Potential Solutions

Throughout the discussion, several potential solutions were brought up.

Polymarket or UMA Refund

Wizard repeatedly pushed for a refund, saying it costs UMA nothing. When others asked about logistics, there was disagreement about implementation.

50Pence suggested having two penalty amounts: one for minor mistakes and one for more serious ones. He didn't define which is which, but this likely relates to proposals submitted too early. Vote discussions or outcomes could help clarify the difference.

For context, over 80% of disputes arise from proposals being submitted P4 (Too Early).

The steep increase in P4 disputes has prompted several proposals for change. Lee from Risk Labs started a discussion suggesting raising the proposal bond cost by at least $250.

Many community members pushed back, saying higher costs punish everyone instead of solving the actual problem. T1Dev floated the idea of heavier penalties specifically for disputes that resolve as P4 too early.

Dynosawr argued the main problem isn't proposal cost, but that the free $5 reward attracts new users who submit most P4 too early proposals. Even with a warning in the UMA Oracle, people still submit proposals too early.

Dynosawr also proposed adding a self-dispute mechanism to the smart contract, which led to more technical discussion about implementation and refunds.

Adding a Data Availability Clause

Some users argued against waiting an unspecified amount of time after an event ends just for official results.

SJ suggested a data availability clause with a clear deadline. For example, if official results don't appear within 24 hours, preliminary results could be accepted.

Other unrelated markets already use this type of clause. For example, the Earthquake 7.0 or above by July 31? market requires waiting for the resolution source to update, even after the event has occurred.

Neg Risk Markets

Sjennings suggested merging all nested markets into a single proposal instead of one for each option. Neg risk markets are winner-takes-all events where only one contestant can resolve to Yes.

The NASCAR market, for example, was a neg risk event with 43 nested options, which meant 43 separate proposals.

Ice explained that the neg risk structure is necessary because some options can resolve early, such as when a participant is disqualified.

Proposal Grace Periods

Domer argues that proposals should be judged based on the moment before a dispute is submitted, not at the time of proposal.

Debosnys questioned his claim. Domer expanded, saying that if a proposal is correct and isn't disputed in real time, especially in sports markets, it should stand since most delayed disputes are just technicalities.

Flame heavily criticized Domer's take, arguing that a proposal is either on time or too early and there shouldn't be exceptions for the oracle. In reality, the timing of a dispute within the challenge period is irrelevant, since it's impossible to know in real time if something turns out to be true "30 seconds later" without hindsight. The oracle considers resolution at the moment the proposal is submitted, not in retrospect.

Domer rejects this claim, saying that people are essentially dispute hunting. This raises questions about whether such behavior is healthy for the oracle.

Flame and 50Pence discussed whether there was any real harm, suggesting that if a proposal isn't immediately disputed, then nobody actually cared. However, in practice, this approach would make the oracle inconsistent.

Car shared his experience with proposing too early and supports staying strict with P4. He argues the oracle should not be liberal in allowing early proposals.

JessicaOnlyChild pointed out that disputes require discussion time and not everyone checks the dispute portal immediately after a proposal.

Crz also shared her experience, describing disputes as less about honest mistakes and more about people actively hunting for errors.

An Improved Rulebook

Several users discussed the need for a better rulebook.

Many Wizard supporters oppose precedent, but precedent is necessary to keep the oracle consistent and predictable. The main issue isn't precedent itself, but making it accessible and transparent.

There was also support for creating a rule writer team and for adding precedents directly into market rules.

Some users suggested adding more warnings with proposals, though there is already no shortage of warnings.

Summary

The discussion brought forward several solutions for improving oracle resolution. These included adding refunds for early or mistaken proposals, introducing tiered penalties based on the severity of mistakes, and increasing the proposal bond to deter careless submissions. Participants also suggested clarifying data availability deadlines, merging neg risk markets to reduce complexity, and establishing proposal grace periods. Many emphasized the need for real-time dispute resolution and stricter enforcement against early proposals. There was shared support for a clearer, more accessible rulebook with precedents. While some called for more warnings, others felt that better enforcement and consistency would have a greater impact.

Coming Up

Now that we've covered the potential solutions, we'll take a look at how Polymarket decided to respond.