ELO is a rating system that assigns a player or team a numerical value reflecting their skill level. The higher the rating, the stronger the opponent is considered to be. The system is named after Arpad Elo, a physicist who originally developed it to calculate the strength of chess players. Later, the principle was adapted for video games and esports because it effectively solves the main task of online matches: pairing opponents of roughly equal skill so that the game is fair and interesting.
The logic of ELO is simple: after each match, the system compares the expected outcome (who, based on rating, should have won) with the actual result. If a strong player defeats a weak one, they receive few points, since this was expected anyway. But if a player with a lower rating beats a stronger opponent, the gain will be noticeably larger, since this is an unexpected result. Rating decreases after a loss work the same way: losing to a weaker opponent takes away more points than losing to someone with a higher rating.
In esports, matchmaking in many games and on many platforms is built on ELO or its modifications — the system pairs players with opponents of similar rating so that a match doesn't turn into a one-sided blowout. Custom extensions of the classic ELO system are common — for example, systems that account not only for a win or loss but also for individual in-match statistics, making the rating more precise.
The term is used not only for individual players but also for teams — for example, when compiling team strength rankings before tournaments or forecasting match outcomes. Players often say "climb ELO" or "lose ELO," referring to a rating's rise or fall after a series of matches. Talking about "low ELO" or "high ELO" is essentially a quick way to assess a player's level without a lengthy explanation of their skill.
It's important to understand that ELO is not an absolute measure of skill but a relative value that reflects results specifically against the opponents a player has actually faced. Because of this, ratings from different games, different platforms, or different regional player pools cannot be directly compared — the numbers are calculated within closed systems and depend on how the specific community of players behaves.