Posts from On-line rating (discussion) Auto-translated

Let's take a look…
Season – Winter 2009…
Table for January 2nd.
According to the table, 113 games were played.
We are looking at the reports thread, page 17, up to post No. 170.

So, according to the table, 113 games were played, and there are only 170 posts in the thread, although there should be at least 226, and that’s considering that about 10% of the posts in the thread are just chatter (but they are offset by the fact that some users have several reports in one post). Yes, respected SS-D took some of the reports, it seems from wfg.ru. Out of all the reports up to page 17 inclusive, 20 reports are not confirmed, i.e., a fifth of the reports are not confirmed!

Question! Guys, where are the confirmations for the reports? How is this possible???

In addition, I set aside time, and a new modified program (without changing the calculation formulas) will be implemented this season to calculate the rating. And the algorithm of the program will no longer allow an unconfirmed report to be counted. So…

I have confirmed all reports. Out of my 2 unconfirmed ones: victory over the evil romantic and Semom-56. /topic/5173/ http://www.wfg.ru/forum//topic/39/

I have confirmed all reports.
Out of my 2 unconfirmed ones: victory over an evil romantic and Semom-56.

/topic/5173/
http://www.wfg.ru/forum//topic/39/

Excellent! So, will the confirmations be provided or not?

And the first report is actually from the previous Season; we won't go back to it. What happened, happened.

I don't understand why people should suffer, wasting their time for no reason, because of some irresponsible players who "forget" to confirm defeats. I suggest the following: if a report isn't confirmed, it should be motivated by something, either in the "Reports..." thread or in a special one. Even if you just played against this person and think they're cheating, why hide it? Write on the forum: "There is evidence (game logs, saves, screenshots, etc.) that the player violated the rules." If there is no such evidence, then apologize – WHY aren't you confirming it? Why did you even start playing? I believe that if you agree to play with someone, you automatically consider that person honest. But if doubts about their honesty arise during the game, and there is no evidence, you can post it on the forum and stop playing with them, but you MUST confirm this game; otherwise, you are acting dishonestly, and whether your opponent is honest or not is another question. P.S. Given the abnormally high number of unconfirmed games, I think it's time to appoint a special person to monitor confirmations, even to the point of nagging those who don't confirm, if confirmations or accusations against the opponent with an official refusal to confirm don't follow within a certain time (at least 2-3 days). It would be best if this were an experienced player with a good reputation, beyond any doubt. But if no one agrees, I can try to take on this thankless task. In general, I offer my candidacy.

(all games are confirmed, draws and technical wins are not included in the list):

Why is that? It seems like a draw in the new system also affects the rating. Correct me if I'm wrong.

Sorry for steering the topic off course, but I agree with Dirty: draws, in my opinion, should be taken into account in the ranking. My reasoning is as follows: firstly, in the most popular sports (like football and chess), draws are already taken into account; secondly, the rankings of the two opponents who played to a draw were likely different before the game – hence, there should be a slight shift in favor of the one who was lower. Isn't that right?

Okay, we can implement a draw count (probably even starting this season). How about this formula: 1. For a player with a higher rating: decrease the rating by M/2 points, where M is the number of points that would be deducted from the rating upon losing. 2. For a player with a higher rating: increase the rating by N/2 points, where N is the number of points that would be added to the rating upon winning. Perhaps there will be reasonable suggestions to replace the coefficient 1/2 with another one (1/3, etc., perhaps only for one of the players)?

The scheme proposed above has a drawback. If a player only slightly exceeds the rating of another player, they will receive points for a draw, while the other player will lose points and become higher in the ranking. However, in theory, a draw should equalize the ratings of the players. Therefore, I propose a slightly different scheme. A - player with a higher rating. B - player with a lower rating x - points for A winning against B y - points for B winning against A z - points for a draw (added to the player with the lower rating and subtracted from the player with the higher rating) Calculation formula: z = (y - x) / 2 If A and B have the same rating, then x = y; z = 0

As a newcomer, I would like to know what a draw is in Heroes?
Is it when both players simply get tired of playing? Or are there other cases – please tell me...

Isn't it often the case when opponents simply get tired of playing further?

That's exactly what happens in 99% of cases. 1% is for situations where, for example, 2 heroes die in a battle with the War of the Roses, and in that case, agreeing to a draw is certainly reasonable, although theoretically, you could continue playing until someone finds a hero in the tavern or levels up a new one.

There are situations where one can speak of a "guaranteed" draw. For example: both main heroes are alive, but one of them has a huge army (in a direct confrontation, he would simply overwhelm the other with sheer numbers), while the other has many movement points and control (through spells, Pandora, the Scholar, or a pyramid of golems), which the first one lacks and will not have (all guilds are built, and the map is explored). In this case, the "agile" hero can endlessly avoid battle, occasionally capturing a single castle to avoid the 7-day rule; this is almost the only draw scenario. And in the case of a clash between powerful heroes, the one whose fallen hero returns to the tavern sooner will likely win.

Oh, finally I'm off the ranking list, thank goodness! Thanks, I was tired of seeing myself in some 100-and-something place. I agree to always play without it affecting the ranking from now on. But what do people in places 54 to 62 have to do with this ranking, especially since they haven't played a single game, as far as I understand! Especially Rubaka Kreg in 59th place! (Those who don't know, he's Rubaka Kreg on the Portal, and Blackpoint here). It seems like Vitaly (whom I respect as a person and moderator...) is deliberately adding all the well-known Heroes III player nicknames so that those who could potentially be in the top 100 end up lower, in the 110-120 range. Seriously, there's no other explanation: confused. And I won't even mention the constant presence of our esteemed holder of the black mark. In short, however you look at it, I'm no longer participating in this ranking. Now, if I lose to someone, I'll confirm it on other resources (the Portal or Dreamcatcher's website). And, of course, I won't participate in World tournaments to avoid any confusion with ranking points (but I'll finish TK3, since it's not ranked). No hard feelings

With all due respect to Vitaly, I really think the rating system needs some reforms. For example, I suggest setting a threshold of 5 games for participation in the rating. Also, it's a bit unclear why the table is called the SEASON table, but the results and rankings are based on all played games. Otherwise, it could happen that the winner of the season is a player who hasn't played a single game in it. I suppose that if the World Cup in football were held according to a similar system, Brazil would have been the undisputed champion for the last 40 years. ;)
P.S. Kostyan, see my signature. ;)

I suggest setting a threshold of 5 games for participation in the ranking.

I agree. IMO, it's unnecessary to display players who have played 0, 1, 2, 3, or 4 games.

Sorry for the off-topic post in this thread.

Oh! Finally, I'm out of the ranking, thank goodness:D!

Hello! I apologize. There is a limit on the length of posts on the forum, and the table didn't fit completely. The rankings did not reflect the places from 164th to 169th. I will fix this today.

Especially Rubaka Kreg in 59th place (for those who didn't know - this is Rubaka Kreg on the Portal, and Blackpoint with us

This is not my mistake. Thank you for the information, we will fix it.

It seems that Vitaly (whom, as a person and moderator, I respect...) is deliberately adding all the known nicknames of players in HeroesIII, so that those who could enter the top hundred, for example, would drop lower to 110-120th places. Really: there is no other way to explain this.

All players whose nicknames were recorded in the reports on HeroesWorld are included in the ranking.

I won't even mention the constant presence of our honorable holder of the black mark.

We will fix it.