Read the full analysis: Could AI Find An Edge In This Viral “100X Trade” Strategy? on ThorstenMeyerAI.com
TL;DR
Thorsten Meyer AI reports that two simulated versions of a viral Polymarket strategy lost money across about 13,000 trading windows on BTC, ETH, SOL and XRP. The paired-bid version completed its intended double-fill just three times in 9,486 attempts; a post-close version also lost on the fills that settled.
Thorsten Meyer AI says two simulated versions of a viral Polymarket “100x trade” strategy lost money in a two-day test spanning about 13,000 windows on four cryptocurrency markets. The report says the paired-bid approach achieved its intended double-fill just 3 times in 9,486 attempts, while a post-close variant also failed to produce a winning settled fill.
The strategy was inspired by a YouTuber’s reported trade on Polymarket’s five-minute BTC Up/Down market. The video showed a $1 stake returning $50, alongside examples described as 48x and 100x. The source says the transactions could be checked on PolygonScan, but the isolated trades did not establish how often the method could be repeated. The YouTuber acknowledged in the video that results were not yet available because the event was rare.
In the source’s reconstruction, a trader places bids for both Up and Down shares at 2 cents each. If 50 shares of each side fill, the cost is $2. Since one side pays $1 per winning share at settlement, redeeming 50 shares would return $50, for a $48 net gain. That outcome depends on both bids filling at the low price. The article’s proposed explanation for the original trade is that a trader captured stale orders left in the book around market close; this is an interpretation of the trade mechanism, not evidence that such fills are common.
The report tested a literal paired-bid version and a second approach that posted only on the side its price feed identified as the winner after the market closed. It says the paper-fill model counted a bid as filled only when a real taker sold into it. In the paired-bid test, 3 of 9,486 attempts filled on both sides, while 1,297 filled only on the losing side and 8,162 did not fill at all. The reported net paper result was minus $280. In the post-close test, 8 of 3,482 posts filled; none of the four fills that reached settlement won.
Simulation report · Week four · May 21, 2026
Could AI Find an Edge in This Viral “100X Trade” Strategy?
Two simulated approaches to a viral Polymarket tactic lost money across about 13,000 trading windows on BTC, ETH, SOL and XRP. The strategy’s key event—a double-fill—was exceptionally rare in the reported test.
The key event
3 / 9,486
Paired-bid attempts filled on both sides—the event the strategy needed to lock in its intended payout.
01 / The idea
Two cheap bids. One winning side. A big “if.”
The viral trade was reconstructed as bids for both Up and Down shares at 2¢ each. If 50 shares on both sides filled, the cost would be $2. At settlement, 50 winning shares paying $1 each would return $50—a $48 net gain. The catch: both bids have to fill at that price.
The setup
Bid both outcomes
Place low-price orders on Up and Down in a five-minute crypto market.
The dependency
Both sides must fill
A fill on only the losing side leaves the trader holding the unfavorable outcome.
The evidence gap
A payout is not a pattern
A reported $1-to-$50 trade, plus examples described as 48× and 100×, does not show how often the setup can be repeated.
02 / From viral trade to repeat test
What the simulation tried to reproduce
The report proposed that the original trade may have captured stale orders around market close. That is an interpretation of the mechanism, not evidence that such fills are common. It tested a literal paired bid and a second, post-close approach.
Offer 2¢ for both Up and Down shares.
Count a paper-fill only when a real taker sold into the bid.
Compare fills with the winning side and market settlement.
Post only on the side a price feed identified as the winner.
“The double-fill … happened 3 times in 9,486 attempts.”
Thorsten Meyer AI report
03 / Reported outcomes
Rare double-fills. More one-sided losses.
The reported paper-fill counts show how often the paired-bid test reached each outcome. Separately, the post-close test produced eight fills, but none of the four fills that reached settlement won.
Paired-bid version
9,486 attempts · reported net paper result: −$280
Post-close version
3,482 posts · eight fills reported
Why fills matter: the intended gain depended on buying both outcomes cheaply. In the paired-bid sample, losing-side-only fills were far more frequent than double-fills.
Paper results ≠ live returns
04 / Limits and interpretation
A useful test of one setup, not a universal verdict
The report’s results speak to the specific simulation, markets and two-day period it describes. They do not prove that every version of the idea must fail, or establish how it would perform with real capital.
Short sample
Two days of data
The test cannot show performance across longer periods or different market conditions.
Model limits
Independent details absent
The supplied account does not independently verify the code, order-book data or paper-fill model.
Settlement risk
Feed and resolution differed
The report says the early winner reading disagreed with eventual Polymarket resolution on the four post-close fills that settled.
Neither tested version made money in the reported simulation; the paired-bid double-fill was rare.
Per-coin breakdowns, cancellation timing, and whether different feed or execution rules would change results.
05 / Key questions
What readers should take away
Did the strategy make money?
No. The paired-bid version reported −$280 in net paper results; the post-close version had no winning fill among four settled fills.
Were real funds used?
No. The report says both approaches used simulated money. These results are not live trading returns.
How often did both bids fill?
Three times in 9,486 paired-bid attempts. The report counted 1,297 losing-side-only fills.
Does this prove it can never work?
No. It is evidence about the tested setup and period, not every future market condition or possible strategy variation.
Source: ThorstenMeyerAI.com · Simulated test · May 21, 2026
Why Rare Fills Matter
The result speaks to the gap between a dramatic individual payout and a repeatable trading method. A trade can be genuine and still be a poor guide to expected results if the conditions that produced it occur rarely or if attempts more often fill on the unfavorable side. In this test, the double-fill that was supposed to lock in a gain was uncommon, while losing-side-only fills were substantially more frequent.
For readers, the practical point is not that every market or trading system will behave the same way. The report is limited to its particular simulation, markets and testing period. It does show why a viral screenshot or a handful of transactions cannot establish that a strategy has a durable edge. The reported losses also underline that simulated results do not show what would happen with real capital, including the effects of execution, fees and changing market conditions.
From Viral Trade to Repeat Test
Polymarket’s crypto Up/Down contracts are binary: one outcome pays and the other does not, according to the market’s resolution. That structure makes the paired-bid idea look attractive on paper if both sides can be bought cheaply. The challenge is execution: the trader needs both orders to fill, and the report’s results suggest fills were not distributed evenly between the winning and losing sides.
The source describes the work as Week Four of an ongoing series and dates the report May 21, 2026. It says the experiments ran for two days across BTC, ETH, SOL and XRP, with 24 parallel experiments. All results were generated with simulated money; the report says no real funds were used. Its findings are therefore a reported paper-trading test, not a record of live trading performance.
“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
Limits of the Simulation
The source does not provide independent verification of its code, order-book data or paper-fill model, and the supplied material does not include a full breakdown of results by coin and cancellation timing. The test lasted two days, so it cannot establish how the strategy would perform across different market conditions or over a longer period.
The post-close results also point to a timing risk: the report says its price feed’s early winner reading disagreed with Polymarket’s eventual resolution on the four fills that settled. The supplied account does not detail the precise settlement data or timing for each trade. It remains unclear whether changes to the feed, order placement or execution rules would alter the outcome; the reported sample does not test every possible variation.
Further Tests and Market Data
The report identifies itself as part of an ongoing series, but the supplied material gives no date or details for a next installment. Readers should look for any follow-up to specify whether it adds testing time, publishes per-market results, or changes the execution and settlement assumptions.
Until further evidence is available, the current finding is limited but clear: in this reported simulation, neither version made money, and the paired approach’s intended double-fill occurred only three times in 9,486 attempts. The source cautions that the work used simulated funds and says attempting such a strategy with real money could result in losses.
Key Questions
What was the viral “100x trade” strategy?
It was inspired by a YouTuber’s reported Polymarket trade and reconstructed as bids for both Up and Down shares at 2 cents. The idea was that if both sides filled, the winning shares would pay out more than the combined cost.
Did the strategy make money in the reported test?
No. Thorsten Meyer AI reports that the paired-bid version had a net paper result of minus $280, and that the post-close version had no winning fill among the four fills that settled.
How often did both sides fill?
The report says both sides filled 3 times in 9,486 paired-bid attempts. It reports that losing-side-only fills occurred 1,297 times in that sample.
Were real funds used?
No. The source says the strategies were tested with simulated money and that no real funds were involved. The reported results are paper-trading results, not live trading returns.
Does the test prove the strategy can never work?
No. It reports losses in a two-day simulation across four coins and about 13,000 windows. That is evidence about the tested setup and period, not proof about every possible market condition or future version of the strategy.
Source: ThorstenMeyerAI.com