I mean, it requires so much of world-building...

I've long wanted to GM something in the fantasy genre, but I've always run into the problem that I'm not sure what I'd like the setting to be like. I always try to come up with my own world and I get lost in the minutiae...

Sometimes, I think I'd like to run something simple and not very deep, with a lot of monster-fighting etc. I have Pathfinder and it seems good for this kind of game... But even with PF, I end up lost in the world-building, as I don't like the default gods from the core rulebook, I try to fiddle with the races...

And starting a game in GURPS is even worse. I love GURPS, but I don't want to use the default magic system. So, I always end up lost in trying to decide what magic should be like in my setting and how to model it. It's so frustrating that I think that, if I ever run a GURPS fantasy game, I'll be disallowing magic users as a playable character type...

Overall: GMing is hard. But GMing fantasy seems double-hard. :P

There's a reason people buy adventure modules.

Then they don't have to worry about that sort of thing.  :D

Assuming they can just leave the greater world as it is and concentrate on the bits concerning just the adventure of course.

I long ago gave up on upfront world building. I work with my players to create what we need to get the game started and to keep it moving. It doesn't result in a realistically consistent world, but that's the key: it doesn't need to be consistent or detailed or any particular thing, as long as the players are bought into it, and people are usually bought into things they helped create.

If you're at all interested in what I mean, I'm happy to discuss it further, elsewhere.

You think fantasy is hard, try Sci-Fi!
Not just a world but a whole Universe.  ;)

I think you can get lost in worldbuilding with any kind of game.  Even with a modern day or near future game, you can spend massive amounts of time detailing NPCs and areas where the PCs will probably never actually go.  This is one of those "GM as an artist" things comes into play.  You really only need to define the things that matter.  The hard part is figuring out what matters.

Caleb Stokes just successfully kickstarted a game called Red Markets.  Part of the multi-year leadup to that was a large presence for both himself and the game on the Role Playing Public Radio podcast.  There was a dedicated podcast to his efforts at publication called Game Designers Workshop.  It's a pretty rich source for a GM that's trying to build something, either from a game perspective or even from a world perspective.

Anyway, I mention that because I want to refer to something he was talking about in the most recent episode (which was also a GenCon panel).  The most useful thing you can have for your game is a tag line.  Everything you put into your game needs to support that tag line.  In his specific example, the tag line for Red Markets is "a game of economic horror".  Everything that might get added to the game was looked at with the questions "Does this make it economic?" and "Does this make it horrific?"  If the answer was no to both, the item didn't get included.

I think that is a powerful tool for even a GM that's looking to build a world.  Distill what you're going for into a tag line (a couple of sentences, tops).  Is what you're working on going to support that tag line?  If not, toss it out.

For instance, this conversation has given me a vague idea to make a GURPS fantasy game.  The tag line for the game would be "a game about small town folks finding their way after an invasion".  Everything I'd prep for the game would have to do with either the small town, the invasion, or the "way" those folks would have to have.  Anything else, like the demographics of the next kingdom over, whether or not there are dragons (aside from as part of the invasion), or ancient history would be a bit of a waste of my time.

TL;DR: Figure out what the focus of the game is, then do the things that support the focus.

All good points.

And keep in mind that a lot of detailed worlds were not only built over many years, but often had multiple people adding their help and two cents to it. Some games, like Dungeon World, are designed for getting the players to help build the world as they go along.

My own gaming world was started by two other people back when I first started playing. It consisted of a roll of yellow construction paper tacked up on the wall of his basement. A dot was put on the map, which was the hometown of one of the players. Then another dot was added about an inch away. That was the hometown of one of that players henchmen; it was a day's ride away.

And that's how the world started.

The biggest thing you'd have to learn is that not every technique works for everyone. The world I (and others) built never had a 'tag line'. I'd even have trouble coming up with one now for a campaign I'd start.

I'd say the trick is to start small... unless you are one of those brilliant designers who can envision the whole world in toto.

Also keep in mind that no matter what you start out with, it will change and evolve just as you do, and depending on the players you have.

Don't try to do it all at the beginning. If you want to keep it simple to start with, then put them in a free city. That way you don't have to worry about the politics and nearby countries and stuff... at least not right off the bat.

As for deities, again, details aren't necessary. If you have a cleric or paladin that worships a warrior god, that's all you need to know to start. He can get a name later.

I'd probably start with how you want to generate characters. And then work on each step.

For example, some systems start with Ability scores. Will PCs be average, above average? How high can they get? How will they improve them? Will the ability scores dictate what race they can be? What class they can be?

Or maybe you want to start on races first. Rather than delve into variants, either give players the option of 4-5 different races, or make them all variants of the same race, such as Elves (High, Grey, Moon, Sun, Wood, Wild, Fire, Frost, Aquatic, Wind, etc.). Or, if you put them in a specific setting, such as the Arctic, then everyone has to have the 'Frost' template.

With every decision you make, you'll be building your world in increments, as each decision leads to another one.