IronAxe
Auto-translated
On a low rating, it will always slowly increase because the oscillator is positive, and the correction is insignificant due to the low rating. There is nothing surprising about this... On a high rating, the correction is already more significant, but you can still manage with normal gameplay... But a 30% drop after just one loss is already Forex, not Heroes!
Where is the correction insignificant due to the low rating?
A player with a rating of 100 takes last place and raises their rating to 130!!!
That's a 30% increase in their rating after losing, with a 1/10 chance of winning!
Where is it more significant on a high rating?
A player with a rating of 900 takes last place and drops to 850!!!
That's only a 6% decrease in their rating, with a 1/10 chance of winning.
And your statements are only true if the decrease is by 1/3.
Then yes, it works as you said. The correction is small on a low rating, and significant on a high rating.
I don't understand why the rating should care about the players, and not the other way around?
State the task clearly again - I'm already confused about what and where to calculate. Ask a specific question.
The question is this. Let's calculate the rating of the Best tournament according to your system, where it's not good to penalize the leaders. And then calculate how the next tournament with the same players would look. Let's say Dmitry had a major setback, Radik played a little worse than before. But VDV significantly improved his game.
That is, what are the chances of the closest rivals overtaking each other in a fair fight if they all participate in the tournaments and don't skip them?