No shots fired.
Sixteen weeks to win a mind.

• A less one-track Standard opponent. Red no longer hammers one area all campaign. When three weeks on an area have gained it nothing, it moves to another, and the longer it stays on one, the more it looks elsewhere.
• It digs in when it falls behind. Once clearly behind on Decision Health it plays more defence, so taking the initiative no longer settles the game on the spot. Veteran and Elite, which already move their aim between crises, are unchanged.
• Tabs in their side's colours. INFLUENCE (SHAPE in Military words) is framed in red and PROTECT (DEFEND) in blue, like the attack and defence cards.

• The manuals' description of Standard is updated.
• Two-player codes work between 0.21 builds.

• Lessons Learned is worth buying. Against a strong Bot Swarm, nodes you lost never came back by themselves, and one Official Statement a turn could not keep up. Now Lessons Learned gives every node of yours a bigger push when you play it, makes them recover faster every week, and lets nodes the other side has taken start to come back by themselves. The opponent's version gets the same boost, still at half strength.
• The five areas are outlined on the map from the first turn, so you can see where each attack will land. Settings › Display › Decision-loop overlay turns the outlines off.
• New players start on Recruit. A difficulty you already picked is kept.
• The in-game Guide matches the PDF. Numbered lists, bullet lists and the screen key on several pages had lost their layout in the game.

• All three manuals, the tutorial and the card text are updated, with every picture retaken.
• Two-player codes from 0.20 builds do not work in 0.21.

• A new hardest level: Elite. Red plays like a Veteran under four harder rules, set out on the start screen when you pick it.
  • It learns your playbook. Play the same attack card three weeks running and its people recognise the trick: the card pushes less until you rest it for a week. Their edge on you warns you a week ahead.
  • Hidden defences. You cannot see its Watch Teams and Prebunks until you attack into one; then the whole group appears.
  • Dug in. A node it wins back from you is harder to push until you hold it again. A bunker marks it.
  • Blowback. Each of your attacks on it that fails gives it a point.

• A smarter opponent. On Standard, Red now guards the nodes you frame, goes after the nodes that pay you and saves its extra moves for crisis weeks; campaigns are as evenly matched as before. On Veteran and Elite it also thinks about your likely answer and keeps its real aim hidden until late.
• An easier week in review. The same move twice on one node reads as one line (×2), each reason is given once, and every attack says what helped it and what held it back.
• Your result card opens flush under Their edge on you and moves up when you fold that panel; Red's move cards stand beside the right side panel. Neither covers the panel you are reading.

• The Field Manual, Practitioner's Manual and Quick Reference cover the four difficulties and the dug-in symbol. Recruit is unchanged.
• Two-player codes from 0.19 builds do not work in 0.20.

• Three manuals, rewritten for playing on a PC.
  • The Field Manual teaches the game screen by screen: the idea, every part of the screen numbered, a turn in four steps, a whole turn played out with pictures and Red's answer, the cards and ten tips, in plain game words. It is also the in-game Guide.
  • The Practitioner's Manual keeps the research and the maths, and now holds every rule with its exact numbers. Its campaign pages show the new crisis order.
  • A two-sided Quick Reference replaces the player aid: mouse and keyboard, a turn in five steps, the cards, symbols and scoring.

• A clearer area of effect. Watch Team, Prebunk and the move preview ring every covered node in sand and draw a sand line along each link between them, pulsing gently. No node is left out.
• The replay shows the real crisis score next to each result, and its two maps name the entry points and Top Leadership with each one's lean.
• The side panel fits in Military words at every text size, and Red's move cards stop above budget and moves.
• The tutorial panel has the same frame as every other window.

• No rule changes. Two-player codes work between 0.19 builds.

• Tidier nodes. A node's statuses (watched, prebunked, framed) now show once, as the row of icons under it. The extra rings, corner brackets, blinking dot and status words on its label are gone; hover the node for the details.
• A node's name never sits on its own status icons, and area names, such as a crisis test or a lost area, step clear of named nodes.

• The Field Manual's symbol key and pictures are updated.
• No rule changes. Two-player codes work between 0.19 builds.

• Every campaign's crises follow the decision loop. Week 4 tests Notice, week 8 Interpret, week 12 Decide and week 16 Act, which counts double. Trust is no longer tested last. Some crisis stories swapped weeks to fit their area. Balance was checked over 200 computer-vs-computer games per campaign: Blue wins between 42% and 60% on Standard, and Collision at Varrin Reef is much fairer than before.
• A clearer screen, with more of the map in view, without going to the Minimal layout. Decision Health is two glowing bars under the crisis bar, Blue and Olvana mirrored, number in the middle, then the state, then the bar. The side panel is narrower and flush to the right edge, budget and moves sit bottom right, and the log sits bottom left.

• Gauge rings show how firmly each node is held. The ring around a node fills clockwise from the top in the colour of the side it leans to; a full ring is 100. A contested node's empty ring is dashed amber.
• Areas taken replaces LEAD 0/2: a bar from Olvana +2 to you +2 shows the race to an early win, and how many areas each side holds of the other's.
• Under the small map, only what to do: each node under pressure and the card that helps. Lessons Learned shows as three diamonds, with what your level buys.

• The log, bottom left. WHAT HAS HAPPENED opens and closes from its header, which shows the latest week. Inside, moves are grouped by week, each week folds, and ◉ on any move opens its full story. New technologies are rows in the week they arrived. This week's moves always sit just above it.
• Their edge on you folds, and now also says where Olvana's attacks have been landing.

• Card tooltips say what a card does, in plain words. The hotkey numbers are gone from the cards; keys 1 to 8 still work.
• The decision-loop overlay is drawn on top of the map, so its outlines and arrows read clearly.
• More info stays reachable: the node preview in the side panel waits while you move the pointer onto it.
• The speed setting is now called Red turn speed.
• The tutorial and all three manuals are updated, with every picture retaken.
• The crises changed, so two-player codes work only between 0.19 builds.

• A new start screen. Campaigns sit in a column on the left, grouped by opponent, with the opponent's mind live on the map behind them. The chosen campaign's story and its four crises, on a 16-week calendar with the area each one tests, sit at the bottom, with Words, Difficulty and START in one bar. Two players, Campaigns, Guide and Settings are four buttons under the list.

• Every dialog in one frame. A short line says what each one is for, then a big title, clean sections and one sand button. Settings now has five sections: Display, Words, Sound, Briefings and Pace.
• The full story shows the spread-grade formula, worked through with the move's own views, likes, comments and shares, and the score and push for every grade from F to A+. The result card after each move stays.

• New technologies show how disruptive the research scores them: each of the four attributes rated 1 to 3 with what it measures, and the total on a meter with the flag at 9.
• Crisis results, Week in review, Two players, Campaigns and the debrief follow the same layout; the debrief keeps every section, including the replay with both minds.

• The tutorial and all three manuals are updated, with every picture retaken.
• No rule changes. Two-player codes work between 0.18 builds.

• A centred scoreboard. The top bar now reads Blue's points, the next crisis, then the other side's points, in the middle of the bar. The week counter is gone: the timeline lights the current week. In the Minimal layout each side's Decision Health sits beside its points.
• The next crisis stands out. A sand badge names the crisis, marks the four crises, counts down the weeks and names the area it tests. It turns amber the week before and pulses in the week itself.

• Clearer health bars. Name and score above a full-width bar from 0.5 to 1, with the crisis cut-offs labelled.
• Dialogs in the same frame. The start screen, Settings, Week in review, new technologies and crises use the framed panels, headings and buttons of the rest of the screen. In a crisis, Blue's result is edged blue and the other side's red.

• This week's orders sit at the foot of the left side instead of over the timeline.
• The small map is bright again, as it was before 0.18.1.
• On small screens the top bar shortens a few labels to stay on one line, and the music and sound buttons move to Settings only.
• The tutorial, Field Manual and Player Aid are updated, with every picture retaken.
• No rule changes. Two-player codes work between 0.18 builds.

• A new screen layout. Panels now run down both sides of the map. The left side holds the mind you are looking at, Decision Health, budget and moves; the right side holds the other mind and What is happening. The map sits centred between them.

• Default or Minimal. Settings › Screen layout. Minimal gives the map the whole screen and keeps only budget and moves in view; two buttons next to the view tabs open the side panels. Narrow screens always use Minimal.

• Move the map. Drag to pan, scroll to zoom, double-click or ⌖ to recentre.
• Their edge on you. A new panel lists every advantage the other side has right now: technologies that make its moves spread further, Bot Swarm hitting every neighbour at full strength, your lost TRUST, and moves that cost it less than they cost you.
• Moves as a number, with a diamond for each move left. Attack cards are edged in red, defence cards in blue, and both health scores sit in the top bar.
• The tutorial and both manuals are updated, with every picture retaken.
• No rule changes. Two-player codes work between 0.18 builds.

• A new layout of the mind on every map. The five areas no longer sit in the same place every game. Trust is always in the middle; Decide is a lobe somewhere on the rim; Act, Notice and Interpret follow it around the ring. The loop always runs in order, Notice → Interpret → Decide → Act, and each area borders the next.
  • Every map has the same area sizes: Notice, Interpret and Act 6 nodes, Decide 8, Trust 4.
  • Entry points sit at the outer edge of their area; Top Leadership sits deep inside Decide.
  • Balance was checked over 200 computer-vs-computer games per campaign: overall win rates, early wins and Veteran are close to 0.17.4's. The Lake Referendum and The Valmora Vote got easier for Blue; Collision at Varrin Reef and Winter Grid harder.

• A shorter, steadier tutorial. Eleven short steps, one idea each, updated for this build.
  • The panel stays in one corner above the timeline instead of moving around the screen.
  • Steps never move on by themselves. A task ticks when it is done; BACK, NEXT and the step dots go to any step, including ones you clicked past.
  • ✕ CLOSE ends it; Settings ⚙ › TUTORIAL reopens it.

• Effects that stack. When a node has more than one effect, the Watch Team and Prebunk tags stack instead of printing over each other, and the node's label gives each status its own line.
• A decision-loop overlay that matches the map. Its outlines are worked out from each map's own nodes, so every node sits inside its own area and no other.

• Immersive Briefings never start by themselves. The broadcast in the week in review waits for ▶ LISTEN. Before, it started as soon as the week in review opened and could carry on into your next turn.
• Browser only. Cognitive Terrain is supported in the browser for now: play it online, or download the HTML file. There are no installers for this version; earlier installers stay on the Releases page.
• Both manuals are updated, with every picture retaken from this build.
• Maps are laid out differently, so two-player codes work only between 0.18 builds.

• A clearer move preview. Hover a node with a card picked and the card now shows how far the move pushes it as one big number, who the node leans toward now and after the move on a red and blue bar with a gold arrow, a verdict such as "Gets a foothold", arrows for what helps or hurts, and the odds in words: Almost guaranteed, Likely, Risky or Long shot, with the chance to fail. The words describe the game's dice, not real-world odds.
  • The push shown is for the usual audience the move would reach, so a watched node shows how much weaker an attack there usually is. The side panel shows the same figure.
  • Each move's result and full story, with the research formula, are unchanged.

• Show the numbers in Settings adds the arithmetic under the preview.

• Numbers that add up. A node's lean is rounded the same way everywhere, and the push in a move's preview and result is the change you see on the map. A push of 22.5 from Olvana 100 now reads +23 and Olvana 77; before, the map showed Olvana 78.
• The Field Manual's move and push pages show the new preview.
• No rule changes. Two-player codes work between 0.17 builds.

• Both manuals rewritten as numbered rulebooks. Every rule has one number (such as 10.3) and is stated once, so you can find it. Tricky rules get a worked example, and advice is kept to one-line player's notes. The in-game Guide is the new Field Manual.
  • How hard a move lands, what moves its audience and what happens at the end of each week are explained in plain words, with pictures of worked examples and the game's own hover card.
  • Field Manual: sections 1.0 to 15.0 in the order a week is played, then reference tables, a glossary with plain and military terms side by side, the player aid, and sources.
  • Practitioner's Manual: the eight campaigns, sessions, the White Cell, the debrief, the model in numbers, the editor, designer's notes keyed to the Field Manual's rules, and a new section on sources and limits, with a facilitator's checklist and debrief questions for each campaign.
  • The Practitioner's Manual's worked examples come with pictures: one week at the White Cell console, the scoring arithmetic, reading a balance check, and a session timeline.

• New: a two-sided player aid. The sequence of play, your turn, the cards and the areas on one sheet. It is the Field Manual's Appendix C, and its own PDF to print.
• Fixes. Humanoid Robots now says it also gives Plant Stories +1 grade, as it always has. General edits to documentation for clarity.
• No rule changes. Two-player codes work between 0.17 builds.

• Every number checked. All the numbers and worked examples in both manuals, the in-game Guide and the game's own text were checked against the rules.
  • Official Statements restore 22 for Blue while its Trust holds, and 7 once it is lost (the card said 14).
  • A framed node takes 40% more from every push while it is framed. Frame then Plant gives you a foothold on a well-held node, not a capture.
  • A Watch Team cuts an attack's reach to a tenth, usually a grade or more; a lost Trust multiplies the other side's reach by ten.
  • Show of Force pushes 24 on ACT and 20 on DECIDE. The income table, grade bands, crisis bands, the area-loss rule and the extra-move condition now say exactly what the game does.
  • The Practitioner's Manual's math pages add the terms they had left out, and use the same balance band as the campaign editor.
• Fixes in the game. A woken dormant narrative is described as hitting twice as hard, as it does. The debrief counts Donovia's nine capabilities. An opponent player's hover cards show their own odds of failure.
• This page. Each picture now sits beside the note it illustrates.
• No rule changes. Two-player codes work between 0.17 builds.

• Every page checked against the game. Both manuals, the in-game Guide and the install notes were read line by line against the game's code, and anything that had drifted was put right.
  • Defence cards work the same in every area of your own mind. The card tooltip, the Field Manual and the quick reference no longer claim an area bonus the rules never applied.
  • The Field Manual lists all seven technologies the campaigns use, and says that a framed node takes 40% more from every push, that the extra-move button needs at least 4 points, that captured nodes don't drift back on their own, and that Recruit gives the opponent 2 fewer points a turn.
  • The Practitioner's Manual describes the "Who decided better" verdict correctly, notes that Show of Force has no audience roll, and explains the budget cap.
• White Cell notices, timed correctly. Blue sees them before its turn and Red sees them when Blue's code arrives. The manuals and the White Cell's console now say so.
• Small fixes in the game's text. The debrief chart names Donovia in Donovia campaigns (it always said Olvana). The opponent page now says Blue's free press makes statements 60% stronger while Trust holds, not 25%.
• Install notes. The README inside every download has the right file and button names, and new sections on structured sessions and Immersive Briefings.
• Documentation. General edits for clarity.
• No rule changes. Two-player codes work between 0.17 builds.

• The White Cell, for structured sessions. A facilitator can now join a two-player game as a third person who sees both sides and steers the exercise once a week, after Red and before Blue.
  • Only when you choose it: Blue starts a New structured session · with White Cell. Ordinary two-player games are unchanged.
  • Codes go round a circle, Blue → Red → White Cell → Blue. The White Cell joins with Red's first code.
  • Its console has eight tabs: the situation (both sides' health, budgets, hidden facts and what the computer would play next), the board, budgets and moves, crises and new technologies, a breaking story, a ruling, a teaching note, and its own limits.

• Point and click. Almost everything in the console is a button.
  • Click a node on the map and the console opens on it: one click shifts control, sets or clears a status, or plants a dormant narrative.
  • Budgets and extra or fewer moves are one click each, for either side.

• Stories with pictures. Browse for a picture, drop it or paste it. It is shrunk to fit inside the code, so it still travels by email or chat.

• Nothing is secret. Every change is listed with a preview of its effect before it is released, and both players see a White Cell notice: Blue before its turn, Red when Blue's code arrives.
  • A move that breaks the exercise's rules can be struck, with a reason. Every copy rebuilds itself without it.
  • The debrief and its replay show every White Cell change and teaching note, week by week.
• Manuals. The Practitioner's Manual has two new pages on running the White Cell, with suggested injects. The Field Manual says what players see in a structured session.
• New code format: two-player codes work between 0.17 builds.

• Guide matches the game. The quick reference at the back of the in-game Guide and the Field Manual now includes the crisis bonus: when one side's crisis goes ERROR, the other side gets 1 extra point. Week 16 doubles it, like everything else.
• Both manuals checked against the code. Every number in the Practitioner's Manual's math section was checked against the game itself.
  • Donovia's dormant narratives hit twice as hard (not 1.5×), its framing lasts three weeks, and prebunking cuts its attacks to 0.45×.
  • The tie-break, flooding and the Show of Force rules are now spelled out, and the Field Manual says how ties are settled.
• Easier-to-read notes. Longer notes on this page now open with one line and list the details underneath. The animation at the top of the page has its full colour back: it had been saved with too few colours.
• No rule changes. Two-player codes work between 0.16 builds.

• Quick reference card, with the game's icons. The cut-out card at the back of the Field Manual (and the last page of the in-game Guide) now uses the same icons as the game.
  • Every card and area has its icon, and a new map legend covers entry points, contested and leadership nodes, nodes under pressure, the watched, prebunked and framed status icons, coverage outlines, this week's orders and the ON AIR tab.
  • Donovia's Show of Force is listed too.

• Practitioner's Manual, easier to read. The design notes are now colour-coded cards.
  • Each has an icon, a "how I got there" line, and tags that separate what comes from the research from the designer's assumptions.
  • A new two-page section, The math, sets out every formula with worked examples: Decision Health, crises, the spread score, odds of failure, the weekly budget, and how far to trust a balance check.
• No rule changes: two-player codes between 0.16 builds still work. 0.16's Immersive Briefings and the two manuals are below.

OPTIONALImmersive Briefings is an optional setting that needs a free voice pack, downloaded separately. The game is complete without it, and nothing changes unless you load it.

• Immersive Briefings · hear the week. With the voice pack loaded, the week in review opens with a short broadcast, read aloud. The broadcaster is whichever side is ahead on Decision Health: Blue Public Broadcast, Olvana Central Television or Donovia First.

• Built from your week. Each broadcast uses that week's real moves: the leader's best story, the other side's failures, areas won or lost, and the crisis ahead. The words appear as captions while it plays.
• On your computer, offline. The voice ("Emma, UK") is synthetic speech, made on your own computer by an open-source model.
  • Nothing is sent anywhere.
  • The game checks every file in the pack before using it.
• How to add it. Download the voice pack below (111 MB), then Settings › Immersive Briefings › Load voice pack…
  • With an installer, you can instead put the zip beside the game before you start it.
  • You can turn it off or remove it at any time.
• It plays on. Continue whenever you like: the broadcast carries on into the next week, with a small ON AIR tab to stop it.
• Two manuals. The Field Manual now holds only what a player needs, plus a new page 20 on Immersive Briefings.
  • The new Practitioner's Manual is for facilitators and instructors.
  • It covers running a session, choosing campaigns, leading the debrief, the campaign editor and setting up the room.
  • It ends with design notes on how each mechanic works and why.
  • Pick either one from Read the manual above.
• No rule changes. Two-player codes work between 0.16 builds.

DESIGNER'S NOTECognitive warfare is a contest of narratives, and every week in this game happens twice: once on the board, and once in how each side tells it. Immersive Briefings let you hear that second version. When you are winning, you hear your own side's calm account. When you are losing, you hear the other side's broadcaster telling your public that its leaders cannot decide, built from the moves you just watched. I made it optional on purpose. The base game has to stay small, quiet and offline for the classrooms, exercise rooms and locked-down networks where it is played. The voice is for players who want the war of narratives in their ears as well as on the map.

Get the voice pack (optional, 111 MB)

• Two players against Donovia. The two-player screen no longer assumes Olvana. You join as Red, and Blue's first code tells your game which campaign and which opponent you are playing.
• Stories name the right opponent. A few of Blue's story lines still said "Olvana" in Donovia games. They now name the opponent you are actually facing.
• No rule changes: two-player codes between 0.15 builds still work.

• A much harder Veteran. On Veteran, Olvana and Donovia now plan two moves ahead: they try their best options, let the week play out, and keep the move that leaves them best placed for the next crisis.
  • They play fair: no extra budget, no hidden information, and their trial runs never peek at the game's own dice.
  • Recruit and Standard are unchanged.

• Node status at a glance. Tiny icons under each node show whether it is watched, prebunked or framed, and its label names them in their own colours.
  • A Watch Team or Prebunk is drawn as one outline around every node it covers, with the weeks left; pick the card and hover a node to preview what it would cover.

• This week's orders. Above the timeline, each move this week stacks up as an icon, like a strategy game's queue. Click one for its story.

• More story per move. Move pop-ups and the opponent's move cards now tell you what was done and what it changed, without opening the full story.
• Decision Health, explained. The side panel shows how your health is made: areas × trust − lost areas − flooding. Difficulty never changes it.
• "Dots" are now nodes everywhere. Two-player codes work between 0.15 builds.

• Source check. Every real-world claim in the game and the field manual was checked again against the cited research.
  • A few lines were tightened to say exactly what the sources say, and anything that is a game assumption is now labelled as one.

• How Olvana fights. Olvana now gets the same treatment as Donovia: its own cards in the game's "How Olvana fights" panel and its own field manual page, showing what each rule does, where it comes from and where Olvana is weak.

• Both opponents side by side. The manual's "Know your enemy" pages now set out the rules both opponents share, then what is unique to each, with tips for beating them and a designer's note on how each should feel to play.

• No rule changes: two-player codes between 0.14 builds still work.

• Music on this page. Night Watch, the game's main theme, now plays here too.
  • It fades in when you arrive, or at your first click if your browser holds sound back until then.
  • The Music button in the top bar turns it off, and the page remembers your choice.
  • It pauses when you switch tabs.
• How the game was made. The start screen, the field manual's credits page and the project notes now set out how the game was made: the concept, design, rules, sources, opponents and campaigns are mine, and AI tools helped with code and drafting under my direction and review.
  • No AI runs while you play.
• No rule changes: two-player codes between 0.14 builds still work.

DESIGNER'S NOTEThis one is purely for fun. A wargame becomes memorable when it is engaging. Night Watch was written to sound like the quiet hours on an operations floor. It seemed a shame to keep it locked behind the start screen. If it isn't your thing, one click turns it off.

• A second opponent: Donovia. Like Olvana, it is a fictional country from the US Army's training environment (DATE), and it plays by its own rules, each drawn from published research by NATO, ISW, IISS and Carnegie: it floods the zone (many attacks in a week slow your decisions), plants premises (its framing lasts longer), hides dormant narratives in your mind that hit twice as hard when switched on, and has a ninth move, Show of Force, that makes your next move cost more but can backfire.
  • It is also predictable and brittle, so prebunking works better against it.

• Four new campaigns, all original fiction: Winter Grid, The Lake Referendum, The Silent Survey Ship and The Vell Valley, a story of annexation.

• Know your opponent. "How Donovia fights" on the start screen explains every rule and its source. Your side panel counts the dormant narratives still hidden in your mind.

• The campaign editor lets you choose Olvana or Donovia. The field manual has two new pages and the new sources.
• Two-player codes work between 0.14 builds.

• 18 real technologies for your campaigns, past, present and future, all from the game's research: the eleven in Rushing, Umeokolo & Xu (2025), from e-mail in 1971 to AI synthetic media, and the paper's predicted Humanoid Robots, plus Autonomous AI Agents, AI Vulnerability Mapping, Brain-Computer Interfaces, Neuro-cognitive Weapons, Quantum Computing and Cognitive Biotechnology (from the NATO Chief Scientist, NIU and NeuroStrike reports).

• Build your own technology, or change a preset: its name and year, the four attribute scores the research uses (with the total and whether the paper's model would flag it), the moves it boosts, by how much, and its special effects.

• A cleaner campaign editor with tabs, a roomier calendar and a tidier side panel.

• Click any picture on this page to see it full size.

• Week in review. At the end of every week: both sides' health, your moves and what each achieved, Olvana's moves with the most likely reason, how each part of your mind moved, and what the next crisis needs.
  • SHOW ON MAP lights up the nodes that changed.
  • You can switch it off in Settings ⚙.

• See what Olvana does. While Olvana moves, the node it hit lights up on the map with a number and the rest of the map dims.
  • Each move card says why Olvana probably chose it, and the cards sit over the side panel so they never hide the map.

• Looking ahead. From two weeks before a crisis, the side panel shows how each side would handle it today and where Olvana has been aiming.

• Replay in the debrief. Step or play through the whole campaign: both minds week by week, the moves, the crises and the turning points.

• Campaign editor. Press CAMPAIGNS on the start screen.
  • Make your own campaign or copy a built-in one, place crises and new technologies on the calendar, write the briefing, set how strong each side starts, check its balance and test play it.
  • Share it as a file or a code; import other people's.

• Updated field manual with two new pages: After each week (16) and Make a campaign (21).

• Cleaner panels. Hovering a node now shows a short card: who it is, a tug-of-war bar for who holds it, and the expected result of your move as three big numbers.
  • The longer real-life explanations fold away in the side panel until you want them.

• A readable log. Each move is one row with its card icon, target and result badge; weeks, new technologies and crisis results have their own markers.

• The field manual is now the in-game Guide, page for page: the same magazine pages as the PDF, with a contents rail and page buttons. It works offline and remembers where you were.

• Updated field manual. Every screenshot is retaken from this build; the new PDF is in Downloads.

• Three new music tracks in Settings ⚙: Undertow (a slow echoing arpeggio), Long Telegram (a lone voice over the drone) and Dead Air (radio pings in the night), alongside the original Night Watch.

• Twelve colour schemes in Settings ⚙, all keeping the star chart: Starmap, High Contrast, Dusk and Chart Room (a light paper chart for projectors and bright rooms), plus eight operations-floor schemes: Clear Signal, Narrative, OODA, Resilience, Friction, Grey Zone, Ground Truth and Cognitive Domain.

• Easier-to-read text. The grey text is brighter everywhere; the dimmest text is now more than twice as readable.
• Colour-safe mode and text size work with every scheme.

• The game fits your screen. On big monitors (1440p, ultrawide, 4K) everything grows instead of shrinking to tiny text; on small laptops nothing overlaps or gets cut off.

• New Interface size setting (Auto, Larger, Largest) in Settings ⚙.
• Card names wrap instead of being cut short, and names on the map no longer print over each other.
• The start screen keeps its Start buttons in view on short screens.

• Two-player games over copy-and-paste turn codes, with resume.
• A sharper computer Olvana that concentrates on one area at a time.
• Realistic engagement numbers (views, likes, comments, shares) behind each grade.
• Installers for Windows, Mac and Linux, and the illustrated field manual.