D&D 5E How fantastic are natural 1's?


log in or register to remove this ad

I never said confirmation rolls should be part of the system, just that it was an option for people who don't want a 1 to fail and a 20 to succeed because they feel 5% for each is too great. And at that point when I was talking about 20 always succeeding and 1 always failing, those weren't "silly" criticals/fumbles as you seemed to think.
Yes it is silly and it is a terrible rule. Allowing some no-skill, low-ability character succeed in a DC 30 task is silly. Allowing a high level character with expertise and high ability score fail at a trivial tasks is a terrible rule.
 

Especially because people as twist the interpretation... like @Oofta said, "there are lies, damn lies, and statistics..." :D
I worked with a guy who was a very numbers driven person. People didnt like him because he could always back up his arguments with factual numbers but everyone always said numbers could be skewd. So who knows.
 

I found this odd when I took a statistics course, didn't understand most of it, and its very possible Im wrong and misremembering, but I seem to recall that that yes if you flip a coin its 50% itll land on heads or tails. But I think that the if you flipped it 100 times statistically it will land on one more than the other. So 50% doesnt make sense to me but then again I got a D in the class so listening to me isnt wise.
Well... No coin is perfectly 50/50, but the variation is extremely small, and that’s a physics thing not a statistics thing. Statistically, each individual flip has pretty damn near a 50% chance of landing on either side. But you flip a coin 100 times, you’re not necessarily going to get 50 heads and 50 tails. The more coin flips you do though, the more likely you are to get close to that 50/50 average. You’re more likely to get closer to 50 heads and 50 tails after 100 flips than you are to get close to 5 heads and 5 tails after 10 flips. And you’re even more likely to get closer to 500 heads and 500 tails after 1,000 flips.
 

Here's my contribution to the "tell me a story about a critical failure in one of your game sessions" story pool:

In a 3.5 game, I had a player running a half-orc barbarian who was destined to one day become the ruler of a small kingdom. Not wanting his kingdom run by a "gods-be-damned orc," a nobleman from that kingdom hired a doppelganger sorcerer/assassin to slay him with a magic dagger that would imprison his soul in its hilt, preventing him from being resurrected. The doppelganger took on the form of the head of the Adventurers Guild and brought the half-orc in to a one-on-one meeting, under the pretext that there had been an accusation made against him that needed to be addressed.

Once they were alone in the Guildmaster's office, the doppelganger activated spell-tokens that arcane locked the door to the room and made it such that no noise would emanate from the room, as the half-orc's dwarven bodyguard was standing just outside, keeping guard. Then, while the half-orc sat and read the complaint (the player had only recently spent the skill points to allow his barbarian to learn to read and write), the doppelganger took out the magic dagger and prepared for a death attack. He hit, but didn't do enough damage to slay the half-orc and imprison his soul - and now he had an enraged barbarian to deal with instead of one preoccupied with reading a trumped-up complaint. To stall for time, he cast an invisibility spell on himself and backed away.

The half-orc was bleeding from a stab wound, locked in an office with an invisible assassin, and had no weapons on him at the moment. So the player asked if he could lift the desk and throw it the length of the room, hopefully hitting the invisible assassin and pinning him in place. I told him to go for it, well aware the barbarian was easily strong enough to do so but I'd be adding penalties for a non-standard weapon and the overall weight of his "improvised weapon" - and then the player proceeded to roll a natural 1.

I took that situation as a particularly appropriate time for a critical failure (given the room size, just about anything else would have succeeded - there wasn't much chance the desk would miss) and had him slip on a piece of paper (the false complaint) he had been unaware he'd been standing on, which upended the desk right on top of him, pinning him in place instead of the doppelganger.

It all worked out in the end, though: he managed to extricate himself while the doppelganger was still lining up his next death strike (which takes three rounds) and tried again, this time knocking over the doppelganger with the hurled desk as originally intended. Eventually, it turned into a one-on-one fight between a (now visible, after his next attack) doppelganger sorcerer/assassin with a magic dagger against a raging half-orc barbarian armed with the splintered leg of the desk's chair. The doppelganger had his throat pierced by the chair leg and that was pretty much that. (And then the PCs hunted down the nobleman who had hired the hit in the first place....)

Johnathan
 



Eventually, it turned into a one-on-one fight between a (now visible, after his next attack) doppelganger sorcerer/assassin with a magic dagger against a raging half-orc barbarian armed with the splintered leg of the desk's chair.
Gave him the old Ron Artest huh?

 

Well... No coin is perfectly 50/50, but the variation is extremely small, and that’s a physics thing not a statistics thing. Statistically, each individual flip has pretty damn near a 50% chance of landing on either side. But you flip a coin 100 times, you’re not necessarily going to get 50 heads and 50 tails. The more coin flips you do though, the more likely you are to get close to that 50/50 average. You’re more likely to get closer to 50 heads and 50 tails after 100 flips than you are to get close to 5 heads and 5 tails after 10 flips. And you’re even more likely to get closer to 500 heads and 500 tails after 1,000 flips.

I think what they are trying to say is that a slight deviation from average is more likely than an exactly average number

Let us take a 2d6 as an analogy, as I am very familiar with its distribution curve. The average value for 2d6 is 7. That is the most likely number to come up (1/6). But 5/6 times, the result will not be 7.

Let us do a formal proof. We shall flip a coin 4 times. There are 16 possible results. They are:



HHHH
HHHT
HHTH
HHTT
HTHH
HTHT
HTTH
HTTT
THHH
THHT
THTH
THTT
TTHH
TTHT
TTTH
TTTT

Of these results, 6 out of 16 have balanced results (50/50). But 10 out of 16 are not balanced.

I could do a demonstration with much more coins but... I don't feel like doing the work ha!
 

Yes it is silly and it is a terrible rule. Allowing some no-skill, low-ability character succeed in a DC 30 task is silly. Allowing a high level character with expertise and high ability score fail at a trivial tasks is a terrible rule.
Ok, you're not talking about the same thing anymore that I am so I'm leaving this alone and your insistence that some things are silly is just insulting. Later.
 

Remove ads

Top