You're taking a reasonable concept (if you can critically hit, you should be able to critically fumble) to an absurd extreme, it only lessens the value of your post.
That's not at all what I did. I could have picked 18 or higher. Or any other restrictive number. The point was to show that the argument that a thing increased challenge or tension isn't a good argument -- many things can do so that are actively bad.
And yet that was your argument. When defending the use of fumbles, you made the argument that they would increase the challenge of an opponent. If you're rescinding that argument, now, then I think we're making progress.
As I have pointed out numerous times (in case you didn't read my prior posts), I agree having a fumble on a flat nat 1 is not appropriate in 5E due to the nature of the game mechanics. As such, I have always stated a confirmation roll with a fixed DC, but which your bonus improves as you level, is viable. It even works if higher level fighters get more attacks per round.
I perhaps misunderstood you earlier, where you said you used the confirmation roll to avoid what you called 'disasters' and that an unconfirmed fumble would be simpler, like dropping a weapon or falling prone. Ah, yes, I was correct, you call these "mishaps" and confirmed fumbles "disasters" in post #154.
So, you do have an show of incompetence occur on a 1, but a worse show of incompetence if they happen to roll a 1 and then roll low again. Not sure this is actually better, as it punishes competence even more. If I'm making 4 attacks a round as a 20th level fighter, I'll hit a "disaster" fairly often. My attack is base +12, my equal tier opponents ACs are often in the mid-20's, so even with magic my confirmation roll is still around 10-20%. Probability-wise, that's about a 1% or so chance of a disaster on every attack. So, statistically, I'm grossly incompetent on average once every 100 attacks. And I'm the pinnacle of martial prowess!
5% also represents the base-level for statistical significance in pretty much any random event. Also, you might be competent but if a roll is actually called for then you are in a situation where failure will likely have consequences. 5E removed the ideas of taking 10 or 20, which IMO was a horrible omission because they made a lot of sense. Without a stressful event, the idea of taking your time and using your full competence is logical. But, in the heat of combat, with injury or your life on the line, in the moment you're about to be discovered picking a lock and facing the temple guards, etc. that tiny, little 5% of something potentially bad happening works well. It represents the negative outlier, just as critical hits represent the positive one. Of course you aren't meant to "feel good" when those moments happen--but if your entire game revolves around you "feeling good" then we have very different D&D experiences. I want challenge and the unexpected, and critical fumbles can represent another way of bringing those elements into the game.
No, 5% doesn't represent the base level for statistical significance. If I had a correlation of around 5%, I'd toss it as noise. I think you're inverting the wee-p concept, where a statistical confidence of 95% or better is considered sound. This is much contested, because it results in bad assumptions. There's the wee-wee-p move, to try to get a higher threshold for statistical confidence but also a move to abandon wee-p in favor of other approaches that are more clear on the assumptions involved.
5e didn't actually remove the concepts of take 10 or take 20, they just moved them into the normal resolution mechanic. If I can reasonably take more time without risk, then there's little actual consequence involved -- you just take the time and succeed.
And, if you're arguing that fights are dangerous, yes, they are -- that's what the other guys trying to kill you represent. If you're arguing that it's very likely to drop your sword (and 5% is pretty darned likely per attack) or fall down without anyone else doing anything to you, then you're already off the reservation. If you think fights should be more dangerous, then increase the danger of the opposition or the environment. You've chosen to increase the incompetence of the PCs.
Again, if the idea is pursuit of farce, go for it. If you like the farcical results, more power to you -- that's awesome. I tend to play Paranoia for my farce, but there's no reason you can't do it in D&D. Just don't argue that the purpose of fumbles is to increase the realism or danger of combat -- there's lots of ways to do this that don't rely on a heaping dose of farce alongside.
LOL whatever path works for you. Fumbles work for me.
I have other peoples' posts in this thread about how bad fumble rules can ruin the game for them.
This is not data that people have largely abandoned outsized fumble results -- the opposite, really.
If the fumble rule supported those odds, I would agree with you, and anyone who uses a flat nat 1 for "dropping" your sword probably is IMO.
Again, I point to your post #154. You say you do this.
That is because they don't add the confirmation rolls I've discussed in this thread. We currently use DC 15, but you could even do DC 10 to make a fumble a 1 in 400 chance and a "disaster" (more severe which our table uses) would then be 1 in 8000.
I return to my golf example. Even pro-golfers will likely hit a bad-ball (i.e. fumble) once or more times in a round averaging 72 strokes.
Hitting a bad ball is a miss, not a fumble. They don't break their club, or send the ball into their caddy, or... really pro-golfer's idea of a bad hit is to lose a ball in a hazard because they tried a super difficult shot -- one a normal player couldn't make except by extreme lucky accident. You're sighting their normal failures at really hard shots as equivalent to fumbles that cause incompetent results (like a 20th level fighter dropping his sword).
Fumbles are, by definition, extra-bad failures, not just failures. Missing a golf shot is not extra-bad, it's just a failure. The ball still goes towards the green. What you're arguing for isn't failure, but extra bad failure. On a flat percentage. Your confirmation roll is just extra dressing -- the result of a fumble is still to describe the PC as being terribly incompetent. You're leveraging farce, but defending it as somehow representative of how otherwise competent people fail. It's not. If you had the chance to just fall down or drop your sword for even tier II fighters that your approach suggests, then serious combatants are falling down and dropping swords quite often. That's not normal failure in a fight (that's getting run through by the other guy's sword) -- it's a farce full of pratfalls and banana peels. A perfectly valid goal, but not one aligned with your arguments.
