METHODOLOGY
How we build
the tier list.
1. Ranked focus, not pro play
Pro and tournament metas drift heavily from what works in solo queue. Our tier rankings are weighted toward ranked-ladder performance at high MMR (Diamond and above) — the bracket where most active players live. Tournament data is referenced for context but is never the deciding factor.
2. Sources we monitor
Each patch a scheduled job scrapes a fixed roster of sources, weighted by how directly they reflect the Thai ROV ranked meta. Editorial tier lists carry twice the weight of pro-play data:
- rov.in.th — official Thai server portal. Patch notes, hero data, and the in-game item builds shown on every hero page.
- Zathong, CommonSenseGamer, FPTShop, RichmanShop — editorial tier lists. These are the primary signal.
- Liquipedia RoV Pro League — pick, ban, and win-rate context for high-skill play. Supporting signal only.
- Garena Thailand patch notes — the authoritative record of what each patch changed.
Sources degrade, and we would rather say so than quietly average them in. A freshness check flags any source that has stopped moving, and its weight drops until it recovers.
3. Scoring
Heroes are scored by aggregating signals across sources. Strong consistent positive signals (multiple creators independently rating a hero highly, high ranked win rate, healthy pick rate) push a hero into S+ or S. Mixed or unimpressive signals land them in A or B. Heroes rarely picked or with poor solo queue performance fall to C. We use Borda-style rank aggregation rather than averaging numeric scores, because the underlying sources don't all use the same scale.
4. How the commentary is written
The written commentary on each hero is drafted by Claude from the aggregated source signals — the tier each source gave, the win/pick/ban numbers, and what the patch notes changed — and published automatically. It is not a translated video transcript, and it is not copied from another site, but it is machine-written, and we would rather tell you that than let you assume otherwise.
What that means in practice: the commentary explains why a hero sits in a given tier this patch, but it is only ever as good as the sources underneath it. The tier placement, the win/pick/ban rates, and the patch-note references are traceable; the prose around them is a summary, not a first-hand match report.
5. Update cadence
A scheduled job re-scrapes the sources and rebuilds the tier list nightly, so a Thai-server patch is normally reflected within a day of going live. Every hero page shows the date its tier was last reviewed, and the active patch number is always shown in the site header.
6. What we don't do
We don't copy/paste tier images from other sites. We don't accept payment to place a hero in a higher tier — sponsors on this site pay for banner placement and nothing else. If we ever ran a sponsored hero spotlight, it would be labeled clearly and would not affect tier placement.