Kleros and Omen


Kleros is a general purpose dispute resolution dapp. It currently has an escrow, a curated registry of tokens, a proof of impact fundraising platform and an oracle using it.
Kleros is used by the DutchX using the ERC20 badge of the Token² Curated Registry (as a result of a previous proposal) to display tokens in its interface. And the Kleros team has combined the buy and sell frontend of the DutchX and fixed a few UI bugs.
Kleros has been integrated as a final arbitrator through realit.io within the conditional token framework and can be selected as an arbitrator on this frontend (you can even try the US presidential election market and the corona deaths markets which are live and have liquidity).
Kleros is also the arbitrator of the 10 000 GNO (~200 000$) grant from Gnosis to the dxDAO on the condition that there will be at least $20,000,000 in Gnosis Protocol trading volume originating from orders created on interfaces controlled by dxDAO.


Omen is a prediction market project ran by the dxDAO based on the conditional framework made by Gnosis. Its primary arbitrator is realitio and supports selecting a final arbitrator.
When a prediction market reaches its resolution date, anyone can report an outcome by staking ETH on realitio. If someone doesn’t agree with the outcome, they can dispute it by staking on another outcome such that there is at least twice as much on this new outcome. This process repeats until:

  • No one dispute the current outcome for a determined period of time. In this case the current outcome is considered the final one.
  • Someone escalates the dispute to the final arbitrator of the market by paying the arbitration fees. In this case the outcome returned by the final arbitrator is considered to be the final one.

People who staked their ETH on the final outcome / escalated an incorrect outcome to the final arbitrator gets the ETH of people who reported incorrect outcomes.
Since the arbitration fees are paid by the party escalating, escalation is only worth it once the amount of value at stake greatly exceeds the arbitration fees. For this reason almost no disputes escalate to this point (Up to now, there hasn’t been any dispute escalated to the final arbitrator on realitio. On the competing prediction market Augur which uses a similar system, we haven’t seen usage of the final arbitrator either).
The idea being that the system requires a really secure arbitrator to work and that the cost of this arbitrator can be high, but it should almost never be used, as the simple threat of someone escalating to it makes people stake on correct outcomes (directly or potentially after a few staking events when it becomes clear which outcome looks like being the consensual one).


We see 3 potential modalities of collaboration on Omen (with Kleros being a secondary arbitrator, Omen being arbitrator neutral or Kleros being the preselected arbitrator).

For Kleros, the benefit is to get more Dapps connected to Kleros as part of the ecosystem bootstraping and potentially get a few cases.
For Omen is to allow its users to select (potentially among other arbitrators) the first dispute resolution Dapp.

In addition, depending of the closeness of the collaboration, the Kleros team can also bring smart contract security skills in reviewing contracts used by Omen. We already did a security review of realitio when we integrated (We found a small issue which fortunately is only affecting commit and reveal which is not implemented in graphic interfaces. So the issue is not present for people interacting with realitio via Omen or realitio interfaces and will be corrected in next version.), we could expend this security review to the conditional token and liquidity contracts in order to have all the Omen code covered. The security review could include a bug bounty program.

We would also set up a reward program for people creating markets. The rewards will be proportional to the open interest (total amount at stake) of their markets. Open interest is a more interesting metric than volume which can easily be faked by wash-trading. As part of the reward program, the dxDAO will award a reward to the creator of the most impactful market which will be used as part of the communication strategy.
Rewarding the first market creators (which will likely be liquidity providers of their markets) is important, as without market creators / liquidity providers, potential users coming to Omen will not be able to trade and quickly lose interest if there isn’t enough liquidity. We can see Omen as a 2 sided marketplace with liquidity providers and traders. The reward program ensures that providing liquidity initially is a financially rewarding activity even before having a high flow of traders.

In addition, on the higher levels of collaboration proposed, the Kleros team would reward financially the dxDAO for launching and reaching some open interest milestones. This would be done via conditional tokens (same as the conditional GNO to the dxDAO).

Finally, the Kleros team will communicate on Omen which should bring some Kleros community members to trade on Omen and potential partners to use Omen as decision helping tool (providing liquidity to markets in order to get insight on the question while not necessarily trying to make profit).

You can find the following doc different modalities of collaboration, don’t hesitate to make comments and indicate the ones you’d prefer. The most consensual of those will result in a formal proposal to the dxDAO.


To me the strength of the dxDAO can be to “wire” individual protocols together.
Ideally the dxDAO itself is also not seen as a individual protocol/project but instead the voters of dxDAO themselfs are mebers/stakeholders of many Ethereum projects. dxDAO is a place to come together and cross collaborate.

In this spirit and the certainly relevant token incentives for dxDAO I would vote for the " High end collaboration".
Also, while the dxDAO can technically act as an arbitrator today this would likely just boil down to a “gut-feeling” decision of the top ~10 dxDAO members. To make the decision accepted by users it certainly helps to have a structured process around “presenting evidence”/“decision steps”. From why I can tell Kleros has already established that while for the dxDAO such a process is so far undefined.


I am also in favour of the High end collaboration.

  • The Dxdao is still a young organisation which will benefit from working with other projects/communities.

  • Once Omen will be released, it will still be an MVP/Alpha product which will need to be maintained. Based on the High end collaboration proposal, Kleros signals that they will make sure to support Omen as much as possible by doing code audits, testing the application, creating markets and making bug reports.

  • Kleros is a more secure arbitrator then the Dxdao.

  • By working together, both communities can make sure that we succeed with our goals especially with the conditional token challenges which comes with the High end collaboration.


For me the high end collaboration sounds best. This will expand the collaboration network and make sure that a more established arbitrator will do the finalisation of a disputed prediction.

What does " 12k$ of value produced to the network per year" mentioned in the estimates on disputes mean?

Thumbs up for this well worked out article and proposal.


Thanks for your comments, apparently “High end collaboration” seems the most popular and I’ll submit a formal proposal.

That was just some heuristic I used to estimate the “value” (arbitration fees minus value of the time spent by jurors) of the “Omen” integration for the Kleros network while determining/discussing of allocating resources to it on the Kleros side.

1 Like

Thanks for the reply. Somehow I don’t get the value. If I understand the assumptions correctly, the final arbitrations will create a revenue of $60k. Arbitration surplus is 20%. The value created for Kleros are $48k (revenue minus costs) or do you mean $60k+20% surplus=$72k of revenue created for Kleros?

1 Like

That’s economic surplus, assuming this surplus is 20% of the total arbitration fees (ex: a juror would be ready to spend 1h of arbitrating for 8$, but he is paid 10$, so his surplus is 2$ or 20% of the arbitration fees). You could link this notion as related to profit (60k$ of revenue but 48k$ of labor cost, so 12k$ surplus) but it’s not exactly the same as profit as a juror doesn’t need to pay himself for his time, he can however “value” his time according to what other “value”, monetary (working on something else) or not (playing with a friend), from the use of his time.
Note that I didn’t spend that much time on approximating values and it was just to give an order of magnitude when I was determining what to put as financial incentives in the proposal.

Yes agree, that this is an approximation, altough I can’t get a grip of the meaning of value in this case. However I am looking forward to vote on the proposal:)


The High End Collaboration proposal has been submitted on alchemy: