What CS2 rarity tiers tell you
Weapon finishes commonly use the following categories:
| Tier | Commonly associated color |
|---|---|
| Consumer Grade | White or gray |
| Industrial Grade | Light blue |
| Mil-Spec | Blue |
| Restricted | Purple |
| Classified | Pink |
| Covert | Red |
Special items such as knives and gloves are commonly represented by a gold special-item category in case-opening interfaces. Contraband is a separate exceptional classification, not the next ordinary tier in a universal case probability ladder.
Not every case contains every tier, and the labels do not determine the probability table of a third-party case. Wear, pattern, demand and other item characteristics can also affect value. A higher rarity category alone does not establish a current market price.
How to read a case probability table
A complete table assigns a probability to each possible outcome or clearly defined outcome group. For mutually exclusive outcomes covering the whole case, those probabilities should total 100%, allowing for disclosed display rounding.
If a table only gives a tier’s probability, it does not tell you the chance of a specific item within that tier. For example, a 2% rare-tier chance split equally among four items would give each item a 0.5% chance. Without the equal-split rule, that conclusion would not follow.
Third-party cases can use different contents and probabilities even when their names or artwork resemble other cases. Read the table for the exact case you are considering.
Worked example: expected value and house edge
Here is a fictional case priced at $10. The dollar amounts are illustrative item valuations, not guaranteed cash proceeds.
| Outcome | Probability | Item valuation | Contribution to average value |
|---|---|---|---|
| A | 80% | $2 | $1.60 |
| B | 18% | $20 | $3.60 |
| C | 2% | $200 | $4.00 |
| Total | 100% | — | $9.20 |
Calculate expected item value by multiplying each value by its probability and adding the results:
Expected value = 0.80 × $2 + 0.18 × $20 + 0.02 × $200 = $9.20.
Relative to the $10 opening cost:
- Modeled return: $9.20 ÷ $10 = 92%.
- Modeled house edge: 1 − 0.92 = 8%.
- Expected net result per opening: $9.20 − $10 = −$0.80.
The average is not a typical item you receive: this case has no $9.20 outcome. Nor does an 8% edge mean that 8% of players lose. It describes average value relative to cost under the assumed probabilities and valuations.
If fees, withdrawal restrictions or realistic sale prices reduce what you can receive, the economic return can be lower than this table suggests.
A 2% chance does not promise one rare item every 50 openings
With independent openings and an unchanged 2% probability, the chance of missing the rare outcome on one opening is 98%.
The probability of missing it on all 50 openings is:
0.98^50 ≈ 36.4%.
So the probability of getting it at least once in those 50 openings is approximately 63.6%, not 100%.
Even after 49 misses, the next opening remains a 2% chance under those assumptions. Earlier losses do not make the next result overdue. This calculation illustrates risk; it is not a recommendation to buy 50 openings.
What changes in a case battle?
The underlying item odds and the battle settlement rules answer different questions:
- Item odds: what can appear in each opening?
- Settlement rules: who receives the resulting pool or payout?
In a highest-total-wins battle, pulling an item does not automatically mean you keep it. Team formats, Shared modes and weighted Jackpot formats can allocate outcomes differently. The chance of a particular case drop is not automatically your chance of winning the battle.
Do not apply a standalone case’s expected value directly to a player’s battle payout without modeling the entry costs, opponents, team allocation and settlement rules.
A short comparison checklist
- Use the exact case’s current probability table.
- Distinguish individual-item odds from grouped tier odds.
- Confirm whether valuations represent site prices or obtainable proceeds.
- Include entry charges and relevant withdrawal costs.
- Read the battle payout rules separately from case-opening odds.
When the full probability table is unavailable, you cannot reconstruct expected value from prize artwork alone. Treat that missing information as a limitation, not an invitation to guess.
Frequently asked questions
Do all CS2 case-opening sites use the same odds?
No. Third-party cases can have different contents and probability tables. Use the odds for the exact case, not a rarity chart from another product.
Does a gold icon tell me the chance of a knife or glove?
No. The icon identifies a category in an interface; the case’s published probability table determines the chance of a particular outcome.
Is a 92% modeled return a promise to get 92% of my money back?
No. It is an average calculated from probabilities and assumed item values. A short session can produce very different results, and item values are not guaranteed cash proceeds.
Can a losing streak improve my next opening’s odds?
Not when openings are independent and the probabilities are unchanged. Previous misses do not increase the next opening’s chance of a rare outcome.
Set a budget you can afford to lose, stop at your limit and do not chase a rare item. Support is available through Gambling Therapy.