GM Advice

What This Chapter Is For

Division 5 is not the hardest tabletop RPG to run, but it is easy to run slightly off-key, and once the tone drifts, the game stops feeling like the one the table sat down to play. A session that drifts into a generic action game is not a bad session, exactly; it is just no longer a Division 5 session. The table will feel the difference without necessarily being able to name it.

This chapter is the toolkit for holding that tone. The first section defines the core GMing stance: the one-sentence discipline that every other section builds on. The sections that follow address the specific places the register tends to drift: the dial between the Le Carré and Fleming poles, the drift problems that accumulate over a campaign, the double-agent reveal and how to handle it without ambushing the table, the Compromise Track as campaign accelerator rather than attrition model, pacing at operation and campaign levels, and a long miscellany of common-ruling edge cases the book does not quite cover anywhere else.

Everything here is a recommendation, not a rule: where it conflicts with an actual Chapter Eight rule, the rule wins, and where it conflicts with what a given table needs, the table wins.

How to Use This Chapter

If you are reading this before your first Division 5 session, read it in order. The sections build on each other, and the stance sections at the front define the vocabulary the later sections use.

If you are reading this in the middle of a campaign that has drifted, skim the table of contents for the section that names the problem you are having. Each section is self-contained; they are designed to be reread under pressure.

If you are reading this after a session that went well and you are trying to figure out why, read the Compromise as Pressure section and the Pacing section. Those are the two places where a Division 5 GM most often does the right thing without knowing they have done it.

Running Division 5: The Core Stance

You are not the opposition. You are the world.

This is the one-sentence discipline. If you remember nothing else from this chapter, remember this.

The default GMing stance in Division 5 is to play the world rather than play the opposition. The opposition, Der Falke, the KGB colonel, the lobby security contractor, exists inside the world. The world is larger. The world includes the weather in Vienna, the bureaucratic mood of the French border police, the fact that the train from Budapest arrives forty minutes late this time of year, the news on the radio from Moscow that nobody is quite sure how to interpret. The world is what gives the opposition texture. The opposition without the world is an action movie. The world without the opposition is a travelogue. Division 5 lives in the intersection.

What “Playing the World” Means at the Table

You are running every NPC, not just the hostile ones. The waiter. The concierge. The customs officer. The cab driver. The asset’s aunt. The retired intelligence officer who runs the bookstore on the Graben. Each of these is a person with their own register, their own reason to be in the scene, their own bearing on the operation. Playing the world means populating the scene with enough of them that the operation feels like a real place, not a set.

You are running the institutions, not just the individuals. A Stasi investigation is a filing system, a pattern of meetings, a habit of cross-referencing, not a single Stasi officer. An operation against the Stasi is an operation against the institution, and playing the world means letting the institution act like one: slowly, procedurally, with the full weight of its bureaucratic competence. The individual officer is the face of the institution, not a substitute for it.

You are running the consequences, not the threats. This is the subtle one. A new GM runs threats, the sentry who might see you, the drone that might find you, the officer who might catch you. An experienced GM runs consequences, the sentry who did see you three operations ago, whose report is now sitting in a file in Berlin, which was pulled yesterday by the case officer assigned to the new operation, who is about to meet you at the café this afternoon. The difference is enormous. The threat-running GM produces tense scenes. The consequence-running GM produces campaigns.

The Opposition Is a Scene Partner, Not a Boss

The opposition in Division 5 is rarely boss-fight opposition. Der Falke is there to be outrun, outmaneuvered, or out-patienced, not killed at the end of the arc. Shcherbakov will be beaten in a Vienna hotel, never in a firefight. The Architect’s arc ends when his firm collapses around him, not when he dies.

This shifts the GMing stance. You are not running a boss. You are running a scene partner who wants to win. The scene partner plays fair. The scene partner has motives, constraints, and reasons. The scene partner makes mistakes: sometimes because they are having a bad day, sometimes because the operative is actually better than they are. The scene partner is capable of being surprised by the operative.

A scene-partner GM runs the opposition at the level of the operation’s actual logic. If the operative’s plan is clever and the opposition’s counter-plan is less clever, the operative wins: cleanly, without a GM thumb on the scale. If the opposite is true, the opposition wins, and the operative deals with the consequences. The table trusts a scene-partner GM because the decisions are legible. The table stops trusting a boss-running GM within three sessions.

Playing the World in One Sentence Per NPC

A practical discipline: every named NPC who enters a scene gets one sentence of inner life the GM holds in mind before speaking. Not a backstory. Not a characteristic array. One sentence.

“The concierge would rather be home with his granddaughter.”

“The Stasi officer is embarrassed about his French.”

“The waiter has three children and a heroin problem the table will never learn about.”

That one sentence is enough. It turns the NPC from a set piece into a person, which is the move the whole chapter is about. The players will feel the difference even if they cannot say why. Over thirty sessions, those one-sentence inner lives accumulate into a city the table believes in.

The Three Registers You Are Running Simultaneously

Every scene in Division 5 happens at three registers at once, and the GM is holding all three in mind while running it.

The Tactical Register. What is happening mechanically in the scene, who is rolling, what the difficulties are, what the Tradecraft Pool looks like, whose Legend is exposed. The rules chapter. This register is the one the GM has the most explicit support for.

The Procedural Register. What a professional intelligence operation looks like. Real tradecraft has shapes, the surveillance detection run, the dead drop, the brush pass, the asset meeting, the exfiltration, and the GM’s job is to make sure those shapes are visible in the scene. A meeting with an asset that is just a conversation is a bad scene. A meeting with an asset in which the operative picks the café two streets from the usual one, arrives twenty minutes early, watches the approach from a second-story window, and leaves through a back alley the asset did not know about is a good scene, even if mechanically nothing has been different.

The Emotional Register. What the scene is about for the people in it. Every scene in Division 5 has an emotional register beneath the tactical and procedural ones: the fear, the boredom, the old affection, the guilt, the professional respect. The GM’s attention to this register is what separates a Division 5 session from a generic thriller. The KGB colonel is not only a mechanical opponent. He is also a man having tea with someone he has been thinking about for eleven years.

A scene firing on all three is the kind worth retelling later; running on just one of them, tactical alone, procedural alone, emotional alone, produces something merely competent, dry, or soap-operatic, respectively. The craft is holding all three.

The Tonal Slider: Le Carré ↔ Fleming

Division 5 supports a tonal range rather than a fixed tone. The range runs between two poles, and the table, not the GM alone, decides where on the slider the campaign sits. The two poles:

The Le Carré register. Procedural and morally compromised, where the operative’s inner life outweighs their skill roll and every operation leaves a mark. The register of Tinker Tailor Soldier Spy, The Spy Who Came in from the Cold, Slow Horses, The Americans.

The Fleming register. Stylish and larger-than-life, where the operative is competent to the point of being mythic and the villain always has a mountain fortress and a nuclear-era MacGuffin. The register of Casino Royale (either version), Goldeneye, Mission: Impossible - Fallout, Atomic Blonde.

Most Division 5 campaigns sit somewhere between these poles, and most good campaigns move along the slider across sessions: a Fleming-register opening operation that sets the campaign hook, then a Le Carré-register mid-season that delivers the emotional weight, then a Fleming-register capstone that pays off the stylistic setup. The slider is a feature, not a problem.

Setting the Slider at Session Zero

At session zero, the table has a short explicit conversation about the tonal register. Two questions are enough:

1. On a scale of one (Le Carré) to ten (Fleming), where do we want the default register?
2. How far do we want to vary from that default over the campaign?

A table that answers “three” and “not much” is signing up for a slow-burn, morally-weighty, low-action campaign. A table that answers “seven” and “widely” is signing up for a Tinker Tailor episode interspersed with Mission: Impossible capstones. Both are valid Division 5 campaigns. They are also very different campaigns, and getting the answer in writing at session zero prevents six months of cross-purposes.

The Slider in Practice: What Changes

The tonal slider moves specific things. Once you know where the slider is, you know what to dial:

Gear. Fleming register: the Q-branch cameo. Operatives get bespoke gadgets, signature weapons, named cars. Le Carré register: gear is institutional surplus. The operative’s pistol is the one the station had. The car is a Volvo.

Violence. Fleming register: violence is choreographed set pieces. A knife fight in a Budapest bathhouse. A shootout in a casino. The operative wins. Le Carré register: violence is rare, ugly, and has aftermath. The operative kills once, and thinks about it for four sessions.

Locations. Fleming register: exotic is the default. A ski lodge in Gstaad. A yacht off Monaco. A casino in Macau. Le Carré register: locations are working places. A damp flat in Battersea. A Stasi office above a bakery in Leipzig. A Langley conference room with bad coffee.

Antagonists. Fleming register: antagonists are signature, named, stylized, armed with at least one distinctive trait. Le Carré register: antagonists are professional, credentialed, procedural, distinguishable by the quality of their work rather than by a scar.

Outcomes. Fleming register: operations succeed, with style. Le Carré register: operations mostly succeed, imperfectly, with costs the mission briefing did not mention.

NPC Voicing. Fleming register: NPCs have flourish. Their language is slightly more formal, their gestures slightly more theatrical, their secrets slightly more cinematic. Le Carré register: NPCs have texture. Their language is the dialect of their training. Their gestures are small. Their secrets are plausibly boring until they are not.

Moving the Slider Mid-Campaign

A campaign that commits to a single slider position for thirty sessions will feel monotonous. The craft is to vary the position across the arc while keeping the overall tone identifiable.

A typical variance pattern across a twelve-operation season:

The slider averages around 5 across the season while swinging widely within it. A table that watches the register move this way will feel the campaign has shape in a way a monotone campaign will not.

Holding the Tone

Tonal drift is the most common failure mode in Division 5 campaigns. It usually happens slowly: session by session, the register slides toward whichever end of the slider the table reaches for under stress, until nobody can quite remember when the campaign stopped feeling like Division 5. This section is the field manual for noticing the drift and correcting it.

The Five Common Drifts

Drift 1: The Firefight Default. The table reaches for violence whenever the operation gets difficult. Within a few sessions, every operation ends in a shootout. This can be a valid high-action campaign if the table chose it, but if it happened by accident the register has moved from spy thriller toward action movie with spy-themed dressing.

Correction: First ask whether the table likes the new register. If yes, keep it and make the consequences sharper: more Public Heat, harder exfiltration, more asset exposure, and clearer attribution risk. If no, audit the last three operations. Where were the points at which violence was the first-order answer? Redesign those points so that violence is an available answer but not the only attractive answer. Reward tradecraft-based solutions with cleaner outcomes than violence-based ones, consistently, until the table has both options in view.

Drift 2: The Fleming Slide. The campaign starts at slider-4 (Le Carré-leaning) and drifts to slider-8 (Fleming-leaning) across a season. The trappings escalate, flashier gear, more exotic locations, higher stakes, while the moral weight that made the early sessions compelling quietly dissipates.

Correction: This is usually a pacing problem wearing a tonal costume. The table is reaching for stylistic escalation because the campaign has stopped generating consequences, and spectacle is filling the gap. Introduce a cost-paid operation: a consequence from three sessions ago comes due, a forgotten asset is exposed and must be protected, a Legend burns at a bad moment. Let the stakes reset the baseline, not a GM decree.

Drift 3: The Procedural Grind. The opposite drift. The campaign becomes only procedural. Every session is three meetings, one surveillance run, and one brush pass. The register is correct but the texture is flat. The table is losing interest without knowing why.

Correction: A procedural grind is almost always missing the personal stakes, not the procedure: the tactical and procedural registers are running beautifully and the emotional one has been left out. Introduce a scene in which the operative’s own history bears on the case: the KGB colonel is the former lover’s new handler; the safe house is the apartment where the operative’s sister died; the asset speaks with the accent of the operative’s mother. That is texture no checklist can supply.

Drift 4: The GM-as-Opposition Slide. The GM starts identifying with the opposition rather than the world. Sessions begin to feel adversarial between GM and players, the opposition is cleverer each session, the counter-operations are tighter, the table is losing trust in the rulings.

Correction: Return to the core stance (see Running Division 5: The Core Stance, above). If you cannot tell which way the world is pushing in a given scene, stop running the opposition tightly for two sessions. Run them sloppily. Let the operation succeed. Let the team win a couple of easy ones. Come back to the opposition once your stance is no longer “I am playing to win.”

Drift 5: The Compromise Spiral. The Compromise Track starts feeling like an attrition grind. The players begin playing defensively, avoiding operations rather than engaging them, because every operation is another Compromise level nobody can afford.

Correction: This is the drift most likely to end a campaign outright, and it has its own full section below: see Compromise as Pressure, Not Punishment for recalibrating the clearance rate.

The Two-Session Rule

When you notice a drift, give yourself two sessions to correct it. Not one. Not five. Two.

One session is too short, a single countermove is read as GM whiplash, and the table does not have time to feel the register change. Five sessions is too long, by the time the correction has landed, the drift has calcified and the campaign has lost sessions to it. Two sessions is the window. Session A introduces the countermove (a cost-paid operation, an emotional scene, a procedural success the players earn cleanly). Session B reinforces it (a second operation in the same register). By session three, the register has reset and the campaign is back on the slider.

The “Who Do You Trust?” Question

The double-agent reveal is the central trope of the spy genre. It is also the trope most likely to blow up at the table if handled badly. A double-agent reveal that lands produces the scene the whole campaign has been building toward. A double-agent reveal that does not land produces player feels ambushed, player quits the campaign, or table-wide reduction in trust that takes months to rebuild. This section is the handling instructions.

The Three Kinds of Double-Agent Reveals

Reveal Kind 1: The NPC Was Always a Double. An NPC the team has worked with for multiple operations, a Handler, an asset, or a friendly officer from another service, is revealed to have been working for the opposition the entire time. This is the classic Le Carré reveal, and it is the one that lands cleanly when handled well.

Reveal Kind 2: The NPC Was Turned Mid-Campaign. An NPC who was genuinely loyal at the start has been turned by the opposition at some point during the campaign, and now their loyalty has shifted. Mechanically similar to Kind 1, but narratively distinct, the operation that ended the NPC’s loyalty becomes a significant event in the campaign’s backstory.

Reveal Kind 3: The Player Character Was Always a Double. One of the player operatives is revealed to have been working for the opposition all along. This is the nuclear option. It can produce the best single session in the entire campaign. It can also destroy a campaign in a single evening. Never attempt this kind of reveal without the player’s explicit consent, secured the same way the tonal slider is (see Setting the Slider at Session Zero, above) and reconfirmed quietly as the campaign runs. The mechanics below assume this has happened.

When to Plant a Reveal

A double-agent reveal requires runway. The NPC (or PC) must have been present in the campaign long enough that their reveal carries emotional weight. Rule of thumb: the reveal requires a minimum of five operations of prior involvement for an NPC, ten operations for a PC. Any less and the reveal is a twist for the sake of twistiness; any more and the reveal risks coming too late to land before the campaign ends.

The runway also requires earned trust. The NPC must have been useful to the team across multiple operations: saving an operative’s life, delivering a critical piece of intelligence, breaking their own rules for the team’s benefit. The trust the team places in the NPC is what makes the reveal cut. Without earned trust, the reveal is merely surprising, not devastating.

When to Fire the Reveal

Three good times to fire a double-agent reveal:

At the end of a mid-campaign arc. The reveal closes the current arc and launches the next. The operation in which the reveal lands ends on a cliffhanger; the next session opens with the team reassessing everything the NPC has ever told them.

At a moment of maximum vulnerability for the team. The operation has gone wrong, the operatives are trapped in a hostile city with no exit, and the NPC who is their only route home reveals their loyalty and offers a deal. This is the scene the entire NPC’s runway has been for.

In the penultimate operation of the campaign. The reveal recontextualizes the whole campaign and sets up the finale. This is the hardest of the three to pull off because it requires the reveal to land and to generate a satisfying finale from the recontextualization. When it works, it is the best ending Division 5 supports. When it does not, the campaign ends on a flat note.

Bad times to fire a reveal: early in the campaign (no runway); without a clear payoff scene (the reveal is an information drop rather than a dramatic beat); in a session where the table is already struggling (the reveal adds load rather than clarifying).

How to Fire the Reveal: The Scene Craft

A good reveal scene has four beats:

Beat 1: The Tell. A small detail lands in the scene that is wrong. A gesture the NPC has never made. A phrase they use that belongs to the other side’s tradecraft dialect. A piece of information they should not have. The players may or may not catch it. The GM has placed it deliberately.

Beat 2: The Confirmation. An action the NPC takes that can only be explained by the reveal. They withhold help at a critical moment. They deliver the team to a location that turns out to be an ambush. They ask a question the answer to which only matters to the opposition.

Beat 3: The Reveal. The NPC, cornered or triumphant, confirms the truth. This beat should be short, two or three lines of dialogue. The reveal is an image, not a speech. “It was always you.” “I know.”

Beat 4: The Aftermath. The scene does not end with the reveal. It ends with a decision the team must make in the moment. Do they arrest the NPC? Kill them? Let them escape? Turn them? The decision is the scene’s point. The reveal is the setup.

Protecting the Players from Feeling Ambushed

The way a double-agent reveal most often goes wrong is that the players feel the GM lied to them, not that the fiction betrayed them. These feel the same in the moment and very different in retrospect: players who feel betrayed by the fiction lean into the payoff, and players who feel betrayed by the GM leave.

Three practices to prevent the ambush feeling:

Practice 1: Seed the tells honestly. The reveal must be retrospectively legible. When the players go back through their session notes afterward, the tells should be waiting for them: the small inconsistencies, the odd phrasings, the information the NPC should not have had. Finding them is what turns a fooled-by-the-GM read into a fooled-by-a-skilled-opponent one.

Practice 2: Raise it at session zero. Add one question to that conversation: “This is a genre with double-agents in it. Is the table comfortable with double-agent reveals as a genre convention?” A yes is your signoff to run them; a no means telling the stories a different way.

Practice 3: Run the reveal at the NPC’s level, not the player’s level. The NPC is revealing themselves to the operatives, not to the players. The emotional register of the scene is the operatives’ reaction, not the players’. Keep the scene focused on the fiction. The players’ emotional response to the scene happens on its own; you do not need to double-underline it.

When the Reveal Fails

Sometimes the reveal does not land. The beat falls flat, the players do not react, the table moves on without the weight the scene was designed to carry. This happens. It is not a campaign-ending problem.

Three options for salvage:

Option A: Accept the fall and move on. The reveal did not land for whatever reason. The campaign continues as though it had, but without the emotional weight you had planned for. Note the event in the campaign’s history; it may produce delayed consequences that land better than the immediate reveal.

Option B: Reinforce the reveal in the next session. Give the reveal a second beat in the following session, an NPC references it, a consequence of the betrayal ripples into the team’s current operation, the revealed double-agent resurfaces with new stakes. The reinforcement often lands when the initial reveal did not.

Option C: Accept that this campaign does not want this kind of reveal. Occasionally a table simply does not enjoy the double-agent trope. If the reveal did not land and the table’s interest is visibly low, adjust. The campaign can continue beautifully without relying on the trope. Not every spy story needs a mole.

Compromise as Pressure, Not Punishment

The Compromise Track is the single most powerful narrative tool in Division 5. It is also the most easily misused. A Compromise Track used well feels like a countdown to the next reinvention, a rhythm that drives the campaign forward. A Compromise Track used poorly feels like an attrition meter, a slow grinding of the characters’ options until the players have no plays left and the campaign stalls.

This section is how to keep the Track on the right side of that line.

The Design Intent

The Compromise Track is designed to produce three specific effects in the campaign:

Effect 1: It rewards tradecraft. Operatives who play well have low Compromise. Operatives who play sloppily have high Compromise. The Track is legible feedback on the quality of the team’s tradecraft, visible to the players and the GM alike.

Effect 2: It generates narrative events. Compromise level changes are scenes: the moment the operative realizes they have been photographed, the moment the case officer’s file gets cross-referenced, the moment the Legend burns.

Effect 3: It drives downtime. High-Compromise operatives need reinvention downtime, new Legends, cooling periods, relocation. This creates the space for the campaign’s non-operation content: the operative’s personal life, the institutional politics of Division 5, the long romance with the asset handler in Vienna, the year-long sabbatical in Marseille that the campaign will reference for fifty sessions afterward.

If all three of these effects are happening, the Track is working. If any one of them is not happening, the Track needs recalibration.

Calibration: How Fast Should Compromise Accumulate?

Division 5 gives baseline Compromise accumulation rates in Chapter Eight, supplemented by Chapter Nine’s threat-environment modifiers. These baselines are designed for a typical campaign pace of roughly one operation per session. If your campaign runs at a different pace, the baselines need adjustment:

Slow-campaign pace (one operation per 2–3 sessions). Reduce Compromise accumulation by half. A slow campaign that applies baseline rates will grind the operatives into retirement before the arc completes.

Fast-campaign pace (two or three operations per session). Increase Compromise accumulation proportionally, but add an equivalent increase to clearance rates. Fast campaigns need to cycle Compromise rapidly, not accumulate it.

Capstone-heavy pace (long multi-session operations). Accumulate Compromise per act rather than per session. A four-session capstone operation should produce roughly the same Compromise accumulation as a four-session string of single-session operations. Do not let a long operation secretly cost the operatives four times the normal rate.

The Clearance Question

Most Compromise-Track problems are actually clearance problems. The accumulation is working; the clearance is not, and the Track is steadily climbing because the operatives never get enough downtime between operations to relocate, build new Legends, and cool off.

The clearance rule of thumb: every three operations should include one that allows meaningful clearance. This can be a deliberate downtime operation (a recuperation arc, a Legend-building operation, a relocation to a new station), an operation in which the team’s objectives require cooling off (a deep-cover insertion that demands months of pre-operational life-building), or a mission-structural downtime moment (the Bureau pulls the team for two months of internal review between arcs).

If your campaign has gone six operations without meaningful clearance, the Track is punishing rather than pressuring. The correction is a deliberate downtime operation, explicitly constructed to clear Compromise across the team. Do this before the Track forces it: a forced downtime (because everyone is too hot to operate) is narratively weaker than a chosen downtime.

The Compromise Scene: How Transitions Play

Every Compromise level transition should produce a scene. Not a sidebar, not a dice roll, not a tracker update: a scene in the fiction that makes the transition concrete.

Compromise 0 → 1 (Clean → Noticed). A small scene. The operative realizes someone looked at them twice at the train station. A colleague at the Bureau mentions they have heard the operative’s cover name in an unrelated context. The scene ends without resolution; the operative logs it and moves on.

Compromise 1 → 2 (Noticed → Surveilled). A medium scene. The operative spots the same face twice in the same day. An asset reports they have seen someone watching the operative’s usual approach. The scene ends with the operative confirming, to their own satisfaction, that they are under active observation.

Compromise 2 → 3 (Surveilled → Flagged). A significant scene. The operative’s Handler calls them in for a quiet conversation. Their name has appeared on a watch list. The Bureau is considering whether to pull them. The scene ends with a decision: continue operating at elevated risk, or stand down.

Compromise 3 → 4 (Flagged → Hunted). A major scene. The opposition has committed resources. A specific officer, named, described, has been assigned to pursue the operative. The scene ends with the operative knowing who is coming.

Compromise 4 → 5 (Hunted → Burned). The scene. The operative is pulled from the field: captured, killed, disavowed, or permanently reassigned. This is the end of the operative as a playable character at their current Legend stack. The scene is as dramatic as the campaign can make it. Some operatives survive this as Ghosts (Chapter Three) who return later. Most do not.

A GM who skips these scenes and just updates the tracker has turned the Track into bookkeeping; a GM who runs them has turned it into the spine of the campaign.

Compromise and Player Agency

A common GM failure: the GM determines when Compromise accrues without consulting the players. The operative did X; the opposition knows; the Track advances. This is mechanically correct and narratively flat.

The stronger approach: the GM announces the risk before the roll, and the player chooses whether to take the risk. “If you make this check and fail, that Compromise level advances. If you make it and succeed, it stays.” The decision to roll is now the player’s, not the GM’s. The Track advances because the player chose a course of action whose risks were visible, not because the GM wrote a consequence into the scene after the fact.

Visible risk, player choice: that is what keeps the Track feeling earned rather than imposed. Players accept consequences they chose to risk, and resent the ones that show up after the fact.

The Campaign’s Compromise Arc

A well-paced campaign has a Compromise arc at the campaign level, not just the operation level. Early operations accumulate slowly and clear easily (Compromise hovers around 0–1). Mid-campaign operations accumulate steadily and clear with effort (Compromise hovers around 1–3). Late-campaign operations accumulate rapidly and clearance becomes a mission objective in its own right (Compromise hovers around 3–4, with rare 4–5 crises).

The final capstone operation should happen at the upper edge of Compromise: the operatives going in hot, with their last clean Legend, with the knowledge that the operation will use everything they have left. This is the mode Division 5 is built for, and an early-campaign approach that exhausts the operatives’ Compromise tolerance by session ten will have nothing left for session twenty-four.

Pacing

A Division 5 session has two pacing levels, and both need attention. Operation-level pacing is the shape of the individual session, Tasking to Exfiltration, how long each act runs, how the scenes stack. Campaign-level pacing is the shape of the arc, how operations combine into arcs, how arcs combine into seasons, how seasons combine into a long campaign.

Operation-Level Pacing

The four-act Operation (Chapter Ten) gives the structural template. This section is about running that template, not designing it.

The Tasking budget is fifteen to twenty minutes. Longer Taskings drain energy before the operation has begun. If you find yourself running a forty-minute Tasking, the problem is usually too much exposition: information the players need is being delivered in monologue rather than through the Intel Phase. Compress. Move exposition out of Tasking.

The Intel Phase budget is one-third of the session. This is the subsystem that differentiates Division 5 from other action games, and it deserves real table time. If the Intel Phase is running under forty-five minutes, you are under-running the core system. If it is running over ninety, the Phase has probably slipped into planning-without-action, which is a different kind of drift: let the players make some plan and proceed to Execution.

The Execution budget is the largest block. Execution should feel like the culmination of the Intel Phase. The Tradecraft Pool dice are spent here. The Legends are deployed here. The operation’s actual stakes land here. An Execution that runs shorter than the Intel Phase is probably a sign that the operation was solved in the Intel Phase (good, sometimes) or over-planned and now the Execution feels anticlimactic (bad, if repeated).

The Exfiltration budget is fifteen to thirty minutes. Exfiltration is decompression: the operation is effectively over, and the scene is the wind-down, the consequence-collection, and the seed for the next operation. Do not shortchange it; Exfiltration is where the weight of the operation accumulates into campaign capital.

The Scene Clock

Within an Operation, pace scenes rather than acts. A scene that has accomplished its purpose should end. A scene that has lost its purpose should be interrupted.

Ending a scene early. The beat has landed, the information has been delivered, the decision has been made. Close the scene. Cut to the next. “The conversation ends there. You leave the café. It is now three in the afternoon. Where do you go?“ The cut is a gift to the table. A GM who runs every scene to its natural conclusion is running twenty-percent more session than the material requires.

Interrupting a scene. The scene is losing energy: the players are spinning, the conversation has gone circular, the tactical question has no resolution. Interrupt. Drop a complication, introduce a new NPC, cut to a different location where something is happening. The scene was not producing the beat it was designed to produce; move on.

Rescuing a scene. Sometimes a scene is almost landing, and interrupting it would lose the beat that was about to happen. Rescue: introduce a new stake that re-focuses the scene. “While you are talking, someone walks into the café. You recognize them.“ The new stake rebuilds the scene’s pressure. Use sparingly; repeated rescues turn every scene into a rising-action treadmill.

Campaign-Level Pacing

The campaign has a shape. The shape is composed of arcs, each of three to five operations. Arcs are composed of operations. Operations are composed of sessions.

A default shape for a single-season Division 5 campaign:

Pacing Across Arcs

Each arc has its own internal pacing. The most common shape:

Arc Opening. A setup operation that establishes the arc’s stakes, introduces the arc’s antagonist or asset, and delivers a hook the team cannot refuse. Run at moderate register, not the hottest operation, not the coldest, but a clear demonstration that the campaign has moved into new territory.

Arc Middle. One or two operations that develop the arc’s theme. This is where the emotional register does most of its work. Complications accumulate. Costs get paid. The team’s initial read on the arc’s stakes is complicated by information the first operation did not reveal.

Arc Turn. A pivot operation that reframes the arc. What the team thought was happening is not, quite, what is actually happening. The turn does not have to be a reveal; it can be a re-prioritization, a stake change, a death, a betrayal. The arc after the turn feels different from the arc before it.

Arc Close. A resolution operation that pays off the arc’s stakes. This is not the campaign’s capstone (that comes at season-close); it is the end of this particular story. The team either succeeds, fails, or pays a partial price. The closer’s register should match the overall register the arc has been running at: a Le Carré arc gets a Le Carré close, a Fleming arc gets a Fleming close.

Pacing Across Seasons

A campaign that runs more than one season needs season breaks. A season break is an in-fiction period of downtime between seasons, weeks to years, narratively, during which the operatives:

• Clear most of their accumulated Compromise through reinvention and relocation.
• Build new Legends, new assets, new networks.
• Pay or recover from the personal costs the previous season imposed.
• Encounter the seeds of the next season’s arc in small, peripheral ways.

A well-handled season break produces a reset without erasing the campaign’s history. The operatives are still the same people, with the same relationships and the same scars, but their operational baseline is fresh enough to carry another season. Skipping season breaks is the most reliable way to collapse a long campaign: the operatives arrive too compromised, too tired, too invested in the previous season’s stakes to meaningfully start the new one.

The Long Campaign

A Division 5 campaign that runs more than three seasons, fifty-plus operations, crosses into long campaign territory, and the pacing challenges shift. Individual operations matter less; the arc of the career matters more. Operatives become institutionally significant within Division 5. Nemeses defeated in early seasons are referenced in later seasons as historical events. New Nemeses emerge whose existence is shaped by the team’s previous decisions.

Long campaigns benefit from:

Season Theming. Each season has an identified theme: a question the season is answering. Season 1: What kind of operatives are we? Season 2: What happens when the Cold War ends? Season 3: What does Division 5 owe the assets who served it? The theme shapes the season’s arc design.

Campaign-Level Compromise Arcs. By the fourth season, operatives may be institutional Compromise rather than personal Compromise: known to every major service by reputation, incapable of fielding a truly clean Legend anywhere. This is a different game, and the long campaign leans into it. The late-career operative is visible. The game becomes about deciding when to walk away.

Succession Planning. Long campaigns produce operative deaths, retirements, and replacements. A long campaign with an eye on pacing plans these deliberately: a season that ends with two operative retirements and the introduction of two successors, a capstone operation that kills a founding operative, a season-two opener that features a junior operative who will, by season five, be the team’s effective lead.

Edge Cases and Common Rulings

The book does not quite cover every situation. This section collects the rulings that a long-term Division 5 GM accumulates across a campaign: the small interpretations that keep the rules honest without writing additional subsystems.

Mixed-Era Campaigns

A single team across multiple eras. Sometimes a table runs a campaign whose operatives exist across multiple eras: through time travel (rare, and not recommended), through cryogenic storage (campaign-breaking, not recommended), or through long lives (very rare but plausible: an operative recruited in 1943 who is still active in 1989 is unusual but not impossible). For the long-life approach, treat the era transitions as major campaign events with full reinvention implications. The operative who survives three decades in the field has paid for the privilege.

Multiple teams across multiple eras. The more common approach. The campaign features two or three separate teams, one per era, with the players rotating between them or playing multiple operatives simultaneously. Each team has its own Compromise tracks, its own Legends, its own arc. The connection between the teams is institutional: they are all Division 5, the Nemesis of Era I has a successor in Era II, the asset recruited in Era II is still alive in Era III and remembers. This is the approach the book is designed for.

Era transition scenes. A campaign that moves from one era to the next benefits from a transition session, not a full operation, but a connective session that carries the campaign’s through-line from the old era to the new. An archive scene in which a modern analyst reviews the old team’s files. A retirement scene. A funeral. A handoff.

Ghost Characters

The Ghost disposition (Chapter Three) covers operatives who have officially “died” and been reactivated. Over a long campaign, multiple operatives may pass through Ghost status: an operative who is killed in a late-arc operation, then brought back in the next season as a Ghost with a new codename. The mechanics are in Chapter Three. The campaign implications are here:

Ghost reactivation should cost something. An operative who dies and comes back without cost cheapens the Compromise-Burned mechanic. The standard cost is that the reactivated operative loses all prior Legends, starts with a fresh Compromise Track at 2 (they are already flagged in the opposition’s files as presumed dead, an anomaly that raises flags), and cannot return to their pre-death relationships without triggering investigation.

A reactivated Ghost is a narrative event, not a reset. The campaign remembers the operative’s death. NPCs comment on it. The operative’s previous handler may not recognize them. The operative’s former asset network is dispersed or dead. The world has moved on, and the Ghost is catching up.

Legend Cooling

When a Legend that has been cold for years is cold enough to be uncompromised again. Use the Chapter Eight cooling table to clear mechanical Burn Track hits. The guidance here is for older narrative suspicion, residual history, and opposition-file memory after the track has already cooled. A Legend that has not been used for ten times the duration it was last used can usually be treated as narratively cold, provided no active pursuit remains in the opposition’s files. A Legend used for two weeks and then shelved for twenty weeks is narratively cold. A Legend used for two years and then shelved for twenty years is narratively cold. A Legend used for six months and shelved for three years may have cleared hits by the Chapter Eight table, but its long pattern may still be remembered by an attentive hostile service.

When a Legend that was burned can be rebuilt. A fully burned Legend cannot be rebuilt; that is the point of the Burn Track. However, a variant Legend, a different identity built around similar cover details, can be built with a significant downtime investment (months to years). The variant is mechanically a new Legend with a new Burn Track, but it carries one hit pre-applied from the residual association with the burned predecessor.

Mid-Operation Recruitment

The team wants to recruit a Minion. Standard approach: the Minion is played through the Minion rules until recruited, at which point they transition to a compact Asset profile instead of gaining individual combat stats: name, Capability 1 (or 2 if the fiction says they are genuinely competent), Access 1, Trust 1, Exposure 1, one Role, one Pressure, one specific useful access point, and a Handler if a PC owns the relationship. If later used on camera as a combatant, use an appropriate Minion or Elite statblock for that scene, but their campaign-facing profile remains the Asset profile. Recruited Minions make excellent recurring NPCs; they are loyal within reason, occasionally brilliant, and frequently the cost of an operation’s failure if things go wrong.

The team wants to recruit an Elite. Much harder. An Elite recruitment is a multi-operation project: essentially, the Elite becomes a recruitment target in their own right, with full tradecraft requirements. The payoff is significant: a recruited Elite shifts the balance of the campaign. Der Falke, recruited, is the most valuable asset Division 5 has ever operated. The likelihood of this outcome is low, but the possibility gives the campaign a horizon.

Burned Operatives Returning

An operative who was burned and pulled from the field comes back. The standard approach: the operative is not playable as the same character. If the player wants to continue playing them, the character returns as a Ghost (see above) with the accompanying cost and transformation. If the player wants to start fresh, a new character is generated; the old operative becomes an NPC, possibly a handler, possibly a villain, possibly a dead letter in a safe house.

Compromise Clearing Edge Cases

A Nemesis is defeated. Does the operative’s Compromise against that Nemesis clear? Yes: one level. The operative is still known to other services, but the specific institutional pursuit led by the defeated Nemesis is closed. This is meaningful in a multi-Nemesis campaign and allows the Compromise Track to function as a relationship tracker rather than just an accumulator.

A service is disbanded. The Stasi dissolves in 1990. The KGB reorganizes in 1991. For campaigns that span these events, what happens to Compromise levels tied to the defunct services? The ruling: Compromise tied to institutional pursuit clears one level. Compromise tied to successor services (the BfV inheriting Stasi files, the SVR inheriting KGB archives) persists. The files outlive the institutions.

An operative relocates. A full relocation to a new city, new country, new cover, distance and time from the operative’s previous theater, clears one Compromise level per year of real-world in-fiction time spent at the new location. This is slow, but it is the primary clearance method for deeply burned operatives.

Rules Conflicts

A talent modifies a subsystem rule. Chapter Eight gives the default procedure. A talent that clearly changes that procedure applies in its own narrow scope. Clean Hands can prevent one Compromise mark when its conditions are met; it does not rewrite the Compromise system. Ghost Socialite changes how a Face interacts with social Compromise and Legend burns; it does not change the rule for other archetypes.

A piece of equipment modifies a subsystem rule. Equipment qualities apply in their own scope. A Deniable-tagged item reduces or obscures its own attribution trail; it does not erase witnesses, bodies, digital records, or another item on the scene. If an equipment entry and a broad subsystem rule appear to differ, apply the equipment rule to that item and the subsystem rule to everything else.

A Universe-specific skill conflicts with a Core Rules skill. The Universe-specific skill takes precedence in its area. Tradecraft is the skill for surveillance detection runs, brush passes, dead drops. Perception is not. Where a skill’s area is genuinely ambiguous, the GM picks the one that produces the better scene.

Player Questions That Have No Rules Answer

“Can I smuggle a pistol through this airport?” Depends on the airport, the era, the operative’s cover, and how much of the Intel Phase went into preparation. Treat as a Tradecraft or Forgery check against a difficulty set by the environment. Failure means the attempt is caught at security and the operative adds Compromise.

“Can I impersonate a specific named person?” Requires Deception or Forgery depending on the medium (face-to-face or paper-based). A brief impersonation, a phone call, a signed document, is plausible. Sustained impersonation, being the person for a week, requires deep preparation and a Deep Legend.

“Can I hack into [system X]?” Nightscreen era only. Cyber check at difficulty set by the system. Success gets data; failure compromises the operative at the system-owner’s level.

“Can I seduce [person Y]?” Charm or Deception check. Success creates an opportunity (not a fait accompli). The actual seduction, if both players are comfortable, is fade-to-black; the consequences are played out in the following scene.

A Final Word

The craft of running Division 5 is the craft of running a world in which professional people do difficult work at serious personal cost. The dice handle the mechanics. The rules handle the structure. The chapters handle the content. This chapter handles the register: the specific tonal discipline that turns the mechanics, structure, and content into the kind of session the table will remember ten years later.

Three things are worth holding onto.

First, trust your players. Division 5 works when the players care about their operatives, their assets, their covers, and the operations. The GM’s job is to make caring possible: by presenting a world that is coherent, by running opposition that is credible, by pacing sessions that produce emotional weight. The players will do the actual caring. Give them the room.

Second, trust the rules. The five core subsystems in Chapter Eight are enough to carry the campaign: Cover & Legends, the Tradecraft Pool, Assets & Networks, Compromise & Attribution, and Era Framing. The later mini-procedures are there to make recurring spy scenes easy to run, not to invite more bookkeeping. Resist the urge to homebrew more until the table has seen the existing procedures work.

Third, trust your instincts. The chapter above is a collection of practices, not commandments. If a practice conflicts with what your table needs on a given evening, the practice is wrong and your instincts are right. Every Division 5 campaign is its own thing. No two run the same. The craft is knowing when to apply the practices and when to set them aside.

The war you never heard of has been running for eighty years. It will be here when you sit down at the table. It will be here when you close the book at the end of the night. Your job is to make the time in between feel like it mattered.

Welcome to the work.