To calculate the expected value (EV) of a CS2 case, multiply each rarity tier's drop chance by the average price of the skins in that tier, then add the results. ROI is that EV divided by what you paid, minus 1. For keyed weapon cases the answer is almost always well below break-even. In late September 2026, SteamAnalyst put the Kilowatt Case at about $0.76 back on roughly $2.62 spent.
This guide shows how to build your own cs2 case opening calculator in a spreadsheet, how StatTrak and wear change the math, and why one lucky knife says nothing about the next thousand openings. It is written for players who want the numbers, not a sales pitch.
What EV and ROI mean for a case opening
Expected value (EV) is the average payout per opening if you could repeat it a huge number of times. You can't. Nobody opens a million cases, so EV describes the long-run average, not your next opening.
ROI (return on investment) compares that average with your cost:
ROI = (EV − cost) ÷ cost
An EV of $0.83 on a $2.62 opening is an ROI of about −68%. In plain words, you would get back roughly 32 cents of item value per dollar spent, before any selling fees.
The cost is the case price plus the key. The key is $2.49 in the in-game store and is the same for every weapon case. Keys bought since October 2019 are bound to your account and can't be sold. A cheap case doesn't make an opening cheap, because the key is the bulk of the bill.
For background on the containers themselves, see the wiki guide to how CS2 cases work.
The formula: tier chance × average skin price, plus the StatTrak and wear adjustments
Standard CS2 weapon cases share one set of published odds. Valve first disclosed them in 2017, and community testing has stayed close to them:
| Tier | Colour | Drop chance |
|---|---|---|
| Mil-Spec | Blue | 79.92% |
| Restricted | Purple | 15.98% |
| Classified | Pink | 3.20% |
| Covert | Red | 0.64% |
| Knife / gloves | Gold | 0.26% |
The core formula:
EV = Σ (tier chance × average price of skins in that tier)
Every skin inside a tier is equally likely. In a tier with two Coverts, each drops 0.32% of the time. Your tier average is therefore the plain mean of that tier's skin prices. The full odds are covered in our CS2 case odds breakdown.
Four adjustments separate a rough estimate from a useful one:
- StatTrak. About 10% of drops in each tier are StatTrak, and they sell at a premium. Use
0.9 × normal price + 0.1 × StatTrak price. Gloves can't be StatTrak. As an example, SkinSniper's May 2026 listings show AK-47 | Inheritance roughly $26–$104 as a normal skin and $53–$226 as StatTrak, depending on wear. More in our StatTrak explainer. - Wear. The float is rolled when you open the case, and wear conditions are not equally likely. Price ranges within one skin can be wide. M4A1-S | Black Lotus lists at roughly $3.25–$10.91 as a normal skin across wears. Weight each wear by how often it drops, or at least use a consistent wear such as Field-Tested for every skin.
- Fees. Selling on the Steam Community Market costs about 15%, and Steam wallet funds can't be withdrawn. Multiply by 0.85 for a Steam-only view. Third-party markets charge less, but they have their own fees and risks.
- Price quality. Use median or recent sale prices, not the lowest listing on a thin market. A skin that shows $80 but sells twice a week isn't worth $80 in your EV.
This formula applies to keyed weapon cases. Souvenir packages, sticker capsules, Armory items and Terminals use different rarity ladders.
Worked example with a budget case
Let's use the Kilowatt Case, a budget case in the active drop pool. It has 17 weapon skins (7 Mil-Spec, 5 Restricted, 3 Classified, 2 Covert) and a Kukri Knife gold slot.
Cost: about $0.13 for the case plus $2.49 for the key, so roughly $2.62 (SteamAnalyst, late September 2026).
Live per-skin prices move daily, so the tier averages below are illustrative round numbers, not live Steam quotes. I chose them to land near what public trackers report, so you can see how the arithmetic works:
| Tier | Chance | Assumed avg price | Contribution to EV |
|---|---|---|---|
| Mil-Spec | 79.92% | $0.15 | $0.120 |
| Restricted | 15.98% | $0.60 | $0.096 |
| Classified | 3.20% | $6.00 | $0.192 |
| Covert | 0.64% | $35.00 | $0.224 |
| Knife (Kukri) | 0.26% | $75.00 | $0.195 |
| Total EV | ≈ $0.83 |
ROI = ($0.83 − $2.62) ÷ $2.62 ≈ −68%. Multiply by 0.85 for Steam fees and you get about $0.70 of spendable wallet credit.
Real trackers use live prices and slightly different methods, so they disagree on the exact figure. They agree on the direction:
| Source (date) | Case | Opening cost | Expected return |
|---|---|---|---|
| SteamAnalyst (late Sep 2026) | Kilowatt | ~$2.62 | ~$0.76 (−71%) |
| SteamDB snapshot (29 Jun 2026) | Kilowatt | ~$2.64 | ~$0.815 (about 31% back) |
| SteamAnalyst (late Sep 2026) | Prisma | ~$3.70 | ~$1.92 (−48%) |
| GGSkins, Steam medians (26 Sep 2026) | Prisma, best of 42 priced cases | case + key | 66% back |
A blunt takeaway from that table: the best cases return about two-thirds of their cost on paper, and that is before selling fees. Older, pricier cases return less.
Also note where the money sits. In this model, Covert and knife drops make up about half the EV, yet they land in under 1 in 100 openings.
Why knife and glove prices skew ROI, and why the median result is worse than the average
The knife slot is a lottery ticket that inflates the average. Kukri Knife finishes from the Kilowatt Case start at roughly $38 and climb to several hundred dollars for the top finishes. Valve doesn't publish odds per knife finish, so you can't price the gold slot precisely.
According to cs2price's July 2026 analysis, gold delivers on average about 29% of a case's total EV. For Dreams & Nightmares it is around 53%. In the other 99.74% of openings you get an item worth pennies to a few dollars.
That is why the average is not the typical result. cs2price simulated 200,000 openings per case and found the median is 5–25× lower than the average. For the Kilowatt Case, the median single opening was about $0.06 against a cost near $2.60.
When you see a headline ROI, ask how much of it depends on gold. If you removed the knife row, the ROI you would actually feel most of the time is much lower.
Variance: what 10, 100 and 1,000 openings look like
EV is a long-run average. Short runs swing wildly around it. Here is the same illustrative Kilowatt model at three sample sizes. The "±" column is one standard deviation, which is a rough range. It ignores price spread inside each tier, so real swings are somewhat wider.
| Openings | Total cost | Expected return | ± 1 SD (rough) | Chance of ≥1 Covert or knife | Chance of ≥1 knife |
|---|---|---|---|---|---|
| 10 | $26.20 | ~$8 | ~$15 | ~8.6% | ~2.6% |
| 100 | $262 | ~$83 | ~$48 | ~59% | ~23% |
| 1,000 | $2,620 | ~$827 | ~$152 | ~99.99% | ~93% |
Three things stand out:
- At 10 openings, one Covert can cover your whole spend. Most sessions end with a few dollars of blues and purples.
- At 100 openings, results are still spread out, but it is now unlikely you get anywhere near your cost back.
- At 1,000 openings, results cluster around the EV, which is far below the $2,620 you spent. Averaging out means losing about what the ROI says.
A knife every 385 openings is only an average. Your chance of at least one knife in 385 openings is about 63%, not 100%. Each opening is independent, so a dry streak doesn't make the next one "due".
Tracking your own openings without chasing losses
Tracking helps only if it makes you stop earlier, not play longer.
- Use a spreadsheet. Log date, container, cost (case + key), and the item's value at the time. Compute your running ROI.
- Try a stats extension. The "CS2 Case Opening Stats" Chrome extension has been shared on the Steam forums for aggregating Steam account openings. It needs your Steam interface set to English. Check any extension's permissions and reviews before installing, and never share login details.
- Use free simulators for the odds, not for hunches. Tools like SteamAnalyst's simulator show how your simulated ROI drifts toward the calculated EV.
- Beware small samples. A run of 50 or 600 openings that looks better or worse than the official odds is normal randomness, not proof that the odds changed.
- Set the stop rule before you start. "I stop when the budget is gone" works. "I stop when I'm back to even" is loss-chasing.
If your real goal is one specific skin, read buy skins vs open cases. Buying the skin directly is usually cheaper than opening for it.
How this applies on CSDROP: house edge and provably fair verification
CSDROP cases are not Valve cases. There are no keys, and the odds and item prices are set by the site and shown on each case page. Don't use the 0.26% knife figure from this article. Use the listed per-item odds and prices from all cases and run the same formula:
House edge ≈ 1 − (EV ÷ case price)
A real-money case site is a business, so the EV of a paid case sits below its price. That gap is the house edge, and it is the same concept as the Steam key and fee drag above. Check the numbers for each case you plan to open instead of assuming. Use the odds displayed on the case, and keep the same price source for every skin.
CSDROP is provably fair. That means each result comes from inputs fixed before the opening, so you can verify it afterwards instead of taking the site's word for it. Verify a few of your own openings at least once. It's the fastest way to confirm that what you saw is what the math says.
If you want to see the mechanics without paying, the free case is a no-deposit way to experience an opening.
Set a budget: responsible play checklist
Case opening is entertainment with a negative expected return. It is not income and not an investment. Treat it that way:
- Decide a fixed monthly budget you can afford to lose completely.
- Write down a stop-loss and a time limit before your session.
- Never open with rent, bills, or borrowed money.
- Don't reopen to "win it back". That's the gambler's fallacy.
- Take a break after a big win. Don't reinvest all of it.
- Use deposit limits and self-exclusion tools if you feel control slipping.
CSDROP is for adults aged 18+. Play responsibly, and never bet more than you can afford to lose. No outcome is guaranteed.
FAQ
What is a good ROI for a CS2 case?
None is positive on average. Recent trackers show even the best-performing cases returning roughly two-thirds to four-fifths of what they cost to open, before Steam's ~15% fee. Treat anything near that as "least bad", not "good".
Are the odds the same for every CS2 case?
Standard weapon cases share the same rarity odds: 79.92% / 15.98% / 3.20% / 0.64% / 0.26%. Souvenir packages, sticker capsules, Armory items and Terminals use different ladders. On CSDROP, use the odds shown on each case page.
Does the key price matter more than the case price?
For budget cases, yes. The key is $2.49 regardless of the case, so a $0.13 case still costs about $2.62 to open. Always cost an opening as case + key.
Is buying the skin cheaper than opening cases for it?
Usually, yes. If you want one particular skin, buying it directly almost always costs less than the expected spend to unbox it. Opening is about the experience, not the price.
Do dry streaks make a knife more likely?
No. Each opening is independent, so a long streak of blues doesn't raise the next chance. The ~385-opening figure is an average, not a promise.
How do I verify a result on CSDROP?
CSDROP is provably fair, so results can be checked after the fact using the verification tools and seeds. Try it on a few of your own openings so you know how it works.
