What he said.
What traded.
His account publicly proposed changing a market’s subscriber-count metric. A separate chain scan shows circular trading dominating three markets for a defined period. Here are the actual records.
Open the first-party comments ↘His account proposed manipulating what the market measured.
In the market MrBeast biggest YouTube channel by June 30?, Car’s account recommended NO and then proposed paying for artificial subscribers as protection if that bet began to lose. His reply described it as:
“Risk free bet holding NO shares!”
This is a public recommendation to inflate the subscriber count used by a prediction market. Four official records supply the sequence:
| UTC | What the account said | First-party record |
|---|---|---|
| 22 Apr 2024 20:59:28 | Threatened to inflate the count just before the market ended. | Comment 29763 ↗ |
| 22 Apr 2024 21:04:51 | Discussed growth projections and recommended NO. | Comment 29764 ↗ |
| 22 Apr 2024 21:05:54 | Replied to that recommendation with the paid-subscriber tactic. | Comment 29765 ↗ |
| 10 May 2024 23:10:16 | Repeated the deadline-botting threat and said it would hurt two other traders. | Comment 32398 ↗ |
Polymarket’s own responses assign the comments to user address 0x1f64f2739a19fc5cdfd7c3f50c42463aa8983e82 and embed Car’s known 0x7c3db723f1d4d8cb9c550095203b686cb11e5c6b wallet. This is platform attribution of the comments.
The market asked whether MrBeast led at any point before June 30. Its official record closed with YES on June 2, 2024; changing counts at the deadline would not undo an earlier qualifying lead. The current trade API returned no trades for the eight disclosed wallets in this market. That bounded lookup does not establish a stake or exclude unknown accounts.
The returned thread contains banter and another trader plays along with a smiley. Car’s returned comments contain no explicit joke marker or retraction. The evidence establishes the recommendation and threats; it does not establish that subscriber bots were purchased or used.
Comment IDs, author binding, context and market rule ↗ · Independent account-binding check ↗ · Official event record ↗
The circular trades were almost the entire observed market activity.
On September 12, 2025, Car and CarComboCup completed twelve mint-and-merge rounds in 33 minutes, four seconds. Every round restored their incremental target-market holdings and shifted cash toward Car. The total was 1,601.50 USDC.e to Car.
We expanded the check beyond those two wallets. Across both then-deployed canonical exchange contracts, the full 18:01:29–18:58:31 UTC window contained 13,046 order-fill logs. Filtering the six outcome tokens in these three markets and counting each settlement once leaves:
| Market | Car / Combo fills · all fills | Car / Combo collateral · all collateral |
|---|---|---|
| GTA 6 launch postponed? | 4 / 4 | 4000 / 4000 |
| EU sanctions on Israel by December 31? | 2 / 2 | 100 / 100 |
| Hypurr NFT airdrop by November 30? | 18 / 21 | 28000 / 28015 |
These are 57 minutes and two seconds in three markets. Turnover counts each actual collateral lock or release once and matches the USDC.e transfers involving Conditional Tokens. It is not a claim about the website’s volume counter or an entire market’s lifetime.
During the narrower first-to-last loop episode, all 24 observed fills in those markets involved Car and CarComboCup. The three other five-set settlements came afterward and each involved Car with another trader.
Four successful transfer queries check both directions and both Conditional Token event types. They show exactly 48 target-token movements during the episode, all belonging to the known opening or closing trades. At the end of every atomic transaction, the group’s added YES and NO positions were equal. After each completed round, its added positions were gone. Nine rounds closed within 32 seconds.
Car was the maker in all 24 settlements. His 24 distinct orders were fully filled. That establishes the quoted prices and cash gains, while leaving a plausible alternative: an independent counterparty repeatedly crossing a legitimate maker’s bid and ask. The records do not establish who directed CarComboCup or when Car’s orders were first posted.
Full market-window calculation ↗ · Complete holding-interval transfers ↗ · All 27 settlements CSV ↗ · Maker roles and order sizes ↗ · All 48 actual order signatures ↗
The acknowledged-alt round trip remains the clearest self-dealing case.
The separate b4k9xj2wh2 case ties Car’s main wallet to an account he later publicly identified as his alt. They opened and closed 28,000 complete sets in 404 seconds, producing 56,000 USDC.e of collateral turnover. Their combined cash and target-token holdings returned to their starting levels, with 840 USDC.e shifted internally to the alt. All four actual authorizations and the archived balances are verified.
This is concrete self-matching with unchanged combined exposure and a strong wash-trading indicator. The September CarComboCup case adds repeated circular-counterparty execution, but its human ownership remains unresolved.
Open the b4 cash, inventory and signature audit ↗
Polymarket’s archived August 2025 terms addressed wash trading and arranged trades without genuine competition. That is historical context for the 2025 trades; it is not retroactively applied to the 2024 comments. Purpose and knowledge still require evidence. Dated terms and archive hashes ↗
Open the record, one link at a time.
Scroll to pan. Nodes and arrows open the corresponding first-party comment, transaction or proof.
The checks that keep the claims precise.
The reward investigation verified actual account-level payments, but did not allocate rewards or rebates to these matched trades. A midnight payout does not identify which market or orders earned it. The matched orders’ recorded fees were zero. Reward attribution findings ↗
The historical Hypurr price endpoint returned 33 samples during the loop episode: 29 at 53.5¢, three at 54¢ and one at 54.5¢. These samples do not establish that the alternating executions moved the public chart. Primary price history ↗
The specific findings are public advocacy of artificial subscriber growth, a verified acknowledged-alt self-trade, and a repeated circular counterparty dominating a bounded trading window. Executed subscriber botting, directed CarComboCup trading and a connection to the DebugBear drainer remain unproved.