Deep dive
Structured debate, formal accountability, and a credibility system that follows you. Here's everything under the hood.
The core mechanic
A duolog is a structured, one-on-one debate between two users. It can be initiated from any post or comment on the platform. Both participants agree on the format before the debate begins, then argue in alternating entries with citations required for non-obvious claims. The finished record is published to the feed for anyone to read.
See a take you disagree with? Challenge it to a duolog directly from the post. You write your opening argument as part of the request.
Both participants agree on word limits, video time limits, and the number of rounds. Either side can propose, counter, or accept. The debate doesn't begin until both agree.
Entries alternate between participants. Each entry must respect the word limit. Citations are built into the composer — you add sources as you write, and they're displayed alongside your text.
If your opponent argues in bad faith — misrepresents your position, makes uncited claims, contradicts themselves — you issue a formal challenge called a flick. They must respond before the debate continues.
When the agreed round limit is reached, either participant can request an extension. The other party can grant or deny. If denied, the duolog ends.
After the duolog ends, each participant can publish the full exchange — entries, flicks, verdicts, and all — to their profile and the public feed.
Accountability
A flick is a formal challenge issued during a duolog against a specific passage in your opponent's entry. It's not a downvote, a report, or a moderation request. It's a rule-specific accusation that something in their argument violates the principles of good-faith debate. There are six types, each color-coded and governed by its own rules.
Flicks can only be issued during your turn in a duolog. You can highlight specific text in your opponent's entry and flick that passage. The opponent must yield or contest the flick before the debate can continue.
Used for non-constructive labels including name-calling, insults, and smears.
"What would someone like you know about economics?" — said to avoid engaging with a fiscal policy argument.
Rule: Applies when a label or descriptive phrase: 1. dodges the argument, or 2. is non-constructive (e.g. "You knucklehead!").
Used when a claim is demonstrably factually incorrect — not merely contested or disagreed with.
"The Berlin Wall fell in 1991." — The Berlin Wall fell in 1989. This is verifiably false.
Rule: Only for claims that can be demonstrated to be false. Not for opinions, interpretations, or contested characterisations.
Used when a claim is unsupported, has an incomplete citation, or cites a source that doesn't actually back the claim.
"Studies show a 40% increase in anxiety among teens." — A specific statistic like this needs a source that actually contains that number.
Rule: Common knowledge does not need citation. Cannot be used to contest a source's credibility — only its absence, incompleteness, or mismatch with the claim.
Used when your position has been misrepresented and the false version is attacked instead.
You say "we need stricter pollution rules" and they respond with "so you want to shut down all businesses."
Rule: The safest approach is to quote your opponent's exact words. Only paraphrase if you can cite a supporting statement.
Used when statements within the same duolog contradict each other.
Round 1: "Free speech should be unrestricted." Round 3: "Hate speech should be banned." — Direct contradiction.
Rule: Only for blatant contradictions within the same duolog. Subtle or ambiguous contradictions are best addressed in the discussion itself.
Used when a question embeds an unestablished assumption that forces you to accept a false premise.
"Why do you keep defending harmful actions?" — assumes the action is harmful, which hasn't been established.
Rule: Only use if you disagree with the embedded assumption.
The lifecycle of a flick
Every flick follows the same path. The flickee — the person whose entry was challenged — must respond before the debate can continue. Their choice determines what happens next. Contested flicks are not adjudicated until the duolog is complete — the jury reviews all contested flicks together once all rounds have been played.
A flick is filed against a specific passage in the opponent's entry. The debate pauses until the flickee responds.
The challenged party must choose one of two paths:
Accept the challenge. No penalty is applied. The debate resumes immediately.
Dispute the challenge. The flick is marked contested and the debate resumes. A jury will review it later.
Once all rounds are complete, a jury of 7 qualified users is randomly selected to review every contested flick from the finished duolog.
Each juror reviews the contested flicks with the full exchange in front of them and votes independently. A strict majority determines the outcome:
The challenge was valid. The flickee receives a −25 point score penalty.
The challenge was invalid. The flicker receives a −25 point score penalty.
Both sides have skin in the game
If a flick is sustained, the flickee loses 25 points. But if the flick is overturned, the flicker loses 25 points. This means filing a frivolous flick is just as costly as arguing in bad faith. You don't flick lightly.Try it yourself
This is a self-contained sandbox — nothing here touches the real platform. Walk through the full flick lifecycle: issue a challenge, respond as the flickee, and vote as a juror.
Click the highlighted passage above to issue a flick against it.
Decentralised adjudication
Elenkia has no moderators. When a flick is contested, the dispute is decided by a panel of seven randomly selected, qualified users. Not employees. Not algorithms. Peers.
After a duolog is completed and all flicks have been issued, the system randomly selects up to seven eligible users who were not participants in the duolog. These jurors are notified and review all contested flicks together, with the full context of the finished exchange in front of them.
Each juror votes independently: Sustain (the flick was valid) or Overturn (it was not). A strict majority of the seven-person jury — four matching votes — determines the outcome. Verdicts are final. If either party chooses to publish the duolog, the results are made public.
Jury eligibility requires an Exceptional rating, at least five completed duologs, and all six tutorials passed. Users who accumulate too many penalties — from their own bad-faith arguing — lose jury eligibility entirely. The system is self-correcting: only users who have demonstrated sustained good-faith participation can adjudicate others.
Reputation
Every user carries a public credibility score derived from their full debate history. It's not a popularity metric. It's a measure of how you argue.
Your final score is your raw score divided by your total number of duologs. This means your score reflects your average performance, not just volume.
Your score is tied to your history — not your account. It's public, persistent, and visible on every entry you write.
Elenkia is currently in private beta. Apply for access and help us build a platform where arguments have real consequences.