Jester David
Hero
Personally, my interest is more toward making skill checks more like combat − robust, balanced, thoughtful, challenging.
When skills use the same math that combat does, then it is easier to adjudicate the difficulty and the appropriate levels when difficult stunts should be more plausible. Consistent math makes it easier to handle skill checks on the fly, as they arise naturally as part of narrative decisions by players.
Also, when skills use the same math as combat, it is easier and more balanced to integrate skill checks as *part* of innovative or situational attack maneuvers.
The difference is in play, you want combat to have a moderate miss chance. And unless you die, you can repeat the check again during your next turn with little or no penalty due to having missed. Missing once in combat doesn't majorly alter the narrative.
Meanwhile, for skills, a single missed check can change the story. And you may not get a second chance. So while there should be a chance of failure, the odds should be different as the results are different.
Plus, all characters have roughly comparable success rates for combat. A mage hitting with a spell and warrior hitting with a weapon have the same odds of success. But skilled characters should have higher chances of success than unskilled. Excluding Expertise, the bonuses of a proficient character are not that much higher than an non-proficient one; for most of the game, Ability scores will matter more, so the Wizard might be better at Religion than the cleric just because they're smarter. Given the swingy nature of the die, if everyone in the party rolls the check (because everyone can) then the odds of the person who chose to specialise and train in that skill getting the highest result, is likely 1:3. It only gets higher if you can synergize with Ability Score and proficiency, but even then a lucky roll can make someone "better" at the task than you.