Platform review
Zeitgeist review
Prediction markets and futarchy on a purpose-built layer 1, denominated in ZTG
Our verdict
On paper, one of the most serious dispute designs in the category; in practice, a chain that looks dormant and documentation three years out of date. Zeitgeist runs its own layer-1 with a bonded per-market oracle, a staked juror Court using commit-reveal voting, exponential appeal bonds and a global-disputes layer behind that. Two things in the source undercut it. The runtime binds the resolve origin to Root, on an extrinsic whose own comment says it “Allows the `ResolveOrigin` to immediately move a reported or disputed market to resolved” — so a privileged account can end a live dispute and settle it, bypassing Court and appeals entirely. And there is no void: the outcome type carries only categorical and scalar variants, with no invalid, so a badly specified market resolves to something and its traders are not refunded, whatever the documentation’s leftover references to `Invalid` still say. That documentation is unreliable on its own mechanism — wrong dispute period, wrong bond token, a disputer flow the current code does not accept — and every page is stamped August 2023. More materially than any of it: markets sit listed months past their end dates, the last commit to the chain repository is 1 December 2025, and the official RPC and indexer did not answer us.
Verified documents & product scope
Check country eligibility
Choose the country you want to check. This lookup does not change your site region or establish individual account approval.
These are document checks, not a funded-account test. Access may depend on residence, physical location, account approval and the contract category. Check the operator’s current terms before depositing.
- Zeitgeist documentation — checked
The documentation describes the prediction-market chain and ZTG. It does not establish current interface eligibility or available liquidity.
Pros and cons
What works
- A bonded per-market oracle with a staked juror Court, commit-reveal voting and exponential appeal bonds
- Runs its own layer-1 rather than deploying on someone else’s chain
- The dispute design is among the most carefully thought through in the category
- The chain source is public and auditable — which is how the Root override below was found, and is more than most venues here allow
What doesn’t
- The runtime binds the resolve origin to Root, which can settle a disputed market and bypass Court and appeals entirely
- No invalid outcome exists, so a badly specified market resolves to something and its traders are not refunded
- The documentation contradicts the code on the dispute period, the bond token and the disputer flow
- Every learn page is stamped “Last updated on Aug 22, 2023”
- The venue looks dormant: markets unresolved nine months past their end dates, last chain commit 1 December 2025, RPC and indexer unresponsive
- No licence, operating entity or jurisdiction stated anywhere
Prices and liquidity
We do not currently publish live quotes for Zeitgeist. Check the platform’s order book before placing an order: the last trade can differ from the price available for your position size.
How to assess liquidity and spreads →What Zeitgeist was built to be
Zeitgeist runs its own layer-1 chain — “an evolving blockchain for prediction markets and futarchy” with “optimized runtime logic” — rather than deploying contracts on someone else’s. The dispute design on paper is among the most serious in the category: a bonded per-market oracle, a staked juror Court using commit-reveal voting, exponential appeal bonds, and a global-disputes layer behind that.
Commit-reveal voting means jurors commit to a hidden answer and reveal it later, so nobody can simply copy the emerging majority; exponential appeal bonds mean each further appeal costs geometrically more, which is a deliberate brake on litigating a market forever. These are not naive mechanisms. Someone thought hard about them.
No licence, registration, operating entity or jurisdiction is stated on zeitgeist.pm or in its documentation, and real value is at risk — the docs state traders “risk loss if their predictions are incorrect”.
Two things in the source undercut the design
The runtime binds the resolve origin to Root, on an extrinsic whose own doc comment reads: “Allows the `ResolveOrigin` to immediately move a reported or disputed market to resolved.” A privileged account can therefore end a live dispute and settle it, bypassing the Court, the appeals and the global-disputes layer entirely. Every protection described above sits below that call in the hierarchy.
There is also no void. The outcome type carries only categorical and scalar variants, with no invalid option, which means a badly specified market resolves to something rather than being annulled, and its traders are not refunded — whatever the documentation’s leftover references to `Invalid` still say. On a venue whose markets can be created with arbitrary wording, the absence of an escape hatch for an unanswerable question is a structural gap rather than an edge case.
The documentation does not describe the code
This is not a matter of a stale sentence or two. The dispute-period page states a flat 24 hours where the source sets a per-market range of 12 hours to 30 days. It names the wrong token for the bond. And it describes a disputer choosing an outcome, which the current extrinsic does not accept.
Every learn page it sits on is stamped “Last updated on Aug 22, 2023”. So a reader doing exactly the right thing — reading the official documentation before trading — would come away with a wrong model of the dispute window, the wrong bond asset and a flow that no longer exists.
The venue looks dormant
More materially than anything above: markets are still listed as Active and Closed more than nine months past their end dates, the last commit to the chain repository is 1 December 2025, and the official RPC endpoint and indexer did not answer us.
We are recording observations rather than declaring a project dead, and a chain can be quiet without being abandoned. But a reader deciding where to put money should weigh those three facts together: markets that have not resolved long after they should have, no development activity for nine months, and infrastructure that does not respond.
We would rather publish this than a review implying you can trade here normally. If the RPC comes back and the markets resolve, tell us and we will re-read it with the date we checked.
User reviews
We have no approved reviews for Zeitgeist. Every submission is read by a moderator before it appears, and none has passed for this platform. The User Score is not published below 3 reviews, so nothing is averaged here either — the Editor Score above is measured against our own criteria and does not depend on it.
New review submissions are temporarily closed. For corrections, contact bestpredictionmarkets@proton.me.
Zeitgeist FAQ
- Can a disputed Zeitgeist market be settled without going through the Court?
- Yes. The runtime binds the resolve origin to Root, on an extrinsic whose own doc comment states it “Allows the `ResolveOrigin` to immediately move a reported or disputed market to resolved”. A privileged account can therefore end a live dispute and settle it, bypassing the juror Court, the exponential appeal bonds and the global-disputes layer entirely. The elaborate dispute design described in the documentation sits below that call in the hierarchy.
- Is Zeitgeist still running?
- We cannot say it has shut down, and we can report what we saw. Markets are still listed as Active and Closed more than nine months past their end dates, the last commit to the chain repository is 1 December 2025, and the official RPC endpoint and indexer did not answer us. Its documentation is stamped “Last updated on Aug 22, 2023” and contradicts the current code on the dispute period, the bond token and the disputer flow. Weigh those together before committing anything.