Player rating for tournaments across all installments of Heroes.

And you are wrong again. Let's say we have players A and B, and they participated in 9 Tournaments together. In the first eight, player A consistently showed a result of 31 days, and player B just as consistently showed 32 days. The winner always had 30 days—basically, BB-type maps.
But in the ninth tournament, B showed the same result as the winner (the organizers decided to change the heroes version for variety, for example, to Heroes 4) with a result of 3 days and took second place. Player A took third place with 4 days.
Who was closer to the winner more often? Correct, player A (8 out of 9 times).
And who gets a higher rating? Correct, player B.
A - 849.19
B - 850.00

:) You're twisting the numbers nicely. The part highlighted in black is "showed the same result as the winner" in your post. Since placements are not taken into account at all in my system, player B, unlike player A, managed to match the map leader's result once, whereas player A simply didn't manage it, which is why B overtook him by 0.81. And what the hell does your example show :) It shows EXACTLY what I said: that players A and B almost always stayed near the leader in any of the Tournaments, they have roughly equal ratings, and player B got lucky that he also grabbed a hundred for leadership, unlike player A. Where is the contradiction? Your version is completely incomprehensible. If someone considers your bonus in Heroes 4 (out of 55 points earned, you got 28 purely for your placement) as a good boost for themselves, then by all means, let them vote for systems that account for placement. In my system, you would have received only 27 points purely for your result relative to the near-limit result, which is fair for offline tournaments.

This is not a flaw, but a virtue.

:) Well, obviously :D

You know, I'm ready to forgive a 0.1 rating point margin of error in any system.

But at 0.81, you're already losing it :) What kind of "0.1 margin of error" is that? I'm telling you that in the 4th stage you got 28 free points :) I didn't get any freebies in that tournament. And you're trying to pretend that I also got a freebie in Heroes 4 :) uh-huh—the limit result there is based on days and points. Players should not be held responsible for the attendance or absence of other players in a tournament with their rating!! Nor should they receive free rating points because other players didn't show up. Attendance is the organizers' headache, not the players'.

Conclusion: beat strong players and you'll get your rating. Especially if we count separately by versions, then everything is fine.

You're quite a pro at drawing conclusions. Then tell me, which strong players did YOU personally beat in the fourth stage? Specifically for that version :) 28 free points—and everything is fine :)

Added after 9 minutes

and? you don't like the first option?
In my opinion, everything is correct. :smile01:
Two sixths and one twelfth is roughly equal to one first place.
IMHO.

Even three twelfth places are better than one first place and nothing else :) Where is the guarantee that if the Prophet had played as Alexei did in the following two tournaments, he wouldn't have grabbed two twentieth places there? But he didn't play, and that's great :) After all, the more tournaments a player plays, the more chances they have to lose; beating the strongest is NOT always possible :)

Added after 9 minutes

The problem is exactly that everyone sees it differently. You see that the placement/days ratio should be 3 to 7, while Alexei sees it as roughly 1 to 2
(including the bonus). Kyle and I believe it should be 1 to 1, and Topor considers placement an evil and wants to ignore it entirely. If the vast majority (or the administration) spoke out specifically for one ratio, then the question could be closed.

As for the placement parameter, I don't just want to ignore it, I've ALREADY ignored it a long time ago :) However, if it comes to a vote and systems that account for placement take over (meaning I see the system without placement failing), then I will give my vote to the one that has a greater ratio in favor of the actual result rather than the placement (for example, 3 to 7 as Kamikaze suggests). And if there are other people who understand the flaw that placement introduces into the rating and suddenly vote for my system, I will urge them to vote for Kamikaze's system instead.
1:1 is a "must-give" freebie :)