Someone shows me a fresh token every few days now. Deployed an hour ago, chart already looks like a staircase, and the pitch is always some version of we are early. Maybe. But early to what is the question. Half the time the first candle is not organic demand at all. It is a cluster of bots that bought in the same block the deployer opened trading, and they are sitting on your unrealized gains waiting for enough real buyers to arrive so they can exit into your bids.
The good news is that this leaves a very legible trail. Unlike a lot of manipulation, sniper bundling happens in public, on chain, in a sequence you can reconstruct block by block. You just have to know where to look and what a clean launch is supposed to look like so the manufactured one stands out.
What a bundle actually is
When a deployer adds liquidity and enables trading, there is usually a specific transaction or block where buying becomes possible. Snipers want to be in that exact block, ideally in the same bundle the deployer submitted, or the first block right after. Being first matters because they are buying at the lowest price the token will ever trade at, before any organic flow moves the curve.
Coordinated snipers do this at scale. One operator will fund a spread of fresh wallets, then have all of them buy within the opening block or two, often through the same contract or router, sometimes as a literal bundle passed to a builder so the transactions land together and cannot be front-run by anyone else. The result is a first candle that looks like a stampede of buyers. It is really one entity wearing a lot of hats.
The tell is that this same entity has to sell eventually, and it tends to sell into the first real wave of retail. So the chart shows a sharp open, a bit of a climb as organic buyers pile into what looks like momentum, and then a distribution that the untrained eye reads as normal volatility.
Reading the opening blocks
Pull up the token on any block explorer and go to the transfers, sorted oldest first. You want to see the moment liquidity was added and then walk forward one block at a time. Here is roughly what I look at.
- Block position clustering. How many distinct addresses bought in the launch block and the one or two blocks after? A handful of buys spread across the first minute is normal. Fifteen or thirty addresses all landing in the opening block is not something humans do. Nobody is that fast by hand, and organic buyers do not arrive perfectly synchronized.
- Gas and ordering. Sniper transactions usually carry aggressive priority fees or sit in suspiciously tidy positions within the block, because getting ordered ahead of everyone else is the whole game. If the earliest buys are all paying to be first and the deployer transaction is right there next to them, that is a bundle, not a crowd.
- Buy uniformity. Real buyers use messy, human amounts. Bundles often show near-identical buy sizes across many wallets, because one script sized them. When ten wallets each buy the same round number of the token in the same block, you are looking at one operator.
- Wallet age and history. Click into the buyers. Fresh wallets funded minutes before launch, with no prior activity, that do nothing after the one buy, are the signature. Organic buyers usually have a trading history and a wallet that predates the token.
Tracing the funding
Clustering in a block tells you something is coordinated. Tracing the money tells you it is one hand. This is the step that actually confirms it, and it is not hard.
Take five or six of the suspicious buyer wallets and look at where each one got its gas and its buy capital. Very often you will find they were all funded from the same source address, or from a short chain of hops that converges on one wallet, within a tight window before launch. A single funder splitting ETH out to twenty fresh wallets an hour before trading opens is about as clear as on-chain evidence gets. If the funder also received the proceeds back after those wallets sold, you have the full loop, and there is no ambiguity left about whether that first candle was demand.
Two honest caveats. Funding through a mixer or a fresh centralized-exchange withdrawal will break the trail, and a genuinely sophisticated operator will use a different funding path per wallet to avoid exactly this. So a clean funding graph is strong confirmation, but a messy one does not prove innocence. Absence of evidence is weaker here than presence of it. When the graph does converge, though, trust it.
A quick pre-entry checklist
When someone sends me a launch and asks if it is real, I run through roughly this before I have an opinion.
- How many unique addresses bought in the first two blocks, and does that number look human or scripted?
- Are the earliest buys uniform in size and paying for priority ordering?
- Are the top holders right now the same wallets that bought in the opening block, or have they already started distributing?
- Do a sample of those early wallets trace back to a shared funder?
- What percentage of supply landed in that opening cluster versus reached organic buyers afterward?
If the opening block is thin, the buys are irregular, and the early holders are still holding, the first candle is probably real demand and you are looking at a normal launch with normal risk. If thirty synchronized wallets from one funder took a large slice in block zero and are already trimming, the chart is a trap dressed as momentum, and the people buying the climb are the exit liquidity.
None of this tells you whether the token goes up. Bundled launches sometimes run for a while because more retail keeps arriving than the snipers can sell into, and clean launches fail constantly. What the opening blocks tell you is who you are actually trading against and how much of that first move was manufactured. I would rather know that going in than find out on the way down. This kind of block-level reconstruction is a big part of what we built the launch and whale tooling at Blockcircle around, because the pattern repeats on almost every chain and the trail is almost always there if you bother to read it.
The habit worth building is simple. Before you read the chart, read the ledger. The chart is a summary someone else drew. The first few blocks are the raw thing, and they are usually more honest than whoever is in your replies telling you it is early.