Trust
Fair play, verifiable
Fairness here isn't a promise — it's a mechanism. Sealed challenges, server-replayed runs, published rules, human review. This page explains each part.
Skill versus chance
What makes this a skill platform
In a game of chance, randomness picks the winner — dice, reels, draws. Here, randomness only creates variety: the target sequence is generated from a sealed seed, then both players face exactly the same sequence. Whatever the layout, the person who aims better, reacts faster and stays composed scores more.
Practice measurably improves performance, outcomes are predictable from prior skill ratings, and identical challenges make results directly comparable — the classic hallmarks of skill competition. We collect exactly this evidence (score progression, rating-band analysis, identical-seed comparisons) for independent expert review.
We deliberately avoid definitive legal claims. The classification of each game and format is assessed with UK counsel before any paid mode could ever open — see competition rules.
Banned by design
- Slots, roulette, dice, crash games
- Random jackpots, loot boxes, lucky draws
- Randomly selected winners or tie-breaks
- House-controlled outcomes of any kind
- Fake opponents or bots posing as humans
- Prediction markets (sport, crypto, elections)
Sealed challenges
Commit, play, reveal
A cryptographic commitment makes 'same challenge for both players' checkable rather than take-our-word-for-it.
1 · Seal
When a match is created, the server generates a 128-bit seed inside the database and publishes its SHA-256 fingerprint on the match — before either player can start.
2 · Play blind
Starting your run discloses the challenge to you and burns your single attempt. The seed itself stays locked away from both players until settlement.
3 · Reveal & audit
After settlement the seed is published on the match. Hash it yourself: it matches the fingerprint from step 1, proving nothing was swapped after scores were known.
Server authority
We replay your whole run
A score you merely claim is worthless. The only thing the client can submit is the run itself.
During a run, the game records every tap — position and millisecond timing. On completion this raw event log is submitted to the verification service, which replays it through the identical deterministic engine against the sealed challenge and recomputes the score from scratch. The number on the leaderboard is always the server's.
The same pass screens for things human hands don't do: impossible reaction times (under 60ms), sustained inhuman tap bursts, machine-perfect precision across a whole run, corrupted or reordered logs, and scores that don't reproduce. Hard violations are rejected outright; borderline patterns are flagged and held for human review — a reviewer replays the log before any result stands. No account is punished by an unexplained algorithm.
Think a result is wrong? Every settled match has a dispute path. Results are never altered silently; every change leaves an audit trail.
Ties share the position
Identical verified scores are a dead heat: both keep the position, ratings adjust symmetrically. There is never a coin-flip.
Versioned rules
Game and scoring versions are stamped on every match. Formula changes create a new version — historical results are immutable.
Honest failure states
Rejected runs, held settlements and reviews are shown to you in plain language — never hidden behind a spinner.
See it hold up
The fastest way to trust the system is to put a run through it.
Play a verified practice run