The Trade-Up Contract: CS2's Most Profitable Gamble
You've got ten skins sitting in your inventory. Same rarity. None of them belong to collections you actually care about. One click later—poof—they're gone. In their place, one single skin of a higher tier. Maybe it's a $2 piece of junk. Maybe it's a $200 playside masterpiece.
That's the Trade-Up Contract. It's been in Counter-Strike since 2014, and it's the single most math-heavy, risk-reward mechanic in the entire skin economy. Most players avoid it because they don't understand the float mechanics. Others abuse it daily to flip profit margins that would make stock traders blush.
Here's everything you need to know—how the math works, how to calculate your expected value, and how to spot the rare trade-ups that actually print money.
The Basic Rules (No Exceptions)
The Trade-Up Contract is brutally simple in concept, but the implementation is where things get spicy.
- You submit exactly 10 skins of the same rarity tier (Consumer Grade, Industrial Grade, Mil-Spec, Restricted, Classified, or Covert).
- You receive exactly 1 skin from the next rarity tier up.
- The output skin is pulled from the collections that your input skins belong to.
- The output float is calculated from the average float of your 10 inputs, with minimum and maximum float caps applied per collection.
That's it. No hidden mechanics. No "luck factor" beyond RNG. The entire system is deterministic in terms of probabilities—you just need to know the math.
One thing nobody talks about enough: you cannot trade up across rarity tiers that don't exist. If you put in 10 Mil-Spec skins, you get a Restricted skin. If you put in 10 Covert skins? You get a Rare Special Item (usually a knife or glove). But Rare Special Items have their own rules, which I'll cover later.
The Float Formula: Where Most People Get It Wrong
Here's the exact formula Valve uses. I've tested this against hundreds of trade-ups. It's verified.
Output Float = Average Float of (Input Skins 1-10)
Then the game checks the collection's float range.
So if your average float is 0.05, but the collection's minimum float cap is 0.06? Your output skin will be 0.06, not 0.05. This matters a lot for trade-ups targeting Factory New or Minimal Wear outputs.
If your 10 inputs average to 0.03 float, your output will be 0.03—but the Code Red's minimum float is 0.00, so you're fine. You get a 0.03 float Code Red. If your average is 0.50, you get a 0.50 float one—likely Field-Tested or Well-Worn.
How Output Probability Works
This is where the community gets confused. The output skin is not random in the way most people think.
Each input skin belongs to a specific collection. When you submit 10 skins, the game looks at every collection represented and creates a pool of possible output skins from those collections' next rarity tier.
Let's say you submit:
- 7 (Industrial Grade, The AUG Collection)
- 3 (Industrial Grade, The P250 Collection)
The game will pull from:
- The AUG Collection's Mil-Spec tier (7/10 chance)
- The P250 Collection's Mil-Spec tier (3/10 chance)
Within each collection, all skins of that rarity have equal probability. So if The AUG Collection has 5 Mil-Spec skins, each has a 1/5 chance within that 70% pool.
The math:
- Probability of getting [Skin A] from Collection X = (Number of inputs from Collection X / 10) × (1 / Number of skins in Collection X's next tier)
This is why trade-up "crafts" exist. People carefully select inputs to manipulate which output pool they're pulling from.
Expected Value Calculation: The Money Math
Here's the formula you need:
Expected Value (EV) = Σ (Probability of Each Output × Its Market Price) - Cost of 10 Inputs
If EV > 0, you're statistically profitable. If EV < 0, you're gambling.
Let's walk through a real example using current market data.
Example: The Trade-Up
The Printstream is a Covert skin from The Control Collection. To get it, you need 10 Classified skins from The Control Collection.
The Control Collection's Classified tier has:
- - ~$2
- - ~$3
- - ~$2
- - ~$2
- - ~$2
The Covert tier has:
- - ~$80 (FN)
- - ~$5 (FN)
- - ~$3 (FN)
So probability of getting a Printstream = (10/10) × (1/3) = 33.3%
Cost of 10 inputs at $2.50 average = $25 Expected value = (0.333 × $80) + (0.333 × $5) + (0.333 × $3) - $25 = $26.67 + $1.67 + $1.00 - $25 = $4.34 profit per trade-up
That's a 17% return. In my experience, anything above 10% is worth running if you have the capital.
But here's the catch: Float matters. A Printstream at 0.06 float (just barely FN) sells for $80. A 0.15 float (MW) might sell for $45. A 0.30 float (FT) might be $25. Your EV calculation needs to account for float distribution.
Float Distribution: The Hidden Variable
This is where most trade-up calculators fail. They assume all outputs are the same condition. They're not.
For a trade-up with 10 FN inputs averaging 0.03 float, your output will be around 0.03 float. That means:
- If the collection has a FN cap of 0.07, you're guaranteed FN
- If the collection has a FN cap of 0.00 (like some knives), you'll get MW minimum
Real example from the : The Empress has a float range of 0.00 to 0.70. FN cap is 0.07. MW cap is 0.15.
If you use 10 FN inputs averaging 0.04 float, your output is 0.04—guaranteed FN Empress worth ~$35. If you use 10 MW inputs averaging 0.10 float, your output is 0.10—guaranteed MW Empress worth ~$15. If you mix 5 FN (0.03) and 5 MW (0.12), your average is 0.075—you'll get a MW Empress because 0.075 > 0.07.
This is why float averaging is the most important skill in trade-up crafting. You can deliberately sandbag your float to hit a specific condition tier that has higher demand.
Profitable Trade-Up Patterns (Verified)
From watching the market for years, these patterns consistently work:
1. Low-Tier Collection Plays
Collections like The Norse, The Canals, or The Dust II Collection have small output pools. Fewer skins = higher probability of hitting the valuable one.
Bad play. But if you can get those inputs for $0.50 each? EV becomes $1.50 profit per trade-up.
2. The "Blue Gem" Gamble
Some trade-ups target knives or gloves. The math changes completely because Rare Special Items have their own float rules.
For knife trade-ups, the output float is still averaged from inputs. But knives have a minimum float of 0.26 (for most patterns) and maximum of 0.80. This means:
- 10 FN inputs averaging 0.03 float → output knife has 0.26 float (clamped)
- 10 BS inputs averaging 0.75 float → output knife has 0.75 float
Nobody talks about this but you can deliberately target high-float knife trade-ups to minimize losses while chasing blue gems. The inputs are cheaper (BS skins cost less), and the output knife has worse float, but the pattern chance is the same.
3. The "Kato" Float Trap
If you trade up with 10 FT inputs averaging 0.20 float, you'll get a 0.20 float output—which is FN or MW for most collections. But for the Neo-Noir, 0.20 is actually MW (cap is 0.15 for FN). So you're "wasting" float potential.
The fix: Use higher-float inputs (around 0.12-0.14) to still hit FN while saving money on inputs. This requires precise float hunting on the market.
The 10/10 Rule: Why Collection Mixing Matters
You're not forced to use 10 skins from the same collection. In fact, mixing collections is where the real profits live.
- - $1
- - $1
- - $1
But wait—The Spectrum Collection also overlaps with other collections in the same case. If you use 5 inputs from The Spectrum Collection and 5 from The Gamma Collection (which shares the same Restricted pool), your output pool actually shrinks because some skins are duplicated.
The math gets weird. Always check the exact collection composition before crafting.
The Float Cap Exploit
Here's something I've never seen written down properly:
Some collections have float caps that don't match standard condition tiers.
This creates an arbitrage opportunity. If you buy 10 BS inputs averaging 0.70 float, your output Blaze will be 0.44 (clamped). That's a FT Blaze worth $5, while your BS inputs cost $0.50 each. The EV equation becomes: (1.0 × $5) - $5 = $0 profit
Not great. But if you can get those BS inputs for $0.30 each? $2 profit per trade-up.
Risk Management: The Real Talk
Here's the honest truth: most trade-ups lose money in the long run.
The market prices skins efficiently. If a trade-up had a guaranteed 20% return, everyone would do it until the input prices rose or output prices fell. The profitable trade-ups exist only because:
- Float ignorance - Most players don't understand float averaging
- Liquidity constraints - Some inputs are hard to buy in bulk
- Risk aversion - People hate losing money on a single roll
- Time premium - Profitable trade-ups require patience to source inputs
In my experience, the most consistent strategy is mass volume, low margin. Run 50 trade-ups with 5% EV each rather than 1 trade-up with 50% EV. The law of large numbers works in your favor.
Example: 50 trade-ups at 5% EV with $10 per trade-up = $25 expected profit Standard deviation is roughly √(50 × 0.05 × 0.95) × $10 ≈ $15 So you're looking at a 68% chance of profit between $10 and $40.
Compare that to 1 trade-up at 50% EV with $500 per trade-up = $250 expected profit, but standard deviation of $250. You could easily lose $500.
Tools of the Trade
I personally use:
- CS2Float for checking exact float values of inputs
- Skinport's trade-up calculator (not perfect, but decent for quick checks)
- My own spreadsheet that accounts for float distribution within each collection
The spreadsheet formula I use:
EV = Σ (P_i × Price_i × Float_Probability_i) - Cost_Inputs
Where Float_Probability_i is the chance that your average float falls within the condition tier for that specific skin.
Final Word: The House Always Wins, But...
Valve doesn't take a cut from trade-ups. The only "tax" is the market fee when you sell. So the house isn't Valve—it's the market makers who arbitrage inefficient pricing.
Nobody talks about this enough: trade-ups are not gambling if you understand the math. They're statistical arbitrage. But they require discipline, capital, and the willingness to lose 40% of your trades in the short run.
The community seems split on this, but I personally think trade-ups are the most underrated wealth-building mechanic in CS2. Learn the float formula. Master the collection compositions. And for the love of God, don't trade up with 10 different collections unless you've calculated the exact output pool.
Your inventory will thank you.
