Read the signal—and its limits.
Every result describes the selected stored sample. It is not a claim about every League player or a prediction of an individual game's outcome.
Data flow
- DoxBot's server retrieves supported match data for configured tracked accounts through Riot's API.
- The server normalizes relevant match, participant, item, rune, spell, role, queue, patch, tier, and region fields.
- The public dashboard queries stored records. A page load does not make an authenticated Riot API request from your browser.
- Results are grouped using the filters shown on the dashboard.
- A browser may reuse a matching response for approximately five minutes. Refresh data bypasses that browser cache and requests the current stored result from DoxBot.
Metrics
Games
The number of matching participant-build records in the selected sample.
Pick rate
The matching build's games divided by all matching games returned for the selected sample.
Win rate
Wins using the matching build divided by its recorded games. Correlation does not establish that a build caused those wins.
Recent-game window
A cap on recent matching records considered by the query. It is not necessarily the number of games played by one person.
Build context and insights
Each build may show the rank, patch, region, and number of tracked players that contributed to that exact item combination. Rank is the ranked context available when DoxBot ingested the match and may be absent for older records. The most-played label identifies the item combination with the largest count in the selected sample. A top-win-rate label is shown only when a build meets the published minimum sample; it is not a recommendation or a prediction.
Rune pages and summoner-spell pairs shown inside a build are grouped only from matches containing that item combination. The broader rune and spell panels summarize the entire filtered sample.
Matchups
Where role information is available, DoxBot pairs a tracked participant with the enemy participant assigned to the same team position in Match-V5 data. Matchup filters use the resulting opponent champion. Role assignment can be missing or imperfect, and matchup results inherit all other sample limitations.
Player groups and contributing games
Tracked-player groups are maintained by DoxBot editors. A Pro label is a directory classification, while Stage 5 requires an active verified roster entry connected to the tracked account by PUUID. Stage 5 reflects the DoxBot coaching verification process and is not inferred from Riot rank.
Expandable contributing-game rows show up to three recent stored participant records for an exact item combination. They may include the public player label, approved portrait or placeholder, opponent champion, result, rank context, patch, items, runes, and spells. Raw match payloads, PUUIDs, Discord IDs, private verification notes, and verifier IDs are not returned.
Limitations
Current data comes from DoxBot's configured tracked-player set, not the complete League population. Small samples, player selection, patch changes, rank distribution, role classification, incomplete matches, and item timing can bias results.
Low-sample results are labeled and should be interpreted cautiously. Items shown represent recorded end-of-match inventory classification; they do not necessarily describe exact purchase order.
Static game data
Champion, item, rune, and summoner-spell names and artwork are resolved from Riot-supported static data where available. Static assets and match data can update on different schedules.