Why sniping a pons launch rarely works

The tax on buying opens at 99% and decays exponentially. At the very first block a purchase is almost entirely consumed. One second in, roughly a quarter is still taken. By two seconds it is a few percent, and by three it is gone.

That shape is deliberate. It does not block anyone from buying — it makes buying first worth less than buying second, which removes the incentive rather than the ability.

A snipe is profitable when the price moves up more than the cost of getting in early. Elsewhere that cost is gas and competition. Here it is a tax that can exceed the entire expected gain.

Buying at one second in and selling later means the price has to rise by roughly a third before breaking even. Buying at half a second needs it to double. On a launch where nobody knows the token yet, that is a bet rather than an edge.

On chains where transactions are ordered by fee, a sniper outbids everyone and lands first. Robinhood Chain does not work that way: the sequencer orders by arrival, so gas buys no position at all.

The only lever left is latency, and latency does not beat the tax. Arriving a hundred milliseconds sooner means paying a higher tax rate, not a lower one.

Addresses on the creator's exemption list pay nothing at any point in the window. That is the whole difference, and it is a documented parameter rather than an exploit — the protocol allows up to 32 of them precisely so a team can take its own opening.

So the honest summary is that the opening belongs to whoever was named before the token existed. Everyone else is choosing between a tax and a worse price.

There is rarely a reason to rush a pons launch. Waiting three seconds costs you a slightly higher price on the curve and saves you the entire tax, which is almost always the better trade.

If you want to know whether a launch was bundled before you buy, the answer is on chain: count the wallets that paid zero tax in the opening blocks.