and the roll happenned, frankly, he's a could. He did something dumb and the rolls punished him for it there would also be the factor of death by massive damage against a villain like that but if the player does something phenominally stupid and the rolls come out that way it's storytelling at it's finest. Even better is when a player wants to change character and has only told you about it that allows you to kill off characters in all manner of interesting and dramatic ways.

In the case of the death I mentioned in the original post it was a trapped corridor in the tomb of horrors, the guy had mamanged to survive for some time huddled in a corner behind a tower shield and, well, doing some rather unpleasant things to survive only to stand up and get killed by the first arrow that comes his way. It was just TOO PERFECT dramatically for him to die then and there and the player was alright with it. I think the way I am going to broach it with my players is - if it makes sense dramatically for you to die, you die. IF it's a goblin with a crossbow and you are level 18 and it happens you have a chance against it.

[Vladislav]

Let's say there are 10 combat encounters per level. And a combat lasts 5 rounds. And an average PC is attacked once per round. So there are a total of 10x5x20 = 800 attack rolls made against your character over the course of his 1-20 level lifetime. Which means there is approximately 10% chance your character will die to a random triple-twenty, and there's no way you can stop it.

Not fun.

[etrpgb]

@etrpgb
You'd like to have a lot more than 20 commoners to put down a big beastie, ...

LOL, you are right of course. I mistyped something. You need 220 commoners!

[Seharvepernfan]

I tweaked this rule and use it as a houserule: "If you roll 2 20s in a row, roll again, if this third attack roll hits, the blow is treated as a coup de grace."
Except that rogues dont suddenly get sneak attack damage added in and things like that.

This goes for pcs and npcs alike.

It's still lethal, but at least both sides have a chance.

[Sith_Happens]

LD50 is 5545 attacks.

Not sure if the number is correct (haven't read the whole thread and don't want to take the time to check myself), but I would just like to comment how hilarious I found the concept of a houserule having an LD50.

I'm not even going to try and calculate one for the "Natural 1 on an attack roll = you hit yourself" rule that most of the groups I've seen use...

[candycorn]

Not sure if the number is correct (haven't read the whole thread and don't want to take the time to check myself), but I would just like to comment how hilarious I found the concept of a houserule having an LD50.

I'm not even going to try and calculate one for the "Natural 1 on an attack roll = you hit yourself" rule that most of the groups I've seen use...

It is correct. It's over 50%. 5544 attacks is under it. However, 5545 is closer to 50%. By Price is Right rules, it'd be 5544 (closest to the 50% threshold without going over). removing the "without going over" part means that 5545 is 5 times closer to 50% than 5544. Bear in mind, the difference is miniscule, on the order of 1 in 1,000,000 or so.

[hustlertwo]

I tweaked this rule and use it as a houserule: "If you roll 2 20s in a row, roll again, if this third attack roll hits, the blow is treated as a coup de grace."
Except that rogues dont suddenly get sneak attack damage added in and things like that.

This goes for pcs and npcs alike.

It's still lethal, but at least both sides have a chance.

That's the same rule the DMs in a game on this board, Aldhaven: Vicious Betrayals, use. Just had it happen with my skeleton companion while fighting a bugbear spirit shaman. Two nat 20's with the masterwork short sword, and he said auto-kill if the third roll confirmed. It certainly did, since it was a third nat-20, so autokill (and a juicy XP reward for my unlikely luck).

However, having that rule work against PCs (unless maybe in PvP combat) is dumb. If you're that gung-ho on killing your players, there's less cheesy ways to do it. Like all of them.

[Seharvepernfan]

That's the same rule the DMs in a game on this board, Aldhaven: Vicious Betrayals, use. Just had it happen with my skeleton companion while fighting a bugbear spirit shaman. Two nat 20's with the masterwork short sword, and he said auto-kill if the third roll confirmed. It certainly did, since it was a third nat-20, so autokill (and a juicy XP reward for my unlikely luck).

It sounds like they're using the one from the book, not mine.

[Thurbane]

That's not really a rule. Or rather, not a printed rule.
DMG p.28, Variant: Instant Kill.

[Jarxlee]

Happened to our felow player level 8 , started from 1 and died by other player hands after that player threw a small rock at him and dm called roll and after 3×20 rolls he is dead. One of dumbes optional rules ever.

[atemu1234]

Except that probability doesn't work that way. The odds of 3 consecutive twenties are (1/20*1/20/*1/20) which is .0125%. The odds increase dramatically with each consecutive twenty, as you need to not only roll twenty, but roll it three times.

Yes, but depending on how much your characters are involved in combat, the odds of that happening at least once increase surprisingly quickly.

[weckar]

0.0125% per attack. Even discounting Haste, attacks of opportunity, and the like, iteratives turn that into the 0.05% per round that was quoted.

Still does not work that way. That's like saying that after 8000 attacks it'd be 100%: Guaranteed. It is not.

[Ashtagon]

Still does not work that way. That's like saying that after 8000 attacks it'd be 100%: Guaranteed. It is not.

There'd be a 28.3% chance of rolling triple 20s after 8000 attacks. Rises to 50% after 16,613 attacks.

Assuming 4 PCs and 4 NPCs per encounter each make 1 attack roll every round, and the encounter lasts 4 rounds, that's 32 rolls per encounter. Assuming a full level1-20 campaign at 13 1/3 encounters per level-up, that's 253 1/3 encounters, and 8107 attack rolls. 28.67% chance of getting a triple 20 over the lifetime of a full campaign.

[weckar]

... putting it at roughly once every 8 years, if you play weekly.

Why is there a rule for this again?

[Vhaidara]

Still does not work that way. That's like saying that after 8000 attacks it'd be 100%: Guaranteed. It is not.

You're right, it doesn't. Once this rule is in place, the odds of someone being instakilled by it each round are somewhere around 1/3.

What? You still think DnD cares about how actual probability works?

[Khedrac]

Does it work on undead, such as this thread?

Let's try:

20, 20, 20 = Autokill Undead Thread.

If not, then it needs a moderator.

[Jarxlee]

Does it work on undead, such as this thread?

Let's try:

20, 20, 20 = Autokill Undead Thread.

If not, then it needs a moderator.

Anything is up on the table for a kill even crit immune monsters. I have seen once lvl 3 player kill Asmodeus with 3x20

[Yogibear41]

Its a variant rule in the DMG. Our group uses it and I HATE it. Mainly because I am the only one who has ever been affected by it, TWICE (In less than 1 year). Although to be fair the 2nd time I had a way to not die, it still took me out of the fight, but my character lived.

Although it works vs monsters and npcs as well, I personally rather get rid of it, because my luck when it comes to dice is pretty horrible.


The rule makes even less sense when even on a critical hit the weapon wouldn't damage the monster, like rolling a critical hit with a steel dagger for 2d4 damage vs a werewolf with a dr of 10/silver.

The level 3 killing asmodeous is just crap.

[atemu1234]

A character could... You'd start with pounce, and slap as many natural attacks as you can on.

Assuming you can get 12 (doable)? 600 rounds of combat yields a 50% or higher chance.

Assuming you can get 26 (hard, but possible)? About 250 rounds of combat.

If accuracy is no object, let's look at a level Thri-kreen Fighter 15/Barbarian 1. The multiweapon fighting tree, 4 held weapons, 4 elbow blades, 2 boot blades, 2 knee blades, a mouthpick weapon, and a prehensile tail weapon?

4 attacks with each, 14 weapons, total 56 attacks per round. That'd be 120ish rounds for a 50% chance. In any one round, it's have about a 0.7% chance of getting one.

Gods, now I just want to add in something like rogue or sneak attack fighter to make the most of those ridiculously numerous attacks.


Sorry i know this is slightly off topic but massive damage fort save 15? wow thats week... umm my book says DC.15 plus the damage over 50. ex 100 dam is fort DC.65 which makes it much much stronger and easier to fail.

To be fair, by the point you're taking fifty points of damage in a single round, you're supposed to be able to shrug it off with relative ease.