Preliminary clan ranking on HW. — Rating systems.

Kyle

19 years ago

Auto-translated

Ladies and Gentlemen,
Based on the formulas discussed here, a preliminary ranking of Clans has been calculated. The following were taken into account:
for Heroes1 - Offline tournament "Evil Lands"
for Heroes2 - SECOND OFFLINE TOURNAMENT FOR HEROES 2 "CLARIFICATION!!!" and NEW YEAR'S OFFLINE TOURNAMENT FOR HEROES 2 «ON THE BRINK»!!!
for Heroes3 - FIRST ANNUAL BEST OF THE BEST TOURNAMENT!
for Heroes4 - "First blood" offline Heroes 4
for Heroes5 - First Offline Tournament for Heroes 5!

The tournament rating was calculated according to the system of the respected IronAxe.
Take a look, comment, if you find any errors, please write; I will be very glad to hear your opinion. WinRar archive 3.40

------------------------------------------
I have made corrections based on the survey. That is, there are three pages in the file with the titles: "Total #1", "Total #2", and "Total #3". for each option.
For option #1, taking into account the wishes of the esteemed Adelaide.
For option #2, the ranking was counted as 3 participants; if there were not enough participants, then the ranking was not counted.
Evaluate!

Sincerely.

Attachments 1

1 file attached • Total size: 10.7 KB

zip

klans

10.7 KB 252

IronAxe

19 years ago

Auto-translated

Kyle;64557

The tournament rating was calculated using the system of the respected IronAxe.
Take a look, comment, and if you find any errors, please let me know; I will be very glad to hear your opinion. Archive WinRar 3.40
Sincerely.

My opinion is positive. You can now see that, using this system, which doesn't take into account the number of players or their placement, but only the result of the leader and the player in relation to him, you calculated ratings for clans across DIFFERENT versions of the game, and you selected only the players you needed from the tables. Taking into account the placement would only distort the results, as the tournament compositions and game versions are different. By the way

Аделаида

19 years ago

Auto-translated

I have some comments regarding the calculation of the average rating in the "EPU" situation. Firstly, it was not taken into account that some players were not registered in a clan (or even on the portal) at the time of the active tournament used for the rating calculation, yet this player's rating in said tournament was counted (i.e., 0), which lowered the clan's rating coefficient.... well, that's fine, it happened to many clans....
Secondly!!!!! one of our clanmates joined us very recently (this is Snaiper) and hasn't had time to participate in a single tournament on this resource, but in the table he is listed as a declared player for the version, meaning his rating in the table is "0", which also unjustifiably lowers the clan's rating.
I hope my comments will be taken into account when creating the final inter-clan rating.
Respectfully....

Kyle

19 years ago

Auto-translated

to Adelaide: You are absolutely right. When I started calculating the rating, I realized that I couldn't avoid similar questions because this problem exists. Therefore, in order to correctly adjust the rating, I need to receive a list of clan members from you for each tournament. Then I will be able to correctly adjust the rating.

Added 3 minutes later

IronAxe;64582

...
By the way, you can add up each player's results horizontally :) This way, you can determine the best all-around fighter in the clan, and also at the inter-clan level :)

When I was creating the table, I also thought about this, and I will

Kyle

19 years ago

Auto-translated

to Adelaide: How is it going? Will I receive a list from you? Or should we leave it as is?

Best regards.

Аделаида

19 years ago

Auto-translated

I promise to send the list by February 5th! Will that deadline work? I just can't do it any sooner; I'm short on time, and I still need to gather the information.

CrasH

19 years ago

Auto-translated

Hmm, hmm, hmm... something doesn't seem quite right... What's the purpose of this list? :confused: And why assign a 0 to a player who is listed in the clan's representative thread but hasn't actually played? And what if they are listed for all versions? Should we assign five zeros and lower the clan's rating in all versions? This will lead to everyone only having one, their favorite version, listed, or just the message "haven't decided yet!" Or, if I play a version that I'm not listed for, will my rating not be taken into account? I perfectly understand that you are trying to combat "ghost players" with this method! I support the fight, but not with this approach... Think about it... The rating should reflect the clan's strength, right? And strength is determined by the players who actually play, not just those who are listed! So, I'm only listed for version 3, and so far, I've only played it... But, if I now add myself to the other versions in the representative thread, will the clan's rating suddenly drop? But will the clan's strength actually change? Shouldn't these two parameters correspond to each other? Conclusion: the clan's rating should be calculated based on actual gameplay! And a zero should be assigned only if a player has registered directly for the tournament but hasn't submitted the final save file. IMO, this would be fairer.

In reply to CrasH

IronAxe

19 years ago

Auto-translated

CrasH;65396

Conclusion: the clan rating should be calculated based on actual gameplay! And a zero should be assigned in cases where a player applied directly to a tournament but did not submit the final save file... IMHO, this would be fairer...

I would suggest calculating the clan rating based on the existing ratings of its members (those with 0 are not included), without limiting the number of players used in the calculation. And then finding the average among them. That is, if a clan has ratings calculated for 4 members, and the rest have zeros, then we divide the sum by 4. If another clan has 10, then their total sum is divided by 10.

Kyle

19 years ago

Auto-translated

to Crash: I understand where you're going with this, and I will try to explain my vision to you.
For example, you claim that: "strength is determined by the players who actually play, not those who are simply listed!" - but I disagree with you, as I believe that a clan's strength is determined by all representatives of that clan. Since a clan is one team, everyone's result should be taken into account. Based on this statement, I hope it is clear why I gave a zero to those who were listed as participants but did not participate in the tournament.

In total, we have three different approaches for calculating the clan rating:
1) Take into account the result (or absence) of the tournament for all clan members who are listed in this version of Heroes. (proposed by Kyle and AlexeyPank)
2) Take into account the tournament results of a specific number of participants. (proposed by AmberSoler)
3) Take into account the tournament results of everyone who played in the tournament. i.e., do not give a zero for missing it (proposed by Crash)

I think we should put up a vote (in this thread) on these three points.

In reply to Kyle

IronAxe

19 years ago

Auto-translated

Kyle;65484

In total, we have three different approaches for calculating the clan rating:
1) Take into account the result (or absence) of the tournament for all clan members who are registered in this version of the game. (suggested by Kyle and AlexeyPank)

Who knows what anyone will declare. I joined a clan and said that I play any version of the game, but immediately said that I don't have enough time for everything – I will watch and choose which tournament to play. That is, if I only manage to play in one tournament, will I bring ZERO points to the clan for all the others? Answer me – am I understanding this point correctly? If so, I will not vote for the first point. I think it's clear why.

Kyle;65484

2) Take into account the results of tournaments for a certain number of participants. (suggested by AmberSoler)

This point is more correct, but it seems a bit unfair if there are more participants from the clan in the tournament than the specified number. We need to count all participants in the tournament, but for each clan, find the arithmetic mean. It would be nice to give a bonus for participation here :) Because the more participants from the clan, the greater the chance that they will get fewer points than if one strong player from the clan performs.

Kyle;65484

3) Take into account the results of tournaments for all who played in the tournament, i.e., do not give zero for missing it (suggested by Crash)

Of course, don't give zero :) Will clan players play and try their best, and then also pay for someone who could not play in this tournament?

In reply to IronAxe

AmberSoler

19 years ago

Auto-translated

Option #1.

IronAxe;65498

Who knows what anyone might say. I joined a Clan and said that I play any version of Heroes, but immediately added that I won't have time for everything – I'll be watching and choosing which Tournament to play. So, if I only have time to play in one Tournament, will I bring zero points to the Clan in all the others? Please answer – have I understood this point correctly? If so, I won't vote for the first option. I think the reason is clear.

By the way, I agree.

Option #2.

IronAxe;65498

...but it somehow doesn't seem right if there are more participants from the Clan in the Tournament than the specified number of participants to be counted...

Why "doesn't seem right"? Count the top three, for example. If only two are playing, then both count, but only two, not three – this is already a problem for the team, i.e., a "penalty" for the absence of the third, but there is an incentive to invite strong players to the Clan. If more than three are playing, then there is a choice – this is a bonus for the team for the activity of its participants.

IronAxe;65498

All participants in the Tournament should be counted, but for each Clan, the arithmetic mean should be found.

This creates the conditions for kicking weak players out of the Clans. Or refusing to accept them because of their low skill level. Or, as an option, a preliminary check of the participants' save files for results in order not to put weak players forward and spoil the Clan's rating. But this will not allow weak players to gain their own rating. There is a contradiction. And if someone thinks that this is not important, then I assure you – it is not important until the rating means something to you... :) As soon as there is an interest in the rating, thoughts about its optimization immediately arise... Only not all optimization methods are good for all members of the Clan. :)

Option #3.

IronAxe;65498

Of course, don't include them :) Will the Clan players play and start, and then also pay for someone who couldn't play in this Tournament?

And again, I agree! :)

In reply to AmberSoler

IronAxe

19 years ago

Auto-translated

AmberSoler;65502

This creates the conditions for weak players to be kicked out of clans. Or for them to be denied entry due to their low skill level.

I agree. However, the clan leader will look very bad in this situation. :)

AmberSoler;65502

Option #2.
Why is this "not good"? For example, count the top three. If only two players participated, then both are counted, but only two, not three – this is already a problem for the team, i.e., a "penalty" for the absence of a third player, but there is an incentive to invite strong players to the clan. If more than three players participated, then there is a choice – this is a bonus for the team for the activity of its members.

I thought about it and decided to vote for your proposal, but I don't even know whether to explain why or not. Moreover, the count should be based on ONE best player from the clan. Why? Otherwise, completely different conditions may arise. And if we assume that the clan leader may go so far as to exile weak players or even deny them entry into the clan, why couldn't such a leader organize what I hint at below?

I'll explain the situation without suspecting anyone of anything yet. Let's say clan X is participating and knows that the count is based on three players. But they only have two. What can the clan leader and players do in this situation if they know that this is their clan's problem and it needs to be solved somehow? The answer is not simple. Especially if the leader is willing to sacrifice all principles for the sake of the clan's victory, up to and including exiling the weak, etc.
If we don't have a third player, we'll invent one. :)
By solving this problem, we can tackle another one: why does one of my clan's fighters perform excellently, the second well, and the third also well, like the second? The solution to the first problem was precisely in this. And couldn't we organize the performance of all three fighters excellently? :)
And actually, nothing prevents this.

AmberSoler, now let's see, if the leader will be kicking out and not accepting weak players, then all this will happen in front of the forum users. But the schemes I described to boost the result for three players will be STRICTLY kept secret. If they do it cleverly, no one will notice. I don't know what you will answer, but I am considering the worst-case scenarios, you already know this, but what can be done? In general, the clan rating is a difficult issue in light of these considerations. In any case, it is necessary to anticipate such situations in order to try to prevent them as much as possible.

In short, I voted for the AmberSoler system. However, I would also put a checkmark next to the third point, but I'm not allowed to. AlexPank and Kyle, if you start giving zeros to those who missed, people will start copying the final saves of others and asking them to simply send them. Moreover, they will do it in such a way that it will be difficult to figure out, they can even manipulate the timestamps by +/- 1-2-3 days, and you won't be able to prove anything! And thereby they will completely ruin my calculation of the Current Rating! And yours too!

AmberSoler

19 years ago

Auto-translated

I won't delve into a debate on this topic because I believe people are inherently decent. By definition. Although I acknowledge the possibility of various cunning maneuvers, I can confidently say that for a long time, all regular players on the portal (and especially members of Clans) will be easily identifiable. And any sudden, significant performance by a player who previously didn't show much ambition can always be verified... To get lost among competitors, there need to be a lot of them... No one will ever play the game on two fronts solely for the sake of the Clan's rating. A strong player will primarily dedicate their efforts to their personal rating, and they'll be more likely to complete an extra run on the map for their own benefit rather than for some abstract "nothingness." It's very difficult to complete two strong, independent runs. If one run is based on another, with the end of the game being replayed, it can be detected in two minutes, and you don't even need to open the save files for that. :) How? I think you know perfectly well. And if not, that's even better... And if the Clan leader provides advice and guidance to all other members without touching their save files, I personally don't see a big problem with that. You can't really help a weak player, and a strong player doesn't need advice. However, if players exchange tactics and tricks and apply them in practice, that's only worthy of respect. That's what Clans are for. And the portal benefits from it. That's what I think...

In reply to AmberSoler

IronAxe

19 years ago

Auto-translated

AmberSoler;65512

I won't delve into a debate on this topic because I believe people are inherently decent.

I understand. But your own assumptions were about the reasons for expelling weak players or refusing to accept new members into the clan, which, you must agree, is not entirely decent on the part of the leader. And in the discussion about the rating and AlexPanka's coefficient, it was also suggested that a player might intentionally lower their rating in order to get a bigger bonus later.

AmberSoler;65512

By definition. Although I admit the possibility of all sorts of cunning moves, I can assure you that for a long time, all the regular players on the portal (and especially the members of clans) will be in the spotlight. And a sudden, strong performance by any player who has not previously shown much ambition can always be checked... To get lost among the competitors, there must be a very large number of them...

I'll have to use myself as an example, but in a negative light. I've played in several tournaments here and managed to get into the top 10. I can be used in any of the scenarios except for the second one, and the result won't raise suspicions. But, ultimately, even a bad result instead of a missing third player is better than nothing. What can you do? If a quorum cannot be reached, that is a problem for the clan.

AmberSoler;65512

No one will ever play on two fronts for the sake of the clan's rating. A strong player will devote all their efforts primarily to their personal rating, and they will rather do an extra run on the map for themselves than for an abstract "nothing"... It is very difficult to make two strong, independent runs. And if one run is based on another, with the end of the game being replayed, it can be detected in two minutes. How? I think you know very well.

It depends on how developed the sense of collectivism is and who stands up for what, some for their personal rating, some for the clan. I'm talking about three strong runs as an extreme case. Well, let's hope for honesty again.
Can anyone tell me - is there a replay function in all versions of Heroes, or not? I don't remember - I rarely use it. I know there is one in the fourth game.

Joke: Vasil Ivanovich arrives in London in a tuxedo and with a suitcase full of money. Petka is very surprised - where did you get it? I played cards with gentlemen in London, Petka. And so what? Well, one of them says - sirs - I have a winning hand. And I say - let me see! Everyone was indignant - what are you doing! We are all gentlemen here - how can you? Petka - and so what? Well, from that moment on, my luck started to change :)
I'll say this for myself - I'm not Vasil, but I don't want to be in the place of the gentlemen either.

In reply to IronAxe

AmberSoler

19 years ago

Auto-translated

IronAxe;65516

I understand. But your own assumptions were about the reasons for expelling weak players or refusing to allow a player to join the Clan by the Clan leader, which, you must agree, is not entirely fair on the part of the leader.

Don't confuse unfairness and self-interest. Here, everything will be open and unbiased. "Here is a save file, play through the map, and if you don't complete it in so many days, we won't accept you into the Clan" - where is the unfairness here? It's just that a weak player is not profitable. That's all. And this does not contribute to the gameplay, because I am in favor of having as many strong players as possible. And a weak player, if he wants, will always become strong, especially if he is not discouraged...

IronAxe;65516

And in the discussion about the rating and the coefficient, AlexPanka also suggested that a player might deliberately lower their rating in order to get a bigger bonus later.

And this is a trick based on an ill-conceived rating calculation. And if someone can use the mathematicians' miscalculations to their advantage, then they can only be applauded. Because everything here is done within the rules. If there are legal loopholes in the law, then using them cannot be called unfair. It can be called a trick and cunning - yes. :) But there is no unfairness here...

But forging a save file is a lie - that is, unfairness.

IronAxe;65516