How do I mechanically handle player's improvisation?

[MrZJunior]

A couple sessions back my players did something I totally did not expect. They managed to totally rearrange the geopolitics of the world in the second session.

I was upfront with them and said I was not at all prepared for this, apologized, and ended the session an hour early.

[quinron]

A couple sessions back my players did something I totally did not expect. They managed to totally rearrange the geopolitics of the world in the second session.

I was upfront with them and said I was not at all prepared for this, apologized, and ended the session an hour early.

Okay, I have to know what they did.

[Yora]

I always feel that everything becomes much more important and interesting if you can say the whole time "remember, all of this is because of you!"

[OldTrees1]

I always feel that everything becomes much more important and interesting if you can say the whole time "remember, all of this is because of you!"

Indeed! I am currently living that up right now.

My PCs decided to release 4 powerful souls that I did not predict they would release. That has basically rewritten the endings and added a new epilogue chapter addressing these 4 powerful entities and the consequences of their release.

So how did I mechanically handle that player improvisation?
I already had some notes about what those entities were and how they were held. I knew what would happen if the players did actions A thru Y. When the players went with action Z, I asked them for details on what they were doing. That gave them time to discuss / speak and give me time to think. I compared action Z against the outcomes I knew and I thought about how these souls were held. Based upon that time and those comparisons, I was able to figure out that the souls being released was my best guess at what action Z should do in this context. So I went with that outcome. The players released the souls and I described what they observed. Cut to a couple months later and they are now dealing with the result of the outcome. In this case the couple months of IRL time helped me figure out important but time consuming things like stats for these entities. Luckily I could delay that prep until after the session because it was not immediately required.

[MrZJunior]

Okay, I have to know what they did.

I was expecting the campaign to revolve around a siege of a particular city. I was going to have the party do tasks relating to helping the defenders. Instead they opened the gates.

Now the centaurs have taken over the imperial capital, their khan is the new emperor, and the party is helping him consolidate his power.

[BRC]

I was expecting the campaign to revolve around a siege of a particular city. I was going to have the party do tasks relating to helping the defenders. Instead they opened the gates.

Now the centaurs have taken over the imperial capital, their khan is the new emperor, and the party is helping him consolidate his power.

Good on you for rolling with that, but here is another note.


It's okay to request that the players buy-in to the premise of the campaign, at least in the most general sense, and especially at the start.


If the GM says "This campaign is about defending this city from a siege, build characters who would involve themselves in the effort to defend this city from a siege", that's fine. As the GM, you are not obligated to accept every rules-legal character concept that arrives at your table, and making sure everybody is on the same page is a good step to running any game.
It's also a good way to check to make sure the players are going to be engaged in the story you're interested in telling, and to avoid frustrating character conflicts.



For example, if one player wants to play a Noble Paladin who Destroys Evil, and the other wants to play a Demon Worshipper, you should make sure that they both understand that they will be sharing a party and table with the other character.

What's more, even if they like the idea of "We will be at each other's throats constantly", you should probably check with the rest of the group to make sure they're okay with "hey, a lot of the game is going to be these two characters fighting each other until they can be wrangled back into cooperation".

[MrZJunior]

I don't intend to sound unhappy, I'm actually quite pleased. I think it's great that my players did something so unexpected. My favorite thing about RPGs is how flexible they are. You can do just about anything. It's not like a video game where you are limited to what the designers thought of and had the time and resources to implement. I like to present people with a scenario and see what they do.

[Calthropstu]

I am responding to the original question.

Short answer: You don't.
Long answer?
You flesh out a realistic high level wizards defenses, complete with deadly traps that can kill the characters. You have them roll a wisdom check to realise how stupidly dangerous it is. Then go all out creating his tower. Have a succubus disguised as a dryad inside a magic circle begging for their help. Add a few deadly traps, hiding most of the wizards truly valuable objects inside secret chest and the like. If they succeed in robbing the tower, all they'll really get is a few grand in gold, enmity from the mage in question, a cursed item left as a trap and a released succubus... whom they will like try to extort in her "dryad" form. (if they extort her for sex boy are they screwed... heh.)

Have the wizard's quasit or imp familiar secretly fly around setting (and manually resetting) traps invisibly. Have a cloudkill effect (which the familiar is immune to) go off. (Any wizard with a bound creature would have such an effect to prevent ants or similar from walking across summon circles).

Stress how stupidly dangerous this is. If they say "we can handle it" and die en masse so be it.