Concentration for Martials

[GeneralVryth]

Really good point, and just one part of the development of skills that was left incredibly underbaked. Some things I would strongly favor/consider
- there should be a limit on untrained checks, something like 10 + modifier, or even just flat 10. I know this kind of parsing was mostly abandoned in 5e, but automatic disadvantage on checks above that DC, or just not being be able to try at all, would really raise the value of proficiency
- guidance on DC's, including examples of various common tasks
- some system of total failure, partial success (near miss?), success, and smashing success (or something). I would favor something like missing by more than 5, missing by 5 or less, just succeeding, and succeeding by more than 5 for mechanical categories

I like all of the above frankly. On the variety of success front another thing you could do with proficiency, is limit any successes to partial or basic success, and require proficiency for the greater successes.

1) AC was better managed
Between casters easily stacking full plate with shields and shield spells, and gishes combining heavy armor with haste or blur, and then randomass builds that have a flat 25 AC...bounded accuracy kinda falls apart. It doesn't even matter how hard a creature hits; if +5 to +7 is supposed to be relevant into t2 and t3, PC's simply can't have AC that's well into the 20's. Most monsters get to attack 4-9 times ever. With a 15% chance to hit (and that's before factoring in further defenses like silvery barbs and runic shield), 4-9 attacks adds up to .6 to 1.35 hits. Over an entire combat. That's simply not bounded accuracy in any meaningful sense.

I agree with all of this. One of advantages of armor as DR is takes some pressure off AC as the primary mitigation option. It should still be harder for casters to get heavy armor prof, but that is another topic. Imagine if the top of line tank Fighter at level 10 was only rocking 18 to 20 AC, but had 10 DR on top of that? While the friendly Wizard is at more like 15 or 16 AC, 3ish DR, and access to something like the Shield spell?

2) Level cap was lower than 20
This is tied to a ton of other problems (casters scaling vs martial scaling), but as far as bounded accuracy goes, I don't want or need hobgoblins to be relevant enemies at level 15. 10, maybe, if there's a whole army of them. Between AC, HP, and sheer number of resources, at some point the characters are basically superheroes, and yah know what? Street thugs don't threaten Spiderman. I'd prefer to not pretend that they should. I really think that for bounded accuracy to remain true, characters would have to scale a whole lot less than they do...in which case, why have 20 levels? Why grind through 20 levels of extremely marginal changes, when 12 or 10 or 8 or 6 would bring you right up to the correct power level?

I don't think lowering the level cap is necessary. The basic Hobgoblin isn't seriously expected to be a threat by level 15, or really likely tier 3 in general. But when you compare to prior editions, where they would be obsolete by level 7 or sooner it's an improvement. Also, I think the place it's more noticeable, are things like the Knights, Veterans, or more "elite" warriors. A CR 2 to 4 stays dangerous for a long time if there is a sufficient number of them. So to use your Spiderman example, street thugs can become obsolete but he still has to respect the SWAT team that gets called in (or following comic book logic maybe not, but you get the idea).