The error is thinking of ratings as a ladder system when it's not.
If you mean what if a lower rated player scored well in a match against a world champ, the champ would lose a lot of rating points, and the low rated player would gain many points, but there would still be a large gap, with the world champ still on top.
Unless they played 100 games and the lower rated player won them all or something like that.
Google around for the ELO and Glicko systems to give you a detailed answer.
The short summary as relates to your question is :
Rating changes for each player are computed post-game or post-event based on the how different the actual result of the games played ( 1-0, 0.5 - 0.5, 0-1) varies from the expectation based on the ratings before the game. This expectation is determined by the rating system (ELO, Glicko etc.).
Your example makes little sense in that the system doesn't care about past performances ... it only cares about the ratings (for both of you) prior to the match you are about to play.
Google around for the ELO and Glicko systems to give you a detailed answer.
The short summary as relates to your question is :
Rating changes for each player are computed post-game or post-event based on the how different the actual result of the games played ( 1-0, 0.5 - 0.5, 0-1) varies from the expectation based on the ratings before the game. This expectation is determined by the rating system (ELO, Glicko etc.).
Your example makes little sense in that the system doesn't care about past performances ... it only cares about the ratings (for both of you) prior to the match you are about to play.
But those ratings would be based on your past performances.
I get what is being said, though. It makes sense.
Google around for the ELO and Glicko systems to give you a detailed answer.
The short summary as relates to your question is :
Rating changes for each player are computed post-game or post-event based on the how different the actual result of the games played ( 1-0, 0.5 - 0.5, 0-1) varies from the expectation based on the ratings before the game. This expectation is determined by the rating system (ELO, Glicko etc.).
Your example makes little sense in that the system doesn't care about past performances ... it only cares about the ratings (for both of you) prior to the match you are about to play.
But those ratings would be based on your past performances.
I get what is being said, though. It makes sense.
This line of thinking is analogous to equating "rich vs. credit history". A bank may not give you a loan for a car/house just because you are rich "TODAY"! You could have won the lottery yesterday but were a deadbeat the weeks and years before that. They'll expect cold cash for that car/house!
However, you could have the same amount of money and have been diligently paying your bills for years and years ... and thus get put in a different category.
The rating system however makes no such distinction ... just tell me "what you were rated BEFORE the tourney/event" and I will adjust your post-game ratings based on this information.
Google around for the ELO and Glicko systems to give you a detailed answer.
The short summary as relates to your question is :
Rating changes for each player are computed post-game or post-event based on the how different the actual result of the games played ( 1-0, 0.5 - 0.5, 0-1) varies from the expectation based on the ratings before the game. This expectation is determined by the rating system (ELO, Glicko etc.).
Your example makes little sense in that the system doesn't care about past performances ... it only cares about the ratings (for both of you) prior to the match you are about to play.
But those ratings would be based on your past performances.
I get what is being said, though. It makes sense.
This line of thinking is analogous to equating "rich vs. credit history". A bank may not give you a loan for a car/house just because you are rich "TODAY"! You could have won the lottery yesterday but were a deadbeat the weeks and years before that. They'll expect cold cash for that car/house!
However, you could have the same amount of money and have been diligently paying your bills for years and years ... and thus get put in a different category.
The rating system however makes no such distinction ... just tell me "what you were rated BEFORE the tourney/event" and I will adjust your post-game ratings based on this information.
That is a BEAUTIFUL analogy Shivsky. Coulden't have said it better myself.
This is unlikely to happen, but I'd like to know anyway:
Supposing one player plays another player, and wins most of their games.
What if the losing player has better results against the rest of the world than the player who won?
What would the ratings look like?