Fortify The Court

Application settings

One court, a route chosen by the application

A jury route lists the first jury and every possible appeal. The requesting application chooses that complete route before submitting the case. A person filing directly makes these choices through the court website. DemoThemis checks the terms, collects payment and draws the jurors. Neither the application nor the person submitting can choose the jurors.

Once accepted, the case keeps its route and terms. A new pricing policy cannot change them. Later appeals receive their own quotes under the case’s agreed pricing policy.

Appeal panel configuration routes are set before the draw

The appeal panel configuration route is set before the first draw. The requesting application follows the route it precommitted. A route may repeat a panel size, but every stage draws fresh jurors.

Read a route from left to right

Large rounds use separate 31-person panels. A 93-seat round has three panels. A 155-seat round has five. Each person serves once on a case, including across appeals.

Experience requirement

Each 31-person panel requires experienced jurors

Every 31-seat panel requires experienced jurors. Their experience requirement is 100 separate cases where they voted against their own round’s ruling and a later appeal produced a final ruling matching their vote. The vote can come from the initial round or an earlier appeal round. It must have been cast before that round’s ruling and the later appeal, not after learning the answer.

A YES vote in a round that ruled NO counts if a later appeal finally rules YES. A YES vote in a round that already ruled YES does not count. With several parallel panels, compare the vote with the overall round ruling. Count each case once per juror across their service, not only within Beta’s recent-case window. There is no separate years-of-service wait. Also require a Beta lower 95% endpoint of at least 90%, 100 scored cases in its current window, and completed work in the past 30 days. Count people who actually meet these requirements when forming a panel.

Before paid jury work: applicants answer 10 separate cases without pay or influence on the rulings. Each answer must be submitted before that case’s ruling is revealed. All 10 must match the final rulings, whether or not anyone appeals. These trial answers do not count toward the experience requirement above.

New work uses recent demand

Daily demand = (90 + 60 + 30) ÷ 3
             = 60 funded jury seats per day

The formula is fixed at filing. Each appeal price is calculated and locked when its funding window opens. Accepted pay and funded terms never change. The minute is a shared cutoff, not a promised quote-delivery time.

A replacement reuses unpaid seat money

No replacement cap, administrator approval or extra reserve is added. Money promised elsewhere cannot be spent. If the difference cannot be funded, the case and its remaining money wait. Weekly rewards use only money not already owed or committed.

From ruling to finality

Appeal finality and later accountability use separate clocks

The initial appeal-funding window can be up to one year. Each later window is at least 15% shorter, with an optional 24-hour minimum. These are payment windows, not the time jurors have to answer. The complete schedule is fixed when the case is accepted.

Connected escrow then follows the agreed result rules. Later juror audits and proven collusion can change penalties and eligibility. They do not reopen that customer ruling or hold its settlement open.

Keep the work payment and the appeal bond separate

Each appeal pays for its new jury, applicable charges and the immediately previous round’s possible audits. Its quote also shows any refundable appeal bond separately. The application may collect contributions, but the court accepts only the full payment.

At finality, compare the final ruling with the ruling each appeal challenged. If they differ, return that appeal’s bond to its requesting account. If they match, send that bond to the shared pool. The application handles payments back to its contributors.

Possible audit funding covers up to three auditors per assessed juror, one operating charge and the hidden-test contribution. Unused audit money goes to the shared pool. It is not an appeal-payer refund.

Appeal-triggered audit

Audit the juror’s work, not just the answer

A juror may attach a private explanation to their ballot. If their work is later audited, the auditors read that explanation alongside the case evidence. An overturned answer or a missing explanation alone does not prove carelessness.

Audits start only after the case is finalized. The appeal route must end, or an appeal-funding deadline must expire without full payment, and the contract must record the final ruling. Earlier appeals can reserve money for possible audits, but they do not start them. Audits assess juror work and never reopen the final customer ruling.

Auditors are randomly selected jurors who assess someone else’s work. They must meet the identity, quality and auditor-test requirements, be available and have no involvement in the case. There is no owner-selected list. They can research the question and see later evidence, rulings and any repeat-audit defence. They judge what the juror could reasonably know when answering, not with hindsight.

Auditors answer Reasonable, Careless or Insufficient information to judge the work. Two auditors assess each juror first. If they disagree, a third decides. Three different answers give Insufficient information. Missing answers leave the audit unfinished.

Count jurors whose votes matched the previous round’s ruling.
For each, reserve enough to pay up to 3 auditors.
Add one operating charge for that round’s audit.
Add the contribution that funds hidden auditor tests.

A Careless notice starts 90 days to request and fully pay for one repeat audit by three fresh auditors. The payment covers their work, the operating charge and a refundable deposit. The deposit returns for Reasonable or Insufficient information. Careless sends it to the shared pool. The separate careless-work deduction applies only after this challenge process ends.

Try one juror’s audit below. The example uses $4 per auditor assessing one juror and $1 to operate the whole audit. These are illustrative prices. The separate contribution for hidden tests is explained after the examples.

A penalty capped by what active jurors earn

After the challenge process, charge the smaller of twice the juror’s accepted pay or the median monthly earnings of active jurors, once. To calculate that median, include only jurors with an active arbitration in the last week and at least seven arbitrations in every week of a continuous month. Sort their monthly earnings and take the middle amount, or the average of the two middle amounts.

Use the careless juror’s available money. Any unpaid amount becomes debt against their future earnings. Never take another person’s protected money. This deduction is separate from the bond for answering on time.

One paid chance to challenge the audit

The refundable deposit equals that same median monthly earnings amount, not the penalty being challenged. Pay it together with the three fresh auditors’ fees and the operating charge. For the example below, suppose the median is $300. With $12 for auditors and $1 for operations, Maya pays $313.

Hidden tests keep checking the auditors

The software mixes past cases with final Reasonable or Careless assessments into auditors’ work without labelling them as tests. Nobody manually chooses an answer key. Tests measure agreement with those past assessments, not independently proven truth.

Extra percentage = tests’ share of all audit work + 10 percentage points
Example: tests are 20% of all audit work.
Add 30% to the appeal’s audit charge, not to the whole appeal.