Connection ErrorBad RequestThis page went staleSession ExpiredSign in requiredForbiddenNot FoundToo Many RequestsServer ErrorBad GatewayService UnavailableGateway TimeoutHTTP Version Not Supported
Session Expired
Your session has expired. Please log in to continue.
Sign in required
You need an account to do that. Sign in or create one — it only takes a minute.
This page went stale
Your security token was rejected, usually because the page sat open too long or you signed in elsewhere. Nothing was saved. Reload the page and try again.
Request Failed
We encountered an issue while processing your request.
Please check your internet connection and try again.
Error Details:
Share Link
Posts from Player rating for tournaments across all installments of Heroes.
Iron, you understand that the skill level of HP players is no lower than ours. I would say it differently... You asked if there are any examples? Yes, there are. And is that a reason to give everyone a zero rating? Team Battles are not a tournament for the weak; it's a tournament for beginners... It's not the same thing...
The rating should behave correctly in all calculation cases, regardless of the circumstances.
What if it's just a very difficult map? Will we immediately assume that everyone is a noob? You have to agree that you should probably take those words back... You didn't think it through...
And what if it's not difficult? Who and how will determine that? A map difficulty coefficient is needed :) Otherwise, you'll end up biting your own tail and nothing will come of it, and you know that yourself, when you wrote that we shouldn't make things more complicated :) But you went ahead and did it. Will you soon start introducing random numbers?
Added 2 minutes later AmberSoler;103925
Iron, you understand that the skill level of players on HP is no lower than ours. I would put it differently... You asked if there are examples? Yes, there are. And should we give everyone a zero rating just because of that? TB is not a tournament for weak players, it's a tournament for beginners... It's not the same thing... The rating should behave correctly in all calculation cases, regardless of the circumstances.
I'm not arguing that there are strong players on HP, but it seems like there weren't any in this tournament :) What? Let's check this map then? To make sure that it can be improved? Surely Dmitry and Radik can't either. It's clear with me :) The point is that the rating should behave correctly and not give beginners a free rating just because there were no strong players in the tournament
But what if it's not complex? Who and how will determine that? A difficulty coefficient for the map is needed :) Otherwise, you'll end up biting your own tail and nothing will come of it, and you yourself understand that when you wrote that let's not make it more complicated :) And then you took it upon yourself. Soon you'll start introducing numbers pulled out of thin air, won't you?
Starting from the sixth stage, the bonus will be calculated. I will try to calculate it according to two options - the old one and the new one (where the strength of a beginner is equal to 1/3 and there is an increasing coefficient of 1.4). After that, the differences in all three systems will be more significant.
Why wasn't the bonus calculated before? Because comparing the strength of players in versions 3 and 5 is meaningless. Now, the comparison will be based on versions.
Until the esteemed Fireball offers his system, I will calculate the VK ratings based on three systems. Later, if necessary, I will add a fourth. I calculate the overall rating ONLY according to my system, so please excuse me - I'm not made of steel.
Starting from the sixth stage, the bonus will be taken into account. I will try to calculate it using two options – the old one and the new one (where the novice's strength is 1/3 and there is an increasing coefficient of 1.4). After that, the differences in all three systems will be more significant. Why wasn't the bonus calculated before? Because comparing the strength of players in stages 3 and 5 is meaningless. Now, the comparison will be based on versions.
Of course, this is your system, and you can decide, but my advice is not to introduce a bonus, as all stages should be EQUIVALENT, at least for the War of the Kingdoms ranking. A player should not suffer because they performed normally in the Sword stage, but no bonus was awarded there. And it was right not to award it. I have criticized the existing systems a lot, but for War of the Kingdoms, an oscillator without a bonus is a much more acceptable system than with it.
2AmberSoler You have already proposed a clearer and more practical version of the formula I provided. Do you have any other comments or suggestions? I would prefer to continue our discussion; perhaps it will lead to further improvements to the system. And when everything is ready, it will be easier to convey the result to the public than to burden them with reasoning. For now, the written formula should be considered only as a starting point.
The winner of a more difficult tournament should not receive fewer points than the winner of an easier one. Especially if their only fault was getting a slower map in their tournament. You want to make it complete randomness. Effectively, whoever gets the fastest map will win the entire WC. But the stages should be equal in value.
What does the WC have to do with this? The rating is calculated based on other Tournaments as well. Let's put it this way: someone who intended to win the WC in the Universal category (across all hero versions combined) understands that they cannot skip a single version of a single stage. Agreed. Someone who wants to win in a nomination for a SPECIFIC version understands that they only need to play great in both stages for that specific version. And wouldn't they not care who grabs the universal win if they win, for example, in their version, say Heroes 3 ??????? Do you not understand this? Fireball;103922
For example, let's say a certain player showed results of 12,7,7,6,9 days respectively across the WC stages, winning 4 out of five stages. Even with a 1-day virtual bonus, he would still lose to you in the rating. If you consider this normal, there is no need to continue the conversation.
I don't understand anything :) What did this player even want? To win the rating for the universal category? From your example, it turns out that both he and I played all stages for all versions. He won everywhere except for one version. It turns out he failed the fifth one so badly that he blew the universal win. Is that possible? Punk calculates the WC using three systems, and what if everyone gives such a result? If so, who is to blame — me? You can't really blame the person who picked the maps either. Comparing ratings obtained from different versions is useless, and that has already been mentioned. Everyone knows they will be summed up, and that's all. Beyond that, everyone chooses their own strategy, or rather, already chose one and played based on it for half the tournament. We'll see what happens. But in my opinion, it's already clear that there will be nominations for each version and one universal one. I can't add anything more to this, only one thing: it seems to me that your example is unrealistic even under the existing systems. And now that half the tournament is over, people have suddenly woken up to develop a new system (where were you before — supposedly here — why were you silent?) — now everyone who feels like it is rushing to invent their own systems, adjusting coefficients to fit their already existing results in the tournament in the hope that the final will be determined by a newly invented system.Fireball;103922
For the WC, Alexey has a bonus of 0 for everyone
Then how the heck does his system differ from Ambera's system — him, not me :) Fireball;103922
Not so. The result is not just days, but also place.
I've already proven a hundred times that in Heroes, place is nothing. Result is EVERYTHING. Because place is so relative in combination with the game version, map difficulty, the player pool and the strength of a given party, whether players of even the same level have free time for this tournament, attendance or absence for valid or invalid reasons, and other factors that you'll never derive a formula that actually calculates the rating under a single denominator across different versions of the Heroes game, unless you bring your entire Research Institute to the task :) Why complicate everything when it's already perfectly clear. Six nominations, one of which is universal, which does NOT determine that the player is the best across all versions, because there will be a separate nomination for each version of the game. What's wrong with that? Fireball;103922
If you don't like it, don't read it. If a system had been chosen, I wouldn't have said a word (at least until the end of the tournament), regardless of whose it was. But as it stands, I prefer to discuss it, listen to criticism, and propose a ready-made option. As for how the choice will be made — that's a separate matter.
Since I'm reading it, it means I like it :) hee-hee :) look at you guys doing math with each other :) You wrote yourself that you'd prefer to listen to criticism — so listen. PS. If Punk is not using his bonuses for the WC calculation right now, then he shouldn't start until the end of the tournament. But then what is he calculating there, I don't understand :) according to his own system. Is the difference from Ambera only in the lower limit or what? Fireball;103922
As for how the choice will be made — that's a separate matter.
:) Well yes — a separate matter. And it should be done after the end of the Kanigem tournament. Because introducing a new rating calculation system after half the tournament has passed is even worse than changing the rules during a tournament on a given map, imho.
I'll explain why there wasn't a bonus before: the players' strength wasn't known.
Why is it needed now? To determine the strength of the tournament and the growth of players' skills. Since I hope the King's Bounty tournament will be an annual event, the bonus will be calculated from the very beginning in the following stages. This King's Bounty tournament is largely a trial run. Based on its results, we can develop new unified rules, a unified rating, and so on. But at the same time, I can also calculate without the bonus.
2AlexeyPank Of course, this is your system, and it's up to you to decide, but my advice is not to introduce a bonus, because all stages should be EQUIVALENT, at least for the War of Kings ranking.
I agree. Fireball;103935
A player should not suffer because they performed normally in the Sword Round, but no bonus was awarded there. And it was right not to award it. I have criticized the existing systems a lot, but for War of Kings, an oscillator without a bonus is a much more acceptable system than with one.
And without an oscillator, it's even better :) Because the ranking doesn't depend on such a relative parameter as the place, but only takes into account the days, that is, the main task on the map - to win in terms of speed. It seems that they have already dealt with bonuses in War of Kings. Thank goodness. The oscillator is still here :) PS. Actually, I wouldn't mind if, for the War of Kings tournament, to determine the winners, they used a PURE calculation of points according to the VDV_forever system. But who would prevent them from calculating each player's game rating both separately and universally. Why did they decide to determine the winners in War of Kings based on the game rating? How is this related? The rating is calculated on the forum and not only for War of Kings games, but also for blitzes, and for skeletons, etc. If Pank started using a bonus, would determining the winner in War of Kings then depend even more on random games or something? They've mixed everything up. I suggested simply calculating game ratings for players on the forum, but not taking them into account when determining the winner of any of the tournaments. Because that's how it happens in separate standard tournaments. First, the winner is determined - places are assigned, and then the rating is calculated. But what happens in War of Kings? For each of the stages, it seems to be the same. And then they decide, based on GAME RATINGS, to determine the winners of the entire tournament and separately for each version :) In general, I don't really care how they determine the winners in War of Kings. The main thing is that all players on the forum continue to have their game rating calculated AND NOT ONLY FOR PARTICIPATING IN WAR OF KINGS. I hope my position is clear?
:smile19: Deleted :smile19:
Fireball;103935
2IronAxe Did you read post 643?
Yes. And I even answered. I just had internet problems at work.
Added 5 minutes ago AlexeyPank;103942
I'll explain why there was no bonus before: the strength of the players was unknown. Why is it needed now? To determine the strength of the tournament, the growth of players' skills. Since I hope that the War of Kings tournament will be annual, then in the following stages the bonus will be counted from the very beginning. This War of Kings tournament is largely a test run. Based on its results, it will be possible to develop new unified rules, a unified rating, and so on. But at the same time, I can also calculate without a bonus.
Well, there was some kind of database for some versions of the games - for 1, 2, 3, 5, several tournaments were held before Kanigem :) You could take this into account. Or will you only take into account the bonus strength of players from the Kanigem tournament stages? And will you assign some additional rating to new players out of thin air? It may turn out to be higher than the rating of some active participants in the tournaments who were unlucky and have a rating from 0 to 10 points :)
I will explain how the bonus in VK will be calculated. Bonus=((M(r)-M(t))/n)*square root of K, where M(r) is the player's position in the tournament rating, M(t) is the place in the tournament, n is the number of players, and K is the average strength of the tournament (defined as the average value of the non-linear rating). Newcomers will be considered to have a position equal to 1/3 of the maximum. That is, if there are 10 players, newcomers will be considered seventh in terms of strength. Additionally, for the VK tournament, an increasing coefficient proposed by the esteemed VDV_forever, equal to 1.4, is being introduced. There will be no additional rating. The results of previous tournaments outside the scope of VK were not taken into account because the majority spoke out against giving anyone an advantage before the tournament. Now that the strength of players in the tournament is beginning to be determined, the bonus can be introduced.
Tell me, what exactly bothers you so much about us discussing a calculation system that isn't accessible to you?
I have some access to it. And some people don't understand it at all. Nothing bothers me – I already said that. I just don't understand why you're all making things so complicated and confusing people. Why are you doing this? AmberSoler;103928
Well, you don't understand it, but no one is blaming you for that.
It might just be the case. No one understands your system except for you, but you're the ones who will be doing the calculations. Then no one will be able to check what you've calculated with your logarithms, sines, and squares. For example, I suspect that there are many errors in Pank's calculations, taking into account the bonuses from previous games, but I'm not only unable to check them, but I don't even want to. How, for example, will they check if the rating for the Boots was calculated correctly? :) I'll have a good laugh when you start explaining to everyone here how to calculate all this using your formulas, so that ANYONE who wants to can be sure that their tournament rating was calculated correctly, and that no one overtook them due to some hidden error. AmberSoler;103928
Maybe you'll figure it out later, and if not, just calculate it your own way; no one is stopping you. There will be twenty different calculation methods on the
Of course, it will be better to calculate it your own way than to constantly recalculate your rating and the ratings of your opponents after each tournament using your formulas to check for errors, and with Pank, you'll also have to keep a database of PREVIOUS tournaments. AmberSoler;103928
But you're only calculating it for your own enjoyment. But you're going to push it as the official one. Then everyone will enjoy it, especially those who are interested in their rating.
Added after 16 minutes AlexeyPank;103946
I'm explaining how the bonus will be calculated in the Cup. Bonus=((M(r)-M(t))/n)*sqrt(K), where M(r) is the player's place in the tournament ranking, M(t) is the place in the tournament, n is the number of players, K is the average strength of the tournament (determined as the average value of the non-linear rating). Beginners will be considered to have a place equal to 1/3 of the maximum. That is, if there are 10 players, beginners will be considered to be in seventh place. In addition, a boosting coefficient proposed by the respected VDV_forever, equal to 1.4, will be introduced for the Cup tournament. There will be no additional rating. And the results of previous tournaments were not taken into account because most people were against giving anyone an advantage before the tournament. Now the strength of the players in the tournament begins to be determined, so a bonus can be introduced.
That's great. And how will all this be calculated outside of the Cup? Will the Cup tournaments be taken into account when calculating non-Cup tournaments, I mean, will the bonuses from it be included? P.S. I'll probably leave this topic anyway :) And only Fireball, Amber, and AlexPank will remain here with their nine-story formulas. And only occasionally someone will drop in here with a post like: what the heck is this – I don't understand anything – how many ratings do I have and how many, and according to what formulas can I check them? Seeing the formulas and calculation systems, they will quickly leave :) with cries of "forget it!" :) P.P.S. Okay, I'll also drop in from time to time, maybe I'll come across some funny thing
The bonus for VK tournaments is no longer considered in the calculations. The results are available, and you can easily check them. VK tournaments contribute to the overall rating. The difference is that in the overall rating, I am not yet using either the increasing coefficient or the novice strength of 1/3. If this bonus option (with coefficients and 1/3) suits the majority, I will recalculate everything based on it without any problems.
I don't understand anything. What did this player even want? To win the General ranking? From your example, it seems that both he and I played all stages in all versions. He won everywhere except for one version. So, did he mess up the fifth version so badly that he failed in the General category? Is that possible? Punk calculates the War of the Kingdoms ranking using three systems, and if everyone gets such a result, will it be counted? If so, who is to blame – me?
Such a player will win by a landslide in all systems, but he will lose in the virtual one, which proves the absurdity of the virtual idea.
And now, when half of the tournament has passed, they remembered to develop a new system (where were you before – you were here, right? – why were you silent?).
I thought that the VDV Forever system was chosen initially, but it turned out that half of the tournament had passed, and they still hadn't figured out how to determine the winner.
Now, because I'm reading this, it means I like it, hi-hi. Look at you, engaging in mathematical calculations with each other. You yourself wrote that you would prefer to listen to criticism – so listen.
You're not offering criticism, but demagoguery. In some ways, Prorock was definitely right.
Because introducing a new ranking system halfway through the tournament is even worse than changing the rules during the tournament on a given map, IMHO.
This is a gross violation, but it seems inevitable, because no system has been officially adopted. And it doesn't matter when the system was invented. The only way to avoid the violation is to choose the VDV Forever system, because its ranking was immediately recorded in the results table.
Actually, I wouldn't mind if the War of the Kingdoms tournament used PURELY the VDV_forever system to determine the winners.
I don't mind either. It would be fair and, let's say, conciliatory. Then, after the War of the Kingdoms, having other tournaments of the year as well, without personal grievances, we could start discussing the system for the next War of the Kingdoms. But such a choice must be made by the official organizers.
Such a player will win in all systems by a landslide, but will lose to a virtual player, which proves the absurdity of the virtual player idea.
Since you state this so confidently, it means you calculated it based on some example, even if it's made up: ----------------- Let's say some player showed the following results in the stages of the competition: 12, 7, 7, 6, 9 days, respectively, winning 4 out of 5 stages. Taking into account the virtual player's time of 1 day, he would still lose to you in the ranking. If you consider this normal, there's no point in continuing the conversation. ----------------- Here you listed the player's results in the order of completing the stages in days. And the second player is supposed to be you? And you need to use your own times for the calculations. Listen, where can I see how many days you took to complete each stage? Give me a link. I don't remember where to look, and I don't want to spend time searching for it right now. I want to see this phenomenon for myself. Fireball;103961
For now, you have the highest chance of winning in one system – your own.
I also have good chances in other systems. The main thing is not to miss a single stage. And with your arguments, I might easily get caught up in this – I haven't even Downloaded the map for the sixth time yet. Fireball;103961
This is not criticism, but demagoguery. Prorock was definitely right about some things.
At least at the time, he personally approved my ranking system in a private message. I have a message from him.Fireball;103961
This is a gross violation, but it is probably inevitable, because no system has been officially adopted. And it doesn't matter when the system was invented. The only way to avoid the violation is to choose the VDV Forever system, because his ranking was immediately written in the results table.
My system was also approved by VDV_forever. I agree that the winner in the competition should be determined by the VDV_forever system. But I don't want to call it a game ranking. Instead, the game ranking should be calculated based on the results obtained in Tournaments, even if they are based on different systems
Here you listed the player's results in order of completing stages, by days. And is the second player supposed to be me?
Exactly.
Listen – where can I see how many days it took me to complete each stage? Give me a link. I don’t remember where to look; I don’t want to search for it right now. I just want to verify this phenomenon myself.
I only remembered the results of the winners (and my own). I looked up your results here: /topic/2077/
I also have good chances in other systems. The main thing is not to miss any stage.
I would even say that you have the highest chances (but higher than mine, admittedly). But we will see. Have you seen Alex R’s signature?
At least at that time, he personally approved my rating system in a private message. There is an email from him.
You probably don’t remember, but I also approved the idea of using days as a metric back then, and I still approve it now. It's great for evaluating systems and verifying them. But I think it’s insufficient. But I wasn’t talking about your system, but about your “criticism.”
I agree that the winner in the VK should be determined by the VDV_forever system. But it shouldn’t be called a game ranking.
You speak the truth. The system for determining the winner in VK should not coincide with the ranking. All stages are equal in VK, but this cannot be said about tournaments
After all, that's how it works. In Competitive events, the winner is determined not by the in-game rating, but by the actual result or the place taken. And it may happen that a player with a high in-game rating might not win. Well, they just got unlucky – anything can happen. Prize-winning places are awarded based on the actual results. But then, various organizations or others calculate in-game ratings for players, possibly using different formulas. And it's not a fact that these in-game ratings reflect the real state of affairs in the next Competitive event. I suggested keeping in-game ratings for all forum players, both separately for each version and as a universal rating. But I never said that these formulas should be used to determine the winner in tournaments. There, it's based on the actual result. But someone decided to take the rating formulas to determine the winners in the ВК tournaments. :) What do I have to do with it? What the hell are these bonuses? It's like in a football game with a score of 4:2, giving one of the teams a bonus of an extra goal: 4:3. And even if it doesn't change the result of this match, this bonus goal can later play a role in distorting the results of the entire championship. Of course, the example is silly, but it's for the sake of simplicity to understand what is happening in the ВК tournament right now. In my system, at least the goals (read: days) are real, without bonuses. If the number of days is the same, it's a draw – they receive the SAME rating in that particular tournament. And what does it mean that secondary parameters affect the place? Like, so what if it's a draw, but those in red shorts ran faster and looked better? I can't stop laughing. :)
AmberSoler
