An ELO rating is the result of statistical calculations designed to measure the playing strength of players. These methods were developed by Arpad Elo and are named after him. The ELO system is in almost universal use and has been further refined by others, notably Glicko, but is still named after its creator.
In devising the ELO scale, some premises were used that are worthy of note. It was decided that a rating of 2000 would be the equivalent to scoring 50% in a US Open Championship. It was also decided that a player's rating would never be negative.
Importantly, the standard deviation was fixed at 200 points. This means that if a player's true strength is 1500, he will score around 68% of his results within the range of 1400 to 1600 (as measured by performance formulae). Another result of fixing the standard deviation at 200 points is that it also defines playing categories. For example, most International Masters and Grandmasters are in the 2400 - 2600 category, most national masters in the 2200 - 2400 category. Those in the 2000 - 2200 category are called Experts, or Candidate Masters.