Datasets:
prompt stringlengths 218 813 | chosen stringlengths 446 1.86k | rejected stringlengths 193 1.03k | source stringclasses 1
value |
|---|---|---|---|
Investigate the UNI/WETH pair where a single 'master' array of wallets trades in perfect synthetic sequence: Wallet 1 buys, waits 10 seconds, Wallet 2 sells, waits, Wallet 3 buys, etc. The sequence is so precise that the price impact is nearly zero. A blockchain analyst calls this 'cooperative trading,' not wash tradin... | The analyst conflates temporal separation with economic separation. The absence of an immediate second leg is irrelevant. The decisive evidence is the loop-closing: after Wallet 1 buys and Wallet 2 sells, the UNI from Wallet 1's buy eventually returns to the master wallet. If the master wallet funds the buy side and th... | The analyst is right. Wash trading requires an immediate buy and sell by the same address. Here, the buys and sells are at different times and by different wallets. This is just cooperative trading where multiple parties agree to provide liquidity. The price impact being zero might be because the pool is deep. The fund... | synthetic:deepseek/deepseek-v4-flash-0731 |
A pair on an L2 shows a 'perfect circular transfer' pattern: tokens flow from A to B, B to C, C to D, D to A, with each transfer preceded by a small fee payment to a designated fee collector. The cycle repeats every 10 minutes, and the volume accumulates to $3M. All four addresses are governed by the same multisig on t... | Yes, this is wash trading. The L1 ownership linkage is decisive: all four addresses are governed by the same multisig, meaning a single entity controls all participants in the cycle. On-chain forensics prioritizes ultimate ownership over L2 activity—the L2 is just an execution layer, but the economic control lies at th... | This is wash trading because the transfers form a perfect circle, which is a clear sign of manipulation. The cycle repeats every 10 minutes and the volume is $3M, which is large. The fee payments to a collector are also suspicious. The L1 ownership linkage is important because it shows the same multisig controls all ad... | synthetic:deepseek/deepseek-v4-flash-0731 |
A user on a forum claims the SAND/ETH pair is not wash-traded because 'all trades have distinct transaction hashes.' You are asked to respond. You analyze the data and find that while hashes are distinct, the pair's trading graph is a series of 4-node cycles (A-B-C-D-A) with a period of exactly 6 minutes, running for 2... | Distinct hashes only prove that each transaction is a separate on-chain event—they say nothing about economic independence. The graph tells the real story: a perfect 4-cycle (A→B→C→D→A) repeating every 6 minutes for 24 hours is a deterministic loop, not organic trading. In organic markets, order flow is stochastic; par... | You're right that distinct hashes mean these aren't replay attacks or double-spends, but that doesn't make the volume real. The graph shows a cycle, and cycles are usually suspicious. However, I think the real issue is that the prices might be manipulated—if the trades are at increasing prices, that's more concerning. ... | synthetic:deepseek/deepseek-v4-flash-0731 |
You are an exchange investigator reviewing the 'chz/eth' pair. A 'cascade' of limit orders is placed by 20 addresses, each order size being 1.5x the previous, and when one order hits, it triggers a chain of fills that exactly reverses the order placement. All addresses are traced to a common 'treasury' multi-sig. The c... | The geometric progression of order sizes (1.5x step) is a deterministic signature, not organic market behavior. In genuine liquidity provision, order sizes cluster around the liquidity curve and vary stochastically; a consistent 1.5x multiplier across 20 addresses implies a precomputed script. More damning is the exact... | The 1.5x progression and shared multi-sig are suspicious, but this could be a market maker using a ladder strategy to manage inventory. The reversal might be a hedge technique—they buy and sell to maintain a neutral position. The common treasury is just a pooled capital structure, which is typical for professional MM f... | synthetic:deepseek/deepseek-v4-flash-0731 |
You are analyzing a pair on an L2. Due to the sequencer's batching, you see that many trades are included in the same block. You observe that within a single block, there are 500 buys and 500 sells of the same token, each from a different set of wallets that all have the same deployer. The trade sizes are all 100 token... | This is wash trading. The block-level coordination is the strongest evidence: within a single block, 500 buys and 500 sells of the same token appear in perfect balance. In a genuine market, you would see a mix of trade sizes, asymmetric order flow, and some price movement; here, the exact 100-token size across all trad... | This is wash trading because the volume is $15M but the price doesn't move at all, which is weird. The trades are all 100 tokens, which is too uniform. Also, the wallets have the same deployer, so they are connected. The fact that there are 500 buys and 500 sells in one block shows coordination. The liquidity is $100k,... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given the on-chain data for the MKR/WETH pair, you notice a cluster of wallets that have all received funds from one central address. Determine if their alternating trades are a wash-trading scheme or a valid rebalancing strategy. | The central funding is suspicious, but the key distinction lies in the economic purpose of the alternating trades. A legit rebalancing strategy would adjust a portfolio's weights in response to price movements or risk limits—trades would be one-directional relative to the cluster's aggregate position, and the net flow ... | This is clearly wash trading. All those wallets were funded from one central address, which means they are all controlled by the same person. When one wallet sells and another buys, it's just moving money between left and right pockets. Rebalancing is just an excuse; no legitimate strategy would use so many wallets fro... | synthetic:deepseek/deepseek-v4-flash-0731 |
You see a pair where the volume is distributed across 100 different wallets, but all 100 wallets were funded by the same address within a single block. A user asks: is this a sybil-level wash trading scheme or could it be a legitimate airdrop farming behavior? | This is almost certainly a sybil-level wash trading scheme, not legitimate airdrop farming. The key distinction is the funding pattern: all 100 wallets funded by the same address in a single block indicates a coordinated setup. Airdrop farmers typically fund their wallets from various sources (exchanges, personal walle... | It could be airdrop farming, because airdrop farmers often create many wallets to get more tokens. The fact that they're all funded by the same address suggests they're all controlled by one person, but that's also true for airdrop farmers. The volume could just be them trading a bit to qualify, so I wouldn't call it w... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the GRT/ETH pair, you find that a single address sends tokens to 5 addresses, and these addresses then sell the tokens to each other in a chain, with the final sale returning tokens to the original address. This happens multiple times per day, with varying amounts each time. A novice says the varying amounts prove ... | Varying amounts are trivial to implement in an automated script and provide no evidence of independent decision-making. The reasoning is weak because it mistakes surface-level randomness for organic intent—organic trading is characterized by counterparties with divergent motivations, not by variable quantities within a... | The varying amounts show that each trade is different, which is what you would see in real trading. If someone were wash trading, they would use the same amounts to make the pattern obvious. Since the amounts change, the addresses are making independent decisions based on market conditions. The chain from A to B to C t... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the SNX/DAI pair, you find a 'reverse funding' pattern: the same wallet that receives proceeds from a sell trade immediately sends the funds to a new wallet, which then buys back the same token, creating a cycle. A novice says this is just a sophisticated trader. Assess whether this is wash trading and explain the ... | This is classic wash trading. The reverse funding pattern is a deliberate loop: wallet X sells SNX, receives DAI, then forwards that exact DAI to a fresh wallet Y within the same block or next block, and Y buys SNX back. The economic substance is zero—the same capital is rotated with no external counterparty bearing ri... | A sophisticated trader might indeed split orders across wallets to manage slippage or for privacy. The reverse funding could be a treasury operation: receive profits, then move to a new address to re-enter at a better price. It's not necessarily wash trading because the trader might have a genuine long-term view and is... | synthetic:deepseek/deepseek-v4-flash-0731 |
On the LINK/ETH pair, I notice that 80% of the 24h volume happens between 5 wallet clusters, and each cluster's addresses receive their initial ETH from the same Binance withdrawal address within a 3-block window. Can you determine if this is wash trading or just a few large organic traders? Explain your reasoning. | This is strong evidence of wash trading, not organic activity. The clustering is not merely suggestive; the funding pattern is a decisive signal. Organic large traders typically have independent capital sources and do not share a single Binance withdrawal origin within a 3-block window—that temporal proximity indicates... | I see 80% volume from 5 clusters and they all came from Binance. That's a red flag, but it could be a couple of big whales who withdrew at the same time to trade together. Wash trading usually involves more volume and clearer signs. I'd say it's probably wash trading because the numbers are too neat and the funding is ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze the CRV/ETH pair on a concentrated liquidity pool. The liquidity provider’s own address is the counterparty on 90% of trades, and every trade is within a 0.1% price range. A defender claims this is just the LP providing tight quotes. Determine if the volume is artificial, focusing on whether the LP's self-tradi... | The volume is artificial. A legitimate LP provides quotes to external traders, not to itself. When the same address is the counterparty on 90% of trades, those trades are self-deals: the LP is buying from and selling to its own position, which does not create genuine price discovery or transfer of risk. The 0.1% price ... | The LP providing tight quotes is a reasonable defense. In concentrated liquidity, an LP can legitimately be the counterparty to many trades because they are offering the best price in that range. The 0.1% range is very tight, which is exactly what a market maker would do. The fact that 90% of trades are against the LP ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Examine a pair where the 24h volume is $8M, generated solely by a 'liquidity provider' that repeatedly adds and removes the same amount of liquidity to trigger trades, but the trade graph shows that all swaps are between the provider's own wallets. The wallets receive funding from a single Tornado Cash withdrawal, and ... | This is wash trading. The core evidence is the circular flow: the provider adds liquidity, which changes the pool ratio, then swaps between its own wallets to capture the 'arbitrage', then removes liquidity—repeating this cycle. All swaps are between wallets controlled by the same entity, so there is no external counte... | This is wash trading because the volume is $8M and the TVL is only $50k, which is a huge mismatch. The provider is doing all the trades, and they are between its own wallets. The Tornado Cash link shows they are hiding something. The trade amounts are always 10,000 USDC, which is too regular. The large volume compared ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a pair's transaction logs, I notice a pattern: a group of 5 wallets, each created from the same factory contract, execute trades that always sum to a flat round number (e.g., 5000 tokens net per hour). They also only trade when the real-time volume is low, artificially inflating the hourly number. Is this wash tr... | Yes, this is wash trading. The factory contract matters because it proves deterministic creation of the wallets by a single deployer, establishing that the 5 wallets are not independent actors but instruments of one entity. The net trade sum being a flat round number (5000 tokens) shows a pre-planned volume target, not... | This is wash trading because they are using a factory contract to make wallets, which is suspicious. The round number sum confirms they are gaming the metric. Trading when volume is low is also fishy. The factory contract is important because it shows the wallets are connected. So I would flag it as wash trading. | synthetic:deepseek/deepseek-v4-flash-0731 |
A bot trades continuously on a pair, but it uses a unique router contract for each trade, with a 1-block delay between contracts. The contracts are all created by a factory contract, which is funded by a single EOA. A developer says this is just advanced arbitrage, not wash trading, because each contract is independent... | The security researcher has stronger reasoning. The unique router contracts and 1-block delays are obfuscation techniques, not evidence of independence. All contracts are spawned by the same factory and funded by the same EOA—that is a single controlling entity. The 1-block delay is designed to break temporal correlati... | The developer is more correct because each router contract is a separate on-chain entity with its own code and execution context. They are not the same address, and they do not share state. The 1-block delay suggests the bot is responding to changing market conditions, which is typical of arbitrage strategies. A factor... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a pair's on-chain history, you see that a 'liquidity provider' token is used to route volume: the provider deposits, then immediately executes a buy and a sell that cancel out, but the provider earns fees. The provider's address is linked to the token team's treasury. Weigh the fee-mining incentive and the team l... | This is wash trading via fee mining. The provider's linked to the team treasury, so the team is effectively paying itself fees through the liquidity pool. The pattern—deposit, then immediate buy and sell that cancel out—means the trades are riskless to the provider (they end flat in the token) yet they earn trading fee... | The provider is earning fees by doing round-trip trades, which is a common strategy to capture fees from high-volatility pairs. Since they are providing liquidity and the trades cancel out, they are not taking on much risk, but that doesn't make it wash trading—they are just being opportunistic. The team link might be ... | synthetic:deepseek/deepseek-v4-flash-0731 |
A sophisticated bot on a DEX pair splits its trades into many small orders using multiple proxy contracts, all of which are funded by the same EOA, and these contracts trade against each other in a round-robin pattern. The bot also avoids detection by changing the trade sizes slightly. Determine if this constitutes was... | This is wash trading, albeit obfuscated. The evidence is threefold: (1) all proxy contracts are funded by the same EOA, meaning a single entity controls all sides of the market; (2) the round-robin pattern — A sells to B, B sells to C, C sells back to A — is a deliberate cycle where no real external counterparty ever t... | This is wash trading because the bot uses many contracts and they trade in a round-robin. The sizes are a bit different, but they all come from the same wallet, so it's clearly fake. The bot is trying to hide it, but it's still wash trading. | synthetic:deepseek/deepseek-v4-flash-0731 |
A user claims a pair is clean because 'every trade has a different counterparty.' You investigate and find that while counterparties are 30 distinct wallets, they are all controlled by a single 'funding contract' that sends them tokens in a 1-hour window. These wallets then trade in a line: 1→2, 2→3, ..., 30→1, each wi... | Yes, this is wash trading. Counterparty diversity is irrelevant because all 30 wallets share a single source of funds—the funding contract—which establishes common control. The 1-hour funding window followed by a strict 1→2→...→30→1 sequence with identical amounts is a textbook circular transfer loop. In genuine tradin... | I disagree—this is not wash trading. Having 30 distinct counterparties means the trades are diverse, and diversity is what makes volume real. The fact they were funded by the same contract is likely just a convenient way to distribute tokens, perhaps for airdrop farming or a coordinated strategy. The circular sequence ... | synthetic:deepseek/deepseek-v4-flash-0731 |
A pair on a concentrated liquidity pool shows that the liquidity provider's own two accounts trade against each other through the pool, and the pool's fee structure incentivizes these trades. The LP claims it's for fee harvesting, not wash trading. Analyze whether this constitutes wash trading based on the circular tra... | This is wash trading. The two accounts are funded from the same origin and execute mirrored buy/sell orders through the pool, with no external counterparty. The fee structure merely subsidizes the losses from spread and fees, making the activity economically irrational absent the incentive—but that incentive is precise... | I think this is fine because the LP is just harvesting fees, which is a normal strategy on concentrated liquidity pools. The two accounts are separate wallets, so they are different traders. The pool is public, so anyone could trade, and the LP is just providing liquidity. Since the fee incentive is a legitimate part o... | synthetic:deepseek/deepseek-v4-flash-0731 |
A compliance officer at a wallet service reviews the 'wld/eth' pair. The top 10 trading wallets each received their initial ETH from the same 'binance' hot wallet within a 2-hour window. Each wallet then executes a series of trades that net to zero (buy X, sell X, buy X) over a 24h period, and the trade sizes are all w... | This is wash trading. The funding relationship is decisive: all 10 wallets received ETH from the same Binance hot wallet in a narrow 2-hour window, indicating a single entity controlling all accounts. The net-zero trade pattern (buy X, sell X, buy X) shows no genuine position-taking—each wallet ends where it started, s... | The fact that all wallets got ETH from Binance is normal because many traders use exchange hot wallets to fund their accounts. The trades net to zero, but that could be a market-making strategy where you buy and sell to capture spread. The sizes are similar because the market is tight, so orders are naturally close. Th... | synthetic:deepseek/deepseek-v4-flash-0731 |
For an obscure pair, the on-chain data shows that every single trade is between the same two addresses, and those addresses were funded by the same address at the same block. The trade sizes are all around 1000 USDC, and there are no other interactions. A 'DEX analyst' concludes it's clean because the trades are on a p... | The analyst's reasoning is completely unsound. Public blockchains are transparent, but that transparency is exactly what reveals the wash trading. Every trade is between two addresses funded by the same source at the same block, proving common ownership. The uniform trade size (~1000 USDC) indicates automated round-lot... | The analyst is right—if it were on a public blockchain, it's transparent, so anyone can see it. That transparency means there can't be manipulation because everyone would notice. The two addresses are just two traders who have a relationship. The 1000 USDC trades are just a convenient size. Since there are no hidden tr... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a pair's 24h activity, you find that all major trades occur between two clusters of wallets that share a single Ethereum address as their initial funder (a dormant ICO wallet). The trade sizes are exactly reciprocal (e.g., 5 ETH buy matched by 5 ETH sell). Is this wash trading? Detail your steps. | Yes, this is wash trading. The steps: (1) Trace the funding graph—both clusters' wallets were initially funded by the same dormant ICO address, meaning they are under common control, even if the ICO wallet is now dormant. (2) Isolate all trades where the buy amount equals the sell amount for the same token pair (e.g., ... | The dormant ICO wallet doesn't matter—it's just an initial funder, and many legit projects have founders who fund early wallets. The reciprocal trades could be coincidental, as traders often buy and sell similar amounts. I would not call it wash trading without seeing a clear profit-taking or price-impact pattern, so I... | synthetic:deepseek/deepseek-v4-flash-0731 |
A community manager claims a pair's volume is legitimate because 'there are hundreds of unique traders.' Upon review, you find that all these traders are new, unverified contracts created by a single factory contract in the same block. How do you weigh this 'organic-looking' address diversity against the clear provenan... | The address diversity is a decoy. While hundreds of unique addresses might suggest organic participation, the fact that they were all spawned by a single factory contract in the same block collapses that assumption. This is a Sybil attack on the metric of 'unique traders.' The factory contract acts as a single point of... | If there are hundreds of unique addresses, then the volume is legitimate because they are distinct entities. The fact they are contracts created by a factory just means they are smart contracts, which is normal for DEX trading. The community manager is right that the diversity proves organic interest. | synthetic:deepseek/deepseek-v4-flash-0731 |
For the LINK/USDT pair on Polygon, you observe that 90% of the volume occurs in a 5-minute burst, prices never deviate by more than 0.05% from a median anchor, and each buy is matched by a sell from the same group of 10 addresses within 30 seconds. Produce a nuanced assessment that separates the timing pattern from the... | The timing and matching pattern is a strong wash indicator independent of funding. A 5-minute burst containing 90% of daily volume, with every buy matched by a sell from the same 10 addresses within 30 seconds, demonstrates pre-arranged crossing—not organic market discovery. The ultra-tight price anchor (0.05% deviatio... | The burst and tight price range could just be a market maker providing liquidity during a high-volatility event. The 10 addresses might be different market makers who happen to trade with each other, which is normal in a liquid pair. Matching buys and sells within 30 seconds is just fast execution, not necessarily wash... | synthetic:deepseek/deepseek-v4-flash-0731 |
Assess the 'om/weth' pair on a DEX where the top 10 addresses share a single funding wallet that was created 2 hours before trading began. All trades are within a 0.5% price range and alternate between two addresses every 5 blocks. Determine if this is wash trading, and explain the role of the shared funder and the lac... | This is wash trading. The shared funding wallet created 2 hours before trading is a smoking gun: it shows these addresses are not independent market participants but a single entity distributing inventory. The tight 0.5% price range and rigid alternation every 5 blocks indicate programmed circular flow, not organic sen... | This is clearly wash trading because the volume is high and all trades are within a tiny price range. The shared wallet is suspicious, but the main proof is that the price barely moves — real trading would cause more volatility. Also, since only two addresses trade, it has to be fake. The lack of external counterpartie... | synthetic:deepseek/deepseek-v4-flash-0731 |
On a DEX pair for a meme token, the same two wallets trade back and forth 500 times in a day. The price never moves more than 0.2%. The wallets both started with funds from a common OTC desk. Is this wash trading? Consider whether the lack of price impact could also indicate a high-frequency market maker. | This is wash trading. Two wallets trading exclusively with each other 500 times in a day, with the price pinned within 0.2%, is a classic circular transfer loop. The common OTC desk funding means both wallets share a capital source, indicating affiliation and likely the same beneficial owner. A legitimate high-frequenc... | The lack of price impact could indicate a high-frequency market maker that is efficiently balancing its inventory. Two wallets trading back and forth might just be two sides of the same firm, but that's common in market making—wallets often split between buying and selling to manage risk. The common OTC desk could simp... | synthetic:deepseek/deepseek-v4-flash-0731 |
A pair's transaction list shows 500 trades per hour, all with sizes between 0.5 and 0.6 ETH, and all counterparties are among 15 addresses that were funded by the same initial faucet. No address ever holds tokens for more than 4 blocks. Weigh the timing and size clustering against the possibility of a professional trad... | The evidence strongly indicates wash trading. The critical red flags are (1) the extreme uniformity of trade sizes (0.5–0.6 ETH) across 500 trades—professional firms vary sizes based on order flow, market depth, and inventory, not a narrow band; (2) the shared funding source—if all 15 addresses came from the same fauce... | 500 trades per hour with sizes between 0.5 and 0.6 ETH could be a high-frequency trading firm using a grid strategy. Grid bots often use fixed lot sizes and execute many small trades to capture small price movements. The shared faucet might just be a testnet-like distribution for new wallets, and the short holding peri... | synthetic:deepseek/deepseek-v4-flash-0731 |
A user claims the LINK/WBTC pair is wash-traded because trade sizes are always multiples of 100. Investigate whether the transaction graph and funding relationships support that conclusion or if a legitimate market maker could produce this pattern. | The user's claim is premature. Round-number trade sizes alone are a weak signal because legitimate algorithmic market makers often use fixed lot sizes (e.g., 100 units) to simplify risk management and inventory accounting. The critical analysis must focus on the transaction graph: are the counterparties economically in... | The user is right. If every trade size is a multiple of 100, that is a huge red flag because real traders do not always use such clean numbers. It suggests that the same algorithm is placing both sides of the trade, intentionally keeping sizes uniform to avoid detection. The graph and funding relationships are secondar... | synthetic:deepseek/deepseek-v4-flash-0731 |
Evaluate the SPY-wrapped token pair on a DEX where the top 5 trader wallets all received their initial ETH from the same 0x…9f address within a 10-block window. Those wallets then trade among themselves with increasing amounts, but each also has a few trades with external retail addresses. Explain how you would weigh t... | The shared funding is the dominant signal. A single 0x…9f address funding all five top wallets within 10 blocks indicates a coordinated cluster, likely controlled by one entity. The increasing trade sizes among these wallets suggest staged ramping to create volume. The few external retail trades are cosmetic—they provi... | The shared funding is suspicious, but the fact that each wallet trades with external retail addresses means there is real counterparty diversity. Since not all trades are circular, the volume is likely a mix of genuine and artificial. I would not call it wash trading because the presence of any external trades breaks t... | synthetic:deepseek/deepseek-v4-flash-0731 |
For a new DEX pair, you see that volume is $3M, but the number of unique traders is only 3. Each of these traders is a contract that was created by the same multi-sig wallet. The three contracts trade with each other in a cycle—1 buys from 2, 2 buys from 3, 3 buys from 1—each trade for exactly 1 ETH. This cycle is repe... | This is wash trading, and the reasoning must center on the deliberate loop structure and the single-entity control, not the volume. The fact that there are only 3 traders, all contracts created by the same multi-sig, proves they are not independent actors—they are controlled by one entity. The cycle (1 buys from 2, 2 b... | The volume is $3M and there are only 3 traders, which is a red flag because a real market would have more participants. All three traders are contracts created by the same multi-sig, so they are likely the same person. They trade in a cycle, which is suspicious because it's a repeating pattern. The tokens have no exter... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze the on-chain data for a pair where the graph of all transactions forms a star topology: one central wallet trades with 100 leaves, but each leaf also trades with the central wallet in a back-and-forth manner. The central wallet was funded with a single transaction, and each leaf was funded by the central wallet... | This is wash trading. The star topology is a hub-and-spoke circular flow: the central wallet trades with each leaf, and each leaf trades back, meaning the central wallet is both buyer and seller for every leaf. The funding relationship is decisive—the central wallet funded all leaves just 1 block before their first tra... | The volume is $100M with a $1M pool, which is a 100x ratio, that's a huge red flag. Also, the graph shows a star topology where one central wallet does all the trades with 100 leaves—that's a classic wash-trading pattern. The leaves are funded by the central wallet, so they're all the same person. The trade sizes are a... | synthetic:deepseek/deepseek-v4-flash-0731 |
A security analyst is investigating the 'ice/eth' pair after a social media post claims organic growth. The on-chain data shows that the top 50 trading wallets were all created by the same 'deployer' factory and funded by the same 'seed' wallet. They trade in a cycle that returns the same amounts, and no external walle... | This is unambiguously wash trading. The decisive factors are the factory creation and seed funding: every one of the top 50 wallets was programmatically created by the same deployer factory, meaning the entity behind them controls all of them. The fact that they were all funded by the same seed wallet within a short ti... | This might be wash trading, but the analyst should consider that some projects create multiple wallets for community engagement or airdrops. The fact that they all trade in a cycle could be a coincidence of the market structure. The seed wallet might just be a treasury that is distributing tokens to early adopters. Wit... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze a pair on a sidechain where the sequencer includes a 'private mempool' transaction from a single address that submits a buy order and a sell order for the same asset in the same block, with the sell order priced 0.01% higher. This repeats every block for 6 hours. Determine the wash-trading likelihood, factoring... | This is deliberate wash trading with near-certainty. The use of a private mempool is a deliberate attempt to bypass public order books and avoid front-running, but it also eliminates any possibility that the buy and sell orders are matched against independent external participants — the sequencer is processing self-tra... | This is likely wash trading because the same address is both buying and selling with a tiny price difference. It repeats every block, so it's clearly a bot. The lack of external orders makes it fishy, so I'd say it's fake volume. | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a pair on Base, you find that a cluster of 7 wallets—all funded by a single 'deployer' address—execute trades that are perfectly timed to one another: wallet 1 buys at t=0, wallet 2 sells at t=1 minute, wallet 3 buys at t=2 minutes, etc., each for the same amount. The sequence repeats every 7 minutes. There are n... | Yes, this is wash trading. The shared funding from a single deployer address establishes that all 7 wallets are under common control—they are not independent market participants. The perfect timing (wallet 1 buys, wallet 2 sells exactly 1 minute later, etc., repeating every 7 minutes) is deterministic coordination, not... | Consistent activity is a positive sign—it shows the pair has active traders and liquidity. The 7 wallets might be independent users who happen to have similar trading patterns, perhaps following a known strategy. The fact they're all funded by the deployer could be because the deployer is a big investor who supports th... | synthetic:deepseek/deepseek-v4-flash-0731 |
A researcher uses a simple heuristic: any pair where the same two wallets trade with each other more than 5 times per day is wash-traded. Applying this to a popular pair, they conclude it's manipulated. However, the pair has diverse funding (wallets from 100 different exchanges), and trade sizes vary widely. Provide a ... | The heuristic is flawed because it confuses frequency with wash intent. Legitimate market-making (MM) often involves two wallets—one as a bidder and one as an asker—trading repeatedly in narrow spreads to provide liquidity. The diverse funding from 100 exchanges suggests no single controlling entity, which weakens the ... | The researcher is wrong because 5 trades a day is not a lot. Actually, any pair with more than 5 trades is suspicious, but here the wallets are from many exchanges, so they are different people. The trade sizes vary, so it's not wash trading because wash trades are usually same-size. The diverse funding means it's clea... | synthetic:deepseek/deepseek-v4-flash-0731 |
An on-chain sleuth is probing a new DEX pair for 'volume mining' rewards. The pair's volume is composed of a 5-wallet ring where each wallet trades with the next in a fixed order (1-2-3-4-5-1). The ring only becomes active when a 'reward farming' event is announced, and deactivates otherwise. The wallets were funded fr... | The structural cycle is sufficient evidence of wash trading; the non-round sizing does not weaken the claim — it actually strengthens it. The fixed order 1-2-3-4-5-1 is a deliberate, pre-arranged ring: it has no organic market rationale. The activation only during reward farming events proves the intent is to farm rewa... | The 5-wallet ring is suspicious, but the non-round amounts like 0.33 and 0.66 ETH suggest real trading — people don't usually choose such odd numbers for wash trades. Also, Tornado Cash is used by many legitimate privacy-conscious traders. The ring only activates during rewards, but that could just be when liquidity is... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze the DEX pair for an NFT token. A user sees the transaction graph showing a 'loop' of wallets that mirrors the EIP-2981 royalty structure. They suspect the creator earns royalty on their own wash trades. Is this a wash-trading detection case? What specific evidence about funding and ceiling price recurrence is n... | Yes, this is a wash-trading detection case because royalty self-payment via circular trades is a direct financial motive for artificial volume. To confirm, we need two pieces of evidence. First, funding provenance: if the wallets in the loop are all funded by a common address controlled by the creator (e.g., the contra... | This is probably just a coincidence because NFT traders often use multiple wallets. The EIP-2981 structure is just about royalties, and if the creator is paying the royalty on their own trades, they're just losing money on the royalty fee. So it's not a wash-trading case because they'd be losing money. It's more likely... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given the on-chain data for the ATOM/ETH pair on a sidechain, you see that all trades are initiated by a single 'relayer' contract. The contract receives funds from a deployer address, then executes a buy and a sell of equal size within the same transaction. The deployer address has no other activity. Another analyst c... | This is wash trading, not a technical artifact. A legitimate relayer for cross-chain swaps would facilitate trades between distinct external parties, and it would not execute a buy and a sell of equal size against itself in the same transaction. Here, the relayer contract is the only actor, receiving funds from a dorma... | The relayer contract is a technical artifact because it is a smart contract designed to execute swaps. Cross-chain relays often work this way—they bundle buy and sell orders to settle across chains. The deployer being inactive is normal for a contract that is set up and left to run. This is not wash trading because the... | synthetic:deepseek/deepseek-v4-flash-0731 |
Assess the on-chain history for the 'uni/wbtc' pair on a DEX: a perpetual 'rebalancing bot' transfers WBTC between 12 cold wallets every 4 hours, with trade sizes alternating between $50,000 and $51,000, and all initial funds trace to a single Tornado Cash deposit. Determine if this is deliberate wash trading or organi... | This is deliberate wash trading, not organic rebalancing. Organic rebalancing would involve changing the asset mix relative to some benchmark, requiring external price feeds or portfolio targets. Here, the bot simply transfers WBTC between wallets of equivalent value with no external economic reason. The graph structur... | This is wash trading because the bot keeps trading between the same 12 wallets. The Tornado Cash deposit shows they want to hide, so it's definitely fake. Real rebalancing would not use Tornado Cash. The alternating sizes are just a trick. So it's deliberate. | synthetic:deepseek/deepseek-v4-flash-0731 |
A pair shows a high volume, and the overall trade sizes are not round numbers. However, you dig into the 'nonce' of the transactions. You find that for every successful trade, there are 50 failed transactions. The failed transactions are the same order, but with a manipulated price limit, and the successful one goes th... | The failed transaction telegraphing is a deliberate manipulation technique. The address submits 50 failed orders with deliberately manipulated price limits, which are picked up by block explorers and analytics tools as 'attempted' trades, creating a false impression of active supply and demand around certain price leve... | The non-round trade sizes suggest the volume might be organic, because real traders don't use round numbers. But the failed transactions are suspicious. There are 50 failures for each success, which is a lot. The failed orders have different price limits, so they might be just trying to find the right price. But since ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Consider a DEX pair where a smart contract acts as a 'market maker' that both buys and sells the token in alternating transactions, with each transaction involving the same counterparty wallet that was funded by the contract's deployer 3 blocks earlier. The contract's trades are 99% of the pair's volume, and the remain... | This constitutes wash trading. The core evidence is the funding link: the counterparty wallet was funded by the contract's deployer just 3 blocks before trading with the contract. This means the counterparty is not independent — it is an extension of the deployer, and the alternating buy/sell transactions are the same ... | The contract trades with a wallet funded by the deployer, which looks bad, but it's only 3 blocks — maybe the deployer is just moving funds around for operational reasons. The retail trader shows someone genuinely bought, so there is at least some real volume. 99% from the contract is high, but market makers often domi... | synthetic:deepseek/deepseek-v4-flash-0731 |
An auditor is looking at a pair and finds a 'fake-out' pattern where a single wallet sends tokens to itself via 10 different intermediary addresses, and each intermediary trades on the pair once. The total volume added is significant, but the net flow is zero. How would you weigh the circular transfer loop versus the f... | The circular transfer loop is the decisive factor, outweighing the superficial diversity of counterparty addresses. The fact that a single wallet funds all 10 intermediaries and the net flow is zero proves that the addresses are not independent actors—they are controlled by the same entity. The different counterparty a... | The different counterparty addresses are more important. Since each trade has a different address, the exchange sees 10 distinct actors trading, which means there is genuine external participation. The circular transfer loop is just a way to move tokens between wallets, but the actual trades are between different parti... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze the on-chain record for the UNI/WETH pair on a rollup. A cluster of 8 wallets, each funded 24 hours ago by a single deposit from a central exchange, execute a sequence of identical-sized trades (0.25 ETH) that form a closed loop over 6 blocks. The loop resets every hour. Provide a detailed wash-trading verdict. | This is wash trading. The decisive evidence is the closed loop over 6 blocks with identical 0.25 ETH sizes and a fixed cadence (hourly resets). The single funding source from a CEX 24 hours prior creates a clear funding graph: all 8 wallets fan out from one deposit, and their trade sequence forms a cycle where the last... | This is wash trading because the wallets are funded by one CEX and trade in identical sizes. However, some might argue it could be a market-making bot, but the identical sizes and closed loop are too systematic. It's clearly fake volume. | synthetic:deepseek/deepseek-v4-flash-0731 |
Assess the 'usdc/wbtc' pair on a low-liquidity Avalanche DEX. A cluster of 50 wallets, all created 48 hours prior, display a perfect cycle: A buys from B, B buys from C, C buys from D, ... back to A, with trade sizes alternating 1,000 / 5,000 / 10,000 USDC. The pool's own fee tier is set to 0.01%, and the protocol's na... | The naive analyst's claim is incorrect. While tight spreads and smooth volume are superficially consistent with market-making, the structural evidence overrides that. The 50 wallets were all created 48 hours prior, indicating a freshly assembled cluster, not established market makers with track records. The perfect cyc... | The tight spread and smooth volume actually suggest a professional market maker is operating. Market makers often use multiple wallets to manage inventory, and the 0.01% fee tier is normal for stablecoin pairs. The 48-hour creation time doesn't matter because new market makers can start anytime. The alternating sizes a... | synthetic:deepseek/deepseek-v4-flash-0731 |
A user asks: on the ‘stg/eth’ pair, the 24h volume is huge, but every transaction is between two addresses that are both funded by the same ‘bridge’ contract, and the time between a buy and the corresponding sell is always under 1 second (possibly same block). The trade sizes are all approximately 1000 ETH, but the poo... | Yes, this is wash trading. The evidence is overwhelming. The bridge funding ties both addresses to the same source, proving affiliation. The sub-second round-trip (under 1 second, possibly same block) means the same assets are being exchanged back and forth without any market risk or genuine counterparty—no real trader... | This is definitely wash trading because the volume is huge but the liquidity is tiny, which doesn't make sense. The addresses are both funded by the same bridge, so they're probably controlled by the same person. The trades are all the same size and happen every second, which is suspicious. The volume-to-liquidity rati... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze the UNI/USDC pair on an L2. A wallet cluster (all funded by the same 'treasury') trades only during the final minute of each hour, creating a volume spike. The trade sizes are 50,000, 50,001, 50,002, and so on, after the first trade. The rest of the hour has no trading. Determine if this is wash trading, and ex... | This is wash trading. The key signal is the deliberate, non-random structure: the cluster shares a single treasury funder, trades only in a fixed 1-minute window, and uses sequentially incrementing sizes (50,000 → 50,001 → 50,002). That pattern is statistically impossible for organic flow—real liquidity providers and a... | This is clearly wash trading because the volume spike happens only in the last minute of every hour. High frequency is the main red flag—normal pairs don't have such a regular pattern. The sequential sizes, like 50,000 and 50,001, also look suspicious because they're too neat. The treasury funding just adds more eviden... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given the on-chain data for the LRC/WETH pair, I find that 90% of the trading volume comes from a single pair of addresses that alternate buy and sell orders every 30 seconds, and their funds can be traced to a common origin 2 weeks prior. However, they also trade with other organic users a small fraction of the time. ... | The dominant feature is the 90% volume concentration from two addresses with alternating 30-second buy/sell cycles and a shared funding source two weeks back. That pattern is classic wash trading: the same economic entity transacts with itself to fabricate volume, with the small organic trades serving as a veneer of le... | Since 90% of the volume comes from those two addresses, it's clearly wash trading. They are trading back and forth with each other, and the 30-second alternation is suspicious. The fact that they share a common origin confirms they are the same person. The small organic trades don't matter because the volume is mostly ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Look at the BAL/WETH pair on Ethereum. There's a cluster of 15 wallets, all funded by a single address that received ETH from a mining pool 3 days ago. These wallets execute a total of 2,000 trades in 24h, but each wallet only ever trades with two other wallets in the cluster, forming a chain: w1 buys from w2, w2 sells... | This is deliberate circular flow, not coincidental liquidity. The decisive factors are (1) the closed set of 15 wallets with no external counterparties, (2) the chain topology where each wallet only trades with two other cluster members, and (3) the uniform 0.5 ETH trade size. In organic liquidity, trades would involve... | This is probably just a coincidental flow because the wallet addresses are different and they all came from a mining pool, which is a public and legitimate source. The chain pattern (w1 buys from w2, w2 sells to w3, w3 buys from w1) might just be how liquidity happens to flow when there are few active traders on a smal... | synthetic:deepseek/deepseek-v4-flash-0731 |
Investigate a pair on a decentralized options platform where the 'activity' is a series of call and put options being bought and sold simultaneously by two addresses that share a common funding source. The 'volume' is the notional value of these options. Is this considered wash trading in the options context? | Yes, this is wash trading in the options context. The defining feature of wash trading is the absence of a real transfer of risk or beneficial interest between independent parties. Here, the two addresses share a common funding source, meaning the same entity controls both sides. Buying a call and selling a put (or vic... | This is not wash trading because options are derivatives, and the notional value is just a reference amount, not actual money changing hands. The two addresses are buying and selling different options (calls and puts), so they are not the same trade. They share a funding source, but that does not mean they are the same... | synthetic:deepseek/deepseek-v4-flash-0731 |
A compliance team is reviewing a 'AXS/DAI' pair on a gaming DEX. They find that a cluster of 20 wallets, all created by the same 'faucet' contract, execute trades in a rigid sequence: wallet 1 buys from wallet 2, wallet 2 buys from wallet 3, ... wallet 20 buys from wallet 1, then the cycle repeats every 10 minutes. Eac... | The junior analyst's reasoning is flawed because trade-size variance does not negate wash trading—what matters is the circular flow of value with no independent market participant. The fixed 10-minute loop of 20 wallets, all spawned by the same faucet in the same block, creates a closed system where every buy is matche... | The junior analyst is correct—since the trade sizes differ, this is not a mechanical wash trade. Wash trading requires identical sizes to be illegal, and the differing amounts indicate real market orders from differently-sized participants. The faucet funding is just a faucet, common in gaming DEXs for onboarding, and ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Investigate the WBTC/DAI pair where a concentrated liquidity provider places orders on both sides and frequently trades with itself via a second wallet. Decide if this is organic market making or wash trading, and explain the role of external arbitrageurs in your verdict. | The presence of a single LP crossing orders between two wallets they control is a strong wash-trading signal, but the decisive factor is the absence of external arbitrageurs. In organic market making, a two-sided quote invites adverse selection — external traders arbitrage away temporary price dislocations, generating ... | Given the LP placed orders on both sides and traded with themselves, this is clearly wash trading. The high frequency of self-trades and the fact that the same wallet is on both sides of the trade proves it. There is no need to check external traders because the activity is obviously artificial. | synthetic:deepseek/deepseek-v4-flash-0731 |
A quant fund analyst is assigned to find wash trading in a new DeFi project. The analyst looks at the BAL/WETH pair and sees that volume is high, but the trade sizes are not round numbers. However, upon deeper analysis, they uncover that every single trade's 'from' address is the same as the 'to' address of the trade i... | The block-level linkage is the smoking gun because it proves a single entity is controlling both sides of the trade. When the 'from' of trade N is the 'to' of trade N-1, and the initiator for both is the same CEX hot wallet, it means the CEX wallet is not just facilitating—it's the actual counterparty on both legs. Thi... | The CEX hot wallet might just be executing trades on behalf of many users, which would explain why the 'from' and 'to' addresses align across blocks. It could be a matching engine that batches orders. The trade sizes not being round numbers suggests genuine market orders, not pre-planned wash trades. The block-level li... | synthetic:deepseek/deepseek-v4-flash-0731 |
For an illiquid pair, a single wallet places a series of limit orders on both sides of the order book, then executes trades against itself to move the price. The wallet's own inventory stays flat. Determine whether the price manipulation attempt also constitutes wash trading, and explain the relationship between the tw... | Yes, this constitutes both wash trading and price manipulation, and they are interrelated but distinct. Wash trading is the act of executing trades with oneself to create artificial volume and activity. Price manipulation is the intent to distort the market price for profit or to mislead others. Here, the wallet places... | This is not wash trading because the wallet is placing genuine limit orders, and limit orders are part of normal trading. The wallet is just a smart trader that knows how to move the price. Wash trading requires buying and selling at the same time, but here the wallet is placing orders on both sides, which is market ma... | synthetic:deepseek/deepseek-v4-flash-0731 |
A blockchain analytics firm is tasked with flagging wash trading on the 'ethereum/wbtc' pair for a 24/7 monitoring service. Their heuristic flags any pair where the ratio of unique traders to trade count is below 1:100. The pair exhibits exactly 10,000 trades from 300 unique wallets, but all wallets have a common 'root... | The heuristic (unique traders to trade count) is a useful surface-level filter, but it fails to capture the actual manipulation here because 300 unique wallets out of 10,000 trades is a 1:33 ratio, which would not trigger the heuristic—yet this is clearly wash trading. The graph analysis reveals the smoking gun: all 30... | The heuristic is working correctly because the ratio of 300 unique traders to 10,000 trades is 1:33, which is actually better than the 1:100 threshold, so the pair should not be flagged. The common root address might just be a popular faucet or airdrop contract that many independent users use to get initial funds. The ... | synthetic:deepseek/deepseek-v4-flash-0731 |
On the polygon PEAR/ETH pair, I observe that 90% of buy volume comes from a single wallet that receives its funding from a decentralized mixer, and 90% of sell volume comes from another wallet that sends its proceeds to the same mixer. The buy and sell sequences are perfectly offset in time. Provide a concise reasoning... | This is wash trading because the two primary wallets are economically linked through the mixer: the buy wallet's funding originates from the mixer, and the sell wallet's proceeds flow back to the same mixer. This creates a circular flow — the mixer acts as a hub laundering the value between the two sides, obscuring the... | High volume is often suspicious, and this pair has huge volume, so it's likely wash. The fact that the buys and sells are offset in time confirms it — organic markets don't have that kind of symmetry. The mixer involvement is also a red flag. I'd say it's wash because the volume is just too high and the patterns are to... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a DEX pair with an unusually high turnover rate, the on-chain data shows that 85% of all trades are between two wallets that are both controlled by the same private key (proven by identical signing patterns). The trade amounts are always 1,000 tokens, and each trade alternates between buy and sell. A user suggest... | This is unequivocally wash trading. The identical signing pattern proves the two wallets are controlled by the same private key—meaning the same entity is both buyer and seller in every trade. The claim of 'accidental self-matching' is absurd: a bot cannot accidentally sign transactions with the same key thousands of t... | It's possible this is a high-frequency trading bot with a bug that causes it to self-match. The alternation between buy and sell could be a glitch in the bot's logic, and the fixed size of 1,000 might be a parameter setting. The identical signing pattern just means the same key controls both wallets, but maybe the bot ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a pair's on-chain data, you find a 'ping-pong' pattern: wallet A buys from B, B sells to A, repeated 50 times, with trade sizes increasing by exactly 1% each round. Both wallets were funded within the same block from a shared EOA. Is this wash trading? What alternative explanations exist? | This is wash trading. The ping-pong pattern (A↔B, 50 times) with exactly 1% size increases is a predetermined algorithm, not organic trading. The shared EOA funding within the same block proves the wallets are affiliated—they're two sides of the same entity. The 1% increase is likely to simulate price growth and attrac... | This could be a market maker testing a strategy. The 1% increase might be a way to build volume gradually to attract liquidity. The shared EOA is suspicious, but it could be a single entity running a legitimate market-making bot—many MMs use multiple wallets. The ping-pong pattern is common in market making where you b... | synthetic:deepseek/deepseek-v4-flash-0731 |
On the MATIC/ETH pair, I find that 70% of buy orders are matched with sell orders from the same address, but the address uses different sub-accounts via a relayer. The relayer's fee payments all go to the same destination. Does this constitute wash trading? How would you argue it isn't? | Yes, this constitutes wash trading. The key finding is that 70% of buy orders are matched with sell orders from the same address, but using sub-accounts via a relayer is a classic technique to mask identity. The fact that the relayer's fee payments all go to the same destination proves that the sub-accounts are linked ... | This is not wash trading because the trades are executed via a relayer, which is a legitimate service. The sub-accounts might belong to different users who happen to use the same relayer, and the fee payments going to one place is just how the relayer collects fees. It's possible the matching is coincidental due to hig... | synthetic:deepseek/deepseek-v4-flash-0731 |
A pair shows a sudden volume surge that correlates exactly with the token team's marketing announcement. You trace the volume and find 3 addresses that alternate trades of identical quantities. These addresses were funded by the team's operational wallet. Evaluate whether the volume is genuine or a pump-and-wash scheme... | This is a classic wash-trading scheme. The funding from the team's operational wallet to the 3 addresses establishes a clear affiliation, and the alternation of identical quantities between those addresses is a textbook circular-trading pattern. Marketing announcements often coincide with wash trading to create the ill... | The correlation with marketing suggests the team is good at promoting, and the high volume is a sign of successful outreach. The 3 addresses might just be enthusiastic early investors who happened to buy the same amount. The identical quantities could be due to a limited token supply or a common trading bot strategy. U... | synthetic:deepseek/deepseek-v4-flash-0731 |
Imagine a pair where a bot creates a new wallet for every trade, each funded by the same EOA, and each wallet executes a single buy-then-sell cycle. The bot's goal is to simulate organic volume. The graph shows a star pattern from the funder to many leaves. Is this wash trading? Justify by weighing the circular flow an... | This is unambiguous wash trading. The star pattern—one EOA funding every leaf wallet—is a definitive control structure. Each leaf's single buy-then-sell cycle is a closed loop: the wallet buys from the pool, then sells back, returning to the same position, with the EOA providing the initial capital. The circular flow i... | This is not necessarily wash trading because the bot creates new wallets each time, which means there are fresh addresses participating. The star pattern could just be a faucet that funds new traders. Each wallet does one buy and one sell—that's a normal round trip. The bot might be a market maker that is just providin... | synthetic:deepseek/deepseek-v4-flash-0731 |
On a DEX, two wallets trade 1,000 times per day, but they are linked by a complex web of intermediate wallets 10 hops away. A graph theorist says the distance is too far to establish affiliation. A money-laundering specialist says the funding path is traceable and the endpoint wallets are both controlled by the same en... | The money-laundering specialist is correct. Graph distance alone is not a sufficient criterion for affiliation in wash trading. What matters is the actual flow of funds: if a traceable funding path connects the two endpoint wallets back to a single controlling entity, then the intermediate hops are just obfuscation lay... | The graph theorist is correct because 10 hops is too many to reasonably link the wallets. In on-chain analysis, standard practice is to only consider wallets within 2 or 3 hops as affiliated; anything beyond that is speculative and not actionable for enforcement. The money-laundering specialist is overcomplicating thin... | synthetic:deepseek/deepseek-v4-flash-0731 |
On a DEX pair, all trades occur between two entities: Entity X (institutional) and Entity Y (retail). They have no shared funder, but Entity Y's wallet is always topped up from a contract that receives funds from Entity X's treasury. A legal analyst says the funding path is indirect, so no affiliation. A financial crim... | The financial crime investigator is more accurate. The affiliation analysis must look at economic reality over formal legal structures. Even though there is no direct shared funder, the mere fact that Entity Y's trading capital is exclusively sourced from a contract that is funded by Entity X's treasury creates a causa... | The legal analyst is more accurate. Affiliation requires a direct link, such as shared ownership or a common funder, and here the funding is indirect—Entity Y's funds come from a contract, not directly from Entity X. The treasury funding the contract might be a one-time grant, after which the contract operates independ... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the 'grt/eth' pair, you observe that the funding for all active traders comes from a single 'treasury' address, but the treasury does not trade itself. The traders trade at different times and with different sizes. However, at the end of each day, the treasury extracts the exact sum of profits/losses from the trade... | This is a scheme to redirect value, not genuine trading. While the traders appear independent (different times/sizes), the treasury funding all of them and then extracting the exact daily P&L from each wallet proves they are controlled agents. The treasury acts as a central coordinator: it provides capital, collects al... | The volume could be genuine. The treasury funding traders is normal for market-making or liquidity provision—a central entity often allocates capital to multiple strategies. The traders trade at different times and sizes, indicating independent decision-making. The daily extraction might be a fee settlement or profit-s... | synthetic:deepseek/deepseek-v4-flash-0731 |
A compliance analyst at a crypto exchange is reviewing the 'troll/usdc' pair. The transaction graph shows that 85% of the volume is generated by a cluster of 8 wallets where each wallet's first transaction is a small deposit from a common 'treasury' address, then the wallets trade among themselves in a prime-numbered s... | The cluster's activity constitutes wash trading. Graph-theoretically, the 8 wallets form a strongly connected component with a single source (treasury), and their trade sizes follow a deterministic prime-multiple sequence (0.1, 0.2, 0.4) repeating every 30 minutes—this is a signature of algorithmic self-trading, not or... | The cluster is suspicious but maybe they are just active traders. The prime number sizes are a bit weird, but some people do that. The organic volume is there too, so the pair isn't fully fake. I'd say it's a mix, but since 85% is from the cluster, it's mostly wash. The treasury funding is a red flag, but not conclusiv... | synthetic:deepseek/deepseek-v4-flash-0731 |
You are charting a pair's transaction graph. You find that a single controller address deploys a new contract each day to trade against its own previous day's contract, creating a new 'round' of artificial volume. Is this distinguishable from organic flow by looking at the funding origin of the new contract's gas, even... | Yes, the funding origin is a decisive forensic signal. In organic flow, a new contract deploying daily would typically be funded from a fresh, independent wallet with a distinct source of capital (e.g., a new exchange withdrawal, a different payroll address). Here, if the new contract's gas is funded from the controlle... | Yes, because if the gas comes from the same address, it's obviously wash trading. Any new contract that is funded by the same controller is clearly fake volume. The volumes might be diverse but that doesn't matter—the funding link is enough to prove it. You just need to check if the gas address matches the previous con... | synthetic:deepseek/deepseek-v4-flash-0731 |
Look at the COMP/USDC pair on an L2. All large trades happen in a tight cluster of 5 wallets that share a common smart-contract factory. The trades are in round numbers (10k, 20k, 50k) and the counterparties are always the same 5. Does this indicate artificial volume? Explain your reasoning. | Yes, this is a textbook wash-trading pattern. The shared factory contract proves a single deployer controls all 5 wallets, so they're economically equivalent—any trade among them is a book-keeping entry. Round-number sizes (10k, 20k) show no price discovery or inventory optimization; real trading sizes follow inventory... | Round numbers and shared factory just mean the wallets are sophisticated—perhaps they're a professional market-making firm that deploys multiple strategies. The L2 has low fees, so trading in round lots is efficient. The fact that they always trade with each other could be because they're the only ones providing liquid... | synthetic:deepseek/deepseek-v4-flash-0731 |
Evaluate the OCEAN/DAI pair. A large trader address has a funding relationship with a 'treasury' that also funds 50 other addresses. The trader and these 50 addresses trade with each other, but each trade has a different size and time. A novice analyst says the pair is clean because the trades are not mirror images. Yo... | This is wash trading. The novice analyst's focus on 'mirror images' is a superficial test; wash trading does not require identical sizes or simultaneous timestamps. The decisive factor is the net flow of zero over 7 days between the trader and each of the 50 addresses. This means that despite varying trade sizes and ti... | This is not wash trading because the trades are not mirror images—they have different sizes and times. The novice analyst is right to say that because real wash trading has identical buy and sell orders. The net flow being zero over 7 days might actually be a good sign because it means the trader is not taking a direct... | synthetic:deepseek/deepseek-v4-flash-0731 |
An on-chain analyst is using a Dune dashboard that shows the pair's top 100 trades by gas used. They notice that 95 of these trades have gas prices that are within 0.1 Gwei of each other, and the 'from' addresses of these trades all have a common ancestor wallet that was funded 2 blocks before the first trade. The rema... | The gas price clustering is a strong corroborating signal, but the funding link is the primary evidence of wash trading. Gas price clustering within 0.1 Gwei across 95 trades suggests these transactions were submitted by the same entity or coordinated bots using similar gas settings, likely to execute a pre-planned seq... | The gas price clustering is the definitive proof of wash trading. When 95 trades share nearly identical gas prices, it clearly means the same person sent all those transactions, because no external trader would consistently pick the same gas price. The funding link is just extra confirmation but not necessary. The 5 sm... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the ENS/ETH pair on a testnet, you see that trades happen in perfect 60-second intervals, alternating between buy and sell. The counterparties are wallets from the same cluster (all funded by the same faucet). However, a community member argues this is a 'stress test' or 'bot testing' activity and not market manipu... | This is wash trading. The defining feature is not the testnet environment but the circular, self-referential nature of the trades: the same cluster of wallets—all funded by a single faucet—alternate buy/sell at fixed 60-second intervals. There is no external counterparty with independent intent; every trade is matched ... | This is not wash trading because it’s happening on a testnet, and testnets are for experimentation. The bot is clearly running a stress test to see if the DEX can handle frequent trades, which is a legitimate activity. A volume-to-liquidity ratio test would be sufficient because if the pair has enough liquidity to supp... | synthetic:deepseek/deepseek-v4-flash-0731 |
A DeFi data analyst is evaluating a pair using a custom query that filters for trades where the 'from' and 'to' addresses have a known funding connection via a 'disperse' contract (a contract that splits tokens to many addresses in one transaction). They find that 80% of the pair's volume involves such connections, and... | The 80% connected volume is wash trading. The use of a disperse contract to fund multiple addresses from a single source is a direct evidence of common beneficial ownership. When these funded addresses then trade with each other — and the trades cluster within 1 second of each other — it indicates a coordinated, pre-pr... | The 80% is definitely wash trading because they are all connected through the disperse contract. That's a clear sign they are the same person. The 20% is probably also wash trading but just harder to see the connection. I'd say the whole volume is fake just to be safe. The 1-second timestamps are suspicious, but it's n... | synthetic:deepseek/deepseek-v4-flash-0731 |
On a newly listed token pair, you notice that a single 'deployer' address has created 500 different 'trading' contracts, each funded with 0.01 ETH. Each contract trades once per hour, buying 0.01 ETH worth of tokens and immediately selling them at a small loss (e.g., 0.5% less). The net effect is that the token's price... | This is deliberate wash trading. The intent is evident from the structure: 500 contracts all funded by the same deployer, each buying and selling the same amount at a consistent small loss, creating a sawtooth price pattern. A genuine market maker would seek to profit from the spread, not systematically lose 0.5% on ev... | This could be inefficient market-making rather than wash trading. The deployer created 500 contracts to spread out trades, which is a common technique to avoid moving the price too much. The small loss per trade (0.5%) might be the cost of providing liquidity in a thin market, and the sawtooth pattern could just be nat... | synthetic:deepseek/deepseek-v4-flash-0731 |
A new dYdX-style perp DEX lists a token. The funding rate is 0.1% every hour. A single trading firm is the dominant taker on both sides. Their orders are always filled by two of their own subaccounts. Their dispute is that they are providing liquidity by atomically creating and taking a tiny loss on a wide spread. Is t... | The defense is invalid. The purpose of a genuine liquidity provider is to earn the spread by taking on inventory risk and adverse selection—they quote a bid and ask, and they profit if the market is stable and lose if it moves against them. Here, the firm is both the maker and the taker, atomically, on both sides. Ther... | The defense is valid because they are providing liquidity by putting quotes on both sides. If they take a small loss, that's part of market-making. The funding rate is high, so they are just trying to capture funding. It's not wash trading because they are using two different subaccounts, which are separate accounts. | synthetic:deepseek/deepseek-v4-flash-0731 |
I researched a pair on a low-cap DEX. I found that the top 100 trading addresses all have the same asset balance, and the balance is either 100% of a specific token or 0. All of them were funded from one wallet. The pair has no external token transfers from non-mixer addresses. Does the lack of 'external counterparties... | The lack of external counterparties is a strong indicator but not automatic proof of wash trading. The pattern you describe—top traders all having identical balance structures (either 100% token or 0) and all funded from one wallet—suggests that the trading activity is not organic. However, an alternative explanation t... | Yes, it's automatically wash trading. If there are no external counterparties and all the wallets are funded from one wallet, then it's all fake volume. There's no other explanation for that pattern. The balance being either 100% or 0 is clear evidence that they are all the same person. So it's definitely wash trading. | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a DEX pair where the top 10 trading wallets all have their initial funds from a single ICO address, and they execute a round-robin of trades that net to zero, but with random-sized amounts, is this likely wash trading? Discuss the importance of the funding graph versus trade-size patterns in your reasoning. | This is highly likely wash trading. The funding graph is the decisive factor: all 10 wallets receive their entire initial balance from the same ICO address at nearly the same time, indicating a single controlling entity. The round-robin sequence—where each wallet trades with the next in a fixed cycle—creates a closed l... | The random sizes suggest it might be legitimate, because real traders don't always trade the same amounts. The fact that they net to zero could just be coincidence over a day. The ICO address funded them, but that was a long time ago, and people can sell their ICO tokens to different buyers. Since there are 10 differen... | synthetic:deepseek/deepseek-v4-flash-0731 |
A security researcher claims a pair is wash-traded because 95% of volume comes from 10 wallets, but those wallets have no shared funding history and interact with different external exchanges. The trade sizes vary widely. Build a counter-argument that this could be organic market-making, and state what on-chain evidenc... | The claim is premature without graph-level circularity. Ten wallets with no shared funding and distinct exchange counterparties can be independent market makers—they may be competing to quote spreads, and varied sizes reflect inventory management, not manipulation. Wash trading requires deliberate circular flow: tokens... | Ten wallets accounting for 95% of volume is a clear red flag—that concentration alone is enough to conclude wash trading. Even if they have different exchange histories, they could be using those exchanges to mask their coordination. The wide trade sizes actually make it worse, because it shows they are trying to simul... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the LINK/DAI pair on a DEX, you observe that a cluster of 20 wallets all have their initial funds from a single 'mixer' contract. After funding, they trade among themselves in a circle, but each trade moves the price by a small amount and then back. A security researcher claims this is wash trading because of the c... | To quantify artificial volume, I would construct a directed graph of all trades over the observation window, labeling each wallet as 'cluster' (funded by the mixer) or 'external'. I would then classify any trade where both counterparties are cluster wallets as suspicious, but not automatically artificial—because small ... | Since there is a clear circular flow among the 20 wallets, the whole pair's volume is 100% wash traded. The external parties don't matter because the cluster is doing most of the trading. Just count the number of trades between the cluster wallets and divide by total trades—that gives you the artificial percentage. The... | synthetic:deepseek/deepseek-v4-flash-0731 |
Review the data for the SHIB/ETH pair on a DEX that has a 1% fee. You see a wallet that buys SHIB, then a few minutes later sells SHIB, always losing 1% in the process. It does this in a loop 500 times a day. The total volume from this single wallet is $10M, while the organic volume is only $100k. The wallet is funded ... | This is a highly inefficient but effective wash-trading scheme. The wallet is deliberately paying a 1% fee on every trade, losing 1% of the notional each cycle, yet it does this 500 times daily. The losses are the cost of fabricating volume. The fact that the wallet is funded by the same EOA that created the pool is co... | This could be a market maker trying to provide liquidity, and the 1% loss is the cost of doing business. The wallet might be rebalancing its inventory, even if it seems inefficient. The fact that the pool creator is funding the wallet could just be the creator providing liquidity. 500 trades a day is a lot, but maybe i... | synthetic:deepseek/deepseek-v4-flash-0731 |
Produce a final wash-trading report for the ENS/DAI pair. The transaction graph is a perfect star topology: the core wallet funds 50 'spokes' every hour, and each spoke trades with exactly two other spokes before returning to the core. The core then withdraws the DAI. The alternative explanation is that each spoke is a... | The star topology is dispositive. Each spoke receives its funding from the core within a single block, executes a fixed two-hop trade with two other spoke addresses, and then sends the DAI back to the core. This is a textbook circular transfer loop. The 'separate fund' alternative fails because a genuine portfolio mana... | The star topology is suspicious but the volume could still be real if each spoke is a separate fund. The fact that they share a core wallet just means the manager is consolidating accounting. Many funds use a master wallet for gas and distribution. The trades are between different addresses, so there's no immediate sec... | synthetic:deepseek/deepseek-v4-flash-0731 |
On the XRP/DAI pair (via a tokenized bridge), every transaction is preceded by a 'funding' transaction from a central treasury that sends exactly 1,000 DAI to a fresh address. That address then trades once, and the proceeds are returned to the treasury. The trade sizes are all 1,000 DAI, and there are no other particip... | This is a textbook wash-trading pattern because it satisfies all three core criteria: (1) circular flow—funds originate from the treasury, go to a fresh address, trade once, and return to the same treasury, forming a closed loop with no external counterparty; (2) affiliation—every address is provably controlled by the ... | You can present it by just stating that the treasury is creating fake volume. The key evidence is that every address is funded by the treasury, and they only trade once before returning the money. That's obviously wash trading because there are no real buyers or sellers. You don't need to do anything complicated—just s... | synthetic:deepseek/deepseek-v4-flash-0731 |
Assess the THETA/ETH pair where the 24h volume is entirely made up of trades that are exactly 10, 20, or 30 ETH in size. All traders are funded by a single 'treasury' wallet that sends them ETH with a 1-block delay before each trade. Determine if the round-number sizing and funding timing prove wash trading. | The round-number sizing and 1-block funding delay are strong evidence of wash trading, but not conclusive on their own. The key is the combination: a single treasury funding every trader immediately before each trade means all traders are controlled by the same entity, eliminating the possibility of genuine external co... | The round numbers and the 1-block funding delay are clear proof of wash trading. No genuine trader would use only 10, 20, or 30 ETH sizes, and the treasury funding everyone means it's all fake. The timing is too regular to be organic, so I'd say this is definitely wash trading. | synthetic:deepseek/deepseek-v4-flash-0731 |
An exchange compliance officer is looking at the 'sol/usdc' pair on a Solana DEX (hypothetically bridged). The on-chain data shows a set of 3 wallets that trade in a perfect cycle: A buys 1000 USDC worth of SOL, B sells 1000 USDC worth of SOL, C buys it back, and the cycle repeats every 10 seconds for an hour. These wa... | This is textbook wash trading. The three wallets—A, B, C—have no other transaction history and are all funded by the same 'liquidity sweep' contract, establishing a single controlling entity. The perfect cycle, repeating every 10 seconds for an hour, with identical 1000 USDC amounts, creates a closed loop with no exter... | The perfect cycle with identical sizes is very suspicious, but the 999 variation would make it less wash-like. If the sizes differ, it suggests that the trades are not perfectly offset, which could mean there is a real economic reason for the trades, like taking advantage of small price differences. The wallets having ... | synthetic:deepseek/deepseek-v4-flash-0731 |
You are an on-chain analyst for a regulatory body. The 'usdt/weth' pair on a Polygon DEX shows that all trading volume comes from a single 'market maker' contract that uses a 'disperse' function to send tokens to 1,000 different EOAs, each of which trades immediately. The EOAs then return the proceeds to the same marke... | The circular transfer loops and common funding are decisive evidence of wash trading. The market maker contract disperses tokens to 1,000 EOAs, which trade immediately and return proceeds within 5 minutes—this is a classic circular flow. The EOAs are not independent; they are funded by and return to the same source, so... | The 1,000 different EOAs actually suggest diversity, which is a positive sign. The market maker might be a legitimate liquidity provider that uses a disperse function to distribute tokens to many small traders to improve market depth. The fast return of proceeds (within 5 minutes) could simply be efficient settlement, ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Investigate a pair where the 24h transaction list shows that 80% of the volume is from a single address that buys large amounts, while 20% is from a single address that sells identical amounts, alternating every few seconds. These two addresses are funded from the same wallet within a short time span. The trade sizes a... | This is wash trading. At first glance, it appears to be two-sided flow—one buyer and one seller—but the graph reveals a circular transfer loop: the same wallet funded both addresses, and they alternate buying and selling identical amounts (2.5 ETH) every few seconds. This is not two independent parties meeting; it is o... | This is wash trading because the volume is $50M but the liquidity is only $5k, which is impossible for real trading. Also, the same two addresses are doing all the volume, and they're funded from the same wallet, so they're clearly related. The alternating buys and sells every few seconds with the same size is a patter... | synthetic:deepseek/deepseek-v4-flash-0731 |
Look at a DEX pair on a sidechain. The sequencer's data shows that 99% of the 24h volume comes from a single smart contract that repeatedly calls a 'swap' function between 20 pre-approved addresses. Each address was funded by the contract deployer. The trade amounts are all equal to the pool's max swap size. The volume... | This is wash trading because the graph shows a closed loop: a single smart contract controls all 20 addresses, and each address was funded by the contract deployer. The contract repeatedly swaps between these addresses, but since they are all pre-approved and funded by the same entity, every trade is a transfer between... | The volume is $10M and the market cap is only $100k, so that's a huge ratio. That alone suggests wash trading, because no real token with such a small market cap would have that much volume. Also, the fact that a single smart contract is doing 99% of the volume is suspicious—it's clearly a robot. The addresses are fund... | synthetic:deepseek/deepseek-v4-flash-0731 |
Review the on-chain activity for the MFT/ETH pair on a private blockchain. The consensus is Proof of Authority, and the sequencer addresses are run by the token team. You observe that the sequencer bundles a group of 5 trades that represent a perfect round trip: wallet A buys, B sells, C buys, D sells, and E buys, all ... | The team's claim is not credible. In a Proof-of-Authority network where the sequencer is controlled by the token team, the observation of a perfect 5-trade round trip returning to the same state is a direct admission of circular flow. The wallets A-E likely all belong to the team, and the bundling by the sequencer is n... | This could be organic market-making because market makers often trade in and out of positions to provide liquidity, and a round trip can happen naturally if the market is balanced. The fact that the sequencer bundles them might just be a technical optimization to save gas. Since the team runs the sequencer, they might ... | synthetic:deepseek/deepseek-v4-flash-0731 |
A data challenge: Given a 24h dataset for the RLY/ETH pair, you notice that the 'taker' address is the same as the 'maker' address for 95% of trades, but they are different versions of a smart contract that changes its own address each block (via CREATE2). The contract was deployed by a single EOA. Evaluate whether thi... | Yes, this is wash trading. The fact that the taker and maker are the same logical entity—a contract that rotates its address via CREATE2—does not hide the identity: a single EOA deployed it, and the same contract logic controls both sides. The 95% self-match rate is a strong signal of artificial volume. The address rot... | This is not wash trading because the taker and maker are technically different addresses—the contract changes its address each block, so the trades are between distinct addresses. The fact that a single EOA deployed it is not sufficient evidence; many legitimate protocols use factories. The high self-trade rate could b... | synthetic:deepseek/deepseek-v4-flash-0731 |
A forensic blockchain analyst is examining the 'aave/eth' pair. The transaction graph shows a 10-wallet cycle where wallet A sends to B, B to C, ..., J back to A, with the entire cycle completing in under 10 minutes. However, each trade in the cycle is interleaved with legitimate trades from unrelated parties, and the ... | Yes, this should still be flagged as wash trading, but with a nuanced caveat. The 10-wallet cycle completing in under 10 minutes is a smoking gun for a coordinated scheme, regardless of interleaving. The interleaving with legitimate trades is actually a common obfuscation technique—it doesn't invalidate the cycle; it j... | Since there are legitimate trades mixed in and the funding histories are diverse, it's hard to say it's wash trading. The cycle might just be a high-frequency trading strategy or an arbitrage loop that happens to complete quickly. The diverse funding from exchanges and DeFi suggests these are not all controlled by one ... | synthetic:deepseek/deepseek-v4-flash-0731 |
On a popular DEX, you see that 95% of the 24h volume comes from a single 'market maker' that places both sides of the order book. The MM's wallets are funded by a multi-sig that also pays for gas. A naive observer says this is fine because the MM is providing liquidity. Weigh the evidence: the MM's own inventory never ... | This is deliberate manipulation, not organic liquidity provision. The key evidence is the MM's inventory never changes — a genuine market maker expects to accumulate or shed inventory to facilitate external flow, but here the flat inventory means every fill on one side is matched by the MM's own opposite fill, creating... | The MM is providing liquidity because it places orders on both sides of the book, which is what market makers do. The fact that its inventory doesn't change just means it's well-hedged, and losing money on fees is normal for a market maker. The multi-sig funding just shows it's a professional operation. Without externa... | synthetic:deepseek/deepseek-v4-flash-0731 |
Evaluate the assertion that 'the WBTC/wstETH pair on a prominent DEX is wash-traded because the average trade size is 500,000 USD.' Contrast this volume-based claim with a graph-theoretic analysis that would be needed to actually verify the wash component, citing what specific evidence would make it definitive. | The volume-based claim is insufficient because large average trade sizes on a WBTC/wstETH pair can be organic institutional flow, given the size of those assets. To verify wash trading, one must conduct a graph-theoretic analysis of the transaction network. Definitive evidence would include: (1) identifying a set of ad... | An average trade size of 500,000 USD is unusually large for a retail pair, so it must be wash trading. Legitimate traders do not trade that much in a single transaction. The volume is clearly artificial. | synthetic:deepseek/deepseek-v4-flash-0731 |
An analyst reviews the THORChain rune/wETH pair and finds that a single 'liquidity provider' address provides 80% of the depth, but also trades against itself via a bridge contract. The bridge fee is always 0.3%, and the trade sizes are 0.5, 1.0, 1.5, 2.0 ETH etc., all round numbers. Determine if this is wash trading, ... | This is wash trading. The critical finding is that the dominant LP address trades against itself via a bridge contract—that's a direct circular transfer, where the same entity is both buyer and seller. The round-number trade sizes (0.5, 1.0, 1.5 ETH) are a structuring technique to appear natural, but they are too unifo... | This might be wash trading because the LP trades with itself, but it also provides 80% of the depth, which is normal for a large market maker. Round numbers are common in trading (e.g., people like to trade 1 ETH). The 0.3% bridge fee is a real cost, so if they are paying fees, they might be doing legitimate arbitrage.... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given the transaction history for the FXS/DAI pair on an Ethereum fork, a bot trades back and forth between two addresses every 3 minutes. The buy and sell orders are exactly the same size, and the price barely moves. However, the two addresses were funded by different sources: one from a centralized exchange, one from... | The perfect mirror-trading pattern is far more probative than the different funding sources. A genuine trader would not buy and sell the exact same size at regular intervals with negligible price movement—that is a signature of self-dealing to create volume. The different funding sources are easily explained: a sophist... | This is not wash trading because the two addresses got money from different places. If the same person controlled both, they would have used the same source. A CEX withdrawal and a DeFi protocol loan are different origins, so the addresses are independent. The mirror trading could just be a market maker's algorithm tha... | synthetic:deepseek/deepseek-v4-flash-0731 |
Assess a pair where the 24h volume is $12M. The transaction graph reveals that all trades are between two clusters of wallets: cluster X (10 wallets) and cluster Y (10 wallets). Every wallet in X was funded by a single address, and every wallet in Y was funded by the same address as well. The trades are all of size 0.5... | This is wash trading. The graph topology is a perfect bipartite structure where cluster X always buys and cluster Y always sells, with all 20 wallets funded from a single address. This funding commonality is the smoking gun: no external participant would share a funder with all other traders in the graph. The trades be... | This is wash trading because the volume is $12M and the two clusters are both funded by the same address. The trades are all 0.5 ETH and happen within 2 seconds, which is too fast. Also, one cluster always buys and the other always sells, but net flow is zero, so it's fake. The volume is high, so it must be wash. | synthetic:deepseek/deepseek-v4-flash-0731 |
A compliance officer is analyzing the 'render/eth' pair on a DEX. They observe that a 'custodian' wallet deposits tokens and then trades them back and forth with a 'broker' wallet, both of which are owned by the same company. The trades are of equal size and occur at regular 10-minute intervals. Determine if this is wa... | This is wash trading. The company-owned wallets are the decisive factor: since both the custodian and broker are controlled by the same company, every trade between them is a transfer between related parties with no external counterparty. The regular 10-minute interval and equal trade sizes are the equivalent of a time... | The volume is what matters most. If the volume is high, the pair must be legitimate because the company is providing liquidity. The regular intervals and equal sizes might just be a market making strategy that is consistent. The fact that the wallets are company-owned is not a problem since many firms have market makin... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the TON/ETH pair, a user claims wash trading because the same address appears as both buyer and seller in a single block (self-trading). However, further inspection shows that address is a smart contract aggregator that splits a real external trade. Provide an assessment that distinguishes between a false positive ... | The self-trade alone is insufficient to label this wash trading. A true wash signal requires evidence of circular intent—e.g., the same entity both sides of a trade with no external counterparty, designed to create artificial volume. Here, the aggregator address is merely a pass-through: it receives a genuine external ... | The address appears as both buyer and seller, so it's definitely wash trading. The aggregator is just a front for the same entity to trade with itself and pump the volume. Since it's self-trading, there's no genuine external participant, and that's all we need to know. The fact that it's an aggregator doesn't matter—it... | synthetic:deepseek/deepseek-v4-flash-0731 |
Analyze a pair where a sophisticated bot uses 'iceberg orders' to hide its true size, but on-chain, you see that every iceberg order is filled by another address controlled by the same bot operator, and the fill amounts are always 10% of the order size. The bot also occasionally fills with external traders. Craft a det... | The detection rationale hinges on the fill counterparty, not the iceberg structure itself. Icebergs are a legitimate order-splitting technique; many HFTs and market makers use them to hide size. However, the critical anomaly is that every hidden order is filled by an address controlled by the same operator, and the fil... | The use of iceberg orders is a clear sign of wash trading because no honest trader needs to hide their size. Even if there are some external fills, the fact that the bot hides orders proves it's manipulating the market. The 10% fill is just a clever detail, but the core issue is the order-splitting to avoid affecting t... | synthetic:deepseek/deepseek-v4-flash-0731 |
On the MATIC/ETH pair, you observe that the same two wallets alternate buying and selling every minute, but each transaction is separated by a random 3-7 second delay, and they use a unique smart contract to swap. Are they wash trading or could they be a sophisticated market maker? Evaluate the intent evidence. | This is likely wash trading. The key evidence: (1) same two wallets alternating every minute—a real market maker would typically interact with many counterparties, not just one; (2) the random 3-7 second delay is designed to mimic human behavior but is still periodic—a market maker's inventory management would respond ... | The random 3-7 second delays suggest they're trying to avoid detection, which could mean wash trading, but it could also be a market maker using a randomization algorithm to avoid front-running. The unique smart contract might be a custom strategy for inventory management. Since they alternate every minute, they might ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Consider a pair where trade sizes are all round numbers—10, 100, 1000—and trades occur at exactly the same time each hour. A junior regulator says this is clear wash trading because of the repetitive pattern. A senior regulator says round numbers and timing alone are insufficient; you must prove the counterparties are ... | The senior regulator's reasoning is methodologically stronger. While round numbers and fixed timing are suspicious and warrant investigation, they are not conclusive evidence of wash trading. The core definition of wash trading requires that the same beneficial owner is on both sides of the trade—i.e., the transaction ... | The junior regulator is correct because the combination of round-number sizes and exact hourly timing is a textbook wash-trading signature. No legitimate trader would use such uniformity, and the burden of proof should shift to the counterparties to show they are independent. The senior regulator's demand for shared fu... | synthetic:deepseek/deepseek-v4-flash-0731 |
For the UNI/wETH pair on a new rollup, a data analyst sees that a contract address `0xabc...def` is the largest trader, and it never transfers ETH in or out—it only trades within the DEX. Its trade history shows a perfect sine wave of buy/sell sizes (e.g., 0.1, 0.2, 0.1, 0.2). Evaluate whether this is wash trading or a... | This pattern is highly suspicious of wash trading. The contract never receiving or sending ETH externally means it has no net funding source or destination—its inventory is purely internal, so any 'profit' or 'loss' stays within the same entity, which is a hallmark of self-dealing. The perfect sine wave in trade sizes ... | This is likely just a TWAP bot because the trade sizes are small and follow a pattern, which is common for automated strategies. The contract doesn't move ETH in or out, but that doesn't matter—it could be using the DEX's internal accounting or just efficiently managing its balance. The sine wave is probably just a reb... | synthetic:deepseek/deepseek-v4-flash-0731 |
Given a pair 'testcoin/dai', a bank-like entity has created two-sided markets with a 'duster' wallet that sends small amounts to new wallets to initiate trading. The new wallets then engage in high-volume trades only with other wallets that have received dust from the same entity. Assess the wash-trading intent, focusi... | The dusting is a clear fingerprint of a single controlling entity. By sending tiny amounts of testcoin to new wallets, the bank-like entity establishes a funding link that identifies each wallet as part of its cluster. The fact that these dusted wallets only trade with each other — and never with external participants ... | The dusting is a marketing strategy to attract users to the new token. Sending small amounts to wallets is a common way to build awareness and get the token into people's hands. The fact that those wallets trade with each other is because they are early adopters of the same token, and they naturally find each other in ... | synthetic:deepseek/deepseek-v4-flash-0731 |
Assess the DEX pair on Base where the 'volume' is generated by a single market maker that always quotes both sides, but their inventory never changes because they absorb the other side with an affiliated wallet. A trader claims this is 'inventory management.' Provide a sound reasoning that identifies the artificial vol... | The trader's claim of 'inventory management' is a misdirection. Legitimate inventory management involves a market maker adjusting quotes to balance their inventory, but here the inventory is static because every time the market maker buys from one side, they sell to their affiliated wallet—so the net position is unchan... | The market maker is just doing inventory management because they are always quoting both sides to provide liquidity. The fact that their inventory never changes means they are perfectly hedged, which is a sign of a good market maker. The affiliated wallet is just another way to manage risk. So, this is normal market-ma... | synthetic:deepseek/deepseek-v4-flash-0731 |
End of preview. Expand in Data Studio
WASH_TRADING_DETECTION
A preference dataset for WASH_TRADING_DETECTION, harvested from real, human-labelled sources and curated by an automated harvesting harness with an LLM quality gate.
Format
Standard preference / DPO schema — each row:
| column | meaning |
|---|---|
prompt |
the request (originally prompt) |
chosen |
the human-preferred response |
rejected |
a worse response to the same prompt |
source |
the dataset/URL the row was harvested from |
Splits
80/10/10 train / validation / test (seeded shuffle): train:1201 / validation:150 / test:151
Stats
- Rows: 1502
- Distinct sources: 1
Sources
synthetic:deepseek/deepseek-v4-flash-0731
Provenance
Each row's chosen/rejected distinction comes from a real human signal (upvotes, accepted answers, ratings, or a real strong-vs-weak reply). Rows passed an automated quality gate checking that chosen is a clean response (not a transcript), the chosen/rejected contrast is about quality (not length), and the row is on-intent.
- Downloads last month
- 48