🔍 Read the full analysis: AI Puts The Viral “100X Trade” Strategy To The Test In Week Four on ThorstenMeyerAI.com
TL;DR
Thorsten Meyer AI reports that a two-day simulated test of a viral Polymarket strategy across four crypto markets was net-negative. The paired-bid version filled both sides in only 3 of 9,486 attempts, while a post-close version had four settled fills and all four lost.
Thorsten Meyer AI says a two-day paper-trading test of a viral Polymarket “100x trade” strategy lost money across four crypto markets, with its core paired-bid mechanism succeeding just 3 times in 9,486 attempts. The report says the test used simulated funds across approximately 13,000 windows, not real trades, and found that both versions tested were net-negative.
The strategy was reconstructed from a YouTube video showing a trader turning $1 into $50 on a five-minute Bitcoin Up/Down market, with other examples described in the video as 48x and 100x returns. The source says the transactions could be checked on PolygonScan. The YouTuber reportedly acknowledged that the trade was a rare event and that the video did not establish whether it could be repeated.
In the paired-bid version, the bot posted bids of 2 cents on both Up and Down. If both 50-share orders filled, the cost would be $2 and the winning side would redeem for $50, yielding a $48 net gain regardless of the outcome. But both orders had to fill. In the test’s 9,486 paired attempts, that happened three times, or 0.032%. The report records 1,297 loser-only fills, 22 winner-only fills and 8,162 attempts with neither side filled; the paired version’s paper P&L was minus $280.
A second version posted only on the side the system read as the winner just after the market closed. It made 3,482 posts and got eight fills, according to the report. Four fills settled, and all four lost. The article attributes those losses to a timing gap between its price reading and Polymarket’s official resolution. It says there were no fills for BTC or XRP in this version. Both tests used a paper-fill model in which simulated bids filled only when a real taker sold into them.
Strategy Test / Week Four
AI Puts The Viral “100X Trade” Strategy To The Test In Week Four
A two-day paper-trading test across four Polymarket crypto markets found that both versions lost money. The trade’s paired-bid mechanism depended on two rare fills landing together.
01 / The test in context
A striking win is not a repeatable edge
The strategy was reconstructed from a YouTube video showing a reported $1-to-$50 win on a five-minute Bitcoin Up/Down market, alongside examples described as 48x and 100x returns. The transaction was described as verifiable on PolygonScan; the creator also characterized it as a rare event and said the video did not show whether it could be repeated.
Big multiples
A successful low-cost position can return many times its purchase price if it resolves on the winning side.
Frequency matters
One verified transaction establishes that it happened once, not how often the setup appears or whether it covers losses over time.
Simulated only
Week Four tested two reconstructed approaches over about 13,000 windows using simulated funds and a paper-fill model.
02 / Two approaches
How the versions worked
The source describes five-minute crypto markets as binary: one outcome resolves to $1 per share and the other to $0. Both strategies relied on order timing and the platform’s eventual settlement.
Buy both sides cheaply
Post bids at 2¢ on both Up and Down. If 50 shares on each side fill, the $2 total cost is intended to return $50 from the winning side, for a $48 net gain.
Exposure: If just one order fills, the position can be left on the side that resolves to $0.
Bid after the window ends
Post only on the side the system reads as the winner just after market close. The test logged eight fills from 3,482 posts; four were reported settled, and all four lost.
Risk: The price reading and official resolution may not align in time. No BTC or XRP fills were reported for this version.
03 / Paired-bid results
Rare double-fills, frequent one-sided fills
Across 9,486 paired attempts, both orders filled only three times. The report recorded a paper P&L of −$280 for this version.
04 / What the findings show
Timing and execution shape the outcome
A high-multiple trade can be genuine while still occurring too rarely to offset more frequent losing positions.
The post-close losses were attributed to a gap between the price reading and Polymarket’s official resolution timing.
Simulated bids filled only when a real taker sold into them. This does not establish live execution costs or delays.
“This is going to take time because this is a rare event… probably I didn’t have anything for you in this video about the results.”The YouTuber, as quoted in the report
05 / Limits and next evidence
What the test cannot establish
The account describes a two-day simulated test, but the supplied material does not give enough detail to estimate performance over longer periods, across different market conditions, or with live orders. The post-close result is especially limited: eight fills were reported, four settled, and all four settled fills lost.
More market windows
A longer record would show whether the observed fill pattern persists in other conditions.
Complete settlement data
Follow-up reporting could clarify all fills, outcomes, timing assumptions, and the treatment of unresolved positions.
Auditable attempt log
A reproducible record of attempts and outcomes would let readers assess the method and its assumptions.
The report is Week Four of an ongoing series. It does not establish how the strategy performs with real funds and is not financial advice.
06 / Key questions
Quick answers
Did the test use real money?
No. The report says the test used simulated money and no real funds.
How often did both paired orders fill?
Three of 9,486 attempts, or 0.032%.
What happened to the post-close version?
Eight fills came from 3,482 posts. Four fills were reported settled, and all four lost.
Does the report prove the original trade was fake?
No. It describes the transaction as real and verifiable, while noting that one example does not show repeatable profitability.
Rare Fills Undercut the Trade
The results address the difference between a striking transaction and a repeatable strategy. A trade that returns many times its cost can be genuine while still depending on an unusual market event that is too rare to offset more frequent losses. In this test, the paired strategy’s planned profit required both low-priced orders to fill; instead, the report recorded far more loser-only fills than double-fills.
The post-close version also highlights a separate risk: identifying a likely winning side immediately after a market window ends may not match the platform’s eventual settlement. The report’s four settled fills are a small sample, and the test was simulated, but the outcomes show why a strategy’s assumed payoff and the actual resolution process need to be tested separately. The findings do not establish how the strategy would perform under every market condition or with real execution.
As an affiliate, we earn on qualifying purchases.
How the Two Versions Worked
Polymarket’s five-minute crypto markets are binary, as described in the source: one side resolves to $1 per share and the other to $0. The paired strategy attempted to buy both outcomes cheaply, making its intended profit independent of whether the price moved up or down—provided both orders filled at the stated price. Its exposure came from being left with only one side, particularly the side that ultimately paid nothing.
The report says the original trade may have benefited from stale orders around a market close: orders that remained available after the outcome appeared decided. A trader able to buy the winning side at a very low price could profit if such liquidity were available. The test was the fourth installment in Thorsten Meyer AI’s ongoing series, published May 21, 2026. The source states that all results were produced with simulated money and no real funds were used.
„This is going to take time because this is a rare event… probably I didn’t have anything for you in this video about the results.“
— The YouTuber, as quoted in the Thorsten Meyer AI report
As an affiliate, we earn on qualifying purchases.
Limits of the Simulated Results
The published account describes a two-day paper test, but does not provide enough detail here to assess how the results might change over longer periods, in different market conditions, or with live orders. No real funds were used, and simulated fills cannot establish the precise costs, execution delays or other effects of live trading.
The post-close result is especially limited: although eight bids filled, only four are reported as settled, and all four lost. The report attributes this to a difference between its price feed and Polymarket’s official resolution timing, but does not provide a full breakdown of the remaining fills in the supplied material. The original YouTube trade is described as real and verifiable, yet one successful example does not establish its frequency or profitability across a larger sample.
As an affiliate, we earn on qualifying purchases.
Further Testing Would Be Needed
Thorsten Meyer AI presents this as Week Four of an ongoing series, but the source does not specify a publication date or design for the next installment. Any follow-up would need to show whether the results persist across a longer test and spell out how order timing, settlement data and simulated fills were handled.
For readers evaluating the claim, the next useful evidence would be a reproducible record of attempts and outcomes, including the full settlement results and the test’s assumptions. The report itself warns that its findings are not financial advice and that all testing described was simulated; it does not establish that the strategy is safe or profitable with real funds.
As an affiliate, we earn on qualifying purchases.
Key Questions
Did the test use real money?
No. Thorsten Meyer AI says the strategy was tested with simulated money and that no real funds were used.
How often did both paired orders fill?
The report says both sides filled in 3 of 9,486 attempts, or 0.032% of those attempts.
What happened to the post-close version?
It produced eight fills from 3,482 posts. The report says four fills settled and all four lost; it attributes the result to timing differences between its price reading and Polymarket’s official resolution.
Does the report prove the original trade was fake?
No. The report describes the original transaction as real and verifiable, but argues that a single striking trade does not show the strategy can be repeated profitably.
Source: ThorstenMeyerAI.com
