Drag the sliders. This is the same formula used on every operator — operators can see exactly which bucket would move them up a band, and readers can see how much weight each input carries.
Avoid unless you have a very specific reason. Multiple buckets failed our threshold.
Use with caution. Has measurable problems we documented — possibly slow payouts, weak FI support, or unclear bonus terms.
Functional but not exceptional. Read the WATCH FOR list before depositing — at least one bucket is weak.
Solid operator. Some watch-outs in the review — read them, decide if they matter for you.
Worth depositing. Best-in-class on multiple buckets — payout, license, complaint handling all check out.
The PUTKI Score (0-100) is a transparent weighted sum of four buckets. Trustpilot enters as a real consumer signal. Our findings dominate the data. The editorial verdict carries judgement. Partnership transparency is the guardrail — operators that hide a commercial tie are docked.
Live consumer rating, normalised 0-5 → 0-100. We fetch the public Trustpilot profile JSON-LD and cache the rating per operator. The signal is real, the freshness is shown.
Hard data we measure ourselves: license tier, regulator complaint volume, measured payout times, game library count from each operator’s lobby. No judgement, just numbers.
Editor’s judgement after using the casino: Finnish-language UX quality, support responsiveness via mystery shop, bonus T&C honesty. Subjective by design, transparent by display.
Guardrail. Non-partners get full marks (no conflict to disclose). Partners that disclose openly get full marks. Partners that hide the tie are docked 10/10 — a publishing bug we want immediately visible.
Each bucket’s sub-components renormalise inside their bucket — a missing payout-speed measurement won’t deflate "Our findings" below its 35% ceiling. A missing Trustpilot rating drops the entire 30% bucket and the remaining buckets renormalise.
Two things you’ll notice are NOT on this list: welcome bonus and commission rate. Bonus quality is scored inside the editorial bucket (terms, wagering, withdrawal limits — not the marketing-headline number). Commission rate is the entire 10% transparency bucket — if we earn from a brand, that fact is disclosed and the brand has to clear a higher bar. Everything below moves the score by exactly zero:
Mittari scores (0-100) tell you in one number how good an operator is for a Finnish player right now. Scores update weekly and react to what actually happens - payout speed, complaints, licensing.
The score does not predict whether you will win. It predicts how likely the operator behaves fairly when something goes wrong.
Community signals, complaint handling, the editorial view of how an operator behaves in edge cases. Commercial deals do not raise the score. Affiliate relationships are listed on affiliate.
Data-driven factors update weekly. Editorial assessments update monthly. Material changes (license loss, complaint surge) update immediately.
Commercial relationships are capped at +5 / 100 and disclosed on every relevant operator review page. In 2027 (Finland’s own license) we will refresh the relationships entirely and announce the new ones upfront.
PUTKI HQ uses artificial intelligence as part of its publishing operations - for monitoring news sources, classifying content, synthesizing articles, and generating editorial material. All published content passes editorial review before publication.
Where AI is used:
Sharpness is a 0-100 score over bookmaker market behaviour. Deterministic, AI-free, computed live from The Odds API. We publish it verbatim so any reader can verify it.
Sharpness = round(
0.50 * implied_prob_score
+ 0.30 * consensus_tightness
+ 0.20 * recency_momentum
)
implied_prob_score = clamp( (1 / best_decimal_odds) * 100, 0, 100 )
consensus_tightness = 100 - clamp( stdev(implied_per_book) * 7, 0, 100 )
recency_momentum = clamp( 50 + (avg_implied_now - avg_implied_24h_ago) * 500, 0, 100 )
# defaults to 50 when no 24 h history yetBands:
A momentum modifier is shown when recency_momentum ≥ 60 ("tightened today") or ≤ 40 ("softened today"). Otherwise none.
PUTKI HQ does not place bets. Sharpness is editorial analysis of market consensus, not a recommendation.
Open the ranking, expand any row, read the 4-bucket breakdown for that specific operator. The math is the same as the slider above.