As someone, somewhere, somewhen said, a knight never fell of his horse until they invented the 'ride' skill. I've had PCs fall over when running because the GM, seduced by itemised skill lists and the possibilities that they present for demanding a dice roll, has called for an 'Athletics' check!
I detest skill systems in RPGs. No, that's not quite right. I like the idea of a skill system in principle, the way in which they can add texture to a PC and give colour to the world in which the PCs live. But I detest the way in which skill systems are usually implemented, either through the advice given to the GM in the rulebook/s, or the way in which published adventures set the precedent for their application. Introductory adventures (in particular) for RPGs with skill systems all to often 'teach' the game by demanding that GMs ask for a whole series of pointless, inconsequential skill rolls. Look at Through the Drakwald for WFRP2e, or Caravan from RQ6's Book of Quests. Whatever the other merits of these scenarios, the extent to which they 'teach the system' involves skill tests being called for at inappropriate moments; moments that are either not the result of player choice - railroad skill rolls - or that have no consequential bearing on the adventure - quantum skill rolls - or, worse BOTH.
But I do run (and sometimes play) a whole host of games which have some kind of skill system - a variety of BRP-derived games, Advanced Fighting Fantasy, Traveller, Warhammer Fantasy Role-Play, even RC D&D etc. - while always feeling a dissatisfied with the way in which I find myself applying the mechanics of the skill system.
And why is that? That's because most skill systems are written as if they are 'rolls to accomplish'. Rolling to accomplish means either a lot of failing on the part of the PCs, or dice being rolled where, whether by way of modifiers or very 'competent' characters, there is little chance of failure. It wasn't until I read this post on Tales to Astound! that I 'got' how I ought to be conceptualising skills systems, and refereeing their application in my games. Take out the Traveller specific stuff and concentrate on this extract:
This! Yes, this. Couple this with an understanding of Saving Throws that stresses player agency (from Courtney Campbell) - "The saving throw versus death, especially at low levels is a roll called for when the player has already made a poor choice that results in certain death. It is a chance to avoid death caused by a bad choice" - and we have a way of interpreting RPGs with skill systems in way that is consonant with 'old school' play, which stresses player agency (and player skill) and which reserves dice rolls made by the players for moments of peril and danger (perhaps not always physical). You're not rolling to accomplish. Rather, as in combat, you are rolling to see if your PC escapes without something terrible happening.
So, under this interpretation, skill systems present a more granular breakdown of Saving Throw categories. Of course, Newt Newport's Crypts & Things, a Swords & Wizardry variant, uses Saving Throws *as* the 'skill' mechanic, which is a neat and simple way of doing things. In a D&D alike such as Swords and Wizardry Saving Throws start off around 15 or so (just over a 25% chance of success) and reach about 5 (a 75% chance of success) as PCs reach name level. Which is just about perfect for a skill percentage, where it is understood that this represents a roll to prevent something awful that would otherwise happen from happening. It puts the skill ratings of starting RuneQuest, and even starting WFRP characters, into perspective.
Now, perhaps none of this is news to you. Great, you've been playing a better game than I have. But given the precedent set by the examples in the rule books and in published adventures, it is something worth putting down in explicit terms, even if only to remind myself to referee skill rolls in a better way.