Locations: To prep or not to prep?

One of the draws of DW is the ability to do freeform roleplay, and not have the game drag as a result. The second session I ran went off without a hitch (mostly), and no prep was involved.

So for the second session, thou shalt prep fronts/dangers, and a steading or two. Sensible enough. But what about locations, or dungeons?

Positives: It is Dungeon World, after all, so having a few on hand seems common sense enough. I've also been advised that freeform RP 24/7 is hard to do consistently, and that without location prep games can descend into madness very quickly. Many moves would seem to depend on locations prepped beforehand too, like "reveal an unwelcome truth", for example.

Negatives: With all prep comes the temptation to railroad a group into said prep, because otherwise work is wasted and we just can't have that. Plus it's satisfying to be lazy.

What do you guys think about prep? Yay or nay, and why?

Low prep. Find some dungeon maps on the internet and fill in a short descriptor for each room. Have a couple of them ready and tied to dangers in a front. Then you are prepared, when your group enters a dungeon. Prep a couple more whenever you have time.

It's a rare group that can go completely freeform in my experience. And I know my spatial orientation is terrible, so having to come up with a coherent dungeon on the fly would likely be a disaster.

I think it's perfectly okay to draw up and prepare a location as long as you're keeping an eye on the Agenda & Principles. I'd add that less is more. A couple scant notes here and there, some impressions, maybe a couple of custom moves, and plenty of blanks is enough to keep it from being a railroad. The less you prep, the less motivated you'll feel to force those elements into the game.

Now, if you're talking about "I've created a dungeon and that's what you're doing tonight," well, that's a little different. I'd really want my players' buy-in on that, either by way of world-building or stuff they said they wanted to do as it came up in the fiction during a previous session.

I've always liked players mapping as they go.  That said, there's no reason not to map out individual locations - places you might have a scene.  These might be the Troll King's throne room, the hafling longrider's encampment, or a cell block from the prison at Coresmouth.  That gives you a lot of flexibility to say, separate the party by narrating a split in a tunnel or a forest fire and still have a couple set piece maps to slap down on the table.  You're prepping but, as a cartographer, you're leaving plenty of blanks.

I do zero prep usually in that regard, but I find that online map generators are useful, so I usually have a couple blank pre-printed dungeons on hand.  Something like a cave complex, something that might be a temple, a couple floors of a castle and some generic town bits.  Anything else we just sketch in a couple seconds on a sheet of paper or two.

I keep blank maps onhand, and fill in the rooms as we go based on what the dungeon should have.  Sometimes bits get crossed out or filled in, sometimes we add more bits ... it's all pretty organic.  I find that coming up with some themes (like the themed monster books) and doing a few custom moves for a place that fits that theme and paper-clipping it to the blank maps is a good idea.

I'm with Marshall. I like to 'leave spaces'. I find that scenes that answer stakes questions are what I like to prep, (these are my flags as GM) then inject them into the unfolding narrative as appropriate.

There are two main parts to this prep. The first is a list of the NPCs, organisations, recurrent monsters and their relationships to each other. The second is a short description of these NPC's and their goals and instincts. I try to make these as antagonistic to the player's flags as I can.

Then I think of a few locations already introduced into the game, does it work to re-incorporate one of them to 'set the scene'? If yes then great! If no, I come up with a vague concept of where that stake may be answered, and allow the players to organically fill in the spaces through explorative play.

This may involve having a few one-page dungeon maps on hand, or a handful of map tiles or simply some cool imagery in my head. So long as I have a firm grasp on the antagonists and what they want, the 'backdrop' writes itself.