How to read your BAR leaderboard rating and OpenSkill score
Where to find your rating on the BAR website, why the leaderboard number differs from matchmaking, and what mu and sigma mean without the math headache.
Tags: beyond all reason, bar rating, openskill, leaderboard, ranking, matchmaking, bar guide
Where to find your rating
Log in at the BAR website and head to the ratings page. That is where the leaderboards live. You will see your OpenSkill rating displayed for different queue types: one-versus-one, small teams, large teams. Each queue tracks its own rating separately.
If you are looking for an exact position like top 100, this is the page to check. Just pick the correct queue and sort descending.
Leaderboard rating versus match rating
Two numbers float around and they serve different purposes. Your leaderboard rating applies three times the uncertainty factor, which inflates the display number. Your matchmaker rating uses one times uncertainty and is what the game actually uses to find opponents.
A player with an OpenSkill score of 39.7 might see 33 on the leaderboard. Neither number is wrong. They are just calculated differently. The matchmaking system uses the tighter uncertainty value so you face fair opponents. The leaderboard uses the wider uncertainty so it looks more generous on a public scoreboard.
OpenSkill terms without the headache
OpenSkill stores two values per player:
- Mu represents your estimated skill level. Winning raises it. Losing drops it. This is the number most players care about.
- Sigma represents how certain the system is about your mu. New accounts have high sigma because there is not enough data yet. After many games, sigma shrinks and your rating stabilizes.
Mu goes up and down based on results. Sigma only trends down unless something unusual happens in your match history. Think of mu as your current standing and sigma as the system confidence bar.
Why ratings change
Every match shifts your mu based on who you faced and whether you won or lost. Beating a higher-rated player moves your rating up more than beating someone below you. The opposite holds for losses. This is standard Elo-style behavior under the hood.
New players see large swings early because sigma is high. After fifty or so games your rating movement per match settles into something more predictable. Do not panic about early drops or spikes. They are just the system finding where you belong.
Season resets
BAR runs seasonal leaderboards. At the start of a new season, ratings carry over but the board resets. Season three leaderboards, for example, show only games played within that season. Your underlying OpenSkill rating does not disappear between seasons. It persists as the basis for the next season's starting point.
Check the ratings page after any patch or season change since queue structures sometimes shift. Queue ratings stay separate, so a good large-teams player might have a very different number in one-versus-one.
What to do with this information
Use ratings to find balanced games, not as a personal scorecard. Queue into modes where you want to improve. Ignore the leaderboard display number for matchmaking decisions since matchmaker ratings are the ones that matter there. If you want to climb, focus on playing more games and learning the meta. The rating takes care of itself.
Creed of Champions
Creed of Champions runs a competitive BAR scene built around teamwork and zero toxicity. Players here treat rating as feedback, not identity. Training sessions are open across skill ranges and time zones. The community welcomes anyone willing to learn, communicate, and leave the team-blame at the door.
[Crd] One of the few places where you can for sure coordinate with people in matches with a good supportive attitude. Everybody tends to be understanding and constructive.