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.
Even so, my silly example shows that someone who participated for 1000 days will get 10 points, and the factor from the first option basically doesn't matter at all, even though it's supposed to reflect the most important metric: how close their participation days are to the leader's! And the position they achieved as a result might be completely meaningless, but the factor from the second option significantly boosts their rating, and not entirely deservedly!
You get ten points like this: (x+20)/2=10, where x is an infinitesimally small value based on the first (non-linear) calculation method. Am I understanding your point correctly? So, the whole thing is that after he gets 10 points, they will be "adjusted" to 20, according to my proposal. That is, all participants at the end of the long list will get 20 points, regardless of how they performed...
Note: We can abandon this idea. I can rework the second formula so that all points start from zero instead of 20... There are no problems with that. It's all feasible. I just consciously want to introduce a minimum threshold "for participation." That's the idea. If it doesn't work, then it doesn't work, and the formula will take on a strict form according to your comments, i.e., the last place in your exaggerated example will get (0.1+0)/2=0.05 rating points. And then they won't participate in tournaments anymore... :)
You get ten points like this: (x+20)/2=10, where x is an infinitesimally small value based on the first (non-linear) calculation method. Am I understanding your idea correctly? So, the point is that after he gets 10 points, they will be "increased" to 20, according to your suggestion. That is, all participants at the end of the long list will receive 20 points, regardless of how they performed...
You understood me correctly, but even 10 points is a lot for that player, and you are suggesting increasing them to 20.
AmberSoler;59986
Note: we can abandon this idea.
I'm not suggesting abandoning all ideas; I just want to draw attention to the fact that the game rating should reflect a player's skill relative to others, and you want to artificially equalize them so that no one gets upset. Grails were introduced so that no one would get upset.
AmberSoler;59986
I will rework the second formula so that all points start from zero, not from 20... There are no problems. It's all feasible. I just consciously want to introduce a minimum threshold "for participation"... That's the idea. If it doesn't work, then the formula will take on a strict form according to your comments, i.e., the last place in your exaggerated version will receive (0.1+0)/2=0.05 rating points. And then no one will want to participate in tournaments anymore... :)
So, give him 1000 grails so he doesn't get upset, BUT why should he have a rating of 20 or 10 if his actual skill level is only 0.05?
On the contrary, in the VDV tournaments, I try to separate the rating, which shows the current skill level. VDV gives everyone grails for oil and descriptions and for everything else, so there are no particularly upset players yet.
IronAxe;59990Just give him 1000 grails so he doesn't feel slighted, BUT why should he have a rating of 20 or 10 if his actual skill level is more like 0.05? On the contrary, in VDV Tournaments I try to separate the rating that shows current skill own. VDV showers everyone with grails for oil and descriptions and everything else, so there don't seem to be any deeply offended people yet.That's wonderful. The main points of all our mathematics are taking shape... :) Later we will simply decide among ourselves which option is more acceptable and that's it. We won't put it to a global vote — everyone interested in this matter will speak up right here. For others, it's probably not critical, so by the start of Van Kanegem, we'll have a finished project. ;)
That's great. The main points of all our calculations are becoming clear... :) Then we'll simply decide among ourselves which option is more acceptable, and that's it. We won't put it to a global vote – everyone interested in this issue will express their opinion here. The rest probably won't care, so by the start of the Van Kanegem tournament, we'll have a ready-made project. ;)
Yes, VDV hasn't uploaded my tables yet. He says he will later. He says they're still finalizing the results. I'll still calculate the tournament results for VDV as we agreed. But it's not necessary for the Kenigem tournaments to have the same system, although it would be desirable to somehow adapt a General Rating for all players on
...although it would be desirable to somehow adapt a General Rating for all players on the resource.
What prevents it? You take one rating, agree on it among yourselves, offer a ready-made and refined project to the moderator, get approval or comments, correct it - and that's it... It takes only a couple of days. Why are we all saying "we need to, we need to..."? :) If the rating that we are currently discussing in this topic gets its final form, what prevents it from being applied in VDV tournaments, starting with the one that has already passed?... We are not suffering from dogmatism, and no papers need to be signed, as far as I know... :)
All that is needed here is a show of will, and it is there...
What prevents us from doing this? We take one rating system, agree on it, offer a ready and refined project to the organizer, get approval or feedback, correct it – and that’s it... It takes just a couple of days. Why are we constantly saying "we should, we should..."? :) If the rating system we are discussing in this thread gets its final form, what prevents us from applying it in VDV Tournaments, starting with the one that just took place?... We are not dogmatic, and we don't need to sign any papers, apparently... :) All that is needed is a show of will, and we have that...
There are many inconsistencies here. 1. I disagree with artificially boosting the rating of someone who doesn't deserve that level. Because in VDV tournaments, a player using "oil" and other exploits will receive additional rewards for it, and they will also get a free rating boost. 2. If VDV or someone else decides to hold tournaments again at different difficulty levels, you disagree with the coefficients and didn't offer your own. 3. The rating decay is designed to automatically remove players who are only active for a short time or those who haven't played in any tournaments on this platform for a long time. The decay period can also be discussed, but you seem to be completely against decay. The rating can't always be increasing for everyone. A rating is a rating because it goes up and down for each player depending on their activity and results in tournaments.
And I also didn't understand the latest changes in your calculation system, regarding the fact that the second component in the second calculation option, depending on the place taken, will not have such a strong influence on the rating (at least I showed that it has a much stronger influence than the component in the first option). You could recalculate that same silly example of mine and show on the example how it will look. The example may be silly, but it clearly shows what it was supposed to show
There are many inconsistencies here. 1. I do not agree with artificially boosting the rating of someone who doesn't belong at the level they are being boosted to. Because in VDV Tournaments, someone playing for oil, etc., will receive additional grails for it, and on top of that, their rating will be boosted for free.
I fully admit that my option regarding a minimum guaranteed bonus may be unacceptable. If it doesn't pass, we will abandon it, but this "purely gameplay" point is worth considering, because a bonus for participating in a tournament is not the same thing as a bonus for a beautiful description or for helping organize a tournament... I completely agree that certain things should be encouraged, but certainly not through the rating. However, participation in a tournament — even unsuccessful at the moment — is directly related to the rating.
The significance of the struggle for the top spots in the rating table is not the same as for the bottom ones. In the "header," there is a fundamental struggle, and any leniency there is a sign of bad taste. Everything there is strict and categorical. In the "tail," different laws prevail; there are fewer disputes and more solidarity. Additionally, the "tail" is necessary to popularize tournaments and support those players who do not aspire to leadership. They are not your competitors — I am speaking plainly — and therefore I do not understand your critical mood regarding the minimum guaranteed bonus for participating in a tournament... Although I fully understand your position. But I do not share it because I see no fundamental problem with a minimum bonus. It does not affect the results of the leaders; it simply maintains a quorum. IronAxe;60012
2. If VDV or someone else still conducts tournaments again at different difficulty levels — you do not agree with the coefficients, yet you haven't proposed your own.
I did not propose my own coefficients because they do not exist in nature. Any coefficients designed to equalize incomparable things are purely contrived and unjustified. I have said this before and will repeat it — everything depends on the map. Firstly, each specific map will have its own individual coefficients, and they cannot be presented in a general form — which is what you are trying to do. Secondly, in principle, I am categorically against multiple starting saves. Now, after the BB results are finalized, I can state this without reservations like "we'll see" or "time will tell"... Time has shown that multiple saves create more disputes and problems than advantages... Therefore, I hope the organizers will abandon this practice. IronAxe;60012
3. Rating decay is intended to automatically erase one-day players or those who haven't played in any tournaments on this resource for a long time.
As if there are no other ways to solve this problem without infringing upon the interests of regular players. One way is to adopt a minimum quota for the number of tournaments per year to filter out accidental results. I don't know if you read my post or not, but I'll repeat — the rating decay mechanism you proposed is mathematically unsound. I provided you with a detailed calculation logic. You have not provided any comments on this yet. So let's be consistent — first, your arguments in defense of the proposed decay mechanism, then we will continue discussing this point. IronAxe;60012
Decay terms can also be negotiated, but you seem to be completely against decay. Well, the rating cannot just keep growing for everyone all the time. A rating is a rating precisely because it goes up and down for each player depending on their activity and tournament results.
I am not against it; it's just that your decay mechanism depends on the time elapsed since the tournament took place, which defies any criticism. A February tournament should not be less significant than a May one, etc. For example, I played six tournaments from January to June and received 6*50=300 points, and you also played 6 tournaments from July to December and received 6*50=300 points. In my view, our annual ratings are equal. What would you say? :) Remove this inconsistency in the calculation and then we will continue. IronAxe;60012
And I still didn't understand the latest edits in your calculation system, regarding the fact that the second component of the second accrual option depending on the place taken will not affect the rating as much (at least I showed that it has a much stronger effect than the component of the first option). You should have recalculated that same stupid example of mine and shown how it would look.
Once again. My last post showed that points can be awarded not from 20 to 100 per tournament, but from 0 to 100, and that's it.
"regarding the fact that the second component of the second accrual option depending on the place taken will not affect the rating as much" — I did not say that.
IronAxe;600123. Rating decay is intended to automatically erase one-hit wonders or those who haven't played in any Tournaments on this resource for a long time. Decay periods can also be negotiated, but you seem to be completely against decay. The rating cannot simply grow for everyone all the time. A rating is called a rating because it goes up and down for each player depending on their activity and results in Tournaments.Rating decay of 1/3 per tournament is completely unacceptable. Rating could be decayed, for example, once a year by 1/2 for everyone, and players who have played only 1 tournament in a year could be removed from the rating. The fact that a person missed several tournaments is not yet a reason to reduce their rating practically to 0. They earned it through their play, even if it was in two or three tournaments, and taking away their rating for missed tournaments is simply not right.
Reducing the rating by 1/3 for each tournament is completely unacceptable. For example, you could reduce the rating once a year by 1/2 for everyone and remove players from the rating who have played only one tournament in a year.
I suggest justifying each number in the calculation and not pulling them out of thin air: why can the rating be reduced once a year by 1/2? In what sense - why not reduce it five times a year by 1/8?
And in general, can anyone say on what basis you propose to reduce the importance of early tournaments in relation to later ones within the calculation of the annual rating? The question is not about the numbers and the frequency of the reduction, but about the legitimacy of applying such a mechanism. Why is the upcoming February tournament worse than others, so that its rating will be reduced (once by 1/2 or several times by 1/8 - it doesn't matter) at the end of the year in relation to the subsequent December tournament? I just don't understand this...
However, I tend to believe that there is a fundamental misunderstanding in our discussion. You are confusing the current rating (which is not tied to time) with the annual rating (which is calculated at the end of the year for a strictly defined period of time). It seems to me that the issue of rating decay has confused our conversation... It is quite reasonable to reduce the current rating, but it is not reset at the end of the year, but only reflects the current state of affairs at each individual point in time. We are talking about the annual rating, which will be awarded at the end of the year and reset after that.
I suggest removing the discussion of rating decay as inappropriate.
I suggest justifying each number in the calculation and not pulling them out of thin air: why can the rating be aged once a year by 1/2? I mean, why not age it five times a year by 1/8?
You yourself explained in previous posts why aging it five times a year by 1/8 is unacceptable.
At the end of the year, there will simply be a final annual rating, as a result of the players' achievements during the year. And aging this rating by 1/2 (or by some other number, or simply awarding a certain number of points towards the new year's rating for leading positions in the final annual rating) after the new year begins is to prevent the leaders from becoming complacent – all gaps will be reduced. And at the same time, to ensure that the achievements of the past year are not forgotten as a result of the rating being reset for the newly started year.
In general, without aging or resetting the rating, the situation could become simply comical: for example, players who have not participated in tournaments for a long time could remain in the top 20 for a long time, but remain in the rating due to their achievements of a year or more ago. This situation has arisen, for example, in the Korean Star Craft league, where a couple of players who ended their professional careers remained in the top ten for quite some time. If you reset the rating, this problem will not occur.
Reducing the rating by 1/3 for each Tournament is completely unacceptable. You could, for example, reduce the rating by 1/2 once a year for everyone and remove players from the rating who have played only one Tournament in a year. The fact that a person has missed several Tournaments is not a reason to reduce their rating to almost zero. They earned it through their gameplay, even if it was in just two or three Tournaments, and it's simply unfair to take away their rating for missed Tournaments.
Sage;60019
In general, without aging or resetting the rating, the situation could become simply comical: for example, in the top twenty players, players who have not participated in Tournaments for a long time could remain for a long time, but remain in the rating due to their achievements of a year or more ago. This situation has arisen, for example, in the Korean Star Craft league, where a couple of players who had ended their professional careers remained in the top ten for quite a long time. If you reset the rating, there will be no such problem.
Sage, have you gotten stuck in your ways? So, should we reset it or not? :)
You yourself explained in previous posts why aging the rating five times a year by 1/8 is unacceptable.
Aging the annual rating is completely unacceptable. I asked you where you got the numbers 1/2 from and why not 1/8, for example?..
:) It feels like we're on different planets. :) Please read carefully what others are writing. Sage;60035
And aging this rating by 1/2... after the new year so that the leaders don't get complacent.
There are only two systems:
1. Current rating. It is indefinite and never reset. The only feature is that old results are aged to reflect the current state of affairs. To avoid the Korean nuances you were talking about.
2. Rating (cup, championship, etc.) for a specified period (year, season, month). Here, aging does not occur. The result accumulates during the period, after which the results are summarized and reset for the next start. The start is from scratch.
I'm talking about the second option. And you about which one?
I'll try to summarize all the suggestions made. 1. The pure gameplay rating will be calculated using the system proposed by the respected IronAxe, without considering aging, but with a minimum quota of tournaments per year (I suggest 6 tournaments - one every two months). Players who have played fewer tournaments can have their rating aged. 2. The "grails" that can be spent on a doll will be calculated using the system proposed by the respected AmberSoler, according to the third option (oscillator). There should be an incentive in the form of grails! 3. In the "Van Kanighem" tournament (and similar ones), the results should be calculated using a cup system - a small cup for each version + the winner of the overall standings (among those who have played at least 3/5 stages). I
Such a situation arose, for example, in the Korean Star Craft league, where a couple of players who had ended their professional careers remained in the top ten for quite a long time. If you reset the Player Rating, this problem will not exist.
I don't want any Korean situations. 1. If a player misses Tournaments, their rating automatically decreases, even if for a valid reason, such as breaking a leg, arm, or even their head - isn't that true? 2. What is Botvinnik's current Player Rating? He's not here; he died a long time ago. But if he were alive and missed some major Tournaments or hadn't played anywhere for several years, what would happen? 3. I suggested not aging a participant's rating if they play at least once every couple of months in at least one Tournament on this platform. Because when the results are summarized, their old rating will automatically decrease by 1/3, just like everyone else participating. If a person misses a couple of months and doesn't participate in any Tournament on this platform, their old rating will also decrease when the results are summarized, just like everyone else who participated. Only, for those who participated, the earned rating for the current Tournament will be added to the decreased rating, while for the one who missed it, zero will be added to the decreased rating! VDV and the Prophet determined a reasonable period - they gave a month. I'm not against increasing it. But I'm going to calculate the CURRENT GAME RATING, not the annual one. Let someone else keep track of all those past achievements in the Hall of Fame; we've already talked about that somewhere.
Added 5 minutes later AmberSoler;60040
Aging the annual rating is completely unacceptable. I asked you where you got the numbers 1/2 from and why not 1/8, for example?..
:) It feels like we're on different planets. :) Please read carefully what others are writing.
There are only two systems:
1. Current rating. It is indefinite and never reset. The only feature is that old results are aged to reflect the current state of affairs. This is to avoid the Korean nuances you were talking about.
2. Rating (cup, championship, etc.) for a specified period (year, season, month). There is no aging here. The result accumulates during the period, after which the results are summarized, and it is reset for the next start. The start is from zero.
I'm talking about the second option. What about you?
I've always been talking about the CURRENT one. But the question is - if a championship is taking place, can't I duplicate the CURRENT rating of the players who are participating in it? I don't care what other rating they might have
1. The pure game rating is calculated using the system proposed by the respected IronAxe, without taking aging into account, but with a minimum quota of tournaments per year (I suggest 6 tournaments - one every two months). The rating of players who have played in fewer tournaments can be adjusted.
You are not taking into account the number of participants in the tournament and the player's position in the tournament standings. That is, you are simply using Option #1 from my post, which was based on IronAxe's suggestion. But I have already analyzed this in detail using the example of summarizing the results of the BB tournament (see the attached file above). AlexeyPank;60044
2. The Grail points that can be spent on a hero are calculated using the system proposed by the respected AmberSoler, according to the third option (oscillator). There should be an incentive in the form of Grail points!
Please note that the oscillator primarily reflects the player's strength in comparison to other participants in the tournament. What do Grail points have to do with it, when they can be earned even without participating in tournaments?... The number of Grail points does not determine the player's status; why use an oscillator there? AlexeyPank;60044
3. In the "Van Kanigem" tournament (and similar ones), use the cup system - a small cup for each version + the winner of the overall standings (among those who have played at least 3/5 stages).
There are no questions here. Only, if 3/5 stages refers to all versions of the game in total, then 3/5 is too much. No one will play that many tournaments in a year - roughly, there are 12 tournaments in each version, for a total of 3/5*(12*4) = 29 tournaments. This is not realistic.
AmberSoler
