Guide
How CS2 Trade-Ups Work
A CS2 trade-up contract takes 10 skins of the same rarity and converts them into a single skin from the next-higher rarity tier. The contract uses a deterministic formula for the output float and a probability split based on which collections your inputs come from. This calculator runs the same math the game uses, so you can simulate any contract before you spend a single cent.
The float averaging formula
CS2 normalizes each input's float to a 0–1 value relative to that skin's own min/max range, then averages all 10 normalized values, then maps the result onto the target skin's float range:
normalizedi = (floati − minFloati) / (maxFloati − minFloati)
avgNormalized = Σ normalizedi / 10
outputFloat = targetMin + avgNormalized × (targetMax − targetMin) Because each input is normalized against its own range, two skins with identical raw floats can contribute different amounts to the average. This is why low-float "filler" skins from forgiving ranges are worth more than they look.
Output rarity progression
Trade-ups walk one rarity at a time: Consumer → Industrial → Mil-Spec → Restricted → Classified → Covert. The probability of getting any specific output skin equals the share of inputs from that collection divided by the number of eligible target skins in it. Six Snakebite inputs with three Snakebite Restricted targets gives each target a 6/10 ÷ 3 = 20% chance.
StatTrak and Souvenir rules
All 10 inputs in a StatTrak trade-up must be StatTrak; mixing StatTrak and normal skins is not allowed. Souvenir skins, on the other hand, became valid inputs in the IEM Cologne 2026 patch (May 2026) and can be mixed freely with normal skins — the gold stickers, autographs, map tag and Souvenir attribute are stripped on craft, so the output is always a normal skin. Knife and glove drops are only produced by Covert-rarity contracts from supported crates, never from collection-only Coverts.
What this calculator does
Drop 10 skins into the slots above and the calculator runs the official CS2 formula in real time. It outputs the per-target probability, the exact output float, the projected exterior, the expected value (EV) using your chosen marketplace, and the win probability — defined as the chance the output sells for more than your input cost after fees.