David Ricardo certainly was no role-player, but his theory of comparative advantage can be applied to role-playing systems. The general idea of this theory is that everybody should do whatever he is *relatively* better at. So how does this apply to role-playing games? Well, if you want to play a RPG, you have a lot of very different choices, from pen & paper to MMORPGs. And if you want to create a fun pen & paper campaign, as I am currently trying, you need to ask yourself at what pen & paper is relatively better than a computer RPG.
If you haven't played any pen & paper adventures yet, or at least no good ones, you might think that computer games are better in all respects. Pen & paper games have lousy graphics, are difficult to set up, and are generally extremely slow. A fight of a full party of players against an even number of orcs of the same level in a computer RPG would take not more than a minute, while in a pen & paper RPG it can take half an hour.
But combat is what a computer RPG is comparatively good at. In a pen & paper game a player might need to look up a rule or description of a power, move over the battlefield while counting squares, roll a die to determine success, roll another die for damage, keep various bonuses to his roll in mind, and keep score. Combat in a computer game is a lot easier, because the player just needs to hit the button of the ability he wants to use, and all the calculations and score-keeping are done by the computer. Of course computer RPG combat has often been simplified, with no zones of control or other effects of positioning, and players in many games even being able to run right through a monster.
The comparative advantage of pen & paper RPGs is more on the non-combat side. Dialogue between a player and a non-player character (NPC) is extremely limited, ranging from a simple option to either accept a quest or not, to a still rather canned choice between 3 to 5 responses. And even a dialogue with 5 steps of 5 dialogue choices each usually ends up having only one or two possible outcomes, not 25. In a pen & paper game the Dungeon Master (DM) who controls all NPCs is a real human being, and thus has a much better "parser". I once was in a group sneaking into a warehouse to investigate its owner, and we were caught by the guard. And one of the players in the group came up with a fantastic bluff, pretending to be a security inspector sent by the owner to test how good the guard was, and even chiding him for taking so long to find us. The DM went along, and we ended up with lots of information and a bribe from the guard to keep mum about his "failing". Unless that particular event is scripted into a computer game, it simply cannot happen there.
This is why I try to put a good amount of role-playing and similar non-combat encounters into my Dungeons & Dragons campaign, at least getting the players to spend half of their time out of combat. Now I think that D&D 4E has quite a good combat system for a pen & paper game, very nicely tactical and all. But the comparative advantage is in the role-playing, in having meaningful choices and consequences to your actions, in puzzles and challenges the players can't look up on YouTube, and in my flexibility as a human DM. If my players just wanted combat, they probably could find a good computer game to serve them faster and better. But in presenting an open world and endless options, not even Skyrim can beat a real human DM.