The honest answer about CS2 case ROI in 2026 is uncomfortable: most cases are negative-expected-value to open, and have been for years. The interesting question isn’t whether opening cases beats holding them — it doesn’t, on average — but which cases beat which, by how much, and why. This guide covers the math, the post-knife-update changes, the right tools to run the numbers, and the strategy decisions worth making.
The expected-value formula in one sentence
The expected return of one case opening is the sum, across every possible drop, of (drop probability × price after fees), minus (case cost + key cost). The probabilities are public — Valve publishes them in the game — and the prices are public — Steam Market is the canonical reference. So this is a tractable arithmetic problem, not a guessing game.
The probabilities in the standard distribution are 79.92% mil-spec, 15.98% restricted, 3.20% classified, 0.64% covert, and 0.26% rare special (knife). Roughly one in every 385 openings produces a knife — the rest produce drops whose value almost never recovers the key cost. Knives carry the entire expected return, which is why ROI analysis is sensitive to which knives are in a case’s rare-special collection.
What the post-knife update changed
The relevant 2024 update changed how probability is allocated within the rare-special tier. The legacy model assumed equal weight across every knife in the rare-special set. The current model uses collection-weighted probability: within a knife drop, the case’s own knife collection drives the internal weighting. This matters because it makes case ROI sensitive to the average price of knives in that case specifically, not the flat average across all knives in the game.
The practical consequence is that cheap-knife-collection cases converge to a lower expected return faster than the legacy model suggested. A case whose knife pool happens to skew toward Falchions and Gut Knives produces a substantially worse ROI than one with Karambits and Butterflies in the pool, even when their non-knife drops are priced identically.
Where to actually run the numbers
The two tools worth bookmarking are CSROI for live ROI per case, and CSStonks for the supply data that explains why a case’s ROI is drifting one way or another over time. They’re complements, not competitors — CSROI gives you the snapshot, CSStonks gives you the trend. Our head-to-head review of the two covers the exact split.
The full best CS2 case ROI tools ranking has the long tail — useful when you want a second data source to sanity-check a number that’s about to drive a buying decision.
A worked example
Consider a case with the following live numbers (illustrative — your actual figures will be different by the time you read this): case price $1.50, key price $2.49, average non-knife drop value $0.80 (after Steam fee), and an average rare-special knife value of $300. Expected revenue per opening: (0.9974 × $0.80) + (0.0026 × $300) = $0.80 + $0.78 = $1.58. Cost per opening: $1.50 + $2.49 = $3.99. Expected return: $1.58 − $3.99 = −$2.41, or roughly −60% per opening.
The path to a less-negative ROI is almost always (a) a discontinued case whose non-knife floor has appreciated, (b) a case whose knife pool happens to have a higher average value, or (c) buying the case at a discount from a marketplace other than Steam — see the arbitrage category for tools that surface those discounts automatically.
The strategy decisions that actually matter
Three decisions dominate the long-run outcome for case investors. First: hold vs. open. Holding historically beats opening for most cases — supply contracts as old cases stop dropping in-game and prices appreciate. CSStonks is built for exactly this analysis. Second: case selection. Investing in cases with shrinking supply curves and large knife collections outperforms opening into them. Third: marketplace selection. Buying cases on Buff163 or Skinport at a 5–8% discount to Steam Market and selling into Steam generates a steady spread when liquidity allows it.
The honest summary
If your question is “should I open this case?”, the answer is almost always no — and CSROI will give you the exact number to confirm it. If your question is “is this case appreciating faster than I’m decaying alternatives?”, that’s the right question, and CSStonks has the supply chart you need. Stop opening, start holding, re-evaluate quarterly.
Further reading
Found something wrong, biased, or out-of-date? Reach the editorial desk via the corrections process.