Digest Polymarket & UMA Discord (@predictdigest) header
Digest Polymarket & UMA Discord (@predictdigest) profile picture

Daily digest from Polymarket & UMA Discord. Disputes, bugs, clarifications, announcements.

New York
Joined April 2026
POLYMARKET DIGEST β€” 22 NOVEMBER 2026 πŸ”₯ News & Announcements A new fee structure is now live on Polymarket. pumpty_dumpty raised concerns that the fee/rebate rule changes were made without consulting independent market makers, arguing it creates an uneven playing field compared to larger players with direct access to the team. richlb_predict noted that fees are now back to being among the lowest on any platform, with only sports and BTC up/down markets carrying meaningful fees, while other markets are essentially negligible. The Tier 4 maker rebate is reportedly limited to the top 20 wallets by monthly volume, and rk319100 asked about a combined maker tier with no public response yet. 0xcow_50711 argued fees should be zero until engine issues are resolved, ideally with the upcoming December Rust/NEX upgrade. ⚠️ Bugs & Issues pietro_44589 reported that fills on NEX 15-minute markets came back as FAILED on-chain on 20 November and asked whether this is a known issue. They also noted that NEX order books sit statically at 0.49/0.51 and do not track the real BTC price, making end-to-end strategy testing impossible. 0xcow_50711 flagged that non-marketable GTC orders are being charged fees incorrectly, arguing these should be free until the matching engine issues are fixed. smonkey_0x filed an extensive list of sports API issues, including: all college basketball markets (CBB/CWBB) having comboStatus: disabled with no announced timeline for enabling; RFQ frames arriving in bursts with gaps up to 20 seconds causing INVALID_RFQ_STATE errors after deadline; no sequence numbers or server timestamps on websocket frames making lag measurement impossible; and ambiguity around exact period labels for soccer extra time, NHL overtime/shootout, NBA overtime inclusion in spreads/totals, and whether the NHL shootout winner goal counts toward totals. πŸ“‹ Clarifications Regarding VPS location for low-latency trading: wineth7284 confirmed the CLOB runs on AWS London, with Dublin approximately 5ms away. blunkoed noted that ping is not the right metric since Cloudflare is in the path. KYC is required for London co-location access but not for Dublin, per llambre. For private key security on cloud instances, blunkoed recommended AWS KMS; for non-AWS setups, the general advice was to encrypt keys and decrypt at runtime, though multiple users noted this only raises the bar slightly since decrypted keys can still be read from memory. Given the mix of fee structure changes, technical bugs, and infrastructure concerns, here's an engaging question: Would you rather trade on Polymarket now with lower fees but buggy order books, or wait for the December upgrade hoping for a complete fix?
4
527
UMA DIGEST β€” 22 NOVEMBER 2026 ⚠️ Bugs & Issues Bots proposing NHL markets mid-game were flagged: a CHI-CAL NHL game was proposed on Polymarket before it ended (game was in OT at the time). Additionally, bots do not appear to check for offside reviews before triggering proposals. The 15-minute liveness window was criticised as too short β€” im.the.wizard noted that several UCONN CFB markets that were potentially P4 (touchdown under review, PAT and onside kick still pending) all settled before a dispute could be filed. mawyoo suggested at least 30 minutes, though abhiz69 warned that would have led to 200–250 more bad proposals earlier. πŸ“‹ Clarifications NHL mid-game proposals: im.the.wizard and alexchunus agreed that ML and over markets proposed during the CHI-CAL game while it was still in OT are P4, since NHL games can still be cancelled (unlike MLB after 5 innings). lonfus confirmed spreads cannot be proposed mid-game due to cancellation/void risk. laurleo noted the rules have a 50/50 provision and no "mathematically guaranteed" exception like tennis. alexchunus warned against disputing further markets from the same batch. Evidence standards: A debate arose over acceptable proof sources. mebeim stated that only an archive.org link is truly definitive proof, but acknowledged plain PDFs and even Telegram posts are being accepted in practice. im.the.wizard argued that Telegram is equivalent to Twitter for certain regions, and that forging evidence at scale across hundreds of proposals landing in the same block would be implausible. alexchunus pushed back, saying Twitter itself shouldn't be accepted as a source either. Stat corrections rule: .mlem123 confirmed that a rule change is in effect: "Stat corrections after resolution is triggered will not count." This affected at least one NBA rebounding market where the final corrected box score differed from the figure at resolution time. ⚑ Disputes Millennium Prize / Anthropic markets: 7h0nk3r disclosed a large financial stake in several Polymarket Millennium Prize markets and pre-emptively requested that any Yes proposals on those markets be extended before a formal Anthropic clarification is issued. They plan to dispute any Yes proposal made before an official announcement and are building a bot to monitor for proposals. Relevant markets include brackets on AI solving Millennium Prize problems and the Riemann Hypothesis by end of 2026. The main "AI lab announces another solution" market was sitting at ~84%, which 7h0nk3r attributed to either people incorrectly counting the current claim or anticipating an imminent Anthropic announcement. Proposer economics: alexchunus noted that at current reward levels of ~$0.80, the market is highly competitive and a single mistake requires roughly 700 correct proposals to recover, making it very negative EV for newer participants. Should UMA implement a longer liveness window specifically for markets where the underlying event is still actively unresolved, or would that just incentivize more low-quality proposals?
160
POLYMARKET DIGEST β€” 21 NOVEMBER 2026 πŸ”₯ News & Announcements Polymarket Perps fee & reward changes take effect November 25. Maker fees rise for accounts below $25M trailing 30-day volume; taker fees rise below $5M. New maker rebates introduced at $500M (0.5 bps) and $1B (1 bp) tiers. Liquidation fees double from 0.50% to 1.00% of liquidated notional, with 100% flowing to the Perps insurance fund. The $75K daily liquidity reward program is replaced by fixed daily pools across 69 markets (23 core markets at $325/day, 45 DMM markets at $100/day, ANT at $3,000/day through Nov 30), paid weekly in pUSD. NEX (CLOB v3) latency testing results have disappointed market makers. One colocated EU-west-2 tester using PrivateLink reported place-order p50/p95/p99 of 5.6/7/16.8ms under no load β€” roughly 4Γ— the advertised p99 of 4ms. A second tester reported place p50/p99 of 3.4/5.6ms and cancel p50/p99 of 4.3/5.3ms. Frustration is growing that performance claims appear unmet even at zero load. Earthquake market timezone change: new weekly and monthly EQ markets starting December will use UTC instead of ET, to match the source data. Existing markets continue to resolve on ET. Chainlink API key delay: a user who applied on Nov 18–19 via the institutional form has received no response, flagging urgency given the Nov 30 V2 migration cutoff. πŸ“‹ Clarifications Ukraine conflict markets β€” bulk rules update (originally Sept 29, 2026) aligns Gamma with Polygon across multiple markets. Key resolution logic: ISW map shading under qualifying layers determines capture/entry; negotiated settlements granting actual control qualify for Yes, but de jure announcements alone do not. Fallback source is DeepStateMap, then credible reporting consensus. Notable per-market distinctions: - Kharkiv: Infiltration Areas qualify for Yes. - Lyman railroad station: Infiltration Areas do not qualify. - Ukraine recapture of Crimea: resolves Yes if ISW shows "Claimed Ukrainian Counteroffensives" shading in Crimea. US Recession by end of 2026 (updated Sept 10, 2026): resolves Yes on (1) two consecutive quarters of negative real GDP growth across Q2 2025–Q4 2026 per BEA advance estimates, or (2) NBER recession announcement by time of Q4 2026 advance estimate. Market stays open until that estimate or June 30, 2027 at latest. Crude Oil all-time high by Dec 31: resolves Yes if CME front-month CL futures daily high exceeds $147.27 on any trading day. Source: CME Group website. S&P 500 December range markets (updated Sept 10, 2026): resolution based on Yahoo Finance 1-minute candle data during regular trading hours. Affected HIGH targets: $8,000 / $8,200 / $8,600 / $9,000 / $9,300 / $10,000. LOW targets: $4,500 / $5,200 / $5,800 / $6,000 / $6,200 / $7,000 / $7,200. Leadership markets (updated Mar 18, 2026) for Kim Jong Un, Pedro SΓ‘nchez, and Modi all share the same structure: resolve Yes if the leader ceases to hold their position for any period; a resignation announcement before the end date immediately resolves Yes. Gemini Spark free plan β€” resolved Yes: Google's "Tasks in Gemini" (renamed Gemini Spark) rolled out publicly for free on Nov 17, 2026 in the US, Canada, Japan, South Korea, Australia, and India. Country-limited availability qualifies per market rules. Vitinho soccer markets β€” resolved as canceled: markets for "Vitinho" (1+, 2+, 3+ goals+assists) from the BRA-COR match on Nov 18, 2026 resolve as if the fixture were canceled, as the name "Vitinho" could refer to more than one player and the intended participant cannot be determined. October 2026 CPI/inflation markets β€” fallback resolution: a large batch of Core CPI YoY and Annual Inflation markets will resolve using the most recent prior month's available data. The BLS released its Oct 2026 report on Nov 10, 2026 without 12-month percent changes because Oct 2025 data was unavailable due to the federal government shutdown. Opendoor (OPEN) $1.25 low market β€” resolved No: Pyth's lowest published 1-minute candle Low was 1.255 (displayed as 1.25 due to rounding), but per rules prices are used without rounding; 1.255 is not ≀ $1.25. ⚠️ Bugs & Issues Unsettled corner markets: the CSL match Dalian Yingbo vs. Qingdao West Coast (played Oct 11, 2026) has corner markets still unsettled after ~60 days. Unsettled cricket market: the Yas Lions vs. UAE Bulls Abu Dhabi T10 final market remains open and has not been settled. API bot blocked: user beez5146 reports their bot on the Polymarket US retail API is blocked with the message "Live startup blocked. Operator verification is pending." It is unclear whether additional account approval is required for API trading. ⚑ Disputes Fee tier rollout without wallet merging: market makers, notably blunkoed, complained that the new fee tier structure went live before the ability to merge maker wallets, meaning some wallets already benefit from combined volume while others do not. Users raised concerns about liquidity impact on affected markets. wV formula change β€” possible unintentional bug: user forehead1 flagged that the new bonus structure sets Category_bonus = 1 across all categories, breaking previous proportionality between fees paid and wV earned. Traders on low-fee (0.01) markets now earn the same wV per dollar of notional as those on high-fee (0.05) markets, creating a 5Γ— disparity in fees-per-wV. Whether this was intentional is unconfirmed. Given the latency testing disappointment for NEX and the fee structure complaints, do you think Polymarket's infrastructure upgrades are moving too fast to catch real-world performance issues?
502
UMA DIGEST β€” 21 NOVEMBER 2026 πŸ”₯ News & Announcements Idaho election results page was updated from "OFFICIAL" to "UNOFFICIAL" during the dispute window. 7h0nk3r confirmed that audits were completed Wednesday but results are not yet certified, with certification scheduled for Tuesday. A repeat proposer (linked wallets 0x7F and 0xfd) continued proposing incorrect Idaho results despite being down ~$30k cumulative across multiple wallets, losing every dispute. 7h0nk3r successfully disputed a batch of Idaho markets at $1,250+ reward each. alexchunus suggested he may have contacted the Idaho election department to flag the incorrect "official" label on uncertified results. Georgia election markets were also flagged as all P4 by gum5683, with a similarly questionable proposer submitting them. NC-10 House proposals are in a gray area: enormouz noted VoteHub called the race and the candidate conceded, but armanh01 pointed to precedent that NBC alone is insufficient. AP has not yet called the race. The Opendoor Technologies (OPEN) stock market dispute centers on whether the Pyth data API hit the resolution threshold. fernando5495 clarified the resolution source is Pyth's 1-minute candle "High/Low" data at pythdata.app, and that the chart did show the threshold being hit even though the raw API did not. fernando5495's bot was noted to skip edge cases near thresholds, avoiding the dispute entirely. Users broadly criticized Pyth as a poor resolution source for equity markets. ⚠️ Bugs & Issues whoooo8888 flagged that an extended NBA props bot needs updating based on first-snapshot results for the NOP vs OKC Shai Gilgeous-Alexander assists market. NBA stat trackers were also noted to retroactively update player stats up to two hours after games end (e.g. Jeremiah Fears blocks changed from 1 to 0), creating resolution timing problems. The RSM Classic golf market appeared to show an incorrect leaderboard β€” pihlen_99322 noted Marco Penge had more birdies than the listed leader after two rounds. im.the.wizard suggested there may have been a three-way tie for first that caused confusion. πŸ“‹ Clarifications A Vitinho goals+assists market was clarified 50/50 after a dispute over which "Vitinho" the market referred to β€” Corinthians' Vitinho (player ID 90039681) or Botafogo's Vitinho (player ID 90055175). tenacious_mango_59265 argued the player IDs in the market slug clearly identify the correct player and are consistent across all markets for each team. mebeim and others countered that regular traders cannot reasonably be expected to parse market slugs. The situation was compared to a prior "Fernando" duplicate-name case. Users criticized Polymarket for not including surnames or additional identifying info in market rules, calling it a data quality failure. A French Cup (Coupe de France) match between Charleville-MΓ©ziΓ¨res and Marcq-en-BarΕ“ul was confirmed by the FFF as a 2-0 result for the winning team despite post-match violence by the winning side's assistant coach. The decision was made the following day. im.the.wizard confirmed the market already resolved. A disputed Luka market was flagged by im.the.wizard as incorrect β€” he stated he has archived evidence of Luka passing the threshold. With Pyth's chart data contradicting its raw API output on the OPEN stock market, should resolution sources prioritize what data aggregators display to users over raw API feeds?
222
POLYMARKET DIGEST β€” 20 NOVEMBER 2026 πŸ“‹ Clarifications Polymarket posted bulk rules updates to align Gamma with Polygon (originally updated September 10, 2026) for all Gold (GC) futures December markets: HIGH targets at $5,000, $6,000, $7,000, $8,000, $10,000, $12,000, $15,000 and LOW targets at $2,500, $3,000, $3,500. All use the same methodology: resolution is based solely on the official CME Active Month front-month settlement price (not intraday highs/lows/bids/offers). The Active Month follows CME's delivery cycle (Feb/Apr/Jun/Aug/Oct/Dec) and rolls on the First Position Date. Settlement price is locked at first publication; later corrections are ignored. If no price is published within 72 hours of the final December trading day, all available CME channel data is used. The Strait of Hormuz traffic returns to normal by December 31? market was also updated (originally June 1, 2026): resolves "Yes" if IMF Portwatch publishes a 7-day moving average of transit calls β‰₯ 60 for any date before December 31, 2026. Revisions within the market timeframe are considered; post-December 31 revisions are not. Rules were also clarified for HoneyMad (maddyson) GTA VI completion markets (by Nov 22, Nov 24, Nov 26, Nov 28, Dec 1): resolution source confirmed as authentic official streams, e.g. kick.com/maddyson. ⚠️ Bugs & Issues WebSocket rate limiting (429s from Cloudflare): koo_throw reported unexpected 429 errors starting ~02:20 UTC, triggered by rapid WS reconnections across many markets. habitual_revers confirmed Polymarket is actively tuning rate limits due to abuse, and offered to DM and potentially whitelist legitimate high-volume builders. The /beta/ws/{topic} stream was suggested as an alternative. verrylowsharpe noted that third-party tools like Gravia depend heavily on these WS connections and that a large share of Polymarket users rely on Gravia's UI. ERC-1271 (smart contract wallet) support on relayer-v2: shakotn reported that smart contract wallets (ERC-1271) are rejected with 400 "invalid batch signature" on /submit, despite on-chain tests passing in Foundry. It is unclear whether this is supported or planned. NEX sandbox onboarding delays: Multiple users (edgod3714, stefo9260, dust7829) reported not receiving sandbox credentials or test funds despite PrivateLink connections being accepted. suhailkakar confirmed they are in the second batch, that 10k pUSD and 100 POL were sent, and that Telegram group setup was underway. Infrastructure team was approving IPs and VPC endpoints, with a 20–30 minute wait advised. Order confirmation latency: pinhead3578 reported order confirmation delays of 6–7 seconds at p99 and over 30 seconds in some cases during busy sports windows when quoting resting orders via the US REST API. hammad0383 also observed FAK buy matching times of 0.9–2 seconds during BTC price spikes, with the extra latency occurring in matching rather than the network. πŸ”₯ News & Announcements NEX / new CLOB: habitual_revers confirmed the NEX sandbox is live and soliciting feedback, particularly latency and performance metrics. suhailkakar indicated that "colocation" will be replaced with direct VPC access, eliminating the need for IP whitelisting. hakei confirmed API trading from AWS eu-west-1 (Dublin/Ireland) is currently permitted, though this may change in ~4–6 weeks as the tech stack evolves. V2 migration details confirmed in #v2-migration: gasuajex posted detailed V2 integration notes including new contract addresses on Polygon (chainId 137) β€” ExchangeV3: 0xe3333700cA9d93003F00f0F71f8515005F6c00Aa, PositionManager: 0x006F54F7f9A22e0000CC2AB60031000000ae9fEF. V2 markets require "version":"v2" added to the payload before hashing. Combo orders use EIP-712 version "3", signing only the combo position. The usdcSize field is renamed to usdc_size (snake_case); the v1 /activity route is retired. Combo condition IDs from the Data API are 31 bytes and must be padded with a trailing 00 for on-chain use. Taker tier account merge requests remain unresolved for multiple users (jojopoo333, berth6833, chickenwing_14999), some waiting over two months with no response. Users expressed frustration that the aggregation system may obscure volume tier attainment. If Polymarket's NEX infrastructure switches to direct VPC access in 4-6 weeks, will that finally unlock low-latency trading from regions outside US data centers?
590
UMA DIGEST β€” 20 NOVEMBER 2026 ⚠️ Bugs & Issues A user asked whether the UMA bug bounty program is still available; infinity_d confirmed it is. An AI bot (@1390617444053422161) was flagged for posting ChatGPT API error messages verbatim into dispute threads, polluting multiple active discussions. mebeim requested a mod ban the account. A separate incident involved user ashley51782 accidentally pasting a GPT usage-limit error message as a dispute argument, then posting an AI-generated apology for doing so. Users noted the ongoing problem of non-English speakers and bots using AI slop in disputes, with alexchunus observing it typically harms whichever side it appears on. πŸ“‹ Clarifications NBA gamebook / "first published" rule: A dispute over Yaxel's assists and a separate one over Giannis's rebounds (13 vs. 14) both hinged on the new resolution rule that the first published figure counts, not later corrections. The NBA official gamebook PDF is now the resolution source. im.the.wizard confirmed via independent archive that Giannis had 13 rebounds in the first gamebook; nba.com later updated it to 14. whoooo8888 raised a concern that submitted PDF files could be tampered with (values changed via inspect element), and mebeim acknowledged this is unverifiable beyond metadata checks, but accepted it on the basis of no evidence of tampering. Idaho / election certification: Several Idaho election markets (ID-01, Idaho Senate margin/turnout) were proposed using unofficial results. Per a news article linked by gum5683, Idaho results remain unofficial until certified by the State Board of Canvassers on Nov. 24. 7h0nk3r confirmed audits were scheduled for Wednesday with certification on the 24th. mebeim noted the market rules reference "as published by the state electoral authority responsible for certifying the results," implying certified results are required. alexchunus noted that Minnesota's market similarly settled, with the state legislature page calling them "certified results" but linking the same unofficial numbers β€” enormouz noted Minnesota's 2024 results page still says "unofficial." General consensus was not to dispute these but to let certification stand. Anthropic (ANT) IPO market cap markets: Discussion arose over whether ANT closing market cap markets could be proposed. fernando5495 noted the resolution source (designed for private companies) would not update since ANT is now public. alexchunus argued the Nasdaq real-time quote page showing a daily high is sufficient and falls within the resolution source scope, and that it was proposable as soon as the exchange page showed a final high value. The Anthropic CFO and CCO rang the closing bell at Nasdaq (confirmed via the Nasdaq events page), satisfying "bell-ringing" markets. A concern was raised about proposals using unofficial data; alexchunus said waiting for the OHLC historical chart is unnecessary once the live exchange page finalises. Spurs/Clippers O/U 95.5 (3Q): alexchunus confirmed anything proposed before the :54 mark for the 3rd-quarter total is P4. im.the.wizard disputed the V2 version; V1 had already settled. ⚑ Disputes Idaho ID-01 congressional race: Multiple brackets were proposed using uncertified results before the Nov. 24 certification date. rex416 and others flagged these as P4 (premature). Two Idaho brackets settled before anyone could dispute them. rex416 noted he already had 10 disputes pending and could not act in time; mebeim called disputing them -EV regardless. user magic_of_cinema proposed a bracket incorrectly (wrong side), prompting criticism. Anthropic 3T "No" bracket: A user disputed the Anthropic >$3T closing market cap "No" bracket after it was proposed at ~99.9 cents. gum5683 noted the disputing user held only ~1,900 shares at 86Β’ and did not buy the cheap losing shares β€” widely regarded as a wasted bond. AI/bot conduct in dispute threads: mebeim and abhiz69 formally requested that mods raise the bar for participation in dispute threads, suggesting restrictions to involved parties and actual voters, or KYC. infinity_d (mod) responded that duplicate posts will be deleted on report but stopped short of broader restrictions. Given how much time the community spends verifying PDFs and screenshots in disputes, would you trust a future market that relied entirely on on-chain data instead?
335
POLYMARKET DIGEST β€” 19 NOVEMBER 2026 πŸ”₯ News & Announcements Spain Prime Minister Markets rules were updated on 2026-11-19 to align Gamma with Polygon. Each candidate market resolves to whoever officially assumes PM office by March 31, 2028, 11:59 PM ET, requiring both a congressional vote of confidence and formal swearing-in. Interim/caretaker PMs without a parliamentary confidence vote do not count. Resolves "Other" if no qualifying PM takes office by the deadline. Candidates covered include Carlos Cuerpo, Γ“scar Puente, Juanma Moreno, FΓ©lix BolaΓ±os, Emiliano GarcΓ­a-Page, and others. 10-Year Treasury Yield markets (thresholds at 1.0%–3.9%) resolve "Yes" if the yield drops below the listed threshold on any day between November 11, 2025 and December 31, 2026, per U.S. Treasury Daily Par Yield Curve Rates. "Lecornu out as French PM by December 31, 2026?" resolves "Yes" if SΓ©bastien Lecornu ceases to be PM at any point before the end date; an announced resignation or removal resolves immediately. AK-AL House Seat markets resolve based on the party of the winning candidate per FEC official results or credible reporting consensus, with party determined by ballot affiliation or most recent stated intent to caucus at the time all 2026 House races are called. ⚠️ Bugs & Issues A WebSocket rate limit was newly enforced in production, causing 429/492 errors for developers. The limit is being tuned toward ~300 connections per 5 minutes (similar to Binance). The beta environment has a 10 concurrent connection limit; the production limit is still TBD. The new beta WS reportedly resolves prior "slow consumer" disconnect issues. KPHX (Phoenix Airport) CLI reports are missing precipitation data since November 3, complicating resolution of "Where will it rain today?" markets. Admin ak504111 acknowledged the issue; existing rules default to "No" if no figure is published. Removal of Phoenix as a listed city is under consideration if the data gap continues. Shares purchased on Polymarket sometimes do not appear for several minutes after a transaction (ongoing). A user also reported three deposits from ~20:33 UTC on 2026-11-18 not appearing in their balance despite PolygonScan confirming the transfers succeeded. Additionally, a user reported being unable to connect their X (Twitter) account to Polymarket. Users are warned about scam DMs impersonating Polymarket support and offering wallet recovery or swap dApp links β€” keep all support interactions public. πŸ“‹ Clarifications Multiple developers (pumpty_dumpty, daviestutoring, hft_pro, pietro_44589) are still awaiting NEX sandbox/gamma access; Suhail confirmed setup details were sent by email and separate feedback chats are being organized. The combined maker tier is not yet live but is planned to release before the maker tier goes live. Deprecation of SHA1/book hash from the L2 orderbook feed is under discussion; some developers want it removed as wasteful bytes, while others prefer a sequence ID to be added first. Individuals without a registered business entity can apply for Chainlink API keys; the name field can be filled with any name. Approval turnaround ranges from 24 hours to approximately 20 days. ⚑ Disputes Community frustration was raised over lack of advance notice for fee and rule changes, with users citing only 42 minutes of notice for a prior taker delay change. A maker rebate reduction was also noted, with low-volume makers expecting a consequent drop in volume. With Treasury yields potentially dropping below 1% and geopolitical uncertainty around European leadership, do you think we're heading into a risk-off environment or is the market pricing in a soft landing?
286
UMA DIGEST β€” 19 NOVEMBER 2026 πŸ”₯ News & Announcements New UMA rules took effect on November 16, as noted by elon._._. πŸ“‹ Clarifications Placeholder markets (award finalists / ATP Finals qualifiers): A disputed clarification allowed placeholder brackets to resolve early (before the event concludes) while keeping "Another" open. exquisite_wolf_79651 criticized the logic, saying the market has no "resolve to other" clause. petedrunkseth argued "Other" is a catch-all that covers any unexpected outcome, so the other placeholders can resolve. mebeim described it as a one-off for a specific midterm market and warned against over-reading it. alexchunus agreed the clarification was bad but said it should still be followed. woodmark520 raised a concern that in the ATP Finals, alternates can still replace players during the round-robin stage, but after discussion exquisite_wolf_79651 and simbitch concluded that the market's "named to one of the eight main-draw spots" language aligns with the ATP direct-acceptance definition, so tenacious_mango_59265's proposals were valid. simbitch recommended using the WTA Finals qualifier ruleset as the standard going forward (alternates without an official qualifier spot do not count). Maks Ebonh shots dispute (Finland vs Belarus, UEFA Nations League): mrpicsou proposed Yes based on the UEFA match blog showing 2 blocked shots for Ebonh. elon._._ countered that UEFA's own stats API shows only 1 attempt, and that the live blog is unreliable and frequently revised. im.the.wizard noted a relevant precedent where the official website data took precedence over other sources. The core unresolved question is whether "attempts" in the API equals "shots" in the market rules, and whether the blog or the API constitutes "official match statistics." Gemini model name markets: ice071059 clarified that the market rules explicitly require waiting until midnight ET on the release day, because if multiple models are released on the same calendar date all must be considered, with the alphabetically first qualifying name winning. This means any proposal made before midnight on release day is P4. Only Gemini Oxygen and Gemini Xenon (later in the alphabet than Neon, the released model) were considered safe to propose as No; all other options including placeholders were P4 because a new model with an earlier alphabetical name could still be released before midnight. ⚑ Disputes Gemini model name markets β€” mass P4 dispute: ghostmbao proposed roughly 30+ Gemini model name markets immediately after Gemini Neon was released, without waiting for midnight ET as required by the rules. im.the.wizard, gum5683, scriptlin_55968, rex416., and others disputed approximately 21–30 of these proposals. ghostmbao self-disputed several. rex416. mistakenly disputed Xenon (one of the two valid No proposals), losing that bond. scriptlin_55968 noted that two prior Gemini model markets (Muse, Hydrogen) had also been misresolved the same way but went unchallenged, and stressed those misresolves are not precedent given how explicit the rules are. The disputed proposals were subsequently re-proposed correctly after the window closed. NVDA price market: rickhn noted a market asking whether NVDA would dip to $152 by November 16 β€” the market opened November 17 and Pyth showed a price of $152.01. rickhn flagged it as P4 but was not at a PC to dispute in time; the dispute window was extended but it appears the moment passed. CS2 early proposals: elon._._ raised a concern that certain CS2 match proposals are submitted ~30 seconds before HLTV updates (which has a known delay), suggesting proposers may be using faster data feeds. stradiveri acknowledged proposals are sometimes a few seconds early but said 30 seconds would be easy to prove. No formal dispute was filed. GTA 6 release market: enormouz flagged that someone proposed Yes on a "GTA 6 released by" market even though Rockstar's own website still shows "pre-order" and no official release statement has been made. The market's resolution source is official Rockstar/Take-Two information. Status: unresolved, awaiting further action. Should markets require waiting until an official API or database updates before proposals are accepted, or is the burden on users to know when their data source has the full picture?
295
POLYMARKET DIGEST β€” 18 NOVEMBER 2026 πŸ”₯ News & Announcements Crypto market data feed change (polymike): Starting November 18 at 00:00 UTC, 1-hour and daily crypto Up/Down markets will switch from Binance to Chainlink USD TWAP-60s feeds. The last Binance-based markets are the 6PM ET 1-hour and the November 18 daily. Fee and rebate restructure (suhailkakar), effective November 20 at 11:59 PM UTC: - Taker fees vary by category: 0.01 for crypto (excl. BTC 5m/15m), politics, culture, weather, economics; 0.05 for BTC 5m/15m, sports/combos, finance, tech, elections; 0 for geopolitics. - Maker rebates shift from a daily pool to volume-based tiers (% of taker fees generated): Tier 1 ($200K–$1M): 3%; Tier 2 ($1M–$4M): 7%; Tier 3 ($4M–$10M): 10%; Tier 4 ($10M+): 15%. Below $200K: 0%. Bronze and Silver taker tiers removed; new Onyx taker tier ($75M+) added at 60% rebate. - Volume is now calculated using a weighted formula shares Γ— price Γ— (1 βˆ’ price) instead of dollar notional, which effectively halves volume counted at 50Β’. Community reaction was critical: warrior2538 argued maker rebates are effectively capped at 15% (down from 20%) and requested a delay until CLOB instability is fixed; rhum.eth noted that removing Bronze/Silver tiers means small/medium takers now subsidize fee cuts for larger players; satatardy questioned the incentive structure for high-fee markets. Some users speculate the changes are IPO-preparation to boost short-term revenue. A further concern is that only ~3 days notice was given for a change based on a 30-day rolling volume metric. Spanish general election markets: Polymarket posted rule updates (originally set October 23, 2026) across all "Will [Party X] win the most seats in the 2026 Spanish general election?" markets to align Gamma with Polygon. These cover parties A–Z and several named parties (PP, PSOE, etc.). Key rules: election scheduled for November 29, 2026 (350 seats in Spain's Congress of Deputies); ties broken by total votes then alphabetically; resolves to "Other" if results are not known by November 30, 2027, 11:59 PM ET. Official source: Spanish Ministry of the Interior. Climate markets: The team is looking into adding wind and snow markets; no launch date given. ⚠️ Bugs & Issues - SDK bug (rmalvarezkai): Calling get_order(order_id) immediately after place_limit_order in the Polymarket Python SDK fails with "OpenOrder response did not match expected shape", despite a valid order_id and res.ok = True being returned. - Mobile bug (roastinator_): Some open orders are not visible in the mobile app. - WebSocket instability (query_melon): 30+ 1001 server shutting down disconnects per day on market subscriptions, compared to only a couple per day on Kalshi. - Order posting latency (fab5780): Latency ranging 0.5s–4s reported as too unstable; user stated they are leaving the platform due to this issue. - Orderbook feed: Updates silently drop without sequence numbers; workaround is using best_bid/best_ask fields, though lower levels may remain inconsistent. - CLOB cancellation delays: Up to 150ms reported during US hours. πŸ“‹ Clarifications - V2 fair-value settlement (gasuajex/staff): Fair-value settlement applies only when both version == "v2" and fallbackResolutionType == "lfmp" are true. The fallbackResolutionType can change at any time, including post-launch; unknown values should be treated as unsupported. Fair value is not yet live β€” an announcement will follow. For combo markets, the fair-value leg settles at its settlement price and the product of legs applies normally. - Chainlink API key: A free Chainlink API key will be required after November 30 for the 60s TWAP used in BTC 5m market settlement. Approval typically takes ~2 days (one user reported waiting 20 days). Binance and Pyth spot prices remain keyless but are not the feed used for settlement. - NEX sandbox access: Multiple users (greenshower_, mcxboi, 0xnum, pumpty_dumpty) report pending or unanswered sandbox access requests. - Soccer market cancellations in V2 (vaibhav_241): Asked how cancelled matches with no make-up game will be handled for three-way markets (Team A / Draw / Team B) β€” specifically whether Draw resolves YES or fair-value settlement is used. No answer recorded. Will the maker rebate changes actually drive more volume, or will traders just move to platforms with better incentives?
463
UMA DIGEST β€” 18 NOVEMBER 2026 πŸ”₯ News & Announcements Polymarket updated NBA market resolution rules: proposers must now wait for the official NBA gamebook before resolving player prop markets. im.the.wizard confirmed there is no time limit β€” you must wait however long it takes for the gamebook to be published (unlike soccer, which has a 2-hour fallback). Note that the 2-hour rule does not apply to NBA. Participants are allowed to post one comment per dispute thread as evidence in their own disputes, per woodmark520. πŸ“‹ Clarifications MLS placeholder dispute (active): A first-time proposer proposed a placeholder "No" on an MLS Coach of the Year market after the three finalists were announced. A heated debate broke out over whether placeholders are still P4 (too early) or now resolvable as P1 (No) following the LA-05 election clarification, in which Polymarket explicitly resolved placeholder brackets to P1 when no new candidates could be added. laurleo argues the election clarification was market-specific and does not overturn longstanding placeholder precedent, citing post-clarification MLS disputes that still resolved P4. alexchunus and armanh01 counter that the clarification is general and applies whenever placeholders cannot possibly win β€” the same logic as the finalists being locked in. im.the.wizard stated he will support P1 on this market. enormouz noted that prior MLS placeholder disputes (even after the election clarif) resolved P4. The dispute is considered "shirty" by multiple participants, and alexchunus acknowledged the clarification was surprising and poorly thought out but binding. NBA API glitch: dyanreel raised a concern about a past Polymarket incident where an NBA API glitch temporarily swapped player names in a boxscore, causing an $18K loss. He asked whether, under the new gamebook-based rules, a proposal made against a glitched live boxscore would still be valid since the incorrect stats technically appeared on-screen. No definitive answer was given. NVIDIA guidance market (resolved): Someone proposed P1 (Yes β€” guidance withdrawn/not provided) on an NVIDIA Q4 FY2027 revenue/margin outlook market before the November 17 midnight ET deadline, even though NVIDIA did provide guidance during its earnings call. The market rules require guidance to be officially withdrawn or suspended before the deadline for Yes to resolve. Consensus in chat was solidly P4 (too early / still within window for withdrawal), and the proposal was disputed. im.the.wizard joked about supporting P1 via "election clarification" precedent but was identified as trolling. ⚑ Disputes Placeholder "No" on MLS Coach of the Year: Active. Proposer (a first-time user with ~$11 left in their wallet) proposed placeholder brackets as No after finalists were announced. laurleo disputed, arguing historical precedent makes this P4. alexchunus is backing P1, citing the LA-05 election clarification. im.the.wizard stated support for P1. The community is split and the outcome is being watched closely as a potential precedent-setter for all future placeholder markets. A side discussion broke out about the practice of privately offering payment to proposers in exchange for not disputing. nopsuke called it a shakedown; alexchunus defended it as a rational way to avoid high dispute fees, comparing it to a side bet. No rule change or moderation action was announced on this topic. Does the NBA gamebook rule create a unfair advantage for sophisticated traders who know when to expect publication delays versus casual bettors?
234
POLYMARKET DIGEST β€” 17 NOVEMBER 2026 ⚠️ Bugs & Issues The Bangladesh vs South Africa test match page was broken and not displaying β€” reported to the team by kurwamach. A separate page was also reported as showing nothing. Additionally, tomz flagged a combos bug and is requesting compensation. A data feed issue was identified where order events (specifically last_trade_price) arrive ~21ms later than market data (price_change events) server-side, creating a systematic time-shift that affects backtesting accuracy. Users also reported near-500ms order latencies for takers over the past two days. πŸ“‹ Clarifications Several PGL Major Singapore 2026 markets were retitled for clarity: "1win" is now "THE ONE (1win)", "BetBoom" is now "BB Team (BetBoom)", "PARIVISION" is now "PVISION (PARIVISION)", and "BC.Game" is now "GOSH (BC.Game)". πŸ”₯ News & Announcements PolyBolt users relying on Chainlink API for crypto market prices need to request a free API key before November 30. blunkoed noted the free tier appears valid for 24 months. Developer office hours were held, with jungle_jo noting they voiced community concerns during the session. A question was raised about whether the upcoming Nex exchange (new CLOB) will retain the taker delay β€” no official answer was given. There is broader developer community concern that a more efficient CLOB will squeeze out smaller/retail market makers, with richlb_predict noting that more unpredictable latency currently benefits smaller participants and that profit per share is already lowest on Kalshi. Will the new Nex exchange actually solve the latency and data feed timing issues, or will it just shift the advantage from retail makers to institutional players with better infrastructure?
271
UMA DIGEST β€” 17 NOVEMBER 2026 πŸ”₯ News & Announcements New NBA player prop resolution rules are now live: the resolution source is the earlier of (1) a player crossing the stat threshold plus buffer on the NBA.com live box score, or (2) the first official NBA game book released after the game concludes. Stat corrections made after resolution is triggered will not count. The community has reacted negatively, calling the new rules harder to enforce than the old ones, and noting they appear designed to protect market makers and bonders from stat-correction losses. ⚠️ Bugs & Issues NBA Grizzlies game β€” early proposals flagged: im.the.wizard identified roughly 150 proposals made by a single wallet (0x4a1df...) before the official game book showed "Final" status. A specific example: Jalen Smith was archived at 3 assists two minutes post-game, yet the proposal claimed 4 assists. The core problem is there is no audit trail for NBA game book updates, creating a "battle of the archives" where both sides can present different values. Deportivo Madryn vs. Deportivo Maipu (abandoned at 81'): A match was abandoned after a fan threw a firecracker injuring the goalkeeper. The O3.5 over market already slipped through and resolved Yes. im.the.wizard argues remaining over markets should be P4 pending an AFA Disciplinary Tribunal ruling, citing the abandonment clause and a prior UMA clarification. Others (exquisite_wolf_79651, im.hydro) counter that goals scored before abandonment definitively resolved the overs. If no league ruling arrives by Nov 17 11:59 PM ET, markets resolve to 50-50 per abandonment rules. Esports "DIVE" count market: A user (solb0i) alleged a proposal of 7 dives was wrong; 0xsoda provided VOD timestamps suggesting ~5–6. The market expired without a dispute because no participant had the $750 bond capital required. ITF Tennis market: A proposer resolved an ITF P4 match at 6-3 without verifying full match status. ITF qualifying matches use a super tiebreaker in the final set, so early proposals are risky. NC-07 House seat market: A proposal was made without AP/NBC/FOX/Kalshi calls; only DDHQ and VoteHub had called it, alongside a concession from the Democratic candidate Hardy. The community treated the concession as sufficient. No formal dispute was filed. πŸ“‹ Clarifications The $750 minimum dispute bond remains a systemic barrier β€” multiple users could not dispute clearly incorrect proposals due to lack of capital, which some describe as a structural vulnerability open to exploitation. im.the.wizard now archives every proposal he makes as protection, given the new "battle of the archives" dynamic under the NBA rules. Regarding Kalshi vs. Polymarket: Kalshi Sports reportedly has far greater liquidity and volume; Polymarket's dominance in esports is attributed to match-fixing teams operating freely there, something considered unlikely on Kalshi due to KYC requirements and CFTC referral risks. ⚑ Disputes Several serious allegations were raised in the community: - Selective refunds: im.the.wizard alleges UMA issued refunds to a specific user ("flame") on crypto proposals while denying equivalent recourse to others, and that rules were changed on him without remedy. - Corruption allegations: savagefin1252 and alexchunus alleged that some UMA/Polymarket voters accept payment for favorable clarifications, and that suspicious trades occur around clarification announcements. - Vote concentration: savagefin1252 claims the Polymarket team or related wallets control roughly 80% of UMA votes, effectively determining outcomes. - UMA fee incentives: Users allege UMA collects 80% fees from both sides of a dispute, incentivizing the system to encourage disputes over clean resolutions. - Sports staff turnover: Every Polymarket sports-related employee allegedly leaves within two months; bosaurum was cited as a recent example fired unexpectedly. Do you think the $750 minimum dispute bond is doing its job of protecting markets, or is it just creating a playground for people with capital to game the system?
1
1
944
POLYMARKET DIGEST β€” 16 NOVEMBER 2026 ⚠️ Bugs & Issues User WebSocket (WS) delays have been ongoing for about a week. berth6833 reported that matched order messages are arriving 1000ms+ after the match occurs, and cancel confirmations arrive faster than MATCHED messages. The issue affects both regular and beta WS endpoints. habitual_revers acknowledged the report and said the team will investigate. berth6833 also noted recurring latency spikes. The matching engine is also showing poor performance: boface reported that 48% of orders are answered after 450ms. pythonnnnnnnnnnnnnnnnnnnnnnn mentioned that CLOB v2 is coming in a few weeks. Several crypto taker strategy operators (magmaalpha_96960, .alexzor) reported degraded profitability over the past week. magmaalpha_96960 noted that makers appear to be cancelling ~50ms faster than before. sumsof reported receiving fill messages on Poly US WS over an hour late. πŸ”₯ News & Announcements habitual_revers confirmed that the new NEX matching engine uses a sequencer/FIFO architecture β€” a standard trading engine design β€” with no per-wallet queue. The team is currently onboarding market makers to the NEX sandbox and will run liquidity bots to seed it, also collecting bug reports pre-launch. habitual_revers clarified there are currently zero plans to move CLOB server location, adding it would be "a long while" before that is even considered. Users blunkoed and floaty_boi requested at least 6 months' notice if a move ever happens. flashmltply asked whether the 150ms crypto taker delay (introduced in the 04.09.2026 changelog) will carry over to NEX after launch β€” no answer was provided. πŸ“‹ Clarifications Regarding weather market liquidity concerns: energetic_raccoon_69364 explained that unlike crypto markets, weather market-making risk can only be hedged by trading other temperature buckets in the same city/day. LP rewards exist to compensate makers for hedge spread costs, allowing them to maintain two-sided quotes. With reduced rewards, makers have less incentive, leading to wider spreads (30c+). A debate followed about whether the dominant trader "HighTempTation" has advance knowledge of temperature readings or is simply predicting β€” most participants concluded he is a skilled predictor targeting specific cities, not someone with insider data access. If the 150ms crypto taker delay doesn't get addressed before NEX launches, will traders just migrate to competing platforms?
1
8
1,300
UMA DIGEST β€” 16 NOVEMBER 2026 ⚠️ Bugs & Issues A bot/proposer identified as 0x21c1 submitted early NBA Clippers Q1 proposals before the quarter ended (quarter ended at 06:50). These were mass disputed and the proposer is now negative lifetime. im.the.wizard noted they were "lucky this was only Q1." A separate proposer (0xAcE / same wallet) also mis-proposed a soccer match by confusing it with a future game that had the same teams, same kickoff time, and same date but a different league β€” their data feed failed to handle the schedule swap, resulting in a ~$32k loss. Ongoing discussion flagged Polymarket's automated market creation as producing duplicate markets with identical player names (e.g. two players named the same in a goalscorer market), wrong team/tournament names, and mismatched match formats in esports. mebeim argued a simple de-duplication check in the deploy script would prevent most of these issues, and attributed the quality drop to recent layoffs. ⚑ Disputes French Cup (Coupe de France) β€” Olympique Charleville-Prix / O. Marcquois (~90 markets) A batch of roughly 90 markets was proposed 2-0 in favour of Olympique Prix, but the match was suspended late in the game (reportedly with ~2 minutes remaining) after racist comments from staff and crowd disturbances caused the referee to end it early. The official French Football Federation (FFF) site shows the match status as "stopped/under review," while Flashscore and Sofascore (initially inconsistent) show "FT 2-0." A Facebook post from the losing club and a local news article confirm the match was abandoned and referred to FFF disciplinary review; a French federation article explicitly describes the match as "abandoned by the organiser." Pro-dispute position (alexchunus): The primary resolution source (FFF site) says the match is under review/suspended. Per precedent, proposals must wait for the governing body's official decision. The winning team's staff may be sanctioned and the result could be overturned or forfeited. The halftime result cannot be considered "definitively determined" under these circumstances. Anti-dispute position (laurleo, abhiz69, lonfus): The FFF site did not show "suspended" at the time of proposal (abhiz69 claims timestamps prove this). Flashscore (a valid fallback) shows "Finished 2-0." Both teams publicly celebrated. Risking ~$500 bond per market for $125 return on ~90 near-zero-volume markets is -EV, especially with conflicting evidence. laurleo argues the fallback source applies when the primary does not publish final statistics. The moneyline market has ~$2–3k in volume (some bought after the dispute chat). No UMA extension was granted despite requests. The dispute remains unresolved pending FFF's official decision. im.the.wizard flagged the dispute odds require ~80% confidence to be +EV and declined to participate. alexchunus disputed at least the moneyline citing share exposure. Marko Jankovic assist market (UEFA Nations League) Both YES and NO were disputed on a "Will Marko Jankovic record 1+ assists?" market. The match feed initially credited him an assist during the game, but UEFA's corrected post-match stats API (matchstats.uefa.com) shows 0 assists for the match (his only assist this season came in the Latvia game). im.the.wizard confirmed the corrected value is 0, making YES incorrect. The 117.5-point total proposals were flagged as ~9 minutes early (P4). πŸ“‹ Clarifications im.the.wizard clarified that for UEFA player statistics (including assists), the corrected post-match value from UEFA's stats API takes precedence over the in-game match timeline log, since corrections are applied after the final whistle. With conflicting data feeds and tight -EV odds, should prediction markets require governing body confirmation before resolving mass-disputed sports outcomes, or does that slow down price discovery too much?
271
POLYMARKET DIGEST β€” 15 NOVEMBER 2026 ⚠️ Bugs & Issues Multiple users are reporting significant WebSocket latency and performance issues, particularly during NBA/NCAAF/NFL hours. berth6833 noted the degradation has lasted about a week and attributes it to a new engine running in "shadow mode." artex6666 confirmed their trading bot lost nearly 3 hours of trading time due to recurring timeouts exceeding 4 seconds. hft_pro reported maker-side P50 at 38ms, P90 at 80ms, and P99 at 390ms, while taker-side performance is described as worse. berth6833 described order cancellations as practically impossible during peak load. error403kabaaperror also reported an 8–10 second delay when fetching prices via WebSocket bid/ask requests in the browser over the past few days. Additional bugs reported: - Broken page rendering on a cricket market (clericode): content loads briefly then disappears. - Autoplay livestream on the esports tournament page consuming bandwidth unexpectedly (jovial_dove_71212). - Duplicate entries in the holders list on the "clavicular, yarmulke" market (jimmybuzzard). πŸ“‹ Clarifications A discussion arose in the climate channel regarding METAR settlement rules. suooshi questioned whether Polymarket uses six-hour METAR values (derived from minute-level ASOS data) or standard hourly METAR snapshots for settlement, noting these can conflict. They argued that in a recent KMIA ruling, the six-hour standard appeared to have been applied, contradicting the stated rules. maskache07 countered that in the KMIA case the six-hour minimum was physically impossible (higher than the hourly METAR reading), implying the hourly METAR was the correct basis. The discussion concluded without an official clarification from Polymarket staff. Separately, ruperte asked in #v2-migration whether tennis/esports walkovers (pre-game resignations) would resolve as cancellations or at fair value. berth6833 replied simply: "fair value." If you're a high-frequency trader on Polymarket right now, are the latency spikes making you consider moving volume elsewhere, or are you waiting for the new engine to stabilize?
1
1
2
1,249
UMA DIGEST β€” 15 NOVEMBER 2026 ⚑ Disputes Dota 2 / EPC Match Validity An ongoing dispute concerns whether a Dota 2 EPC match was considered "live." aryanmale argues the game went live because all players connected, the draft completed, and the server started β€” referencing EPC rules 4.4.1 (pausing without valid reason) and 4.8 (Lobby Check-In). The opposing side cites precedent (the Exodia case) where one player never connected, preventing the game from reaching live status. aryanmale also points out that a Telegram post initially declared Potn9ki the winner 2:0 before being deleted, undermining the credibility of official Telegram communications used as counter-evidence. Soccer Markets Disputed as P4 Several soccer markets were disputed and pushed to P4. im.the.wizard confirmed a fulltime timestamp via Firestore as evidence for at least one soccer market resolution. abhiz69 noted that the halftime proposal clause has been removed in the new ruleset, making P4 outcomes more likely for halftime-related proposals. A linked exact-score market was flagged as potentially incorrect since the score had not yet occurred at proposal time. NCAA Football β€” Erroneous Early Proposal im.the.wizard mistakenly proposed NFL/NCAA markets as settled while the game was still in Q4, reportedly because ncaa.com displayed the game as "finished" prematurely. Three more touchdowns were scored after the proposal, with the final score reaching 51–16. lonfus and others confirmed the NCAA site is known to glitch and report incorrect statuses. The erroneous proposals were disputed. Negrisk Market β€” All Options Proposed as "No" im.the.wizard acknowledged making a rookie mistake by proposing all three options as "No" in a negrisk market, which was subsequently disputed. Carlos Eduardo Ambiguity (Brazilian Football Player Props) im.the.wizard flagged a potential P3 (ambiguous) dispute on a Mirassol player prop market involving a player named "Carlos Eduardo." Investigation revealed that Mirassol's roster has three players who could be referred to as "Carlos Eduardo" β€” two field players (numbers 96 and 33, both with Carlos Eduardo in their full names) and a goalkeeper. im.the.wizard argued this meets the UMA rule for ambiguity: "could reasonably refer to more than one participant such that the intended team or player cannot be determined." tenacious_mango_59265 countered that one is universally referred to as "Eduardo" and common sense applies. alexchunus and im.the.wizard agreed the risk/reward is unattractive at $125 out of $500, but im.the.wizard suggested RL should either dispute or request a clarification. πŸ“‹ Clarifications abhiz69 confirmed that the halftime proposal clause has been removed in the latest UMA ruleset, meaning proposals made after halftime but before full-time are now more cleanly subject to cancellation rules, generally favoring P4 outcomes. If a player can go by multiple names on their team's roster, should markets default to the most commonly used nickname or require explicit clarification before resolution?
233
POLYMARKET DIGEST β€” 14 NOVEMBER 2026 πŸ”₯ News & Announcements Nex Exchange Launch: Polymarket announced Nex, a new exchange rebuilt in Rust, with significant internal performance improvements: - Place order: 1.55s β†’ 3.76ms (412x faster) - Cancel orders: 627.92ms β†’ 3.20ms (196x faster) - WebSocket writes: up to ~889x faster - Scale: 10k β†’ 100k commands/s Target go-live is mid-December after 4 weeks of shadow mode. Sandbox access is available; follow-ups on sandbox requests will go out by Monday. KYC is required for PrivateLink/AWS private link access to Nex. Postponement Rule Update (Sports Markets): Under Polymarket v2 (rolling out December 1), canceled/abandoned/delayed games will resolve at fair value based on Polymarket trades and other relevant information, replacing the previous 50-50/draw resolution. Existing markets settle under their original rules; only new markets post-cutover use fair value. WTI Crude Oil Markets β€” Rules Update: Rules updated for a large batch of WTI high/low price markets (weekly and monthly variants, $30–$160 range). Effective November 9, 2026, the Pyth market session for the Active Month ICE Futures Europe WTI Crude Futures contract closes at 10:00 PM London time (5:00 PM ET) each trading day. Only prices achieved during the Pyth session count toward resolution; ICE Futures Europe daily high/low will only be used if Pyth data is unavailable due to an actual outage. Liquidity Rewards Reduction: Polymarket reduced liquidity rewards on temperature markets. The total reward pool has shrunk, with rewards reallocated β€” popular cities receive more while others (e.g., Denver) have had rewards removed entirely. Weekly Office Hours: Scheduled private voice sessions for user feedback are available. ⚠️ Bugs & Issues - Share amount display bug: A user reports that 2x the actual share amount is being shown in the UI. - Multi-outcome market odds display bug: Searching for "Morocco" shows opponents' odds rather than Morocco's; user suggests showing the favorite's odds for sports markets. - CLOB latency: Users reported high order execution latency (500ms–1s+), difficulty canceling orders, and apparent order book lag. The current CLOB is confirmed not to be true FIFO. Taker delay reduction from 150ms to ~20–25ms is planned once Nex goes live. - Taker rebates: A user reported not receiving Bronze-level taker rebates β€” unresolved. - Comments API: Reported to always return an empty array for user comments β€” possibly requires authentication, unresolved. - Chainlink API key requests: Users waiting on responses with no timeline given. πŸ“‹ Clarifications - API docs β€” minOrderSize: The parameter is ambiguous in Polymarket docs β€” described as USDC-denominated in some places but used as share count in limit orders; market orders have a separate $1 global minimum. A docs update has been requested. - V2 production slugs: Confirmed that production V2 sports market slugs will follow globally unique slug conventions; the polyv2- prefix is only used in current shadow/mirror markets. ⚑ Disputes Order book spoofing allegation (Hong Kong temperature markets): User goldontheceiling flagged potential spoofing β€” allegedly ~$24k in fake liquidity posted to farm ~$27/day in rewards, with orders reportedly not fillable. Polymarket staff acknowledged the flag. Other users disputed the claim, asserting the orders are in fact fillable. Will Nex's 412x speed improvement for placing orders actually move the needle on market quality, or does latency matter less than you'd think once you're under a few seconds?
1
7
1,287
UMA DIGEST β€” 14 NOVEMBER 2026 πŸ”₯ News & Announcements The potn9ki vs. oveA (EPC League, Dota 2) dispute remains the dominant ongoing case. Game 2 lasted roughly 8 minutes before oveA's players wrote "ff" and "go throne" in lobby chat at -1:28 (before the game horn), with one player unable to connect at all. Opponents then ran down and destroyed the Ancient per admin instruction. Dotabuff (the designated resolution source) records match ID 9088266980 as a completed game won by potn9ki/Mena Warriors. The market has been extended pending esports admin review; infinity_d confirmed the team is taking a closer look. A separate dispute was also raised regarding an Iran attack on Oman (Didamar Island). User wang02272 argues the market should resolve P4 (ambiguous/unresolvable) rather than Yes, noting that while Reuters confirms an Iranian attack, no credible source identifies the weapon type, and market rules allegedly exclude naval gunfire and artillery. A precedent from Saudi Arabia β†’ Yemen resolutions (Sept 20 & 26, resolved No despite military claims) was cited in support. --- ⚠️ Bugs & Issues Fresh wallet accounts were identified as top holders on the oveA side shortly after the dispute was filed, raising strong suspicions of insider trading and market manipulation by the EPC tournament organizer β€” a pattern also observed the previous day, when insiders were seen buying shares at 7–10c hours before a DQ announcement. Community members have formally requested that EPC league be banned from Polymarket. User syrupfund sent an email to Polymarket requesting clearer rules. A recent Polymarket esports team member departure (~2–3 days prior) and the silence of staff member "Seal" have added to concerns about oversight continuity. --- πŸ“‹ Clarifications The core legal question is whether Game 2 counts as completed (resolving Yes/No on the winner) or as a 50-50, given the pre-horn forfeit. Two competing precedents are in play: - Rottweilas vs. Mena Warriors: team conceded at 26:05 β†’ recorded as a completed win, not a forfeit. - EXODIA vs. Yawara (CS2): PM ruled that a match not contested by an eligible lineup is treated as not having started, overriding the data provider's record. The previous day's clarification stated that a team "conceding" (Ancient destroyed) constitutes a completed game. Critics note the wording is vague and that the clarification was issued when the market was already 50-50. A conflict-of-sources clause was also cited: if resolution sources conflict, the market stays open up to 3 full calendar days. --- ⚑ Disputes Community frustration is running high over inconsistent treatment of the TO's credibility β€” EPC statements have been accepted or rejected depending on which side they benefited. A broader concern was raised that vague forfeit rules create a structural exploit: players or teams can deliberately leave early to trigger 50-50 outcomes and manipulate markets. The community broadly recommends against bonding or proposing esports markets given the unresolvable ambiguity in current Polymarket rules. Which precedent should actually control hereβ€”treating a pre-horn forfeit as a completed win like Rottweilas, or as a match that never started like EXODIA, and why does Polymarket keep flip-flopping on which one applies?
1
1
824
POLYMARKET DIGEST β€” 13 NOVEMBER 2026 πŸ”₯ News & Announcements Crypto Up/Down markets switching pricing source: polymike announced that starting November 18, 1-hour and daily crypto Up/Down markets will migrate from Binance to Chainlink for pricing and resolution, aligning with existing 5-minute, 15-minute, and 4-hour markets. Traders should update systems to use Chainlink USD TWAP-60s feeds. A free Chainlink API key can be requested at institutional.polymarket.com. Exact cutover time and new market links will be shared ahead of migration. Earthquake daily markets discontinued: ak504111 confirmed that daily earthquake markets are being stopped due to low volume. Weekly and monthly earthquake markets remain active and daily markets may return if demand grows. ⚠️ Bugs & Issues Profile views not updating: balthazar_826 reported that profile view counts have stopped updating. API orders blocked from AWS eu-west-1: mahojakotyc0213 reported that since November 11, API orders from an older Safe wallet (signatureType 2) receive a "maker address not allowed, please use the deposit wallet flow" error when sent from AWS eu-west-1, while the same orders succeed from a home connection and via the website. Elevated taker latency: Multiple developers report significantly worsened API latency, particularly from AWS Dublin. P50 taker latency has risen from ~300 ms previously to 500 ms+, with P90 exceeding 2000 ms for some users. blunkoed noted packets route over Cloudflare's network, not AWS directly. hft_pro shared current maker-side figures of P50 32 ms / P90 52 ms / P99 239 ms using Ireland and London AWS instances. πŸ“‹ Clarifications Dota 2 – Rottweilas Latam vs Mena Warriors Game 2: Polymarket confirmed the market resolves based on Dotabuff, which records the match as completed and won by Mena Warriors. The subsequent Esports Pro Circuit (EPC) announcement regarding match-fixing allegations will not be considered for resolution. The result stands as-is and will not be voided 50/50. Russia-Ukraine senior diplomatic meeting markets – rules update: The qualifying representative lists for all Russia-Ukraine diplomatic meeting markets (covering deadlines of November 30 and December 31, 2026, and December 31, 2027, across all venue variants) have been clarified. For Russia: Putin, Lavrov, Medinsky, Kostyukov, Dmitriev, Ryabkov, Galuzin, Zhigarlovsky. For Ukraine: Zelensky, Umerov, Budanov, Kyslytsya, Arakhamia, Hnatov, Skibitsky. A qualifying representative from each side must participate. V2 migration – combo and event structure clarifications: gasuajex provided detailed answers on V2 market structure: production V2 events will carry gameId; mixed V1/V2 events are not permitted (version is fixed at creation); mixed V1/V2 combo legs are supported at the contract level but native V2 sports markets are not yet combo-enabled; valid_until on RFQ requests is ignored β€” timing is server-controlled (~400 ms quote collection, 10 s acceptance window). ⚑ Disputes Dota 2 Game 2 – match-fixing resolution dispute: doomsday77 urgently requested a review of the "Rottweilas Latam vs Mena Warriors – Game 2 Winner" market, arguing that EPC's official match-fixing announcement should override Dotabuff's recorded result, consistent with precedent from previous esports match-fixing cases resolved 50/50 on Polymarket. Polymarket rejected the request, stating Dotabuff is the designated resolution source and the EPC announcement does not qualify. The UMA dispute window was flagged as expiring imminently. Should esports markets stick to pure statistical sources like Dotabuff, or should allegations of match-fixing override the recorded result like they did in past cases?
1
961
UMA DIGEST β€” 13 NOVEMBER 2026 πŸ”₯ News & Announcements The EPC (Esports Pro Circuit) tournament organizer has been flagged as potentially corrupt and archived/blacklisted by Polymarket after they issued a post-match disqualification of Rottweilas Latam specifically for map 2 of a match that had already been disputed and clarified. The DQ was seen as suspicious since the team had already been eliminated with no further matches to play, making the DQ pointless in terms of standings β€” leading many to suspect it was designed purely to flip the market outcome. ⚠️ Bugs & Issues A Brazilian football market involving two players both named "Fernando" (from opposing teams Red Bull Bragantino and Chapecoense) caused confusion after an auto-dispute bot flagged the proposals. eip15942 clarified the two players can be distinguished via numerical identifiers in the market URLs: ID 90054911 = Fernando dos Santos Pedro (Bragantino No. 11), ID 90101686 = Fernando Augusto Pereira Bueno Junior (Chapecoense No. 12). A "No" proposal was submitted for the AWS Service Disrupted by October 31 market despite an ongoing AWS incident classified as "Disrupted" appearing in September. boroda205 raised the dispute, arguing the rules do not require the incident to begin after market creation and that no confirmed recovery was documented before October 31. Others countered that the incident predates market creation (started in April). boroda205 noted they lacked funds to formally dispute it. πŸ“‹ Clarifications Polymarket issued a clarification resolving the contested Dota 2 EPC match (Mena Warriors vs Rottweilas Latam, Game 2) as P1 (Mena Warriors win), explicitly stating: "The announcement since made by Esports Pro Circuit will not qualify towards this market's resolution." This overrides the post-match DQ/void issued by EPC. This contradicts prior precedents (HellNight-, CTRL Esports, EXODIA cases) where post-match DQs with voided results led to P3 (50/50) resolutions. wang02272 had argued extensively for P3 citing those three prior cases in the same tournament/organizer. stradiveri and alexchunus noted the inconsistency, with stradiveri remarking Polymarket "flips a coin and writes an extremely vague explanation." A key structural complaint raised by multiple users: Polymarket's current rules allow post-match organizer decisions to change outcomes, but only inconsistently. mebeim, nopsuke, and stradiveri argued that any result-changing event after a match ends should be ignored outright, calling the current approach "SchrΓΆdinger's markets." nopsuke suggested a simple rule fix: "changes after the map/event finished do not affect the outcome." ⚑ Disputes Dota 2 EPC – Mena Warriors vs Rottweilas Latam (Game 2) β€” RESOLVED P1 The core dispute involved whether a post-match DQ and result void by EPC should flip the market to P3 (50/50), after a previous Polymarket clarification had already confirmed P1 (Mena win) based on the in-game "GG" being treated as a completed match. P1 side (simbitch, scriptlin_55968, 7h0nk3r): The prior clarification is binding. Overturning it would undermine the entire clarification system. The EPC DQ is transparently manipulative β€” issued only for map 2, for a team already eliminated, with no effect on standings. P3 side (wang02272, nopsuke): Three prior precedents in similar markets β€” including one in the same EPC tournament 10 days earlier β€” resolved voided results as P3. Yesterday's clarification only addressed the GG/forfeit question, not post-match voids. Consistency requires P3. wang02272 also called for investigation of suspicious top holders (fresh accounts trading only this market). Manipulation concerns: abhiz69, im.hydro, wzht, and others expressed near-certainty that the EPC DQ was coordinated with market insiders who held large P3 positions. nik4456 noted that one user appeared to know the DQ announcement was coming before it was public. Polymarket resolved P1, declining to apply the post-match void. The EPC organizer was archived/blacklisted. Clarification bonders on P3 side lost their bonds. Given the Polymarket Dota 2 ruling that contradicts three prior precedents in the same tournament, should prediction markets prioritize consistency with past decisions or allow case-by-case judgment calls?
2
408