Posts from Player rating for tournaments across all installments of Heroes.

I agree with that! I'm talking about maps where the best result is 10 days, and the worst is 14, with 20 participants. In that case, you can get a rating of 71.4 points for 20th place! But if you introduce a component that takes the place into account, it becomes more logical. :)

This, as you rightly pointed out, is another (no less sad) side of the coin.
The place taken in the tournament, in my opinion, is the most universal, if you will, standard unit, operating in almost any competition. And we are trying to reinvent the wheel.

How should I put it...
But what if, for example, there are 6 or 7 participants? The first completes it in 4 days, the second in 20, the third in 25, and the last in 109. And how, tell me, should we not take the result coefficient into account? ;)

That's why a two-component scheme was proposed: place and result. They decided, as far as I understand, to equalize their importance, which I fully agree with, after analyzing different tournaments.
In such a system, components dependent on the result (the leader's result is the benchmark) and on the place are taken into account, according to a curvilinear function (more concave than in biathlon).
In my opinion, this is the most balanced and simple system!

Firstly, for this, a certain quorum of participants must be gathered, below which the tournament cannot "fall" (this number must be determined empirically).

He will find his high place one way or another.

Otherwise, in one tournament, due to the different possibilities of the maps, one can get so far ahead of the others that the rest will be just a formality. The principle of competition is lost.

So, you've basically decided what needs to be included in the calculations, but you haven't yet come to a unanimous agreement on how to do it in the most rational way?

What quorum are we talking about? The number of participants is always spontaneous. Any Tournament, even an unofficial one, becomes official once a certain number of players are gathered, the quorum is calculated, and put into effect. What does it mean: "below which the Tournament cannot 'fall'?"

Here again, it's not quite clear: if there are several players of this level, what - in your opinion, will they periodically change places, giving another player a chance to rise? That is, if a strong player A takes 1st place in Tournament X, and another equally strong player B takes 2nd place, then in the next Tournament Y, B will become first, and A will take second place? In order for them, as you say, to "find their high place," they simply need to make mistakes in turn. In this example, there are only 2 players, but if we take a real example, i.e., if we consider 4-5 such "guys"? And if a newcomer arrives, but his level of play is exactly the same, how will he then be able to find that very - high place?

That very "empty formality" will remain in any case - only it will apply to those who occupy places below 10.

Hello,
Excuse me for butting into your conversation, but:

I looked at this topic and the topic about the rating: and came to the conclusion that you will never reach a compromise, and even less come to a unanimous opinion.
In my opinion, someone with authority needs to choose one of the options proposed here. Otherwise, this dispute will continue forever... because it is impossible to come up with a completely objective rating that would take into account the interests of both the first and the second...
because your Tournaments are not big sports, where all the rules are clearly defined.

I agree with that! I'm talking about the fact that there are maps where the best result is 10 days, and the worst is 14, and there are 20 participants. This way, you can get 71.4 points in the ranking for 20th place! But if you introduce a place component, it becomes more logical. :)

How can I put it...
But what if, for example, there are 6 or 7 participants. The first one finishes in 4 days, the second in 20, the third in 25, and the last in 109. And how, tell me, should we not take into account the result coefficient? ;)

This, as you rightly noted, is another (no less sad) side of the coin.
The place in the tournament, in my opinion, is the most universal, if you will, standard unit, operating in almost any competition. And we are trying to "reinvent the wheel".

It's very simple. Let's say we set a limit of 15-20 participants, and once that limit is exceeded, the tournament becomes a ranked one. And competing in a tournament "against myself" isn't very interesting to me, for example. This would create a direct incentive for participants to attract new or well-known old players. This is possible, believe me. You just need to make the participants interested.

Kamikaze, you say that "the most compromising and simple system, IMHO," but you haven't really compared the result coefficient with the place a person took, which means what is the most compromising? Or do you mean that the principle is already calculated, but you just haven't figured out the nuances of this system yet...? So, you've already decided what needs to be taken into account in the calculations, but you haven't come to a unanimous agreement on how to do it in the most rational way?

Give an example of such a situation from one of the previous stages of the Kanigem tournament. Of course, the top results may be close, but they still differ by 1-2 days.

Quote:
Message from Ley
That "empty formality" will remain in any case, - only it will apply to those who take places below 10.

I am more concerned about the "formal upper limit," the absence of which makes the championship interesting.

Why is the place taken a reference unit?? For determining the winner of a tournament?? I agree. But what does this have to do with the rating calculation? What, are ratings calculated for all athletes, even amateurs? Who needs to calculate ratings for scrubs.

Therefore, when you speak of professional sports and ratings, we imply that athletes of the PROPER level participate in the competitions, those who have gone through a ton of qualifying stages and won many competitions taking first places. Therefore, when they gather for competitions TOGETHER, their results among themselves will be close in MAINLY. Therefore, the place there has significant importance—if not the PRIMARY one. But in our tournaments, there are no qualifying matches, the composition of participants and the result are unpredictable—therefore, it is exactly here that the place is SUBJECTIVE in nature. Our situation is more like throwing athletes of the lower, middle, and higher leagues into one general pile and forcing them to compete with each other—AND AT THE SAME TIME the RATING will be calculated based on the place taken, moreover, attendance at the competitions is not mandatory, and it may turn out that one from the higher league comes, three from the middle, and the rest from the lower. Now, if we proceed from the place taken, there are immediately a bunch of flaws. Because the attending athletes from the middle and lower leagues will receive a hefty boost to their rating thanks to the absent athletes of the higher league—effectively, they will steal the rating that corresponds to higher league players. But if, in this case, the rating were considered only relative to the result, there would be no flaw. Here, the subjective place would not influence—and the rating would be calculated for each relative to the result of the leader from the higher league. In this case, a flaw can occur only in ONE RARE CASE, when NOT A SINGLE player from the higher league shows up for the competition, which, you'll agree, will be much rarer, if it happens at all. Practice has shown that at any of the Kanigem stages, we received maximum results from the LEADERS!

Stage number 7. The top ten differ by only one day!

We are calculating the rating, not determining the winner, as Axe correctly noted. And therefore we will consider both the place and the result!!!

If I'm not mistaken, the rating will determine the final standings for the ranks (for a certain period of time - it's not yet clear what period). That is, the rating will determine the winner of the overall championship, which is essentially a Tournament. And you're asking, what does the Tournament have to do with it.

Secondly, the absence of strong participants will inevitably lead, as you rightly noted, to the fact that others will take their places.
Why this bothers you is unclear. There are no irreplaceable people. And dragging

Thirdly, it is very likely that the winner in the end, say, of the year will be a player who has not won the most prize-winning places, but has received a higher rating (i.e., has created a gap for himself once). This is what you will achieve. And no matter how much others "tear themselves apart" later, they will most likely not have the opportunity to catch up with the leader.
Are you familiar with the word "competitiveness"? Or, Shurik, is this not our method?

Excuse me, but you are not far-sighted. Let's take the best in account. Plus, we haven't had large maps yet. But who said that there won't be any.
You are unfairly referring to practice, which can "end" at any moment, and this moment is obviously not far off.

Excuse me, but what is the rating for then? If the rating is for the SAKE of the rating, then I "wash my hands". Although it seems that everything is different.

Compared: one to one!

Since the tournament is considered official if at least 20 participants apply to participate in it,

...all participants who make it into the top 20 receive points (1st place - 20 points, 2nd place - 19 points... 20th place - 1 point). We have many people who can show good results, and to avoid offending anyone, it is quite acceptable to award points to the top 20.

That is, 50/50. But Iron said that "equally" is not suitable... - see his post just above. Who should we "believe"? :)

In general, maybe we should calculate the rating depending on the tournament, i.e., directly from the task itself (speed, experience, skeletons)? Maybe it would be more convenient to do the calculations separately for each tournament; in my humble opinion, it would be easier to calculate "such things" in total.

I don't understand what you don't understand here :) A place in a particular tournament is one thing. And a rating based on the results of several tournaments is another. The participation of players is random. Therefore, the subjectivity of the places occupied by the participants is higher than their ACTUAL results in days. I think Kamikadze understands this and provided you with two posts as examples, where he says that the RESULT must be taken into account when calculating the rating.

1st tournament
Player A - 3 days - 1st place
Player B - 4 days - 2nd place

2nd tournament
Player A - 10 days - 1st place
Player C - 100 days - 2nd place

Player B simply didn't participate in the 2nd tournament.
As a result, according to the system that takes the result into account, Player A will have a higher rating in the final rating, Player B will have a lower rating, and Player C will have a lower rating than Player B. And if only the PLACE is taken into account, Player B and Player C will have the same rating - is this correct in your opinion? Many have already realized that this is not correct, so Kamikadze came up with another system and now calculates according to it, although I don't know if it has been officially adopted.

What ears :) In my system, the leader is limited to a bar of 100 points. That is, you always get 100 for first place and the leader's jersey in this tournament. All other participants receive a rating relative to the leader's result according to the main criterion of completion (for example, this is speed - days). The leader will not receive an undeservedly higher rating - only if there are NO worthy players in the tournament (this is practically impossible - someone will definitely be there). And in the system that takes into account the place occupied, a very common situation in practice occurs. Players of the middle level do not participate, and the lower level receives a high rating for their TERRIBLE results in days, BUT for good places (due to the absence of some of the players). This is what I call pulling the rating by the ears :) And that's exactly what you're suggesting to do :) Because look at my example. In no way does the second place correspond to a result of 100 days, especially considering that it is 10 times behind the leader in days. This is just funny :) And if you think a little, you will understand everything yourself.

Just take and count how many times each player took what place in the tournament. This is essentially what your proposal boils down to :) As Nikolai Ozerov would say, we don't need such hockey :)

Whether it suits or not is a purely personal opinion. I wouldn't even consider the "location" factor. But many insist on taking "location" into account as well. I agreed that the "location" factor shouldn't overwhelm the "result" factor.

It seems to work that way :) The formula is suitable for calculating both and the third. The leader's result is taken as the basis – that's the main thing, and it doesn't matter what he collected, days, skeletons, or experience :)