HeroesWorld
  • Log in
  • Create account
  • Mode
    Interface scale

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

Rating systems.
5.0 (1 rating)
Open
Which rating system should I choose? (see the text for details)
33 voters
Single choice
Sign in to participate in this poll.
Poll results will be visible after you vote or when the poll closes.
In reply to GriB_o.k.
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#287
Auto-translated
tokha;63220
If I didn't need to, I wouldn't express my opinion.

I thought that here everyone's opinions are heard, and not just told to them...
I'm not going to dispute your right to express your position. But it sounds like this: "Why are you doing this? Do it differently." I explained to you that it is impossible to do it differently, based on your suggestion, because it doesn't make sense. You're back to square one - "my head hurts." Question: what should I have answered? Should I say that we will immediately forget about all our ideas and redo everything according to your suggestion?

Understand that the discussion of the calculation method is open only because it may be of interest to any participant in the tournaments. And all the comments are accepted and discussed. Also, adjustments are made taking into account these wishes - you can browse the topic and see for yourself. However, if you are against our calculation method in principle, then your comment goes beyond the discussion of the issue, and all that remains is for us to abandon our idea, right?

If I offended you with my remark, please forgive me. I certainly did not intend to do so. But I can't say anything else about the essence of the discussed issue.

Added 9 minutes later
Kyle;63226
No, you didn't understand me... I'm not talking about the annual rating. Once a year, you can reset the current rating of a player if they have not participated in tournaments, but if they have participated, then there is no reset. This is needed to remove "inactive" players. The General rating is never reset!
In January, a player played one tournament and won, and leads the list until December. The rest compete for second place all year. Is that right? At the end of the year, he is removed from the rating. He comes in January and takes first place again. Should I continue, or is the idea clear now? :) Aging will gradually push the inactive player down (halving the winner's rating after 6 tournaments), allowing the others to continue competing for first place.
Kyle;63226
This issue, of course, is debatable... I, for example, suggest once a year, because if a player hasn't shown up for a year, it means he's not interested in Heroes and this portal...
One player played in January, the second in February, the third in August. They didn't appear again. Which of them will you leave in December, or will you remove them all? After all, each of them was in a tournament only once a year. It seems that they can all be removed. And another one participated for the first time only in December - also one tournament a year. So, we should remove him too... Or not? Then, please add additional selection criteria. I'm just curious - will you be able to draw up a complete scheme or not...
Do you think Iron and I haven't discussed these issues? :)
Kyle;63226

...it is simply impossible to achieve a single universal formula (IMHO). Therefore, we need to look for a compromise. And this compromise is the introduction of an additional rule.
I agree, there is no optimal formula. Suggest rules that will take into account the interests of all tournament participants equally. That would be very good...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
User Avatar
Offline GriB_o.k.
Registered:
Posts: 106
#288
Auto-translated
I expected you to read my post and provide some specific feedback on why this scheme doesn't work, as I don't see the point of your system.

But since you're not planning to change anything, I'll drop the subject and remain silent.
Get What You Want
0
⚖️
0
In reply to GriB_o.k.
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#289
Auto-translated
tokha;63239
I expected you to read my post and write something specific about why this scheme doesn't work, because I don't see the point in your system.
Specifically:

1. All our ideas were based on not pulling numbers out of thin air...

2. The number of points the winner receives doesn't matter. We have 100, you have 50. This is not essential. But the number of points the player who came in second receives is important. In our calculation, everything depends on the player's result compared to the winner's result. In your case, it's simply 40 points. Justify it.

3. That's why I said that your option is unacceptable. Why? See point 1. How much more specific can I be? :)

Added 7 minutes later
tokha;63239
but since you're not going to change anything, the question is closed, I'll keep quiet.
Don't get upset. :) We are happy with all comments and listen to them very carefully. Just don't ask us to change the foundation. Unfortunately, this is impossible...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
User Avatar
Offline GriB_o.k.
Registered:
Posts: 106
#290
Auto-translated
Yes, the numbers aren't the main thing. I was proposing a system.

As for the numbers... the numbers should be chosen so that if a player takes, say, first and tenth place in two Tournaments, they will ultimately rank higher than a player who took fifth place both times. For example:
1st place - 75 points
2nd place - 55 points
3rd place - 40 points
4th place - 30 points
5th place - 23 points
6th place - 17 points
7th place - 12 points
8th place - 8 points
9th place - 5 points
10th place - 3 points

But the main thing is that the points do not depend on either the number of players or the level of the tournament. And those who didn't make it into the top ten receive no points.
Get What You Want
0
⚖️
0
In reply to AmberSoler
User Avatar
Offline Kyle
Registered:
Posts: 2,153
#291
Auto-translated
AmberSoler;63233
I agree, there is no optimal formula. Suggest rules that will take into account the interests of all participants in the Tournament equally. That would be very good...
We can come up with rules, but first we need to determine what we want from the ranking? i.e., initial conditions are needed, otherwise any system can be broken by changing the initial conditions.
Aging will gradually move an inactive player down (halving the winner's rating after 6 Tournaments), allowing others to continue competing for first place.
i.e., it implies some kind of competition?
Personally, I thought that the rating simply has an informative character:smile32:, like, a player played so many Tournaments, scored so many points, and on average he plays at such a level... In your version:
In January, a player played one Tournament, won - and leads the list until December.
it will be clear that the player played only one Tournament! and that's it!
But, of course, if there is a competition for the maximum rating, then yes, the system does not work...
Therefore, what do we want from the rating? competition? or a reflection of the real strength of the players?

Added after 8 minutes
tokha;63246
yes, numbers are not the main thing. I suggested a system.

well, as for the numbers... the numbers should be taken in such a way that if a player takes, say, first and tenth places in two Tournaments, then he will eventually be higher than a player who took fifth place both times. for example:
1st place - 75 points
2nd place - 55 points
3rd place - 40 points
4th place - 30 points
5th place - 23 points
6th place - 17 points
7th place - 12 points
8th place - 8 points
9th place - 5 points
10th place - 3 points

but the main thing is that the points do not depend on either the number of players or the level of the Tournament. and those who did not make it into the top ten remain without points.
read this post. There is an example there... plug your numbers into that example and see what happens....
0
⚖️
0
User Avatar
Offline GriB_o.k.
Registered:
Posts: 106
#292
Auto-translated
I plugged in my own numbers, and it turns out there was no point in skipping the second tournament if they wanted to earn more points. This system is for the total rating across all versions of Heroes, which means a player should play all versions. If they can't play a particular version or aren't good at it, they'll get zero points for it. I can't play version 5, but if I take first place in all the other versions, I'll still be in first place overall, even though there's a player who took second place after me in versions 1-4 and won version 5. They'll still be in second place.
Get What You Want
0
⚖️
0
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#293
Auto-translated
To Kyle. Rating is not just about a player's strength. The question is: is strength a constant concept, or does it change from tournament to tournament? Most likely, it's a constant, established at a given point in time. Of course, a player's strength gradually increases – they gain experience. But this is, on average, not very fast. However, a rating can change from tournament to tournament, which means that, in addition to the player's strength, it also takes into account their tournament activity. And tournament statistics provide informative data. But the results can be displayed according to an alphabetical list of players. And being first in the list doesn't mean that this player is the strongest at the moment. This is where you can see how many times a particular player participated in tournaments during the season. But you shouldn't compare them to each other. The current rating has a different purpose – to compare players at a given point in time. This is the current rating, and this is what we are discussing. All your ideas are applicable to an annual rating. Only, instead of using the arithmetic mean, you need to consider the sum of all results, and it will depend on the number of games that will be taken into account. For example, the 10 best games of the season are included in the rating. Those who have played fewer than 10 games do not participate in the annual rating. Added 3 minutes ago
toha;63246
yes, the numbers aren't the main thing. I was suggesting a system.
That's it! These are the key words! You are suggesting a system. A different system. I'm not against it – feel free to suggest it. Let the tournament organizers consider it and decide whether it suits them or not. I won't argue with you, and especially not with them. We're not saying that one is better and the other is worse. They are just different, but we can't compare two completely different systems. And it won't be possible to combine them... My calculations don't have room for the coefficients you are suggesting... Because the rating in my system and in Iron's system is calculated each time from scratch, based on the results of the tournament. That was the idea. We are simply offering an alternative to the rating that VDV_forever proposed at one time. You are offering the same thing that he had. That is, there is nothing new in your system. What is the point of returning to something that has already been rejected once? And the rating is calculated separately for each version of the game. We are not combining the third and fifth versions yet... How Alex will calculate in the ВК tournament – ask him.
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline Kyle
Registered:
Posts: 2,153
#295
Auto-translated
AmberSoler;63259
To Kyle.
The current rating system has a different purpose - to compare players at a given point in time.
Here, if we go by that, then yes, as I already said, my system is not suitable... :smile34:
But in turn, I would like to point out that the system you and IronAxe are proposing is not viable, or rather, it is viable, but it will not accurately reflect the strength of the players, because you are trying to combine incompatible things... for example: a 100-meter dash and a 400-meter hurdles race! You can compare them if the runners run both races, but if not? Then how?
0
⚖️
0
In reply to Kyle
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#296
Auto-translated
Kyle;63264
You are trying to combine the incompatible... for example: running 100 meters and 400 meters with hurdles! You can compare if the runners run both races, but if not? How?
No system will be able to compare the results and strength of these runners. Neither yours, nor ours, nor anyone else's... So, try to provide a more accurate argument...

Added 4 minutes later
Kyle;63264
there will be no correct reflection of the players' strength
Explain what you mean by "correct reflection of players' strength"... Provide a specific and clear example so as not to waste our time with empty talk.

That is, the strongest player is:

- the one who has played the most Tournaments
- the one who has the most wins
- the one who has beaten everyone (there are no such players) :)
- the one who has never lost (give a definition of the term "loss")
- the one who is the strongest :D, etc.

You understand that the concept of "strength" is very conditional. Our rating is just one of many ways to reflect it... There are ratings that reflect it roughly, there are those that reflect it very roughly, and there are those that reflect it more or less realistically... But there is not a single one that reflects it accurately... That's why this topic exists. Otherwise, everyone would have done it long ago and forgotten about it. Our goal is to find a more or less close option, and so far all the proposals boil down to taking a few steps back... Why? It seems to me that it is better to move forward...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline Kyle
Registered:
Posts: 2,153
#297
Auto-translated
AmberSoler;63270
No system will be able to compare the results and strength of these runners. Not yours, not ours, nor anyone else's... So try to provide a more correct argument...
This is exactly what I'm talking about!:smile12: They cannot be compared!
If a comparison is to be made, the following conditions must be met:
1) A constant number of participants. i.e., no one joins and no one leaves... or if someone joins halfway through, count losses for the Tournaments that have already taken place.. and if they leave, zero for those yet to come.
2) A time limit. i.e., from 01.01.2001 to 03.03.2003.
3) Equal Competitive conditions. i.e., the use of a single save, a single game version (The Shadow of Death and TE are different versions), etc.
and maybe something else... that I've missed right now...

Everything else, sorry, is NONSENSE!

An information rating is another matter... but that's a completely different story...

Respectfully.
0
⚖️
0
In reply to Kyle
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#298
Auto-translated
Kyle;63275
If we are to compare, the following conditions must be met:
1) a constant number of participants, i.e., no one joins or leaves... or if someone joins in the middle, their losses in the Tournaments that have already taken place should be counted... and if they leave, they get zero points for the ones that will take place.
That's right! :) If someone joins in the middle, all the Tournaments that have already taken place are worth zero for them, which means their rating is zero. This is self-explanatory. If they leave, we add zero to their rating for each Tournament, and then we update the rating. And so on - until it is completely reset... You probably haven't fully understood the calculation method, which is why this confuses you so much... Please, take a closer look at the issue...
Kyle;63275

2) a time limit, i.e., from 01.01.2001 to 03.03.2003.
This is not important at all. If you manage to play the required number of Tournaments during this period, then it is included in the cyclic rating (annual, quarterly, etc.). If you don't manage, then it is not included. It's very simple. And for the current rating, the period is not important at all, because the aging procedure reduces the importance of past Tournaments for all players... And there comes a moment when a Tournament that you played six months ago no longer gives you any advantage over someone who didn't play it... Everything is well thought out! :)
Kyle;63275

3) equal competition conditions, i.e., one save, one version of the game (The Shadow of Death and TE are different versions), etc.
and maybe something else... that I'm missing right now...
This is a fundamental condition for rated Tournaments... BB is a test, and the "normal" save will not be included in the rating. We only play on SoD. This is self-explanatory... And it is simply necessary to adjust the starting conditions of the Tournaments for this.
Kyle;63275

Everything else, I'm sorry, is NONSENSE!
There is no "everything else"!!! :smile13:
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
User Avatar
Offline GriB_o.k.
Registered:
Posts: 106
#299
Auto-translated
VDV_forever suggested it some time ago

so, I guess I missed something. Never mind.

Is your system ready yet?
Get What You Want
0
⚖️
0
In reply to GriB_o.k.
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#300
Auto-translated
toha;63282
Well, it seems I missed something. Never mind.
Yes... and was it worth arguing about? :D It all started with the VDV rating... He was the one who suggested developing an alternative version a long time ago...
toha;63282
Is your system ready?
In its main aspects - yes. Details still need to be clarified, but we can only do that in practice. Without tournaments, there is no material for analysis... We are waiting for the statistics...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
User Avatar
Offline GriB_o.k.
Registered:
Posts: 106
#301
Auto-translated
Okay, let's move on. How are the preparations for the first Heroes of Might and Magic tournament going? Is it worth revisiting that game?
Get What You Want
0
⚖️
0
User Avatar
Offline AlexeyPank
Registered:
Posts: 2,396
#302
Auto-translated
It's in the works, it's in the works. I suggest we return to a specific discussion about the ranking after the first round of the King's Bounty tournament is over. Once we have data from five stages, each using a different version of Heroes, we can calculate everything and determine the pros and cons of all the ranking versions. For now, I'll calculate based on both, and we'll see what happens.
Мы спасемся? Я не знаю. Верьте...
Геройский Интернационал
Клан Homo homini - основатель клана
Член РКСМ с 07.07.07.
ФЛУДОВСКИЕ ССЫЛКИ!- сообщить о бесполезной теме/сообщении.
0
⚖️
0