Float Sniping: Catching Low-Float CS2 Skins Before Anyone Else
Two Factory New copies of the same skin can differ in price by multiples, and the only difference is a decimal most sellers never look at. Float sniping is the practice of finding listings where that decimal is rare and the price doesn't know it. It is slower, nerdier, and more durable than racing new listings — because the edge comes from knowledge, not latency.
The decimal that does the work
Every skin carries a float value between 0 and 1, assigned at unboxing and fixed forever. It determines how worn the finish looks, and it maps to the five wear tiers as fixed game mechanics:
| Wear tier | Float range |
|---|---|
| Factory New | 0.00–0.07 |
| Minimal Wear | 0.07–0.15 |
| Field-Tested | 0.15–0.38 |
| Well-Worn | 0.38–0.45 |
| Battle-Scarred | 0.45–1.00 |
Marketplaces price by tier; collectors price by float. That mismatch is the entire opportunity. A 0.06 Factory New and a 0.003 Factory New sit in the same tier, share the same market page, and commonly sell for very different money — the second one just often doesn't, because the seller priced the tier, not the float. The full mechanics are covered in our float explainer; this post is about turning them into buys.
Two wrinkles matter before you hunt. First, many skins have capped float ranges — a skin whose minimum float is 0.06 can never produce a true low float, so "low" is always relative to what's possible for that finish. Second, the premium curve is not linear: per third-party trackers, the jump from 0.01 to 0.001 commonly matters far more than the jump from 0.05 to 0.04. Extremes are where the money is; the merely-good is commonly priced in already.
Where floats are published (and where they aren't)
You can only snipe what you can see. As of late 2026 the float landscape looks like this: venues like CSFloat expose float, paint seed, and rank-style context natively — which makes them great for valuation and brutal for edge, since every buyer sees what you see. Other marketplaces expose float in the listing data but don't emphasize it in the UI, and that half-lit middle ground is where float snipers commonly do their best work. Overlay extensions such as BetterFloat surface float and reference prices on top of pages that don't, and standalone float checkers resolve any inspect link on demand.
The rule of thumb: the more prominently a venue displays float, the more efficiently floats are priced there. Edge lives where the data exists but the audience isn't looking at it.
The filter setup that finds mispriced floats
Float sniping without filters is scrolling. The working setup is a small number of precise rules per item, not a broad net:
- An item allowlist. Pick skins where the float premium is documented — popular AKs, AWPs, knives with visible wear differences — and where you know the float cap. Skins whose finish barely changes with wear commonly carry no premium at all, whatever the decimal says.
- A float threshold per item. Not "low float" but a number: below 0.01, below 0.005, whatever the collector market for that skin actually rewards. Sales history on float-aware venues tells you where the premium starts.
- A price ceiling near base value. The snipe is a rare float at a base-float price. Set the ceiling at or barely above the tier's ordinary going rate — tight filters that rarely fire beat loose ones that page you hourly.
- Both ends of the range. High-float hunting — 0.99+ Battle-Scarred pieces — is a real niche with its own collectors, commonly with less competition than the low end.
Serious hunters also learn to buy around the float tax: the best fills are commonly at tier boundaries, where a 0.071 Minimal Wear looks nearly Factory New but is priced a tier down.
Valuing the catch — before you pay for it
The classic float-sniping mistake is paying a premium the exit won't return. Float premiums are collector premiums: real, but thin. A top-percentile float on a liquid skin can wait weeks for its buyer; the same float on an unpopular skin may never find one. Before buying, sanity-check three things: recent sales of comparable floats (not listings — sales), the number of similar floats currently listed, and whether the exit is liquid enough to matter. If you can't find a comparable sale, you're not sniping — you're making a market, and valuation discipline is what separates that from guessing.
Trade locks apply to your fill like any other purchase, so a flip thesis has to survive a week of market movement plus two fee events. Plenty of float snipers skip the flip entirely and simply accumulate skins they wanted anyway, at floats better than they paid for.
Where float sniping sits in the wider game
Compared with new-listing racing, float sniping is slow-twitch. Bots dominate rule-clean discounts on liquid items, but a bot only knows what its author encoded — and float-premium curves per skin, per collector market, are exactly the kind of knowledge that's expensive to encode and cheap for a human to carry. That's why float and pattern hunting remain the most human-friendly corners of the sniping economy as of late 2026.
It stacks cleanly with automation, too: let standing rules accumulate your base positions at cross-venue lows, and spend your human attention only on the listings where a decimal makes the price wrong. One buys you cost basis; the other buys you the occasional outlier. Both beat refreshing tabs.