Data Methodology
Overview
MetaBot.GG provides comprehensive statistics and analysis for League of Legends, Teamfight Tactics, and other games. Our data methodology is designed to give players accurate, actionable insights based on real match data. In order to do this, we collect data from a variety of sources depending on the game that we are analyzing. In the spirit of open transparency, this page is dedicated to provide details on how we collect our data and how that allows us to provide you with accurate stats that get updated on a daily basis.
Data Sources
We collect data from a variety of sources to provide you with the most accurate and competitive information that we can.
League of Legends
Our main source of data for League of Legends comes from the publicly available Riot Games Developer API that Riot provides for the community. This data comes from the development team for League of Legends themselves, which ensures that the match data we collect is both useful and accurate.
In order to collect match data, we first begin by collecting player account data so that we can create a registry of League of Legends users. We update this list over time, continually growing our list of players we find matches for. From there, we routinely scan their latest matches and filter out any matches that aren't in the latest patch. This way, we only include match data from the latest League of Legends patches and keep our stats up to date with the latest changes from the Riot development team.
We don't change our statistics over to the latest League patch until we have collected data on at least 100,000 matches. We do this to ensure that we have a sizeable pool of data before running out analytics and providing players with statistics. Collecting this many matches doesn't take long, and we usually are able to update our site to the latest patch within a day or two of the new patch launching.
We continue to collect match data for new patches throughout the entire patch, and update our aggregated statistics on MetaBot several times a day. This ensures our analytics stay up-to-date with fresh aggregations.
TFT
Our main source of data for TFT comes from the publicly available Riot Games Developer API that Riot provides for the community. This data comes from the development team for TFT themselves, which ensures that the match data we collect is both useful and accurate.
The primary Riot Games Developer API doesn't provide all the data we need for TFT though, so we supplement our data with a more thorough community-ran API called Community Dragon. This allows us to get more thorough information about unit stats (attack damage, ability power, etc), assets, and more.
Similar to League of Legends, we start by compiling a list of players that we continually add to. We use this player list to scan match data from the Riot API, and filter out the match data so that we only save the matches that come from the latest TFT patch. This ensures that our data updates between patches and only contains accurate data from the most recent patch changes.
After a new patch releases we wait until we collect data on at least 100,000 matches before updating the site to the newest patch, which usually takes a day or two after a new patch launches. We continue to collect data on the current patch afterwards to grow our match data pool, and update our aggregated statistics several times throughout the day.
Statistical Analysis
Our platform calculates various metrics including win rates, pick rates, ban rates, performance statistics, and more. We use statistical methods to account for sample size and ensure that our data is statistically significant before making recommendations.
MetaBot is scheduled to perform analytics on our data several times throughout the day, and is designed to only use the latest data in its calculations. Once we believe that we can offer stats with a high level of statistical confidence, we update the MetaBot website to display our most up-to-date analytics.
MetaBot Tier Rankings
Our proprietary MetaBot tier system analyzes performance across multiple dimensions including win rate, pick rate, average placement, consistency, etc. We use a wide variety of relevant metrics to determining our rankings to provide strong recommendations on where characters fit in the current meta.
Privacy & Compliance
MetaBot.GG complies with Riot Games' API Terms of Service and respects player privacy. We do not collect or store personal information beyond what is publicly available through Riot's API. All data is aggregated and anonymized for statistical analysis.
We are careful to stay fully compliant with the developer's policies regarding any game that we collect data on.
All of the data from players and matches that we collect is used anonymously, and we only hold on to it temporarily and delete it from our databases once it is no longer relevant. We do not collect data that contains any personal identifiable information.
