Rust Upgrade Game Odds and Target Multiplier Worksheet

Build the target multiplier first

An upgrade game usually risks one balance amount or item value for a chance to receive a higher credited value. Define:

• S = credited stake value;
• T = credited target value;
• m = T / S = target multiplier;
• p = displayed success probability as a decimal.

If the failed result returns nothing and a success returns the target including the economic replacement for the stake, expected credited return is p × T. Expressed per unit staked, expected return is p × m. The implied edge is 1 - (p × m).

This model does not apply unchanged if a failed upgrade returns part of the stake, if a fee is charged separately, or if the displayed target is not the amount actually credited.

Target-odds worksheet

The following values demonstrate the formula. “1% edge” means a 99% expected return in the simplified all-or-nothing model.

To audit a displayed chance, multiply it by the target multiplier. A 24% chance at a 4× target produces 0.24 × 4 = 0.96, or 96% expected credited return. Under the simplified model, the implied edge is 4%. This is arithmetic from example inputs, not a live operator finding.

Valuation can change the apparent multiplier

Suppose a deposited skin receives 90 balance units while the desired withdrawal item costs 200 units. The interface may describe the target as a 2× upgrade from a 100-unit external reference, but the account economics use 200 / 90 = 2.2222× if those are the actual credited values attached to the wager and target.

Do not mix Steam Wallet listing value, a cash-market quote and operator balance in one ratio. Record each in its original unit and UTC observation time. The skin valuation guide explains why a transparent internal calculation still may not equal external recoverable value.

Inspect the roll mapping

A result method may convert a seed-derived output into a roll interval. The success region should correspond to the displayed chance after accounting for boundary and rounding conventions. Save the precise inequality: roll < threshold can differ at the endpoint from roll ≤ threshold.

For one completed round, record the pre-round commitment, revealed seed, client input, nonce, calculation and target threshold. Reproducing the roll verifies only that round under the published algorithm. It does not show that the displayed target was attractively valued or that a future withdrawal will be available.

Boundary tests catch quiet calculation errors

When the success interval sits at a threshold, test the values immediately below, exactly at and immediately above that boundary. A formula using four decimal places can disagree with an interface rounded to two. Preserve the unrounded displayed probability where available and document whether the operator floors, rounds or caps the threshold.

Also check the target after settlement. If the winner receives internal balance based on a refreshed price instead of the selected item or locked value, the pre-round multiplier and post-round credit may use different inputs. That is a settlement question, not evidence that the random roll itself failed.

Copyable audit model

Use the game-type guide to confirm who funds and settles the round, then compare the result with the house-edge framework.

Use the best Rust upgrader comparison for the fixed cross-site rubric and the RustSnowball vs RustClash comparison for like-for-like operator evidence. The RustSnowball review records its published formula and rules; none of those pages turns this generic worksheet into a live transaction test.