So I’m thinking about confict resolution in RPGs (which includes most RPGs’ favorite form of conflict, combat) and considering certain approaches which essentially eliminate the possibility of “one-hit wins” (or losses), that is, where you simply cannot win a conflict with a single roll of the dice. Kind of a reverse of the “minion” type rules popular in some games, where you can (and do) win conflicts with inconsequential characters in one go, something that ensures major conflicts in the game have a certain back-and-forth exchange to them before one side decisively wins. This connects up to earlier thoughts on Building Towards Success on a smaller scale (encounter rather than entire story/adventure).
The fun thing about it is you can have a “sliding scale” of conflict based on just how important the encounter really is: from “it’s not worth rolling dice, you just do it” to “it’s going to take several rolls and completing multiple minor steps/challenges to achieve success.” Plus similar types of encounters don’t all need to be resolved in the same way. Some fights (such as with the aforementioned inconsequential minions) may be summarized as “so, how do you whup their butts?” without a single die thrown to tense back-and-forth conflicts as both sides jockey for advantage and eventual victory.
To use d20 as an example, it would be like having (at least) three combat systems: the standard attack roll + damage roll, with all the fixin’s (hit points, combat maneuvers and such), a simple opposed check (with the higher result winning the fight), and a simple comparison, with the higher value winning automatically, if it comes down to a fight, all based on how significant the fight is to the overall adventure. Of course, the issue where d20 is concerned is that the middle option might be too variable in outcome if it relies on a single d20 roll. This whole approach may work better in systems with bell-shaped probability curves.