CS2 Case Odds: The Real Math Behind Every Drop

The Drop Rates Valve Published

Valve has publicly disclosed the rarity distribution for CS2 weapon cases since 2018. The rates are identical across every weapon case (sticker capsules and souvenir packages follow slightly different rules). Here are the real drop rates:

Put differently: you open 385 cases on average before you get a knife. That's $958 in key costs alone, before any case prices. The probability of a StatTrak version is 1 in 10 (10% chance independent of rarity tier), and the wear is rolled independently — so even rolling a knife doesn't guarantee you walked away rich. A Battle-Scarred knife is often a small fraction of a Factory New one.

Expected Value: The One Number That Matters

"Expected value" (EV) is the average amount you'd earn per case if you opened millions of them. To calculate it for a specific case:

EV = (0.7992 × avg_milspec_price)
   + (0.1598 × avg_restricted_price)
   + (0.0320 × avg_classified_price)
   + (0.0064 × avg_covert_price)
   + (0.0026 × avg_knife_price)

Then subtract the case price and the $2.49 key cost:

Profit per case = EV − case_price − $2.49

If that number is negative, you lose money on average. For every CS2 case currently on the market, it's negative. Below is the calculation across all 42 active cases using live prices from the CSDB Case Odds Calculator.

Top 10 "Least Bad" Cases (Highest EV, Still Negative)

Translation: the "best" case in CS2 right now returns 70 cents for every dollar you spend. You are guaranteed to lose 30% on average. The worst cases return closer to 30 cents per dollar. The math only works if you ignore the key cost — and you can't, because Valve makes you pay it to open the case.

"But what about the dopamine?"

I hear you. Case opening is entertaining. You're paying $0.93 (on the Kilowatt case) for a roughly 5-minute experience that sometimes ends in a cool skin. If you'd happily pay $0.93 to spin a slot machine that 99.74% of the time gives you a consolation prize, go for it. Just stop calling it an investment.

Why "One in 385" Is So Much Worse Than It Sounds

The 0.26% knife drop chance sounds small but tractable. It isn't. Probability theory gets weird at small numbers.

If you open 385 cases, your expected number of knives is 1. But your actual probability of getting at least one knife is not 100% — it's about 63%. You still have a 37% chance of walking away with zero knives after $2,500+ in key costs.

If you want a 95% probability of pulling at least one knife, you need to open 1,153 cases. That's roughly $7,000 to $8,000 depending on case prices. And even then, 5% of players in your shoes still walk away with nothing.

These numbers come from the binomial distribution:

• P(≥1 knife in N cases) = 1 − (1 − 0.0026)^N
• For N = 385: P ≈ 0.632
• For N = 1153: P ≈ 0.950

You can check the live numbers here — plug in a case count and see the odds update in real time.

What Does the Rare Case Actually Pay?

For the 0.26% of cases that DO drop a knife, the payoff varies by case. The Kilowatt Case drops the Kukri Knife family, which sells for $350 to $4,500 depending on finish and wear. The Revolution Case drops newer knives in the $500 to $3,500 range. Older cases can drop legacy knives that now sell for five figures.

That's the lure. The average knife drop from Kilowatt is about $650. But remember: you only see that average after hundreds of openings. The median result is $0 (because you didn't get a knife) and the modal result is "a Mil-Spec skin worth 5–50 cents."

Why averages deceive you.

The Kilowatt case has a mean return near the case + key cost but a median return of $0.12. Half the people who open one get nothing worth talking about. A handful of lucky players get $650+ and bring up the average. This is exactly how lotteries work — the headline jackpot tells you nothing about your actual likely outcome.

The Three Cases Worth Opening (Sort Of)

If you're going to open cases anyway — and that's fine, as long as you know the math — here are the three that are least bad in 2026:

1. Kilowatt Case — newest knife/glove case, highest Covert prices, lowest net loss. Case + key = $3.08, expected return $2.15. Treat it as paying $0.93 to watch a slot machine roll.
2. Revolution Case — second-best ROI, and if you like the knife models this is your pick.
3. Kilowatt StatTrak rolls — if you specifically want to farm StatTrak Covert rolls (about 0.064% per case), Kilowatt has the highest ST Covert prices.

If you want to treat cases as an investment instead of entertainment: don't. Just buy the skin you want directly on the marketplace. You'll pay less than the key cost and you actually get what you paid for.

If You Want Odds That Make Sense: Investing, Not Gambling

CS2 skins do appreciate over time — specifically: major tournament stickers from discontinued events, original-release knives with rare patterns, and old Souvenir packages from closed tournaments. A Katowice 2014 Holo Titan sticker that sold for under $10 in 2014 now trades for over $140,000. That's a legitimate (if illiquid) asset class.

But opening cases is the opposite of that. You pay market rate for a random outcome with negative expected value. Sticker investing and knife investing work because you're buying known items at known prices and holding them. Case opening works because it's entertainment that feels like investing.

If you want a guide to what actually appreciates, see our skins investing guide and trading guide.

Methodology & Sources

Every number in this article comes from three public sources:

• Drop rates: Valve's official disclosure and extracted from items_game.txt in the CS2 game files
• Live prices: PriceEmpire aggregated across Buff163, Skinport, CSFloat, Steam Market, and 10 other marketplaces
• ROI math: the CSDB case odds calculator computes these live

No affiliate links. No sponsored mentions. CSDB.gg is a free, independent database built by a long-time CS player for long-time CS players. If you spot an error in the numbers, email hello@csdb.gg — we fix mistakes fast.

Last updated: April 15, 2026