Trade-Up Contracts: Does the Maths Work?

Trade-ups look like free money to people who have not done the arithmetic, and like a scam to people who did it badly. The truth depends entirely on the specific contract. Here are the mechanics, then a way to reason about them.

How a trade-up actually works

You put in ten skins. They must all be the same rarity grade and the same category — normal with normal, StatTrak with StatTrak. Out comes one skin of the next rarity grade up.

The output is not random across the whole game. It comes from the collections your inputs belonged to. Each input contributes one entry, so if six of your ten came from collection A and four from collection B, the result is drawn from those two collections in that six-to-four proportion.

Within whichever collection gets selected, every skin at the next rarity tier is an equally likely result. That is the key structural fact. A collection with two possible outputs gives you a much tighter result than one with five.

The inputs are consumed. They are gone. There is no undo, no refund, and no partial return if the outcome disappoints.

Float averaging, precisely

This is where most confusion lives. The output float is not the average of your input floats, though the average is the starting point.

Take the mean of the ten input float values. That mean is then mapped onto the output skin's own float range. Every finish has its own minimum and maximum float, and many do not span the full 0.00 to 1.00. The mean acts as a position within that range, not as an absolute value.

So a low average pushes you towards the bottom of whatever the output's range happens to be. If that output has a restricted range that starts above 0.07, no amount of pristine inputs will produce a Factory New. Conversely, a skin whose range starts at 0.00 can reach Factory New from a fairly ordinary average.

The standard wear bands still apply to the result: 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. What changes is which of those bands are reachable for a given output.

The practical consequence: check every possible output's float range before you build the contract, not just the one you are hoping for. A contract can be profitable on one outcome and worthless on another purely because of wear.

Thinking about expected value honestly

The framework is simple even though the inputs are not. Total cost is what you paid for all ten skins. Expected return is the value of each possible outcome, weighted by how likely it is, minus what it costs you to sell that outcome.

Three things people leave out and should not:

• Selling costs. The output has to be sold, and the marketplace takes its cut. A contract that breaks even on gross is a loss on net.
• Liquidity of the outputs. A high-value outcome you cannot sell for months is not worth its listed price to you.
• Wear on the output. The same skin at different wear tiers can be worth wildly different amounts, so you need the value of each outcome at the wear your contract will actually produce.

No numbers here, because prices change constantly and any figure would be wrong by the time you read it. Work with live prices at the moment you build the contract. That is the only version of the maths that means anything.

Where people go wrong

The most common mistake is counting only the good outcome. People find one profitable result in the pool, get attached to it, and mentally treat the contract as a purchase of that skin. It is not. It is a weighted draw across every possible result, and the bad ones happen at exactly their stated frequency.

The second mistake is ignoring variance. Even a contract with genuinely positive expected value can lose several times in a row. Expected value describes the long run, and the long run needs enough attempts and enough capital to survive bad stretches. Most people run a handful and draw conclusions from noise.

The third is chasing floats without checking ranges. Hunting extremely low-float inputs is expensive, and wasted if the outputs cannot reach the wear tier you are paying for.

The fourth is forgetting that markets adjust. When a contract is widely known to be worth running, people bid up the inputs and sell down the outputs until the edge closes. Public profitable trade-ups tend not to stay profitable, the same dynamic that shapes liquidity in skin trading generally.

When trade-ups make sense

They make sense as a way to convert a pile of unwanted low-tier skins into one thing you might actually use. That is the original purpose and still the best one. If the inputs were going to sit in your inventory forever, the effective cost is low.

They also make sense when you want a specific output and the contract route genuinely costs less than buying it outright, counting all possible results rather than just the one you want.

They do not work as an income strategy for most people. The edges are thin, publicly known, competed away fast, and need volume to realise.

Practical takeaway: before you build a contract, list every possible output, check each one's float range against your planned input average, price each outcome at the wear you will actually get, weight them by collection proportion, and subtract selling costs. If the result is not clearly positive with room to spare, do not run it. And treat every contract as money spent rather than money invested, because the inputs are destroyed and there is no path back.