Published: 31 October 2026 · Updated: 31 October 2026
SpinMetrics provides AI-assisted market trend research on slot-game popularity in Indonesia and Thailand, using public YouTube search samples. The rankings use lifetime worldwide YouTube views, not local player counts or revenue. The site has no user accounts and provides no gambling services. The intended refresh is weekly on Mondays around 09:00 UTC+8. SpinMetrics Editorial uses AI assistance and checks the content against dated sources.
Views are lifetime worldwide YouTube counts, not in-country views, player counts or revenue. Scores cannot be compared across markets. How we rank games.
What the rankings cover
The current rankings cover Indonesia and Thailand; Vietnam is not covered yet. Both rankings are provisional snapshots of game exposure in localized YouTube search results.
Public sources and weekly window
We use public data only: market-localized YouTube searches collected via SearchAPI. The current video publication window is 3 August–31 October 2026 and rolls forward weekly. The last day was partly elapsed at collection. View counts are observed at collection time, rather than counting only views gained during that publication window.
Each game entry carries source-video URLs and publication dates, an update date and its own confidence note. Original publisher titles remain in the research records; catalog citation labels are neutral. The research uses metadata and does not include watching full videos or verifying their gameplay or outcome claims. The published ranking records retain the research scores.
How the score is built
The same weights apply in every market:
| Component | Weight |
|---|---|
| Video count | 25% |
| Clipped views | 25% |
| Distinct channels | 20% |
| Median views | 15% |
| Recency | 15% |
The method selects up to three videos per channel per game and clips view contributions at the 95th percentile. Candidate sets and normalization are specific to each market. The result is an editorial index of search exposure.
A game qualifies with at least three eligible videos across two identifiable channels. Only videos from that game's own matched ranking searches count; discovery-only results do not. Unrelated or mixed-game videos, active live streams, records missing views, channels or usable dates, and records outside the publication window are excluded.
For each game, select up to three highest-view eligible videos per channel, breaking view-count ties by video ID. Let N be the number selected, V their summed views after clipping each video's views at the pooled eligible-video 95th percentile, C the number of distinct eligible channels, and M the median of the selected videos' original views. The percentile is the value at position ceil(0.95 × eligible-video count) in the market's ascending list of eligible view counts.
For each of N, V, C and M, use L(x) = ln(1 + x) / max ln(1 + x), with the maximum taken over all qualifying candidates in that market, including those outside the top 15. R is the mean of 2^(−age_days/30) across the selected videos. Age is measured against 31 October 2026 for this snapshot; unresolved in-window relative dates use their interval midpoint for recency. Recency weight halves every 30 days.
Score = 100 × [0.25 L(N) + 0.25 L(V) + 0.20 L(C) + 0.15 L(M) + 0.15 R]. Equal scores are ordered by more distinct channels, then more eligible videos before the channel cap, then game name. The displayed raw video counts, total views and median views can differ from the selected subset used in scoring.
Per-entry confidence and data gaps
Every entry has a confidence assessment. Sample confidence and country relevance are separate judgments; current sample ratings are medium or low. Localized searches can include international and promotional material. Language cues and country-filtered search interest provide only limited corroboration and do not establish audience location.
Entry notes retain specific issues such as channel concentration, ambiguous game editions, launch-related exposure bursts, rank sensitivity and relative publication-date intervals. Gaps are flagged, not filled: unsupported specifications remain marked as unverified, and sparse or missing data does not become a growth claim. Google Trends is a separate check and is not part of the YouTube score.
Refreshes and editorial checks
The editorial refresh schedule is weekly, on Mondays around 9am UTC+8. The visible update date identifies the available snapshot; a schedule does not itself establish that newer data exists.
SpinMetrics Editorial prepares the content with AI assistance and checks it against dated ranking records, source metadata and research notes.
Reporting an error
For a correction, include the page address, market, game name, disputed statement and a dated source supporting the change.
Send corrections to contact_us@spinmetrics.net.