Roobet Provably Fair Audit: Clean Math, Open Questions

TL;DR: Roobet's Math Is Clean - But That's Not the Whole Story

Score: 10.0 / 10 - EXCELLENT
Tests passed: 8/8
Game tested: Crash
Total outcomes analyzed: 50,000
Methodology: NIST SP 800-22 + Custom Gaming Tests
Data source: Roobet public hash chain (independently verified)

This is a mathematical audit, not a recommendation. Roobet's Crash algorithm is clean - the numbers don't lie. But a fair RNG doesn't make a fair casino. We publish these findings so you can verify the math yourself, while making your own judgment about the business. If you don't believe our numbers - good. Grab the hash chain and check yourself.

Data Source & Verification

Transparency about where our test data comes from is non-negotiable. Roobet uses a fundamentally different provably fair system than most crypto casinos - a public hash chain instead of individual seed pairs. This actually makes verification easier and more transparent.

Hash Chain Origin

How Roobet's Hash Chain Works

Unlike Stake or BC.Game which use individual seed pairs per player, Roobet uses a single public hash chain for all Crash games. This means:

1. Every game result is predetermined by the hash chain before any player places a bet
2. The chain is public - anyone can verify any past game using the starting hash
3. Each game hash becomes the next game's input: SHA-256(currentHash) = nextGameHash
4. The crash multiplier is calculated from: HMAC-SHA256(key=gameHash, msg=salt)

The house edge is implemented through a modulo check: if the full HMAC value modulo 20 equals zero, the game instantly crashes at 1.00x. This gives approximately a 5% base instant-crash rate, plus additional rounding effects, resulting in a combined house edge of approximately 5.95%.

Independence Statement

• FairPlay Audit had no contact with Roobet before or during this audit. No payment was received. No affiliate relationship existed at the time of testing.
• The hash chain is publicly verifiable. We used the same starting hash that Roobet publishes for all players. No special access was required.
• Our implementation was independently built using the open-source Roobet verification library by drewnicolette (GitHub) as reference, cross-checked against Roobet's published documentation.
• Anyone can reproduce our results using the same starting hash and public salt. The algorithm is deterministic.

Honest Limitations

• We tested Crash only. Roobet offers other provably fair games that would require separate audits with different algorithms.
• Hash chain analysis covers one chain segment. We analyzed 50,000 consecutive games from the published starting hash. Roobet may use multiple chains or reset chains periodically.
• The public hash chain proves game outcomes are predetermined. However, it does not prove that Roobet displays the correct results to every player in real-time. A separate real-time verification would require monitoring live game feeds.

What We Tested

Using our Scorecard Tool v2.0, we generated 50,000 Crash outcomes by walking the Roobet hash chain from the published starting hash. We then subjected the raw uniform floats to 8 statistical tests - 7 standard tests from the NIST SP 800-22 framework plus a Roobet-specific house edge verification test.

The Roobet-specific test checks that the observed 1.00x crash rate (instant crash) matches the expected rate of approximately 5.95%. This combined rate comes from two sources: the BigInt modulo 20 check (5% base rate) plus additional crashes from multiplier rounding (approximately 0.95% additional). If Roobet were manipulating crash points to increase their edge, this test would detect it.

Test Results: Crash (50,000 rounds) - PASS 8/8

House Edge Verification Detail

The observed house edge is virtually identical to the mathematical expectation. The tiny deviation of 0.03 percentage points is well within normal statistical variation for a 50,000-round sample. There is no evidence of inflated house edge.

What This Means

• The RNG is fair. All NIST SP 800-22 tests pass at 99% confidence. The raw bytes from HMAC-SHA256 are uniformly distributed.
• No detectable manipulation. Serial correlation tests show no dependency between consecutive crash points.
• House edge is exactly as expected. The 5.98% observed instant-crash rate matches the mathematical prediction of 5.95% almost perfectly.
• The hash chain is verifiable. Every crash point can be independently calculated from the public starting hash and salt.
• No trend manipulation. The Runs Up/Down test confirms crash points don't follow predictable patterns - no artificial "hot streaks" or "cold streaks."

What This Does NOT Mean

• This does not guarantee future fairness. Our analysis covers one segment of the hash chain.
• This covers Crash only. Other Roobet games use different algorithms.
• This does not mean you will win. The approximately 6% house edge means Roobet profits over time. That is math, not manipulation.
• This does not replace personal verification. Always verify your own rounds using the hash chain.

Our Methodology

We follow the FairPlay Audit Fairness Testing Framework, based on:

1. NIST SP 800-22 - The U.S. National Institute of Standards and Technology's standard for testing random number generators.
2. Raw Float Testing - We test the raw uniform floats from the HMAC output, not just crash multipliers.
3. Significance Level alpha = 0.01 - 99% confidence. A test passes only if the p-value exceeds 1%.
4. Minimum 50,000 rounds per game - Ensures statistical significance.
5. Game-specific house edge tests - We verify that the observed house edge matches the mathematical expectation.

Our Scorecard Tool is open source. The code, methodology, and raw data will be published on GitHub.

Scoring System

Roobet Crash: 8/8 tests passed = 10.0/10 = EXCELLENT

About This Audit

This audit was conducted independently. FairPlay Audit received no compensation from Roobet. We have no affiliate relationship with Roobet at the time of this audit.

The Ex Pit Boss has 23 years of casino industry experience, including work in European casinos and international gaming operations.

For comparison, see how Stake performed when they expanded provably fair to 1,200+ games in our Stake expanded audit - same methodology, different platform, different results.

Related: Why "Provably Fair" Doesn't Mean Fair | Stake.com Audit (250,000 rounds) | Our Independence Policy

Clean math doesn't always mean clean business. When casinos push back against independent verification, their excuses follow a predictable pattern.

More from FairPlay Audit

• How to Save Your Casino Seeds on Mobile
• Telegram Casino Scams: What to Watch Out For
• The Mathematics of Losing Streaks
• How to Export Your Bet History for Audit

More from FairPlay Audit

• Why "Provably Fair" Doesn't Mean Fair
• Provably Fair Explained: The Full Guide
• Provably Fair Cheat Sheet
• Use Our Free Verifier Tool
• 5 Seed Rotation Mistakes
• Bustabit Audit Report - 500,000 Rounds Tested
• Hash Chain vs Per-User Seeds - Architecture Explained