A rating difference of 400 points means that the better player will score 0.9 on average (where 0=loss, 0.5=draw, 1=win).
A typical hobby player might have a rating of 1600. A good player at a chess club will have a rating of around 2000, meaning that they beat the hobby player around 90% of the time. A low-level grandmaster will have a rating around 2400, meaning that they beat the strong club player 90% of the time. The top few chess players in the world have ratings around 2800, meaning that they beat low-level grandmasters 90% of the time.
You might think the top human players are nearly perfect, but the top chess engines running on an average desktop computer are probably >800 Elo points above the top humans, and they're getting about 100 Elo points stronger every year, based purely on algorithmic improvements (not computing power).
It's more likely for the stronger player to have 8 wins and 2 draws in that line-up. If you're up against such a big divide you should play safe and aim for the draw, it's quite attainable.
I simplified a bit by saying the stronger player wins 90% of the time. They score 0.9 on average. That could be 9 wins and one loss, or as you say, 8 wins and 2 draws.
You would think that a low-level GM would be able to achieve more draws against the super GMs, but somehow the ratings indicate they don't. It may be that the super GMs deliberately avoid drawish lines against much weaker players and try to keep the game complicated.
A rating difference of 400 points means that the better player will score 0.9 on average (where 0=loss, 0.5=draw, 1=win).
A typical hobby player might have a rating of 1600. A good player at a chess club will have a rating of around 2000, meaning that they beat the hobby player around 90% of the time. A low-level grandmaster will have a rating around 2400, meaning that they beat the strong club player 90% of the time. The top few chess players in the world have ratings around 2800, meaning that they beat low-level grandmasters 90% of the time.
You might think the top human players are nearly perfect, but the top chess engines running on an average desktop computer are probably >800 Elo points above the top humans, and they're getting about 100 Elo points stronger every year, based purely on algorithmic improvements (not computing power).