Here are the traditional reason I would say no to my players and why I shouldn't in Dungeon World:
- Because doing so would ruin my plans
- In my head, this is physically impossible or there's not enough time etc.
- Because the action would cause sudden PvP combat
Here's why I would be wrong to say no for those reasons in Dungeon World
1. Because doing so would ruin my plans
In Dungeon World, you have no personal agenda to railroad the players in your stories and your intentions. GMing Dungeon World is about letting go of selfishness and let the player be the master of their own destiny. That's hard for me because I often prepare stuff in advance and will unconsciously lead the players into my preps or simply make them happen regardless of the players' actions.
It's really hard for me coming from a Dungeons and Dragons background to let go of that. I usually plan my setting and often the players will ask me if their cool idea would work in my setting and I would often say no...no you can't play a Drow they are evil and you'll be attacked on sight in the first village you visit. No you can't play a Druid because the game is about political intrigue and a Druid wouldn't fit in the court of the king. Plus..there are no forest around the starting area to host a druid circle.
That's just bad. For some reasons it's assumed in many games that you should ask the GM if your character fits in his world. In Dungeon World you should feed from the players' ideas and background. When they asked me details about the starting area, I told them nothing more than simple stuff that wouldn't ruin their character concept. We'll be starting in a small village near the coast where pirates are constantly threatening the locals. It respects the principles of the game because those are broad strokes and it leaves place for player inputs.
2. In my head, this is physically impossible
The problem there is I would often describe the situation and the options but didn't mention some details that seemed obvious to me. By lack of description, the players will often assume nothing prevents them from doing something.
In your example, I would have asked if he understood the leader was surrounded by guards and passing through that line of defense would be dangerous. If yes, you ask how he reaches the leader. If it makes sense, you call for a move (usually Defy danger) based on the description. Description is the key and will often set the options available for the players.
3. Because the action would cause sudden PvP combat
In my games, I usually forbid the players to start physical conflict between them because the system is not balanced for players to fight other players (like 4E). Also, I think conflict would simply threaten the future of the game (read 1) so I would say: No pvp combat. Fix this out of game.
Big..mistake.
With the bond system and the way alignments work in Dungeon World I realize that player conflicts could be really interesting. I remember playing Warhammer Fantasy Roleplay 3rd edition and they have a party sheet with a conflict tracker. I loved it and included it in my Dungeon World game after the players playing the elf and the dwarf said they didn't like each other. I made a GM move out of it.
I'm the MC that SevenSidedDie was referring to. In our game, we had a Hocus and a Hardholder who were, at times, allies, enemies and mistrustful participants in the advancement of the community they both needed to survive.
The way I handled it was with the PC-NPC-PC triangle. I put all the major NPCs in between the two characters. I put the characters at odds with each other via the NPCs and their pressures / desires / actions. Look at the things they both have stakes in and ask them to solve those problems. For us, it was an easy job - divide the community by Church and State. Let's say Kettle doesn't have enough food, right? She wants to eat. So she goes to the Hocus and says "Hey, Want, I need some food, man." and Want says "let me meditate on that" and then she goes to Mom and says "Mom, I didn't get my rations last week - I need food!" and Mom says "Okay baby, let me see what I can do" and then see what happens.
Find weak points in the status quo, represent them with real human problems and don't be scared to pit the PCs against each other that way. Give them lots of things to agree on, sure, but complicate them. They both want Kettle to get fed, but why does it matter who does the feeding? Want would sure look good if he could scrounge up some rations - that'd make Mom look like a pretty poor leader and paint Want as the real boss of this place.
Think of the NPCs as human beings - simple for the most part, but desperate, sad, in love, hungry, sweaty messy idiots, too. They need things and they turn to the PCs for them. they're manipulative and earnest and nothing complicates things more than people being people.
I find that the most successful games of AW aren't about mutant hordes or unexploded nuclear ordinance but the way those scary things force people together or apart. The question you ALWAYS want to be asking is "what do these people want out of life?" and then let that guide you forward.
Best Answer
There are events happening in Apocalypse World that the players don't know about (yet), and fronts are how the GM collects these happenings and reasons about them. Fronts' details fall under the GM principles of Think offscreen too and Sometimes, disclaim decision-making. The players don't even need to know that fronts exist mechanically, but they're absolutely something that also exist in play that they can be concerned about. “Is Tum Tum's gang going to be a problem?” “Well, Dou says their waterhole dried up, so what do you think?”
If the players want to know more, let them go investigate. AW has lots of ways for the players to get more information.