CS2 Case Opening Odds: Every Official Number Explained

For the first four years of weapon cases, drop rates were a black box. That ended in 2017, when China's regulator began requiring games to publish loot-box probabilities; Valve complied and the numbers have applied globally ever since. The full breakdown, documented in SteamAnalyst's case odds guide, looks like this for every standard container:

Notice the shape: each tier is roughly five times rarer than the one above it. Four out of five opens land on blue, and the two tiers people open for — red and gold — together account for less than 1% of outcomes. The same table applies whether the container costs three cents or thirty dollars, which is worth internalizing early: no case in the game offers better or worse odds than any other, only different items behind the same probabilities. These are the only odds this site covers; third-party opening sites publish their own numbers under no comparable scrutiny, and we do not review or recommend them. For how cases, keys, and drops fit together more broadly, the CS2 cases hub covers the whole system.

Landing a tier is only half the roll. The moment you open, the game also generates the skin's float value — its wear — randomly. Nothing about the case, your account, or the sealed container predetermines it; a Covert can arrive Factory New or Battle-Scarred from the same open, and the price difference between those outcomes is often larger than the difference between tiers. The CS2 float chart maps the exact ranges, and our comparison of Factory New vs Minimal Wear vs Field-Tested shows what each band does to resale value. Floats skew toward the middle wears, which is why a Field-Tested unbox is the typical experience and a true low-float is its own small lottery inside the larger one. In practice this means two players can pull the identical Covert from the identical case and walk away with items whose market values differ by 3x or more — a variance layer that most people never factor into their mental picture of what an open is worth.

StatTrak is the other sub-roll: within each rarity tier, roughly 10% of drops come out StatTrak. Multiply through and the compound odds get thin fast — a StatTrak Covert is about 0.064% (1 in ~1,560), and a StatTrak gold roughly 0.026% (1 in ~3,850). When you see those pulls in highlight reels, remember the denominator: the reel exists because thousands of opens did not make it.

Human intuition is terrible at small percentages, so let's compute instead. Each open is an independent 0.26% chance, which means the probability of at least one gold in n opens is 1 − (1 − 0.0026)ⁿ:

This is the part of CS2 case opening odds that intuition reliably gets wrong, so two rows deserve a stare. At 100 opens — $249 of keys — you are still more than three-to-one against seeing a single gold. And at 385 opens, the famous "expected" break-even point (~$958 in keys), your odds are only 63.3%: more than a third of players who open 385 cases get zero knives. Expected value describes the average across millions of players; it is not a promise to you. Even at 1,000 opens and roughly $2,490 in keys, about 1 in 13 players still walks away goldless. That asymmetry — guaranteed cost, probabilistic reward — is the entire case for buying instead of opening.

Let's price a real open, using the Kilowatt Case at around $0.70 as of mid-2026, plus the fixed $2.49 key — about $3.19 per open. Using rough mid-2026 tier averages from my spreadsheet (blue ~$0.15, purple ~$0.65, pink ~$3.50, red ~$18, gold ~$180 for this case's pools), the expected gross return is:

• Blue: 79.92% × $0.15 ≈ $0.12
• Purple: 15.98% × $0.65 ≈ $0.10
• Pink: 3.20% × $3.50 ≈ $0.11
• Red: 0.64% × $18 ≈ $0.12
• Gold: 0.26% × $180 ≈ $0.47

Total: roughly $0.92 of skin value per $3.19 open — and after the Steam Market's 15% fee, about $0.78 of actual wallet balance, or 24 cents back per dollar spent. Notice where the value hides: over half the expected return sits in the 0.26% gold slot, meaning the typical open returns far less than even this thin average — most sessions are a parade of fifteen-cent blues, and the average is only rescued by a knife almost nobody in the session actually gets. The exact figures wobble with market prices, but I have run this calculation on a dozen containers since 2023 and the recovery rate lands in the 20–35% band every time. If the goal is owning skins rather than buying suspense, our ranking of the best CS2 cases to open starts from exactly this math — and CS2 Skin Prices exists largely to help you take the buying route instead.

The most expensive misunderstanding in case opening is not about the odds themselves — it is the belief that they accumulate. They do not. Every open is an independent roll at the same 0.26%; the game keeps no pity counter, and 384 goldless cases do nothing to make case 385 special. If you have opened 100 cases dry, your chance on the next one is exactly what it was on your first: 0.26%.

This matters because "I'm due" is precisely the thought that turns a $10 entertainment budget into a $200 hole. The cumulative table above describes your position before a series of opens begins; the moment a case is opened, that roll is spent and the future resets. The sunk keys are gone whether you stop or continue, so the only question that ever matters is whether the next $3 open is worth $3 of fun to you — not whether the last twenty owe you something. Streaks of blues are not a signal, they are the baseline: at 79.92%, ten blues in a row happens about once every 9.4 sequences, which is to say constantly. Set a budget in dollars, decide it before the first open, and let the math — not the itch — decide when you stop.

If you take one habit from this page, make it this: before any batch of opens, write down the cumulative probability for that batch size from the table above and ask whether you would pay the key total for that percentage, stated plainly. "$124.50 for a 12.2% chance at a knife" is the same transaction as fifty hopeful opens — it just sounds like what it is.