- CS:GO Chances

How CS:GO Chances calculates return of investment

1. Where the drop-odds come from

Valve first published official case odds to satisfy Chinese loot-box disclosure rules. In weapon cases the “rarity ladder” looks like this:

Exceedingly Rare (Gold — knives / gloves) — 0.26 %
Covert (Red) — 0.64 %
Classified (Pink) — 3.2 %
Restricted (Purple) — 15.98 %
Mil-Spec (Blue) — 79.92 %

Those figures imply a simple ticket model: give the rarest grade one ticket, the next grade five tickets, the next 25, and so on; knife/glove tickets are then doubled to reach Valve’s published 0.26 %. When a capsule skips a colour (some sticker capsules have no Pinks) we still multiply by 5 for the missing step, so odds jump 25× to the next grade.

Within each grade every skin or sticker shares the pool equally, so a Classified tier with three items gives each one ≈ 1.07 %.

Valve’s filing also states that any drop has a 10 % chance to roll a StatTrak™ version. The multiplier is applied after the grade is chosen, making a StatTrak Classified item roughly 0.32 %.

Wear (float) distribution

CSFLOAT sampled millions of listings on skins whose float range spans the full 0 – 1 window and found the following long-run proportions:

Factory New ≈ 3 %
Minimal Wear ≈ 24 %
Field-Tested ≈ 33 %
Well-Worn ≈ 24 %
Battle-Scarred ≈ 16 %

These percentages are remapped into the actual min-max float window of each skin, producing the probability for every wear variant given the item is selected. The distribution is visualised in each item’s detail view.

csgoskins.gg: case odds pcgamesn.com: case odds csgofloat.com: wear distribution

2. Price data & daily refresh

Steam Community Market prices in euros are pulled daily via the ByMykel/CsgoApi feed, providing a Steam-wallet baseline for costs and returns.
When no Steam sale occurred in the last 24 h - or when an item price exceeds Steam’s wallet cap, prices from Skinport, Buff163 and CS:GO Empire are fetched through their public APIs and using the csgotrader service to fill the gap.

CSGO-API on Github Skinport API CSGO Trader App

Third-party sites usually list lower prices because they pay out real money. A daily median exchange-rate (Steam €/Site €) is calculated across items sold on both platforms so those external prices can be converted upward to their Steam-equivalent value.

Note: Steams ~13 % sales fee is not subtracted from prices. Think of the listed figure as the price you would have to pay to buy the skin, not the wallet balance you would see from selling it.

3. Expected Return — what the number means

For each case the model sums over every item variant:

Expected Return = Σ (P_i × Price_i) – (Case Cost + Key Cost)

P_i – probability of variant i (grade × StatTrak × wear, and divided equally among items within a grade).
Price_i – Steam-equivalent selling price from the daily snapshot.

Over a very large number of openings the average profit (or loss) per opening tends toward this figure, although real-world results vary widely around the mean.

4. Why the site exists & how to help

Gambling has become deeply embedded in Counter-Strike; many players have lost far more than they could afford. This project aims to add transparency by showing the cost, probabilities and expected outcome in simple numbers and by letting anyone compare opening odds with the price of buying the skin outright. If that information helps even one person avoid harmful gambling, the effort is worthwhile.

HOUNGOUNGAGNE made great videos about the gambling epidemic, which I recommend anyone to watch:
The Dark Reality behind CSGO. (Illegal Gambling, lies and addiction) Part 1 .

If this site helped you decide - either way - I would love to hear about it. Drop feedback, bug reports or suggestions to my steam profile comment section or at feedback@csgochances.com.

5. Limitations & disclaimer

Figures are estimates based on public data and can drift with market volatility or unannounced game updates. Nothing here constitutes financial advice or a guarantee of future returns.