Tournaments and player rating.

http://heroes3stat.tk/tournament_rating/
Without seasonal games. Start date: 01.01.2014 (Everyone starts with 1500).
p.s. Sorry, but Ive didn't make it into the top 10 :(
p.s.2 I have a feeling that now they will start looking for "unworthy" tournaments.

First of all, thank you very much!
Secondly, I want to say that I have no intention of outperforming everyone, and I'm not going to look for anything to "push through" my arguments.
I want to see the most objective rating, and if the most objective option is tournaments + season, I'm all for it, but for now, it seems that this is not the case. And Ive isn't paying me to get him into the top 10 :D

http://heroes3stat.tk/tournament_rating/

And this is for 2 years; it seems that the objectivity here is not much less than in the main forum rating. What if the period is 4-5 years?
I remember the rating for 2 years according to the current scheme; it was painful to look at. Since we're already going down this road, maybe we can look at a similar period using the tournaments + season system for comparison?

If you think that having an unlimited amount of time, you can farm a good rating without having skill, you are wrong. To move up the rating, you need to beat strong players because you will get a small increase in rating for weak players, but at the same time, there is a huge risk of losing a lot in case of a loss (and in Heroes, in one specific game, anyone can lose).

But calculate, winning against players with a rating 200 points lower, you get approximately +8 points. Playing 2-3 games a day, 8 points per win, how much time will it take? Well, let's say there will be a couple of losses, and the event will happen a week later. The point is that theoretically, this is possible, but it shouldn't be.

The main problem is that the rating of some individual players does not reflect reality because they deliberately do not write down the games they won but confirm the games they lost, thereby artificially lowering their rating. This "problem" can be solved by taking the tournament rating (it would be interesting to see it, by the way), but as already written, the period will be too short. Therefore, for some, the season is the only way to play it.

http://heroes3stat.tk/tournament_rating/
Here, for 2 years, many have less than a hundred games. What if we take 5 years? Will the period be too short?

The main argument against the "tournament rating" is only the lack of a period; no one suggests calculating it based on 2 tournaments. Let's imagine that the rating is calculated over a long period, with everyone having 100 games, as the system requires. Yes, there are easy games in tournaments that distort the picture, but overall, it's more or less a clean product.

Next, we add:
- part of the reports from the "Season" tournament (and only part, because someone wrote a report, and someone didn't), similar to the situation that we add games from the group stage of the tournament, but not the playoffs (roughly, but I hope the point is clear)
- games mainly with lower-rated opponents, similar to the situation: you got lucky, reached the final in a double-elimination tournament, and all the games were easy; there are many games, but little resistance.
- games mainly with higher-rated opponents. Similar situation: you play in a tournament, lose a dozen games, but they don't kick you out; it seems like you're playing no less than the previous person, only he's getting a plus, and you're getting a minus, although you think you're playing just as well.

What do we get as a result?

Well, and of course, there should be a coefficient for tournaments, but that's another story...