thoth · uma oracle court · records 2,692 · updated 2026-07-09
polymarket outsources truth to an oracle.
this records the judges.
every disputed polymarket outcome is decided by uma token-holders voting with wallets - and the payout follows their vote. per dispute: who voted, their weight, whether they matched the official outcome, and how concentrated the vote was. every record is hash-chained before it can age - it cannot be quietly rewritten.
latest disputes
all 2,687 →read this first
- on-chain facts only. this states how a wallet voted and whether it matched the official outcome. it never links a wallet to a real-world identity and never asserts fraud - those are conclusions this ledger does not draw.
- correctness is vs uma's own resolved price, not ground truth. a captured vote that resolves 'wrong' still scores 'correct' here. concentration and correctness together are the signal; neither alone is an accusation.
- the conflict-of-interest flag is not in this build. whether a voter also held a position in the market it ruled needs a cross-chain positions join - phase 2. the day-1 board is concentration + correctness.
- re-derive every number yourself. source is the public log stream of uma's votingv2 on ethereum mainnet; the aggregate json and the hash-chained ledger are downloadable below. trust the chain, not this site.
most contested
all 2,687 →voter rap sheet
all 4,881 →| no. | wallet | votes | graded | correct | rate | weight |
|---|---|---|---|---|---|---|
| 0001 | 0xfc20..1e1c | 2398 | 1203 | 1194 | 99.3% | 4339.0M |
| 0002 | 0xd2a7..9f7f | 2704 | 1325 | 1307 | 98.6% | 3991.3M |
| 0003 | 0xb524..3715 | 1465 | 675 | 674 | 99.9% | 3746.6M |
| 0004 | 0x8e34..c97a | 1471 | 682 | 672 | 98.5% | 3287.0M |
| 0005 | 0xc60a..fea0 | 1474 | 684 | 674 | 98.5% | 3269.0M |
| 0006 | 0x95cd..29b8 | 2715 | 1334 | 1314 | 98.5% | 2769.7M |
| 0007 | 0xfec2..04a5 | 2576 | 1269 | 1236 | 97.4% | 2597.4M |
| 0008 | 0x05d1..f8c4 | 2689 | 1316 | 1298 | 98.6% | 2558.7M |
| 0009 | 0x0afc..0b38 | 2727 | 1340 | 1330 | 99.3% | 2126.5M |
| 0010 | 0x487a..2538 | 2521 | 1234 | 1199 | 97.2% | 1667.6M |
| 0011 | 0x8548..e67e | 2668 | 1317 | 1253 | 95.1% | 1132.0M |
| 0012 | 0x326c..4205 | 2667 | 1303 | 1285 | 98.6% | 1122.7M |
| 0013 | 0x6564..340c | 2519 | 1236 | 1210 | 97.9% | 1115.8M |
| 0014 | 0x66ca..225a | 2700 | 1322 | 1303 | 98.6% | 1032.7M |
| 0015 | 0x996d..aff1 | 987 | 421 | 401 | 95.2% | 835.6M |
| 0016 | 0xdb28..b80a | 2409 | 1144 | 1082 | 94.6% | 757.1M |
| 0017 | 0xf659..d48b | 2716 | 1332 | 1316 | 98.8% | 749.3M |
| 0018 | 0xbd9e..baa7 | 1738 | 783 | 775 | 99.0% | 697.4M |
| 0019 | 0x20ec..56c0 | 2581 | 1262 | 1243 | 98.5% | 651.6M |
| 0020 | 0xf002..bfb2 | 2712 | 1338 | 1324 | 99.0% | 639.6M |
| 0021 | 0x9928..8cee | 2708 | 1325 | 1315 | 99.2% | 629.2M |
| 0022 | 0x840b..946f | 2472 | 1187 | 1171 | 98.7% | 611.2M |
| 0023 | 0x4863..e44f | 2363 | 1170 | 1152 | 98.5% | 592.6M |
| 0024 | 0xbec3..0ff4 | 2317 | 1198 | 1191 | 99.4% | 568.0M |
| 0025 | 0xb731..5790 | 2618 | 1269 | 1249 | 98.4% | 566.5M |
method
source: the public log stream of uma's votingv2 contract on ethereum mainnet,
0x004395edb43efca9885cedad51ec9faf93bd34ac, read directly over rpc. each revealed vote
(VoteRevealed: voter, price, numTokens) is joined to its request's official outcome
(RequestResolved: price) on (identifier, time, ancillaryData).
correctness = |vote - resolved| ≤ 0.01. weight = numTokens / 1e18. concentration = the top wallet's share of a request's revealed weight. question titles resolve deterministically from the polygon oo side of each request.
the chain: disputes are appended once each to a hash-chained ledger in resolution order; each record carries the hash of the previous one. edit any past record and every hash after it breaks.