Welcome to The Great Tribunal. This is a platform where you embody historical figures and argue cases that shaped monetary history. You're not playing a game about money — you're putting real economic arguments on trial, hearing both sides, and deciding guilt or innocence based on evidence and reason.
The red pill: step into a courtroom where the rules are clear, the arguments are rigorous, and the outcome depends on your ability to make a case. No hand-holding. No predetermined winner. Your words matter.
When you enter the Tribunal, you choose one of six roles. Each plays differently and earns IA (Intellectual Authority) by different means.
If you don't fill all roles, AI characters take the remaining seats. They argue in character, respond to you, and can shift the burden of proof on you just like a human opponent would.
Every trial progresses through five phases. You control the pace — advance whenever you're ready.
Each side makes their opening statement. Set expectations, frame the narrative. First impressions count.
The side with burden (usually prosecution) presents evidence and builds their case. Direct questions and responses.
The opposing side pokes holes. This is where overclaimed arguments collapse and burden shifts happen.
Final arguments. No new evidence — just summation. Wrap up your strongest points.
Jurors vote. The verdict is revealed. The trial is over.
You can advance phases whenever you choose. There's no timer. Take your time to think, research, and craft arguments.
This is the core mechanic. Every case has a burden of proof — a standard the prosecution (or a side) must meet to prove their case. If they meet it, they win. If not, the accused goes free.
Here's the trap: If you make an extraordinary claim (something unexpected or extreme), the AI Judge will shift the burden to YOU. Now you must meet a higher standard. If you can't, you lose.
Case: Was Newton a criminal for counterfeiting?
Initial Burden: "Prosecution must prove Newton knowingly counterfeited."
What happens: Prosecution argues Newton was the Mint Master and couldn't have counterfeited.
Defense overreaches: "No, Newton was secretly running a counterfeiting ring that swindled the entire British economy."
Judge response: "That's an extraordinary claim. Burden shifts to Defense. You must now prove Newton ran an illegal ring with documentary evidence."
Result: Defense can't meet that standard. Burden was a trap.
The lesson: make solid arguments, not wild accusations. If you claim too much, the burden moves to you and you have to back it up.
When someone (you or the AI) makes a claim the judge considers extraordinary, the courtroom pauses. An amber spotlight banner mounts above the transcript, the claim is quoted verbatim, and the rest of the screen dims. Jurors visibly destabilise — some flip from green or red into orange (uncertain). IA is auto-bonded against the claim (5–50 IA depending on severity). The claimant is on the spot.
✓ PROVE — cite specific evidence or a verifiable record that meets the judge's stated standard. If accepted, the bond returns to you with a bonus, jurors come home, and the claim is promoted to an established fact in the trial.
↩ WITHDRAW — formally retract the claim. You forfeit half the bond, but the strike is averted and the doubt lingers on the jury rather than crystallising against you.
⚔ DEFEND — argue substantively without introducing a new extraordinary claim. Full bond at stake. If the judge accepts your defense, jurors flip back to you. If not, the claim is struck from the record.
Struck claims are never deleted — they remain in the transcript with a red STRUCK watermark over them. The film cut preserves the moment forever. You watched the defendant fumble, and the audience will too.
Every time you trap the same defendant into an extraordinary-claim failure, the count ticks up. On the 3rd failure, you earn an Inquisitor's Edge token — a glowing amber badge in your HUD. It's consumed automatically on your next winning argument: the jury swing doubles, and you gain a +5 IA bonus. Land more failures, stack more tokens.
When your argument caused the defendant to overreach into a struck claim, the question you asked becomes valuable. Other players will pay to know what you said. The marketplace turns great prosecutorial questions into compounding income.
Right after a successful trap resolves, a 🪤 amber toast appears: "You trapped them. List this argument as a TRAP (up to 100 IA)?" Regular argument listings cap at 50 IA — trap listings double that ceiling because they have proven adversarial value.
If you've already listed the same argument as a regular item, the system auto-promotes that listing to TRAP status when the trap resolves. You get an "Upgraded to TRAP" notice; no second sale needed.
If you dismiss the toast — or open the list modal and cancel out — the offer doesn't disappear. A 🪤 TRAP READY TO LIST badge mounts next to your IA display and stays there until the trial ends, or until you list it. You can come back to it whenever you're ready.
Other players who buy your trap argument can deploy it against any defendant. If their use of your question lands another trap, you earn royalties on every unlock and resale forever (capped at 1000 IA per argument).
→ Marketplace pricing, royalty splits, and the throne mechanic
Every argument used in every trial is recorded, scored, and made browsable. The marketplace and the post-trial archives are the entry points to a living corpus that gets smarter as more trials happen.
Each marketplace listing now shows an effectiveness bar (0–100%), the use count across all trials, the trap conversion ratio (lands/triggers) for trap items, and a decay arrow indicating whether the argument is still landing or being learned around. Fresh listings pitch as "✨ Fresh — never reused" — premium pricing, no decay yet.
Stuck in a live trial? Hit the CONSULT button below the INTERN. For 15 IA, the system fetches the top 3 archived attacks against your current defendant that you haven't yet tried this trial, with full verbatim text. One-click "USE THIS →" autofills your argument input. If the archive has nothing for you, your IA is refunded automatically.
After your trial, visit the Archive. Each per-case panel has an ⚔ ATTACK ARCHIVE — every attack ever recorded against that character, ranked by effectiveness. Stats are free. Full verbatim text is paywalled per-item at 5–25 IA (scaled by effectiveness). Killer attacks (effectiveness ≥ 70% with 3+ uses) get an amber border and a 🗡 marker.
Above each Attack Archive, the DEFENSE PACKS shelf shows curated bundles of proved defenses for that character — typically 20 defenses for 50 IA. Buying a pack unlocks read access to every defense in it AND tags your intern so future draft prompts can lean on the same material. The Locke Defense Pack, for instance, hands your intern Locke's actual best refutations to draw from.
→ How effectiveness, decay, and the AI's anti-repetition all work under the hood
Two currencies power the Tribunal.
You start with 100 energy. Each argument costs 10 energy. Submit evidence and recover 30. When energy hits zero, you can't argue anymore — but you can still vote.
The global scoring currency. Earn IA by: winning an exchange (+5), submitting high-quality evidence (+5 to +50 depending on accuracy/relevance), completing a trial (+10), and voting correctly (+5). IA appears on the leaderboard. The top players globally are the ones with the most IA.
Energy is tactical — you must pace yourself. IA is strategic — it's your global reputation. Play to build both.
During any trial, click the 📁 Evidence Locker to submit historical evidence or testimony. This is how you contribute primary sources, quotes, statistical analysis, or expert commentary.
1. Write your evidence (title + content, minimum 30 characters).
2. Submit. The AI Court Assessor evaluates it in seconds.
3. You get scored: Accuracy (1–10), Relevance (1–10), Rhetorical Strength (1–10).
4. Earn 5–50 IA based on the scores.
5. Recover 30 energy.
6. The opposing AI generates an automatic rebuttal, posted to the transcript.
Quality matters. One strong piece of evidence beats five weak guesses. The AI judges accuracy and relevance carefully.
After the final phase, the trial ends and you move to the Verdict page. Here, all participants vote: Guilty / Not Guilty / Systemic Failure.
Click one of the three buttons. Majority wins. The outcome is recorded.
Claude claude-opus-4-6 (the Reason model) writes the official verdict — a formal document with findings, legal reasoning, historical lesson, and a creative sentence. A performance score (0–100) grades how well each side argued.
Get a permanent URL. Share with friends. Invite others to argue the same case. Each trial stands as a record in the Archives.
Play alone (AI fills empty seats) or invite real players to argue alongside you.
Start a trial. A room code is generated. Share it with friends via WhatsApp, Telegram, or copy-paste.
Click an invite link, pick your role, and enter. Your arguments sync in real time with other players. The transcript updates live.
After entering a trial, you can generate role-specific invite links: "Join as Prosecutor," "Join as Defense," etc. Share these and let others pick their seats.
Don't see a case you want to argue? Submit your own. Go to ✍ Submit a Case and describe a moral dilemma.
1. Write a brief description of the dilemma (100+ chars).
2. AI structures it into a full case: title, charge, side-by-side arguments, balance score.
3. Edit any field. Strengthen weak sides if needed.
4. Choose: Save as Private (share via link only) or Submit for Public Review (editorial approval before listing).
5. Your case goes live. Other players can argue it.
The balance score (1–10) tells you if the case is fair. Score below 5? Strengthen the weaker side before submission.
Before entering a trial, the Studio lets you configure the difficulty and briefings.
Choose your role and see which characters the AI will play. Edit their briefs if you want (stronger/weaker arguments).
Expandable cards for each character. You can rewrite their background, motivation, strongest argument, or key weakness before the trial starts.
Toggle it ON to make the AI argue the absolute strongest version of the opposing case with zero concessions. Designed for lawyers and researchers who need to be genuinely challenged.
After a trial completes, admin can curate it into a cinematic audio episode. These appear in the Podcast archive and are playable via the Playback page.
Full-screen cinematic experience. Character faces animate in the background. Subtitles sync to audio word-by-word. Verdict mode animates a scale of justice as the verdict narrates. You can toggle subtitles, skip scenes, download the transcript as an .srt file.
Browse all produced episodes. Filter by case. Play short cuts (~60 seconds) or full trials (5–15 minutes). Each episode is a polished, narrated courtroom experience.
The goal: turn real arguments into educational media. Listen to how great lawyers frame cases. Hear the AI judge's reasoning. Relive history.
Ready to argue? Enter the Tribunal →