Request a Review
Case WorkMSR Decode

Prediction markets, explained

MSR Decode traces prediction-market trades, payouts and profit claims. These reports explain what happened, how the money moved and what the evidence establishes.

  1. Polymarket · Profit and lossCASE FILE 15 · Published 2026-10-07 · 9 min read

    When Polymarket leaderboard profit diverged from trading profit

    Finding
    Five cases from May through September 2026 show payouts, historical values and unsupported credits entering public profit rankings. The evidence supports distinct accounting problems and leaves some ranking values unexplained.
    Five historical casesMay–September 2026 · distinct accounting problems
    Profit and lossMarket dataCash flows
    Read the case →
  2. Polymarket · Trading patternsCASE FILE 14 · Published 2026-10-05 · 8 min read

    Paired Polymarket wallets traded opposite outcomes and moved through markets together

    Finding
    In a review of 97 selected XFM profiles, 40 of 47 complete pairs entered the same markets in the same order. Their opposite positions reveal a repeatable trading routine, while its economic purpose remains unresolved.
    40 of 47 pairssame ordered market-entry sequence in the selected capture
    Trading patternsPaired walletsEvidence
    Read the case →
  3. Polymarket · Trading and SettlementCASE FILE 13 · Published 2026-09-27 · 3 min read

    An account named in a Polymarket payout proposal bought more shares 14 seconds later

    Finding
    An account named in a Bitcoin market’s payout proposal bought more No shares 14 seconds later. Its 61,713-pUSD surplus came from its full trading history in that market.
    14 secondsproposal to next No purchase
    Trading patternsSettlement rulesCash flows
    Read the case →
  4. Trading records · Cash OutcomesCASE FILE 12 · Published 2026-09-23 · 4 min read

    Two wallets traded $388,000 with each other and lost $71 combined

    Finding
    Their trades supplied 99.75% of reconstructed dollar volume during a 47-hour Los Angeles election-market episode. Two earlier markets showed a similar pattern in a separate three-wallet group.
    99.75%reconstructed dollar volume · 47-hour episode
    Trading patternsWallet resultsEvidence
    Read the case →
  5. Market Data · Volume MeasurementKALSHI SERIES · PART II · Published 2026-09-22 · 4 min read

    Two ways to count Kalshi’s growth

    Finding
    Annualized volume grew about 238% at contracts’ $1 face value and 170% at their trade prices. Both measures rose; they count different dollar amounts.
    238% / 170%annualized growth · face value / trade prices
    Market dataVolume measurementProduct mix
    Read Part II →
  6. Market Data · Trading recordsKALSHI SERIES · PART I · Published 2026-09-22 · 4 min read

    Trades near $5,500 dominated Kalshi’s ether sample

    Finding
    A narrow dollar range supplied 59% of ether trading value in a two-week sample. Further tests did not support suspicions of self-trading or common control.
    59.0%ether trading value · September 5–18 sample
    Trading patternsMarket dataEvidence
    Read Part I →
  7. Market Reconstruction · SettlementCASE FILE 11 · Published 2026-09-20 · 3 min read

    A wallet received the winning payout in the reportedly removed Drake market

    Finding
    The Journal reported Polymarket’s removal of the ICEMAN market. Separately, MSR traced a wallet’s 5,024-pUSD purchase cost and 7,828-pUSD payout.
    +2,804 pUSDone-position surplus · platform token units
    Wallet resultsSettlement rulesEvidence
    Read the case →
  8. Contract Terms · SettlementCASE FILE 10 · Published 2026-09-20 · 4 min read

    Kalshi changed payout terms after 183 election markets had begun trading

    Finding
    New rules let qualifying bets settle on news-source agreement before official certification. The review found changes to timing and terms, without a wrong winner or payout reversal.
    183 marketstraded before the October amendment
    Contract rulesSettlementMarket data
    Read the case →
  9. Market Data · Trading records CASE FILE 09 · Published 2026-09-17 · 5 min read

    Repeated trading dominated 928 Polymarket markets

    Finding
    A screen of 1,351,870 markets found $506.4 million in qualifying trades inside the flagged group. A small share of markets held a much larger concentration of dollars.
    928 of 1,351,870markets crossing the 50% screen threshold
    Trading patterns Market data Evidence
    Read the case →
  10. Payout rules · Cross-Venue CASE FILE 08 · Published 2026-09-10 · 4 min read

    The same unplayed tennis match paid different amounts

    Finding
    Polymarket’s international markets paid 50 cents per side after two walkovers. Its U.S. markets paid 33/67 cents and 26/74 cents under different contract terms.
    2 verified matchesdifferent payouts across the two platforms
    Settlement rules Cross-venue Evidence
    Read the case →
  11. Payout rules · Contract Language CASE FILE 07 · Published 2026-09-10 · 4 min read

    Most trading continued after a tennis withdrawal report

    Finding
    About $103,000, or two-thirds of the market’s lifetime dollar volume, traded after MSR’s stated news cutoff. The final trade alone did not explain the payout.
    66.7%lifetime dollar volume after the stated cutoff
    Settlement rules Trading records Evidence
    Read the case →
  12. Trading records · Market Data CASE FILE 06 · Published 2026-08-19, updated 2026-08-27 · 5 min read

    Two wallets drove $117,000 in cartel-arrest trading volume

    Finding
    Their roughly $172 combined loss concealed much larger individual gains and losses. Separate checks found no trading in two specified pre-announcement windows.
    $117,153.35displayed volume · original two-contract pair
    Trading patterns Wallet results Market data
    Read the case →
  13. Market Data Integrity · Profit and loss CASE FILE 05 · Published 2026-07-28 · 5 min read

    Four promoted weather-trading wallets lost $2.35 million combined

    Finding
    Three selected wallets were positive, but one $2.42 million loss outweighed their gains. A Hong Kong monthly profit claim checked out; the largest screenshot remains unreconciled.
    −$2.35 millioncombined reconstructed result · four selected wallets
    Wallet results Market data
    Read the case →
  14. Market Data · API Integrity CASE FILE 04 · Method Report · Published 2026-07-25 · 4 min read

    Missing purchases created a false Polymarket anomaly

    Finding
    A default feed showed one of a wallet’s 80 purchases. Recovering ordinary trades explained its payout and all 150 deliberately selected follow-up flags.
    1 → 80 purchasesdefault view → maker-inclusive view · example market
    Market data Data feeds
    Read the case →
  15. Trading records · Profit and loss CASE FILE 03 · Published 2026-07-24 · 4 min read

    The $917,000 BNB profit held up; the 23-win claim did not

    Finding
    One wallet made $917,311 before rebates across 24 five-minute markets. The cash ledger showed gains in 21 markets, and later rebates raised the total to $921,179.
    21 of 24markets with positive net cash results
    Wallet results Trading patterns
    Read the case →
  16. Profit and loss CASE FILE 02 · Sample decode report · Published 2026-07-11 · 3 min read

    Argentina won later, but a bet against its 90-minute win made $1.22 million

    Finding
    The match was tied at the contract’s deadline. The wallet sold every No share as prices rose, then received rewards and rebates attributed to the position.
    $1.22 millionfee-inclusive result · one closed position
    Wallet results Settlement rules
    Read the case →

  17. Archived case fileSuperseded research — evidence and corrections retained

    Trading records · Historical CASE FILE 01 · Archived — superseded research · 4 min read

    Archived: missing data undermined MSR’s wallet rankings

    Finding
    A later retrieval found 2.9 million fills for one wallet, compared with roughly 14,000 first reported. The rankings were withdrawn; a few specific purchase receipts remain checkable.
    Archivedsuperseded rankings · correction record retained
    Archive
    Read the case →

Ask MSR to investigate

Tell us which wallet, market or claim you want checked. MSR accepts commissioned investigations using the evidence standards shown in these reports.

Request a Review

← Back to Research