Tải ứng dụng Stake cho Android APK và iOS

Each factor below is scored from 1 (poor) to 5 (excellent) by our editorial team.

Licensing & jurisdiction fit

Scope: score the exact domain/app for the target jurisdiction. Verify the licence in the regulator's public register and confirm that the registered domains include the reviewed domain. Use BetAppStore's maintained, versioned licence-tier list. 1 = No licence, or the relevant licence is revoked or suspended. 2 = A licence is claimed but is expired, cannot be verified, or the validation seal does not resolve to the operator/domain. 3 = A valid Tier-3/offshore licence is verifiable in a public register, but player recourse and market-specific protection are limited. 4 = A valid Tier-1 or Tier-2 licence exists but was not issued for the target market, or a target-market Tier-2 licence applies. 5 = A valid Tier-1 licence explicitly covers the target market, operator and exact reviewed domain/app. Child/local versions are always checked separately for local authorization; a parent licence is not automatically inherited.

Ownership & track record

Check the operating company, parent, corporate registry, beneficial ownership, years since launch, regulatory history and supported unresolved confiscation/non-payment cases from the last 24 months. 1 = Operator identity cannot be verified; or the brand is blacklisted; or there is a confirmed mass non-payment event or more than 15 supported unresolved fund cases. 2 = Company is named but not verifiable, ownership is materially opaque/nominee-only, the brand is under 1 year old, or there are 6–15 supported unresolved cases. 3 = Company is registry-verified, owners may be undisclosed, brand has operated 2–4 years, and there is no mass incident; no more than 5 supported unresolved cases. 4 = Company and controlling owners are verifiable, brand has operated 5–7 years, is not blacklisted, and has no more than 2 supported unresolved cases. 5 = Public/audited company or named beneficial owners, 8+ years operating, no material regulator sanction and no supported unresolved fund case in 24 months. A child brand may inherit group ownership, age and group-wide incidents; its own local incidents override inherited history.

Player funds & game fairness

Count four safeguards: (A) independent RNG/game audit or a public provably-fair verifier; (B) per-game RTP/rules disclosure; (C) segregated player funds or public proof of reserves; (D) independent ADR plus a supported complaint-resolution rate of at least 80% over 24 months where a meaningful sample exists. 1 = None of the four; or a failed unresolved BetAppStore withdrawal; or a regulator finding for withheld player funds or unfair games. 2 = Exactly 1 safeguard, with no verified non-payment event in the last 24 months. 3 = Exactly 2 safeguards, including A. 4 = Exactly 3 safeguards, including A. 5 = All 4 safeguards are current and independently verifiable. Operator marketing claims do not count as evidence. Official records and instrumented tests override lower-level sources.

Account security

Advanced controls are: login alerts, active device/session list, withdrawal-address whitelist, withdrawal lock/delay, and a clearly documented account-recovery process. Score controls available to ordinary users on the reviewed domain/app. 1 = Password-only protection, or a serious confirmed security failure remains unresolved. 2 = No user 2FA, but every withdrawal requires an email or SMS confirmation. 3 = User-controlled 2FA is available, with no additional advanced control. 4 = 2FA plus 1 advanced control. 5 = 2FA plus at least 2 advanced controls; sensitive changes and withdrawals are re-authenticated. HTTPS alone is mandatory baseline security and does not raise the score.

Responsible gambling

Count only working account tools: deposit limit, loss limit, wager limit, session/time limit, reality check, cooling-off/time-out, self-exclusion, self-assessment, and a target-market help-organization link. 1 = 0 working tools. 2 = 1 working tool. 3 = 2–3 working tools. 4 = 4–5 working tools. 5 = 6 or more working tools, including self-exclusion and at least one money limit. A policy page or marketing claim does not count unless the control is usable from the reviewed account.

Payout reliability & speed

Measure request-to-funds-received time using the median of 3 comparable withdrawals on the fastest normally available method. Record method, amount, timestamps, KYC status and failures. Measured results override 'instant payout' advertising. 1 = More than 5 business days; or a failed unresolved test; or a supported pattern of systematic delays/confiscation. Missing verifiable payout evidence also scores 1, not N/A. 2 = More than 48 hours and up to 5 business days, or a fixed pending period above 48 hours. 3 = More than 12 hours and up to 48 hours. 4 = More than 1 hour and up to 12 hours. 5 = 1 hour or less, processed 24/7, with all 3 tests completed and no unexplained reversal or failure.

Fees, minimums & limits

Score four applicable subtests separately, then average without intermediate rounding and round the final factor to the nearest whole score: operator fees on main methods; minimum withdrawal; monthly withdrawal cap; maximum sportsbook payout or casino win cap. Convert to EUR at the snapshot date. Mark a genuinely irrelevant subtest N/A. 1 = Fees above 5% or undisclosed; minimum above €100; monthly cap below €5,000; or maximum payout below €25,000. 2 = Main-method fees above 2% to 5%; minimum €51–100; cap €5,000–19,999; or maximum payout €25,000–99,999. 3 = Fees up to 2% or a material dormancy fee; minimum €31–50; cap €20,000–49,999; or maximum payout €100,000–249,999. 4 = No operator fee on main methods; minimum €11–30; cap €50,000–99,999; or maximum payout €250,000–999,999. 5 = No operator fees on any normal method; minimum €10 or less; no monthly cap or cap at least €100,000; and maximum payout at least €1,000,000 or no cap. If subtests land in different bands, use their mean; do not choose only the best condition.

KYC predictability & privacy

Measure verification effort, speed, disclosure and secure data handling—not whether KYC exists. Lawful, predictable KYC can score highly. Record requirements disclosed before deposit, accepted documents, triggers, secure upload route and median completion time. 1 = Supported evidence that KYC is used to stall/refuse legitimate payouts; verification exceeds 7 days without cause; documents are handled insecurely; or no meaningful privacy policy exists. 2 = More than 72 hours; repeated or duplicate requests; unexplained source-of-funds checks on withdrawals below €5,000; or broad marketing-partner data sharing. 3 = Required at signup or first withdrawal, completed within 72 hours, with a clear document list and privacy policy. 4 = Required only at a disclosed trigger/first withdrawal, completed within 24 hours, with secure upload and specific rejection reasons. 5 = No mandatory document KYC below a stated cumulative threshold of at least €2,000, OR automated e-ID completes within 15 minutes; triggers and privacy terms are clear and data collection is minimized.

Payment methods & market fit

Count only payment categories that users in the target market can use for both deposit and withdrawal: cards; bank/instant banking; e-wallets; domestic methods; crypto; prepaid/vouchers. For crypto-first sites also record supported coins/networks and a fiat on/off-ramp. 1 = Only 1 usable category, or deposit and withdrawal routes are incompatible, or networks/fees are materially undisclosed. 2 = 2 usable categories. 3 = 3 usable categories; for crypto-first sites, 3–4 supported coins with disclosed networks. 4 = 4 usable categories including a leading domestic method in the primary market; for crypto-first sites, 5–7 coins. 5 = At least 5 usable categories including a leading domestic method in every declared target market; for crypto-first sites, 8+ coins plus a usable fiat on/off-ramp. Do not reward a long global list when those methods are unavailable in the reviewed market.

Customer support & dispute resolution

Run 3 timed tests using standard questions about KYC, withdrawals, settlement, complaints and self-exclusion. Compare answers with written terms. Record hours, language, first-response time, accuracy, complaint escalation and independent ADR/regulator access. 1 = No working support channel, median response exceeds 24 hours, or no public complaint route. 2 = Email/ticket only, answered within 24 hours, with a basic complaint route. 3 = Working live chat with limited hours or median response within 15 minutes, plus documented escalation. 4 = 24/7 live chat in the primary-market language, median response within 5 minutes, accurate answers and a formal complaint process. 5 = 24/7 target-language live chat, median response within 2 minutes, accurate answers, written escalation and an independent ADR/regulator route. A support agent's statement never overrides current written terms.

Mobile usability & accessibility

Count seven passed checks: (A) registration, deposit, KYC, play/bet and withdrawal all work on mobile; (B) mobile LCP is 2.5s or less; (C) every offered product is reachable within 3 taps; (D) search/filtering works; (E) no blocking pop-ups; (F) Lighthouse accessibility is 80+; (G) no critical bug in a 60-minute scripted session. 1 = 0–1 checks, or any core flow in A is broken. 2 = 2–3 checks. 3 = 4–5 checks. 4 = 6 checks, with all core flows in A working. 5 = All 7 checks. Native iOS/Android, PWA and responsive web are treated equally when functionality, stability and accessibility are equal. App-store ratings are supporting evidence, not a substitute for testing.

Localization & market relevance

Count six target-market checks: (A) complete UI in target language; (B) wallet in target currency or clear conversion; (C) 2+ domestic payment methods; (D) target-language support at least 12 hours/day; (E) leading domestic sports/leagues or relevant local content; (F) local licence or market-specific terms/promotions. 1 = 0 checks. 2 = 1–2 checks. 3 = 3–4 checks. 4 = 5 checks. 5 = All 6 checks. For an English-language target market, A is automatically met. A lawful single-country operator is not penalized for lacking global reach; country count and restricted-country count are metadata, not quality points.

Bonus & promotion fairness

Where promotions are offered, count five fair-value conditions: (A) sportsbook wagering 5x or less on deposit+bonus, or casino wagering 35x or less on bonus; (B) at least 30 days to complete; (C) maximum bonus bet at least €5; (D) no cap on legitimate withdrawable winnings; (E) 5+ recurring offers or a transparent loyalty program. Terms must be linked from the offer. 1 = Predatory term: wagering above 60x, deposit locked to bonus, hidden material terms, retroactive changes, or undefined 'irregular play' used to confiscate winnings. 2 = Only 1 fair-value condition, or material terms are unclear. 3 = 2–3 fair-value conditions. 4 = 4 fair-value conditions. 5 = All 5 fair-value conditions and no predatory clause. N/A = Bonuses are prohibited in the target market or genuinely not offered. Leave unscored; absence alone is not a quality failure.

Primary product value & pricing

Score the declared primary product only and use the same snapshot time across peers in the same target market. Sportsbook: median pre-match overround on at least 20 comparable top-league events (5=≤4%; 4=>4–6%; 3=>6–8%; 2=>8–11%; 1=>11%). Esports: 10+ tier-1 match-winner markets (5=≤4%; 4=>4–5.5%; 3=>5.5–7%; 2=>7–9%; 1=>9%). Exchange: commission-adjusted effective odds. Poker: rake/cap minus effective rewards. Casino/custom: published RTP, house edge, commission or fees versus provider defaults and same-market peers. For products without the absolute bands above, convert the measured value to a same-market, same-product peer percentile: 1 = Below the 15th percentile, materially worse than peers, or pricing is undisclosed/unverifiable. 2 = 15th–39.99th percentile. 3 = 40th–69.99th percentile. 4 = 70th–89.99th percentile. 5 = 90th percentile or higher. Use the worse applicable outcome where a published price conflicts with instrumented measurement.

Primary product depth & liquidity

Score the declared primary product against operators serving the same target market. Use a fixed snapshot protocol. Sportsbook: covered events/sports/leagues and markets per sampled event. Casino: playable, non-duplicate games, identifiable providers and live tables. Poker: median concurrent players/tables/tournaments at fixed weekday/weekend peak and off-peak times, plus variants/formats. Esports: titles, tournaments and markets per tier-1 match. Custom: usable inventory, prize pools, events or liquidity. Average the applicable depth measures first, then convert the result to the peer distribution: 1 = Below the 15th percentile, or the product is listed but materially unusable. 2 = 15th–39.99th percentile. 3 = 40th–69.99th percentile. 4 = 70th–89.99th percentile. 5 = 90th percentile or higher. Local specialists are compared with the same market/product peer group, not with global all-in-one operators. Shared poker-network liquidity must be disclosed and must not be presented as exclusive to one skin.

Primary product integrity & features

Choose the five tests for the declared primary product. Sportsbook: accurate settlement, cash-out, partial cash-out, bet builder, stats/streaming/live tracker. Casino: independent audit/provably fair, RTP shown, providers identifiable, strong search/filters, useful tournaments/jackpots/demo/originals. Poker: anti-bot/collusion controls, published rake, hand histories/HUD policy, stable mobile software, security and table tools. Esports: clear settlement rules, roster/map/walkover handling, stats/streams, live tools, accurate settlement. Custom: clear rules, verifiable fairness, disclosed limits, reliable settlement and stable full-featured UI. A test passes only when it works on the reviewed domain/app; marketing claims do not count. 1 = 0–1 of the 5 tests pass, or a critical integrity failure is confirmed. 2 = 2 tests pass. 3 = 3 tests pass. 4 = 4 tests pass. 5 = All 5 tests pass and no material integrity warning applies.

Secondary products & useful features

Apply the three primary-product factors—value/pricing, depth/liquidity, and integrity/features—to every offered non-primary product. Average those three scores for each product, then average all secondary-product module scores without intermediate rounding. Products not offered are N/A, not zero. 1 = Final secondary-module average below 1.5, or any listed secondary product is materially unusable. 2 = Average 1.5–2.49. 3 = Average 2.5–3.49. 4 = Average 3.5–4.49. 5 = Average at least 4.5, with no offered secondary product below 3. N/A = The site is a genuine single-product specialist with no secondary product. Product count alone, popularity, advertising, awards and vague 'industry leader' claims never earn points; measurable quality does.