Combat in Threshold — Structure and Flow
Before getting into the details, a quick note on scope.
The previous post focused on what combat is meant to be in Threshold—its role in the game, how it fits into the broader experience, and what kinds of outcomes it should produce.
This post is different.
This is about how combat actually runs at the table.
How Different Games Structure Combat
Once you move from intent to implementation, you start to see how differently games approach the same problem.
Some systems lean into structured tactics. Combat is broken into discrete turns, positioning is precise, and the rules define exactly what a character can do at any given moment. Dungeons & Dragons and Pathfinder are the most familiar examples of this approach. It provides clarity and depth, but it can slow the game as complexity increases.
Others treat combat as another form of conflict, using the same core mechanics that handle everything else. In games like Blades in the Dark and FATE Core, a fight may resolve in a small number of rolls, with the focus placed on consequences rather than step-by-step execution. This keeps pacing tight, but trades away some tactical detail.
There are also systems that experiment with alternative turn structures. The Cosmere RPG, for example, uses a fast turn / slow turn model where players choose between acting early with fewer options or later with greater flexibility. It creates meaningful tradeoffs and keeps momentum high.
I strongly considered that approach.
In the end, I didn't use it—not because it doesn't work, but because it solves a different problem. Fast/slow optimizes for individual heroics—the skilled warrior who reads the moment and acts at exactly the right time. That's a protagonist-centered fantasy. One person making the perfect call.
Threshold prioritizes collective competence instead—a group of people who aren't individually superhuman using communication and planning to punch above their weight. The initiative system needed to reflect that thesis about what heroism looks like.
Initiative — Coordination Over Timing
Combat begins with an initiative roll.
Each character rolls Clarity at the start of the encounter. I chose Clarity over Agility deliberately—awareness of a situation is what lets you prepare mentally and physically, which felt more coherent than pure physical reflex speed as the determinant of who acts first.
From there, the system branches based on the situation:
If the party is prepared, players may freely rearrange their initiative order before the first round begins. This allows tactical coordination—the rogue can go first to create an opening, the defender can position before enemies close, the caster can delay until the right moment.
If the party is surprised or ambushed, initiative remains in rolled order. Characters react in the moment rather than coordinating beforehand.
Enemies always act in their rolled order.
After the first round, the order becomes fixed.
There is one exception: Seize Initiative. A character may spend 2 Strain to act immediately, interrupting the current order. This can be used once per encounter.
I evaluated several alternatives before landing here. Static initiative (roll once, fixed order) was too passive—it removes coordination as a meaningful choice entirely. Contested initiative each round (like the Cosmere model) created too much overhead. A popcorn/side initiative system where players choose to pass to an ally or enemy after each turn maximized coordination but required a decision every single turn, compounding drag across the encounter. It also assumed clean two-sided conflicts, which doesn't hold for the morally complex, multi-faction encounters Threshold's setting implies.
What I kept from those alternatives was the tactical coordination of popcorn—players can arrange their order to set up combos—without the per-turn overhead. You make that decision once, before round one, and then execute. Seize Initiative preserves the emergency override that popcorn builds in structurally, but gates it behind a resource cost and a once-per-encounter limit so it doesn't become routine.
The surprise condition matters because coordination is the mechanical expression of Threshold's core fantasy. If coordination is always available regardless of circumstances, it loses meaning. Being caught unprepared doesn't just mean you act in a bad order—it means you lose the thing that makes your group effective. That's a more meaningful consequence than "you rolled low initiative."
Fast/slow systems have no equivalent vulnerability. Your timing trade-off is always available because it's an individual decision that doesn't depend on your relationship to your allies.
Turn Structure — Three Categories That Do Different Jobs
Each turn consists of three parts:
1 Action 1 Maneuver
1 Reaction
Actions represent meaningful commitments—attacking, casting a spell, or taking a decisive step.
Maneuvers cover movement and smaller adjustments—repositioning, taking cover, or interacting with the environment.
Reactions allow characters to respond outside their turn.
I tested several alternatives before settling on this structure.
Pathfinder 2e's three-action pool offers maximum flexibility but creates decision paralysis and makes ability balancing harder—if everything costs actions, you have to carefully price every technique across the entire system. For a game being built through playtest refinement, that's a fragile foundation.
D&D 5e's Action + Bonus Action + Move + Reaction has too many categories, and the Bonus Action creates feat tax problems where certain builds depend on having something to do with it.
A single "one meaningful thing per turn" model collapses the distinction between acting and existing in space. A character who crosses a room and then attacks isn't doing one thing—they're doing two things that flow together. A single-action model either makes movement free (removing positioning as a meaningful decision) or makes it cost your only action (creating fundamentally different experiences for mobile versus stationary fighters with no mechanical acknowledgment).
The three-category split maps directly onto three distinct questions a player answers each turn:
What do I do? (Action — the meaningful, stakes-carrying activity)
How do I position? (Maneuver — tactical adjustment that doesn't carry stakes on its own)
How do I respond? (Reaction — off-turn agency)
That third category is load-bearing in a way specific to Threshold's design. Reactions provide a structural slot for off-turn agency—interposing yourself to protect an ally, triggering a readied action, responding to something unexpected in the moment. They create opportunities for dramatic choices that fall outside the normal turn sequence.
Separating Maneuvers from Actions lets positioning be a real decision without making it costly. Under minimal resource pressure, players shouldn't feel like moving costs them their turn. Movement is always available, but going beyond standard positioning costs Strain. That's minimal-resource-pressure design with tactical depth preserved.
In practice, these categories are easy to use but harder to define precisely for edge cases. The system is currently driven by examples rather than strict classification rules, with the underlying structure still being refined through playtesting.
Positioning Without Precision Counting
Movement is handled through a zone-based system rather than exact measurement.
Horizontally, positions are divided into:
Engaged
Close
Near
Far
Extreme
Vertically:
Ground
Elevated
Airborne
Beyond
A standard Maneuver allows a character to move one zone.
This approach is grounded in how distance is actually experienced during action. Through training and repetition, you develop an intuitive sense of what is within reach and what isn't. Whether that's striking distance, effective range with a weapon, or how far you can move before someone can respond, those judgments come from familiarity rather than measurement.
My military training and time in combat arts taught me this: when you're moving and fighting, distance is understood intuitively rather than calculated.
That's the perspective I wanted Threshold to capture.
Zones preserve the importance of positioning without requiring constant calculation. Players can make decisions based on intent and relative positioning rather than exact distances.
There are still open questions around zone density and edge cases, but the core system is already functioning as intended.
Strain — Pushing Beyond Limits
Strain serves two roles simultaneously, and the tension between them is intentional.
Within an encounter, Strain is a tactical resource—a budget for pushing past your baseline capabilities. Every spend is a decision: burn Strain now to move an extra zone or seize the initiative, or hold it in reserve because the fight isn't over. That spend-now-versus-save-for-later pressure creates meaningful choices without adding mechanical complexity to every single action.
Across encounters, Strain becomes an attrition resource. It doesn't fully reset between fights, which means the cost of the last encounter is visible in the next one. A party that burned through Strain taking down a patrol arrives at the fortress gate already depleted. That's a design choice rooted in the world of Threshold—this is a setting defined by scarcity and slow collapse, and the resource system should reflect that.
Without Strain, combat becomes a series of isolated tactical puzzles with no connective tissue between them. Characters would be functionally fresh every fight, which contradicts everything Threshold is trying to say about the weight of living in a dying world.
What Strain Does
At this stage of development, Strain's primary combat uses are:
Take additional Maneuvers (2 Strain each) — Extra movement or tactical adjustments beyond your base allotment.
Seize Initiative (2 Strain, once per encounter) — Act immediately, interrupting the current turn order.
The full scope of Strain is still being refined through playtesting. The core principle is clear—it creates meaningful decisions about when to push beyond your baseline—but the exact range of applications is still in active development. I'm also evaluating how Strain and techniques should interact: whether techniques have their own costs, whether they consume Strain, or whether they operate independently. That relationship will shape how advancement feels in practice.
Pool Size and Scaling
Strain pools run 10 to 14 for most characters, derived from 8 + Essence + Force—the two Quality scores underlying physical endurance and mental discipline.
The pool is intentionally flat. Base stats are capped at 1-3 for most characters, with investment pushing toward 4-6 only through deliberate building choices. The difference between an uninvested character (pool of 10) and a heavily invested one (pool of 14) is real but not dramatic. A maximum-investment character hits 20, but that requires significant resource commitment across both relevant stats.
This flatness is a feature. Strain represents your innate threshold for pushing past what your body and mind want to do—the kind of grit that's more about who you are than how hard you've trained. Working out every day doesn't fundamentally change how far you can push yourself past your limits; it changes what your baseline limits are.
Advancement in Threshold gives characters better tools for managing Strain—techniques that cost less, abilities that trigger at low Strain, more efficient recovery—not a dramatically larger pool. That's horizontal progression in practice.
Recovery
Strain recovers through Threshold's four-rest progression, and the amounts are deliberately modest:
1st Rest (10 seconds, combat action): Half Vitality, round down, minimum 1. A battlefield breather—enough to matter, not enough to reset.
2nd Rest (10 minutes): Vitality × 2 if Strain is your primary recovery choice; Half Vitality minimum 1 if you prioritized Stress instead. The choice between resources is itself a meaningful decision.
3rd Rest (1 hour): Full Strain recovery if chosen as primary; Half Vitality minimum 1 if secondary.
4th Rest (8-10 hours): Full recovery of both Strain and Stress, progression resets.
The key principle here is that the 2nd and 3rd rests force a choice between recovering Strain or Stress—you can't fully address both until a full night's sleep. A character depleted in both resources has to make a real decision about which gap matters more before the next fight. That's the grounded protagonist experience in mechanical form: you don't fully reset between fights, and managing that depletion is part of what skilled play looks like in Threshold.
Why Strain Supports the Design
Grounded protagonists: Strain doesn't fully reset between fights. The cost of every desperate sprint and every seized initiative accumulates and carries forward. Characters are capable people operating under real limits, not action heroes who shake off exertion between scenes. The four-rest progression ensures that recovery is earned through fictional action—taking shelter, catching your breath, tending to each other—rather than assumed.
Horizontal progression: The pool stays flat. Veterans don't have dramatically more Strain than fresh characters—they have better techniques, more efficient costs, and harder-won experience with when to spend and when to hold. That's the distinction Threshold draws between experience and power: you get more capable, not more superhuman.
Fast resolution, lasting consequences: Strain keeps the action economy from becoming a spam problem—extra maneuvers cost 2 Strain each, which means movement decisions have real weight without requiring complex opportunity attack rules or movement restrictions. At the same time, the attrition model ensures those costs don't disappear at the end of the encounter. The fight was fast; the consequences linger.
Flow of a Round
A round of combat follows a consistent structure:
Initiative is established. Players act in their determined order. Each character takes a turn using their Action and Maneuver. Reactions occur as triggers arise. The round advances.
The system provides enough structure to support coordination while avoiding unnecessary procedural overhead.
When Combat Ends
Combat does not end because a specific condition is met.
It ends when the situation changes.
One side may retreat, surrender, or disengage. The objective may be achieved. Continuing may no longer be viable.
As established previously, combat is not assumed to be a fight to the death. The structure supports that by allowing outcomes to emerge from the situation rather than enforcing a fixed endpoint.
What I Worked On
This phase focused on translating combat from a design goal into a usable structure.
The emphasis was on clarity, flow, and keeping players engaged without introducing unnecessary complexity.
What Went Wrong
The main challenge was classification.
Dividing actions into Actions, Maneuvers, and Reactions works well in play, but defining those categories cleanly—especially for edge cases—has proven more difficult.
Some systems also revealed dependencies that weren't immediately obvious, particularly around reactions and partial outcomes.
What I Changed
Earlier versions relied more heavily on rigid sequencing and precise positioning.
These were replaced with:
Flexible initiative that rewards coordination
Zone-based movement
Player-driven flow
Each change reduced friction and better aligned the system with its intended pacing.
What I'm Working On Next
The next step is to define how combat resolves at a mechanical level—what players actually do on their turns and how those choices create tactical depth:
Attack and defense interactions
The action economy in practice
How decisions during combat create meaningful consequences
Several parts of the system are functional but still being refined:
The exact definition of partial success in combat
Formal rules for engagement and reactions
Full mechanical definitions for conditions
The complete scope and scaling of Strain
These are active areas of development and will be clarified through playtesting.
This post establishes how combat runs. The next will establish what those actions produce.
Questions
Does coordinated initiative feel like meaningful flexibility, or unnecessary complexity?
How much structure do you prefer in turn order before it begins to slow the game down?
Do zone-based systems feel intuitive in play for you, or do they require additional scaffolding to remain consistent?
