How I Plan Sessions

Continuing from my previous post on campaign design, once I have a campaign outline and a full player roster, I start planning individual sessions. I approach nearly all of them the same way—except for Session 1, which I’ll cover in a future post.

For this walkthrough, I’ll use Session 3 of my campaign Shadows Over Siqram as an example.

Where Are We Starting, and Where Are We Going?

Because I run roleplay-heavy games with deep PC integration, I want each session to feel open-ended and player-driven—even though I’m always steering toward a specific beat. My villains keep moving even if the players don’t, but my goal is to crash the PCs into that moving train.

In order to do that, I start with… well the start. For Session 3, the opening scene was:

There’s a couple of hooks in this scene already - we have the infernal book, we have the fiancé, and we have a bunch of dead cultists who may have some clues on them about what’s going on here. I make note of that, but don’t dwell on it for too long. Then I zoom out: where does the session need to end?

At this point in the campaign, Zakaroth (my BBEG) has completed Steps 1–3 of his plan and is working on Step 4: corrupting a glyph so that, when reinserted into the boundary protecting the city, it destroys it instead.  At this point in my planning, I didn’t really have a good end point. So, I went back to the plan and started to interrogate it.

I know why Zakaroth wants to corrupt this glyph. But I hadn’t figured out how he was going to do it. So I looked through my setting and my PCs backstories for inspiration, and I noticed that one of my PCs, Arranis, was an artifact hunter with ties to the black market.

That clicked: what if Zakaroth needed a powerful infernal artifact to to use in some infernal ritual to corrupt the glyph? Arranis could be tipped off about it entering the city.

That raised another question: how do you smuggle a magical artifact into a city full of powerful mages and magical defenses? Surely some wizard would notice its magical presence, right?

Answer: Hide it in a crowd.

If the cultists could surround the artifact with so many other magical items that it’s hard to detect any one item, they could slip it in unnoticed. But how?

Again, I went back to my PCs backstories for the answer. Ash had a noble mother—Lady Elithor. We’d seen her briefly in Session 1. The players hated her. She was cold, calculating, powerful, and had a chip on her shoulder about the city failing to investigate her husband’s death.

That was it. Lady Elithor would host a gala. Officially, it’s a fundraiser for the cult’s victims. Secretly, it’s cover for the smuggling operation. All those wealthy nobles, all those magical items—they provide the perfect magical camouflage.

I wasn't planning on making Lady Elithor a major NPC, but the players hated her so much, I turned her into one of the main villains.

With that, I had my ending:

Before I go off and figure out how the players will get there, I do a little more work to figure out what else needs to happen in this session.

What Needs to Move Forward?

I keep a “plot tracker” active for all of my campaigns. It’s nothing complicated, it’s just a document that has a list of where each different arc is so that I can make sure I’m pushing them all forward each session. In this campaign I only had 8 sessions to work with, and 4 of them would be mostly combat. That left me with less than 4 sessions to move everything else forward, including player arcs.

At the end of Session 2 when I started planning Session 3, my plot tracker looked like this:

I already knew how far Zakaroth’s arc was going to go - I have the end point of the session. For the rest I just kinda thought through the next logical step:

For most of these player arcs I really didn’t know where they were going - I just kinda thought through “what comes next”. And that’s okay! I really wanted to see how things evolved naturally at the table and then only push the campaign in the direction the players thought would be cool.

So now I know where I start, where I end and what else needs to happen this session. Next I look at what information the PCs need to get there.

What Information Needs to Be Delivered?

Before I start designing scenes, I'll take stock of what information I need to deliver. In addition to the information to drive player arcs forward, I also look at what may have happened off screen in the BBEG's plan. For this session, that included Zakaroth stealing corrupted glyph.

And then I needed to figure out how to give the information for everything else that needs to move forward this session. To do this, I just work backwards from the “what needs to move forward” bullet points and try to fill in the gap between it and “what do the players know now”. For example, Arranis knew they were stuck in the Material Plane, but they didn’t know why or how. To move Arranis’s arc forward, they needed to find out there’s a reason. So I ask myself, what information do I need to give, and who is going to give it?

To find the answer I looked through my setting, the previous session notes, and player backstories, and decided on two things. The first was that the BBEG knew why Arranis was stuck, and had been planning on using it as leverage to get Arranis to cooperate. The second was that Arranis was part of The Raven Queen’s plans for Siqram. Once I decided on that, I had to decide how to deliver this new information to Arranis. And here’s what I came up with:

Lady Elithor felt like a natural choice, as she was emerging as the hated face of the BBEG. I already knew she was going to throw a gala and that there’d be plenty of opportunity for her and Arranis to interact.

And in that way, I continued on down my list. Here’s what the completed “what information do I need to give” section looked like in my notes:

In this step I don’t spend a ton of time trying to figure out how I’m going to get this information to the players, just what needs to be delivered. This gives me a good grasp of where the session will go, but lets me stay flexible at the table. My players never stick to my plans; they almost always derail them. And I love that! By spending my time in the setting and the goals for the session, I can adapt.

If I anticipate a PC will want to go north and look in the forest for clues, but instead they head east to the swamp, I can on-the-fly figure out how to get the needed information to them in the swamp. If a dryad was going to warn them of some danger in the forest, maybe a will-o-the-wisp is going to do it in the swamp instead.

I could, if needed, stop my session planning here and just improvise the session. It would be a bit bumpy, and wouldn’t flow as well as I want, but I’d feel confident I could get the PCs where I needed them and everyone would have fun.

That said, once I’ve got this list of information I need to deliver, I start to work on thinking through the scenes that would be useful for delivering said information.

What Scenes Do I Anticipate?

Like I said before, I find this step to be totally optional, but it helps me ratchet up the drama in the game, so I spend a decent amount of time on it. It works well for me and my players.

So, in this step, I look at the information I need to give, think about where the players currently are, and try to anticipate where a natural delivery point will be. I get this right about 50% of the time, and because I only get it right 50% of the time, I try to brainstorm more than one way to give each piece of information, and if the information is really important, I’ll make sure I give it to more than one player. My golden rule is: If my campaign depends on the players knowing something, tell them about it three times.

So, starting at the top, how are the PCs going to find out about the gala? The gala is a critical piece of information, so I need to tell them about it three times, and preferably, to three different players. There’s an obvious starting point: Lady Elithor is both the host of the party and the mother of one of my PCs. Of course she personally invites him.

That means I need to design a scene that takes place in Ash’s office. (Fun fact, when actually playing this session, Ash never returned to his office. Being the fantastic commander he is, he set up a field tent so he could be on the frontline with his troops. So, I pivoted the scene to the field tent on the fly - that’s why the high level prep is so important).

That takes care of one way this information is delivered. Who else is a natural target for it? Well, Arranis is an Indiana Jones type  - they even carry a whip. They’re probably aware of when important artifacts are on the move; maybe Arranis gets a tip that something is going down at the gala. Nothing concrete, but enough to get Arranis’s interest piqued about attending.

So this implies a scene at Arranis’s home. I chose their home because presumably Arranis would want to get their fiancée home after her grisly ordeal in the ritual.

For the last one, I decided to kill two birds with one stone. In session one we set a precedent of the Marshal’s field officers giving daily briefings to the marshal. So, it seemed reasonable that after such a calamitous event the field officers would give a briefing. In this briefing Ash could be informed of both the missing glyph and the gala.

And so I prepped a scene at the Unified Guard’s HQ. (Fun fact, this one also had to be moved last minute. Ash held this briefing in his command tent!)

And on down the list I went. Sometimes I will give information in a more “meta” way. For example, one of my bits of information to deliver was that Ash was going to be up for nomination to the Unified Guard’s office. One of Ash’s main character goals was to prove that he is a capable leader, and to get out from under the shadow of his deceased father, the previous Marshal of the Guard. In session 1 I made a big deal out of how most of the soldiers of the Unified Guard didn’t respect Ash - they thought he was a nepo-baby. And so throughout the sessions I kept track of how Ash responded to certain situations, and whether or not he was a selfless leader. And I decided to surface that hidden mechanic in a scene in session 3.

To set up the scene, I had one of Izzy’s mentees (Kael)  come to visit. Kael was now on the Council of Voices, and would be voting on whether or not to nominate Ash. Kael had reservations about Ash’s ability to lead. And knowing that Izzy and Ash were friends, Kael wanted to know what Izzy thought of Ash.

So I had a scene prepped for this. The information is delivered to Ash’s player via the meta (he’s at the table while this is going on), but not directly to him. Some folks might disagree with this, but I definitely lean into metagaming at my table. It gives my players the ability to lean into tension and drama between their characters, and makes for really fun scenes. Here’s the one I jotted down for session 3:

This scene was AMAZING. Izzy’s player knew that Ash’s player was sitting at the table. And Izzy spoke openly about her opinion of Ash, and Ash’s player got to peek behind the curtain. Of course Izzy gave Ash a glowing review, and Ash’s player loved it. And now Ash “knows” that his position on the council is uncertain.

I believe I also had Ash’s personal assistant let him know that the vote would be happening, just to close the in-game loop.

At this point I’ve got everything I need to run a fairly in-depth, four hour roleplaying session. Each player has a plot that moves forward, the BBEG’s plan progresses, and the players have some meaningful scenes and decisions to make.

The very last step, which I’ll cover in an upcoming post, is to design the individual scenes so they hit super hard.

Is This "Railroading"?

Really quick, I want to talk about “railroading”. “Railroading”, to me, means taking away player agency during a session. An example might be: When I plan on a player going to the forest, and they instead go to the swamp, I find some way to force them to the forest instead. That’s my definition. 

By some reading, you could argue that my style of campaign and session design is railroading. Some GMs and players would rather be dropped into a world and to just explore that world and see what happens. The GM may have some BBEG lined up, or some plot hooks, but in general the game is freeform. Those games are generally called “sandbox” games.

In my mind, my style of game is somewhere between the two - perhaps a bit closer to railroad than sandbox. And that’s by design. My players are all over the age of 35. We all have full time jobs, some of us have families. Scheduling 4-hour blocks is hard. A game that has no known end date isn’t viable. And on top of that, my player’s aren’t incredibly decisive. Left to their own devices, they tend to hem-and-haw for a really long time about incredibly simple decisions (like which fork to take in a dungeon - spoiler! they end up in the same place). 

So, to facilitate a really deep immersion that will end by a certain date, I design my sessions with very specific stopping points. These are must-hit story beats. And I’ll make adjustments throughout the session to make sure we more-or-less land there by the end. If we’re moving too fast, I’ll slow the pace and give some more lore or more space for player roleplay. If we’re moving too slow, I’ll cut content and narrate things that were supposed to be more interactive. 

This strategy works for my table, with my players. Based on the feedback I’ve gotten, they’re getting meaningful roleplay and choice in a game that moves at a sustainable pace for their personal lives. It might not be the right strategy for you and your table, and that’s totally fine.

Take what works and use it and ignore the rest.