Sniping, explained — and why you probably lose to bots

Sniping is buying a token in the first block or two of its pool existing, using software that was watching for the pool to be created. It is not a strategy a human executes. If you are buying at the launch moment and a bot is selling to you, the bot has already decided that price is worth letting go — which is the definition of adverse selection.

What a sniper actually does

A sniper is a program with three parts: a detector, a pre-signed transaction and a bid for inclusion.

The detector watches for the on-chain event that means a token is tradable — a pool-creation instruction, a launchpad program emitting a new-token log, or a first liquidity deposit into a pair. On EVM chains it can also watch the pending transaction queue and act before the pool exists. Detection happens in software, on infrastructure as close to block production as the operator can afford, in milliseconds.

The transaction is built and often signed before the trigger fires, so the only remaining step is submission. Nothing is typed, nothing is confirmed, nothing renders.

The bid is whatever the chain requires to be ordered first: a compute-unit price on Solana, a priority tip on EVM, or a bundle submitted atomically so the buy either lands in the intended position or not at all — the same mechanics covered in how ordering-based extraction is paid for.

That is the whole design, and there is no analysis in it. Many snipers do not evaluate the token at all: they buy anything matching a filter, sell into whatever demand arrives over the next few minutes, and accept that most positions go to zero because the ones that do not pay for them.

Why speed is not something a human competes with

Solana produces a block roughly every 400ms. Robinhood Chain, live on mainnet since 1 July 2026, produces one roughly every 250ms. Those are the units the contest is measured in.

Your side: seeing a post, reading it, opening an interface, pasting an address, waiting for a quote, approving in a wallet, and reaching a block producer over a public RPC endpoint. Best case that is several seconds — ten to twenty blocks. The bot's entire loop completes inside the time your screen takes to repaint. Terminals advertising launch-moment speed compress your side of the loop, which is worth something, and do not put you in the same race.

The consequence is not that you buy slightly later. It is who you buy from. If a pool is minutes old, the counterparty to your order is disproportionately someone who bought in block one, has a lower basis, and has full information about the position they are exiting. You have none of that. That asymmetry is the cost, and it dwarfs any fee on the page.

Where sniping and bundling overlap

Sniping is external — a third party racing to buy. Bundling is internal — the creator buying their own token in the same atomic transaction group as the deployment, so supply lands in wallets they control before anyone else can bid.

They look similar on a chart and in the first block's transaction list, which is why the terms get used interchangeably. A sniper is a fast stranger; a bundler is the person who wrote the supply distribution, holding a large fraction of it through wallets designed not to look connected.

Detecting either afterwards is a wallet-clustering problem rather than a speed problem, and it is tractable: first-block buyer counts, funding-source overlap and holder concentration all leave traces. The method is in how first-block buying and bundled supply show up on-chain, and the concentration side in reading a holder distribution properly.

Practically: a token where much of the supply was acquired in the first block by wallets sharing a funding source has an overhang, whoever those wallets belong to. The overhang is the fact; the holder's identity changes only how you feel about it.

The venue-level fix: launch windows instead of launch moments

Because sniping is a consequence of the launch being an instant, the only structural fix is to make it not an instant.

Pools.trade, launched by Uniswap Labs on 5 August 2026 on Robinhood Chain, is the clearest current implementation. Its Crowd Launch mode replaces the block-one scramble with a four-hour window in which bids are collected and executed as a time-weighted average price, explicitly to defeat bundling. It also imposes a $10,000 minimum FDV to graduate, refunding participants if the threshold is not met. Its fee model is a zero launchpad fee and a 0.25% LP fee that autocompounds into liquidity the creator cannot remove, against the roughly 1% charged by most venues.

A TWAP window works because it removes what speed buys. If the price is averaged across four hours, being first by 250 milliseconds is worth almost nothing and the advantage of co-located infrastructure evaporates. Bundling weakens for the same reason: the creator cannot capture a favourable opening price by transacting atomically with deployment, because there is no opening price to capture.

It is not free of trade-offs. The window means committing capital for four hours with the state of the book public, and it removes the early-price upside that made launch participation attractive in the first place. Pools.trade also offers an Instant Launch mode — a conventional bonding curve with no minimum and no window — which carries the ordinary launch-moment dynamics in full. The full mechanism is in how Uniswap's Robinhood Chain launchpad is built.

Traction suggests the design is not merely theoretical: within days of launch Pools.trade accounted for roughly 50% of Robinhood Chain launchpad volume and about 40% of new tokens, with roughly $73.6M of day-one Uniswap v4 volume on the chain — more than Uniswap v4 on Ethereum mainnet the same day. Per-venue launch and volume data is tracked on the venue analytics pages.

What this doesn't tell you

On-chain data cannot reliably tell you whether a first-block buyer was a sniper, an insider or a well-configured retail bot. Wallets rotate, funding is laundered through intermediaries, and clustering tools produce probabilities rather than proof. Treat "bundled" as a strong inference from concentration and funding overlap, not an established fact about a named person.

There is also no public figure for what share of launch-moment buying is automated. Estimates circulate; none of them survive scrutiny about method, and we are not going to print one. What is defensible is directional: on high-throughput chains with public launch events, first-block participation is dominated by software, and that has been visible since bonding-curve launchpads became the default issuance path.

Finally, a TWAP window defeats sniping specifically. It does not make a token good, does not stop the creator selling later, and does not prevent the ordinary outcome for launched tokens, which is that they go to zero.

Frequently asked questions

What is memecoin sniping?

Buying a token within the first block or two of its liquidity pool existing, using software that detects the pool-creation event and submits a pre-signed transaction with a bid for priority inclusion. The buyer is typically indifferent to the token itself and is trading the predictable inflow of manual buyers that follows a launch.

Can a normal person snipe a launch?

Not competitively. The contest is decided in the sub-second range on chains producing blocks every 250 to 400 milliseconds, against pre-signed transactions on infrastructure positioned close to block production. Interfaces that advertise fast launch buying shorten your side of the loop; they do not put you in the same race, and any fill you get at the launch moment is a fill somebody faster declined.

How do I tell if a token was sniped or bundled?

Look at the first block's buyers rather than the chart. Count how many distinct wallets bought before anyone could have reacted manually, check whether they share a funding source, and check what share of supply they collectively hold. A high concentration acquired instantly is an overhang whether the holders are the creator or strangers.

Does a TWAP launch window actually stop sniping?

It removes the advantage speed confers, which is the mechanism sniping depends on. If the execution price is averaged over hours, arriving first is worth nothing and bundling cannot capture a favourable opening. It does not address anything else about the token, and it introduces its own trade-off: capital committed for the duration of the window.

---

What a locked pool changes about the launch moment

Nothing above helps if liquidity can be pulled once the buyers arrive. Team Finance — built by TrustSwap, which also builds Meme Central — locks LP tokens for a fixed term across Ethereum, Robinhood Chain, Polygon, Base and BNB, and the lock appears as a verified badge on the token's page in the cross-chain feed of new launches. It does not slow a sniper down by a single block, and it does not stop a creator or a first-block wallet selling their allocation into you. It only removes one specific failure: the pool disappearing.

---

Nothing here is financial, legal or tax advice. Memecoins are extremely high-risk: most lose most of their value, and the majority of tokens launched never reach a decentralised exchange at all. Never spend money you cannot afford to lose entirely. Meme Central does not recommend any specific token. Data described as Meme Central's own reflects tokens indexed by Meme Central and is not whole-market data.