Quote Originally Posted by Skrum View Post
So I know you're just spitballing and I just wanna be clear that I'm not trying to like bang the table about something offered here, but your ideas (and 5e as well) really makes me consider the value of "highly balanced" systems. 5e as we know isn't really balanced, or at least not well, but to some degree it is driven by bounded accuracy. That kind of system essentially by necessity puts a big damper on scaling, as bigger numbers can screw up the carefully balanced numbers.

My thought/concern though is like...how do characters differentiate themselves in a system that doesn't allow scaling. Ability checks are something of an example; like imagine if scaling in 5e was even more constrained, and ability scores was the only factor for skill checks. Well, my character might've been a blacksmith before he become an adventurer. So I made sure he has good str, as that's what he'd roll for smithing. Another character could have even more str...and just be "better" at blacksmithing, even if they've never narratively done blacksmithing.

That's just like a super simple example, but that's the idea. It makes me rethink the wisdom of something like bounded accuracy, or tightly balanced game systems.

*this problem is even worse in a game like DND which tends not to give some classes very notable or distinct abilities
You're right to a degree. Though I also think the point of Bounded Accuracy gets forgotten to a degree as well. Bounded Accuracy isn't about removing scaling, it's about keeping scaling from driving things (and combat abilities in particular) off the edge of the die range. With the goal of keeping lower level enemies relevant for a much greater scope of the game. In essence if an attack without circumstantial modifiers can only hit on a nat 20, or miss on a nat 1, or if DR is involved, damage is reduced to 0, that is bounded accuracy failing. Skills suffer from this because there should be more nuance and less randomness there, the problem is the game doesn't properly divorce skills from combat, so you can't just use a separate system for them or allow much higher bonuses (and 5e does do that some, and you can see some of the issues with the grappler with Expertise in Athletics). Ironically the Saga System I mentioned earlier had this a major pain point, because using the Force was tied to skill use, and you could scale skills faster than normal combat bonuses, and force abilities could allow you to substitute your Force skill for some offense or defense.

The fix for skills is to completely remove from any combat use (or at least make it some they can't be used against someone, or be used to evade or get out of attack). Then you can allow larger bonuses to better differentiate characters. Though for your Blacksmithing example, that just sounds like you should have had Smithing Tool's prof which would have helped in the comparison. Another thing you can do if you want to minimize changes is make skills roll 3d6, that significantly reduces outlier rolls so bonus matters more (and revert to d20 for combat related pieces), another is borrow from BG3 and only allow certain check if a character is proficient, or provide bonuses or alternative check options when proficient.

At its core, I really do like Bounded Accuracy, I like the idea that plain Hobgoblin Soldier can still sting a dragon if there are enough of them, and especially if they are led well. On the differentiation front, I do think that is possible, but you need to have a clear idea of what the themes of your classes are and make sure you have enough levers where you can achieve the same result through different means and making different classes good at those different means. That's part of why I like the DR thing, it's a lever for differentiation, it also has a nice verisimilitude bonus.