Every standard CS2 weapon case uses the same published odds: 0.26% for a knife or gloves, about 1 in 385 openings. That figure covers any gold item in the case, not one specific knife, and after 385 openings you have only about a 63% chance of having hit one. The rest of the ladder is Mil-Spec 79.92%, Restricted 15.98%, Classified 3.20% and Covert 0.64%, and the expected return is negative.
What are the official CS2 case opening odds?
Weapon cases follow a fixed five-tier ladder. The first four tiers are each five times rarer than the one before, and gold is about 2.5 times rarer than Covert. The same rates apply to every weapon case, so a Kilowatt Case and a Dreams & Nightmares Case share one distribution.
| Rarity (colour) | Chance per opening | Roughly 1 in |
|---|---|---|
| Mil-Spec (blue) | 79.92% | 1.25 |
| Restricted (purple) | 15.98% | 6 |
| Classified (pink) | 3.20% | 31 |
| Covert (red) | 0.64% | 156 |
| Rare Special (gold: knife or gloves) | 0.26% | 385 |
Valve's Chinese partner, Perfect World, published these rates in 2017 after China required loot-box disclosure. A community thread on Steam from September 2017 reposted them, including the 1-in-10 StatTrak rate. The two-decimal figures are rounded. Valve's stated 2:5 ratio of gold to Covert implies a gold rate of about 0.256%, or roughly 1 in 391, and the difference is negligible in practice.
As of October 2026, I found no sign that Valve has changed this table since CS:GO. What changes from case to case is the item pool in each tier, and with it the market value.
What are the odds of one specific knife or pair of gloves?
They are far lower than 0.26%, because the game rolls rarity first and then picks an item from that tier. Items in the same tier of the same pool are generally treated as equally likely, so a case with two Covert skins splits the 0.64% between them.
The gold slot works the same way. Take a hypothetical gold pool of 20 items: each lands at about 0.013%, roughly 1 in 7,700 openings. Real pool sizes differ by case, so check the contents first. Finish, wear and pattern are further random steps on top.
Chasing one exact knife is a very different bet from chasing any knife. If you want a particular one, buying it is almost always cheaper and more predictable than hoping to unbox it.
How many cases until a knife? 385 is an average, not a guarantee
At 385 openings you have about a 63% chance of having hit gold, so roughly 37% of people are still empty-handed. The 385 figure is the average number of openings per gold drop over a very large sample.
The chance of at least one gold item in n openings is 1 − (1 − 0.0026)^n. The table applies that formula and prices the Steam keys alone at $2.49 each, the figure widely quoted for CS2 keys. Confirm the current price in-game, and note the cost column excludes the cases themselves.
| Openings | Chance of at least one gold item | Keys alone at $2.49 |
|---|---|---|
| 100 | about 23% | $249 |
| 267 | 50% | $665 |
| 385 | about 63% | $959 |
| 885 | 90% | $2,204 |
| 1,151 | 95% | $2,866 |
| 1,769 | 99% | $4,405 |
Opening counts are rounded up to the first whole number that reaches each chance.
Key takeaway: Streaks in both directions are normal. By the same formula, about 7% of people open 1,000 cases without a single gold item.
Does a losing streak make a knife more likely?
No. There is no pity timer, and every opening is independent. Your last 300 blue drops do not make gold more likely, and a gold drop does not make the next one less likely. Believing otherwise is the gambler's fallacy: the chance is 0.26% on opening number 1 and on opening number 1,001.
How do StatTrak, float and pattern change the result?
They are extra rolls on the item you already got, so they never change your chance of a knife. Rarity decides the tier, then the game rolls the properties.
StatTrak is a flat 10% chance on eligible items. Gloves have no StatTrak version, so the StatTrak knife figure below applies to cases whose gold pool contains knives.
| Drop | Chance | Roughly 1 in |
|---|---|---|
| Any Covert | 0.64% | 156 |
| StatTrak Covert | 0.064% | 1,560 |
| Any knife or gloves | 0.26% | 385 |
| StatTrak knife | 0.026% | 3,850 |
- Float: wear is a number from 0 to 1 rolled when the item drops. The ranges are Factory New 0.00–0.07, Minimal Wear 0.07–0.15, Field-Tested 0.15–0.38, Well-Worn 0.38–0.45 and Battle-Scarred 0.45–1.00. The roll is not spread evenly: one third-party estimate puts Factory New at only about 3% of drops.
- Float caps: each finish has its own minimum and maximum. Karambit | Doppler, for example, tops out at 0.08, so it can never be Field-Tested or worse.
- Pattern seed: for pattern-dependent finishes such as Case Hardened or Fade, the seed is rolled too. That is why some copies of the same skin sell for far more than others.
Do terminals, Souvenir packages and trade-ups use different odds?
Yes. The weapon-case table applies to key-opened weapon cases, and the other routes work differently.
| Route | Key needed? | Knife or gloves? | Odds status |
|---|---|---|---|
| Weapon case | Yes | Yes, 0.26% | Published in 2017 |
| Souvenir package | No | No | Covert 0.0256% (1 in 3,906) |
| Terminal (Genesis, Dead Hand) | No | Dead Hand: gloves. Genesis: none | No official table found |
| Trade Up Contract | No | Yes, from 5 Covert items | No rarity roll; inputs set the pool |
Terminals
Terminals are buy-or-skip offers rather than pulls. The Sealed Genesis Terminal arrived in September 2025, and Valve's Dead Hand update of 11 March 2026 added 17 community finishes and 22 all-new gloves as rare special items. Terminals can drop weekly and need no key.
As bo3.gg's explainer describes, a terminal presents up to five sequential offers, each a specific skin with a price. If you decline all five, it disappears. I could not find a Valve-published odds table for terminals. Some third-party sites assume case rates apply, so treat any per-offer glove figure as unconfirmed.
Souvenir packages
Souvenir packages use six rarity tiers instead of five, need no key and cannot contain knives or gloves. Covert sits at 0.0256%, about 25 times rarer than a Covert from a weapon case. Souvenir items cannot be StatTrak.
Trade Up Contracts
Trade-ups are a separate mechanic from case odds. Since the October 2025 update, five regular Covert items can be exchanged for one regular knife or gloves from a collection of one of the items provided. Five StatTrak Covert items give one StatTrak knife. You cannot choose the finish, and a contract is not a profit guarantee. The HLTV report on the update quotes the patch-note wording, .
What does provably fair prove on CSDROP, and what doesn't it?
Provably fair proves a result was fixed before you played and not changed afterwards. On CSDROP, case openings and games use HMAC-SHA256 built from a server seed shown to you only as a hash before play, your own client seed and a nonce. In case battles, the client seed comes from a future EOS blockchain block ID.
To check a result:
- Before you play, note the hashed server seed and your client seed.
- Open your cases as normal. Each opening uses the nonce along with those seeds.
- Find the opening in your game history.
- Verify it with the method on the CSDROP provably fair page.
Provably fair does not mean you are likely to profit, and it does not remove the house edge. It covers how a result was generated, not whether the odds on offer are good for you.
Do these odds apply on CSDROP, and what does the house edge mean for me?
Not automatically. Valve's table covers Steam weapon cases, while each CSDROP case shows its own item pool and drop rates before you open it, so read those. You can browse them in the CSDROP case catalogue. Cases open from your balance with no key, and items won are credited to the balance at market value.
Expected value (EV) is the average return per opening if you could repeat it forever: each tier's chance times the average value of its items. The table below uses made-up round prices for a Steam-style case to show the shape of the maths. It is not a CSDROP case, and it ignores StatTrak and wear.
| Tier | Chance | Assumed avg. value | Contribution |
|---|---|---|---|
| Mil-Spec | 79.92% | $0.10 | $0.080 |
| Restricted | 15.98% | $0.50 | $0.080 |
| Classified | 3.20% | $3.00 | $0.096 |
| Covert | 0.64% | $20 | $0.128 |
| Gold | 0.26% | $600 | $1.560 |
| EV | about $1.94 |
Against an assumed $2.99 cost (a $2.49 key plus a $0.50 case), that returns about 65 cents per dollar before any Steam Market fee. About 80% of the EV sits in a 1-in-385 event, so most sessions return well under the average and a few return far more.
Note: Real prices move daily, so use your own numbers for any real case. Higher-volatility cases mean longer dry spells and bigger swings, not better value.
If you want to see how an opening works before depositing, CSDROP's free cases are the low-stakes place to start. New accounts get free cases with no deposit needed, though you may first have to verify your email or link Discord or Steam.
How do I keep case opening under control?
Treat it as paid entertainment, not income. CSDROP is for adults only: 18+, or the age of majority where you live if higher.
- Set a budget before you start and treat it as already spent.
- Never chase losses. A losing streak does not mean a win is due.
- Stop when your time or deposit limit runs out, and take a break if you are opening to win money back.
If it stops being fun, stop and contact a support service in your country.
FAQ
What are the exact CS2 case opening odds?
For standard weapon cases: 79.92% Mil-Spec, 15.98% Restricted, 3.20% Classified, 0.64% Covert and 0.26% Rare Special (knife or gloves). The same table applies to every Valve weapon case, and only the item pools differ. Souvenir packages, terminals and trade-ups work differently.
Does 1 in 385 mean I will get a knife by case 385?
No. It means one gold item per 385 openings on average. At 385 openings the chance of having hit at least one is about 63%, so roughly 37% of people still have nothing. Reaching 95% takes about 1,151 openings.
What are the odds of a specific knife, like a Karambit?
Much lower than 0.26%. That figure covers every knife and glove in the case and is split across the gold pool. Finish, wear and pattern add further random steps, so buying the exact knife is usually cheaper than chasing it.
Is there a pity timer in CS2 cases?
No. Each opening is independent, so previous results do not change the next one. A long dry spell does not make a knife "due", and a gold drop does not make another one less likely.
Can I get a knife without opening cases?
Yes. Since the October 2025 update, you can exchange five Covert items for a knife or gloves through a Trade Up Contract, or buy the exact knife you want on a marketplace. The contract does not let you pick the finish, and neither route guarantees a profit.
Does provably fair mean I can win money?
No. It only lets you verify that a result was fixed before you played and not altered afterwards. The house edge still applies, so expect to lose money over time and play only with money you can afford to lose. CSDROP is 18+.
