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 AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#91
Auto-translated
AmberSoler;60160
If you calculate the average rating based on the number of tournaments each player has participated in, then there's no need to "age" the ratings... And "occasional" players shouldn't be reflected in the current ranking at all... If they don't play, they're not in the ranking. If they return, their rating is restored to the table... How does this idea sound? It would make the system simpler and show the number of tournaments a player has participated in...
I don't quite understand. How would you know if a player is just "dropping in" or if they're playing in a duo tournament on this same platform, or if they simply decided to take a month off for legitimate reasons? And it gets worse. Someone plays one great tournament and then disappears. Two months later, they come back and start complaining about where they are in the ranking table. Who will bother searching for where they were and what their rating was? Not me, and I doubt anyone else will. If you want to take on that task, go ahead :) The simplicity that VDV requested is slipping away like water through your fingers. I'll only track those who are more or less consistently active here. If someone disappears for a long time and then comes back, whether they complain or not, their rating is gone – bye-bye. I don't need to remember every "jerk." :) Let the moderators take care of the Hall of Fame – let them go there and revel in their past victories and get motivated to fight for a new rating. And there are probably players who like to change their nicknames for every tournament. It will be quite a challenge to track them, even without them. :)
0
⚖️
0
In reply to IronAxe
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#92
Auto-translated
IronAxe;60163
I don't quite understand. How do you know if he's a newcomer, or if he just played a couple of games on this site, or if he simply decided to take a month off? And it gets worse. Someone played one great tournament and then disappeared. Two months later, they return and start complaining about where they are in the ranking table on the site. Who will bother looking for who is where with what rating? Not me, I can't.
Don't rush... :)

There's no need to find anything out about him. He plays once, and he's permanently in the ranking. If he doesn't play, the game counter stays put, which means his rating isn't recalculated, and the player gradually slides to the bottom of the table, even if he played well. For regular players, the rating increases depending not only on their tournament results but also on the number of tournaments (we'll make this parameter less significant, of course)... You'll never have to look for anyone. Everything will be calculated automatically, and it will be much easier for you – you won't have to do any manual work. The inactive players will slide to the bottom of the table on their own... :)

All of this is, of course, just a rough idea, but it's too early to dismiss it. We just need to find that golden mean that will satisfy everyone...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#93
Auto-translated
AmberSoler;60160

It would be easy to add a multiplier that boosts a player's score based on the number of tournaments they've played, to avoid a situation where one player has played two tournaments well and stops playing, while another has both successes and failures, is active, but has a lower average score than the first... then the activity coefficient would gradually increase his score...
And when should this cool player with two games be removed from the ranking, and on what basis, if his rating doesn't decrease, even though it doesn't increase. Otherwise, he'll just get old, old, and eventually stop playing, freeing up a spot for others.
Also, there won't be a situation where someone sees a formidable opponent in the announcements and, even though they have a decent rating, thinks, "Nah, I'll go play in another tournament where this guy isn't." Or they'll just skip it. But the aging system will, on the contrary, encourage them to play a tournament on this platform and do it quickly, so that their rating doesn't decrease.

Added 4 minutes later
AmberSoler;60164
Don't rush... :)

There's no need to find out anything about him. He plays once - he's constantly in the ranking. If he doesn't play, the game counter stays in place, which means his rating isn't calculated over time, and the player gradually slides to the bottom of the table, even if he played well. For active players, the rating increases depending not only on their result in the tournament, but also on the number of tournaments (we'll make this parameter less significant, of course)... You'll never have to look for anyone. Everything will be calculated automatically and it will be much easier for you - there's no need to "age" anyone... The "empty" players will slide to the bottom of the table on their own... :)

All of this is, of course, just ideas and concepts, but it's too early to dismiss them... we just need to find the golden mean that will satisfy everyone...

I think we discussed such a system and AlexPank rejected it. It's hard to tell who will slide down - the "empty" players or an unlucky player who scores 2 points in 10 tournaments, 20/10=2. And the "empty" player will stay at the top for a long time, 100/1=100 :) So, let's adjust the coefficient so that the guy with two points is happy and overtakes the "empty" player :)
0
⚖️
0
In reply to IronAxe
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#94
Auto-translated
IronAxe;60165
I think this system was discussed, and AlexPank rejected it. It's hard to tell who will drop out – inactive players or a weak player who scores 2 points in 10 tournaments (20/10=2). But an inactive player will stay at the top for a long time (100/1=100). So, let's adjust the coefficient so that a player with two points is satisfied and surpasses the inactive player.
Okay :) So, for now, let's adopt a decay of 1/10 for each tournament that passes... Whether they play or not doesn't matter. In half a year, the rating will be halved, so inactive players will naturally drop down... But, this is the current rating, which is not subject to any rewards at the end of the period... And it will only be used to identify active players, to create a current TOP-10, for example, and to calculate the qualification coefficient for each tournament.

This coefficient will affect the number of rating points for a tournament for each player who participated in that tournament. How will it affect it? I think simply by increasing it proportionally. So, the maximum for a tournament will ultimately be not 100 points, but 100*k, where k=sum.av(TOP5)/sum.av(ALL).

Now, let's estimate how much this coefficient will be if everyone who participated in BB plays in the next tournament. So, k=72/38=1.895. Hmm... it distorts things a lot... although the distortion is only now, because there is too much variance based on just one tournament. Over time, this coefficient will approach one, so we can leave it as a standby idea, and we'll see... By the way, I would like to note that this will be a solution to the problem of a strong player appearing in a tournament, who will lower the bar with their performance... The stronger the players, the higher the compensation... and everything is interconnected, and there isn't a single coefficient pulled out of thin air... which is pleasing... :)

I'm just gathering thoughts and ideas... Otherwise, they're just engaging in sophistry :)
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#95
Auto-translated
AmberSoler;60167
Okay :) So, for now, let's assume a rating decrease of 1/10 for each tournament that passes... It doesn't matter if they play or not. In half a year, the rating will be halved, so the inactive players will naturally fall down... But, this is the current rating, which is not subject to any rewards at the end of the period...
It will still take a long time to fall. But I agree with this because we are going to reduce the rating for everyone.
100 100 2 2
90 92 3.8 3.3
81 85 5.4 4.2
73 78 6.9 4.8
66 72
59 67
53
48
43
The first column is the natural decrease, the second is a player who is extremely unlucky and receives 2 points in the remaining tournaments after a victory. The third column is the unlucky player :) who always gets 2 points. Calculate the rest yourself :) And with 1/3, even with a monthly bonus. And these wonderful people who are trying to play actively will observe the slow fall of the eagle :) FOR A LONG TIME :) Well, maybe the coefficient will somehow accelerate the decline :) or increase the catch-up rate.

Added after 14 minutes
AmberSoler;60167
Okay :) So, for now, let's assume a rating decrease of 1/10 for each tournament that passes... It doesn't matter if they play or not.
The table also shows that a player who plays well once and then plays poorly for a long time also cannot fall quickly. In VGAPlanets, such tricks didn't work; if you played well once, maintain your level. If you fall to last place in a tournament after being first in the previous one, you will fall down just as quickly in the overall current rating :) precisely because of the 1/3 decrease. The dynamics of the rankings in space strategies were amazing :) And look how well the third player is doing. Even a 1/3 decrease wouldn't hurt him. The only thing hindering him is the "parachutes" of the first two :)
0
⚖️
0
In reply to IronAxe
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#96
Auto-translated
IronAxe;60168
...And so these wonderful people who are trying to play actively will watch the eagle's slow descent :) FOR A LONG TIME :) Well, maybe some coefficient will speed up gravity :) or increase the catch-up rate.
I disagree.
1. You took two extreme values. Why didn't you add a third player of "stable average strength" to the table? Again, not the full picture... This is not a method for analysis, this has already been discussed...
2. We calculate the player's strength rating taking into account their activity. But strength is still primary, not activity. Aging is intended to remove inactive players from the table, correct? But through aging, you are trying to punish a player who made a mistake "for nothing." Therefore, one lost or missed tournament should not hit the result hard... this is not a lottery and the rating should be elastic...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#97
Auto-translated
AmberSoler;60172
I disagree.
1. You took two extreme values. Why didn't you include a third player with "stable average strength" in the table? Again, it's not a complete picture... This isn't a method for analysis; we've already discussed this...
2. We calculate a player's strength rating based on their activity. But strength is still the primary factor, not activity. The aging system is designed to remove inactive players from the table, right? But with aging, you're trying to "unfairly" punish a player who made a mistake. Therefore, a single lost or missed tournament shouldn't significantly affect the result... this isn't a lottery, and the rating should be flexible...
To understand how the average value will behave, you need to consider two extremes; I looked at one, and it was bad. I'm not considering the second, which is when a player performs exceptionally well and takes high places in a tournament. With aging, I'm not just trying to punish a player who made a mistake, but also to give less strong players a chance to be at the top, even if only for a moment. Why aren't you talking about that?
The rating is, in fact, a lottery. A tire blew out – a pole was hit. Took last place – the rating dropped significantly. Do you think anyone will feel sorry? Or will this be an excuse in Formula 1? No way – don't reduce the rating; it wasn't me; it was the tire's fault :)
And there, they can cut you off and hit you – who cares? :)

Added 5 minutes later
AmberSoler;60172
I disagree.
1. You took two extreme values. Why didn't you include a third player with "stable average strength" in the table? Again, it's not a complete picture... This isn't a method for analysis; we've already discussed this...
2. We calculate a player's strength rating based on their activity. But strength is still the primary factor, not activity. The aging system is designed to remove inactive players from the table, right? But with aging, you're trying to "unfairly" punish a player who made a mistake. Therefore, a single lost or missed tournament shouldn't significantly affect the result... this isn't a lottery, and the rating should be flexible...
And in the example, I'm showing a player who, after taking first place once, then takes last place five times in a row, and you say it's a coincidence. He's giving it his all, but the rating doesn't seem to want to drop :) Although he's already taken the same place five times in a row with a guy from the third column – their ratings would have been equalized a long time ago in reality :)
0
⚖️
0
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#98
Auto-translated
Well, we have different approaches to ranking... :) A person who is at the top of the ranking, which can change drastically in just one tournament, doesn't inspire respect as a player... The ranking should be such that it takes a long time and consistent effort to reach the top, and luck shouldn't knock a player down... And if you're suggesting organizing a "King of the Hill" or "Caliph for an Hour" type of game, then I'm out... :) I'm not interested, to be honest. I advocate for creating motivation for long-term tournament competition... You're suggesting playing here and now... :) I'm willing to work towards the top for a year, proving my strength and significance as a player... But what's the point of putting in effort all year if a ranking system allows someone like Vasya Pupkin to jump to the top because someone else had a flat tire (i.e., a random setback)? And no one will be able to stay at the top for long with such a system – again, due to the law of averages and the theory of probability... No matter how hard you try, everyone will have setbacks. And the player's skill has nothing to do with it – it's all decided by blind chance... It seems to me that you're trying to model a ranking system that captivated you in space racing games :) But is it worth it? Here, everything is different, I think... The calculation mechanism is the same in both cases, but the coefficient of randomness (determined by the magnitude and speed of decay) can be different. I agree that decay should gradually remove weak players from the ranking. But as I see it, you're striving to create an opportunity for random players to get into the TOP of the ranking... why? I think a random player shouldn't be among the best, no matter how lucky they are, and the best player shouldn't lose their position significantly due to a series of bad luck... But if they stop playing or simply don't try to play... then they'll fall down, and a more persistent and hardworking player will take their place... Then – yes! Just my opinion, as always... :)
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#99
Auto-translated
AmberSoler;60172
N
Therefore, losing or missing one Tournament should not significantly affect the result... it's not a lottery, and the rating should be flexible...
Missing one shouldn't immediately have a big impact – a grace period is given, and the deadline can still be negotiated, but losing one is a different story; the rating severely penalizes losers, especially those who were recently among the champions, isn't that right? But I'm saying that I'm not against it, but look at the examples given and think for yourselves.

Added 9 minutes ago
AmberSoler;60175
Well, we have different approaches to the rating... :)

A person who is at the top of the rating, which changes drastically after just one Tournament, doesn't inspire respect... The rating should be such that it takes a long and persistent effort to reach it, and randomness shouldn't knock a player down...
And if you're suggesting organizing a "King of the Hill" or "Caliph for an Hour" game, then I'm out... :) I'm not interested, I'm being honest. I advocate for creating motivation for Tournament competition in the long term... You, on the other hand, are suggesting playing here and now... :) I'm willing to work towards the top spots for a year, proving my strength and significance as a player... But a rating that would allow Vasya Pupkin to jump to the top spot because someone's tire burst (i.e., a random mishap) – what's the point of putting in effort all year?..
And no one will be able to stay at the top for long, again, according to Murphy's Law and the theory of probability... And the player's skill doesn't matter – it's all decided by blind chance...
Don't you notice how persistently Vasya Pupkin is moving up in the third row, but he has few chances :) He's just hoping that the leaders' tires will burst. And we're talking about the Current rating, and you're talking about the Annual rating again. We were talking about how the current ratings, in a way, have a minor impact on the Annual ratings of each player. So why not take care of the tires :) You know that in sports, the rating is unlikely to fall smoothly if a player loses repeatedly; I'm just showing an example – the player has lost five times, and five times a tire has burst. But Vasya Pupkin's tire doesn't burst; he just drives poorly. But the ratings can't equalize them. Why doesn't anyone want to stand up for Vasya Pupkin? :)

Added 8 minutes ago
AmberSoler;60175
And no one will be able to stay at the top for long, again, according to Murphy's Law and the theory of probability... And the player's skill doesn't matter – it's all decided by blind chance... I think you're trying to model a rating system that has captivated you in space races :) But is it worth it? Here, everything is different, I think...
If we treat the game of Heroes as a sports game like chess, then the rating is the rating. Today it exists, and tomorrow it falls, and you need to pull yourself together and continue the fight if you want to stay in the TOP. But I'm saying, even if you do it this way, I'm not against it. We've already complicated things enough; VDV will come and be surprised. Aging was accepted by both the Prophet and VDV at 1/3. But I think they would agree to reduce it; then we'll remove all sorts of bonuses. But that cosmic system has probably been working since 1992 – it's been tested by time :) Yes, strong players have fallen, but in the next Tournament, they would rise again. But again, in the Tournaments, the strength of the Tournaments was taken into account. That is, they went where there were strong players and tried to win – if they succeeded, the status quo was restored :) If they didn't succeed, they slowly climbed up in the next Tournaments :)
0
⚖️
0
In reply to IronAxe
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#100
Auto-translated
IronAxe;60176
...But Vasya Pupkin doesn't "implode," he just plays poorly. But they can't equalize the ratings. Why doesn't anyone want to stand up for Vasya Pupkin? :)
:D :D :D

Essentially, your position is understandable, and therefore acceptable. I was just responding to your nostalgia for the aging coefficient of 1/3... That's a lot, and I explained why I think so... 1/10 is closer to reality. And your calculation of a gradual (but not rapid) decrease in rating is correct... Yes, an unlucky player will halve their rating in five tournaments, but even that is a catastrophe for a normal player! And you want to lose half your rating in two tournaments?

Five tournaments with a tight schedule is two months of play. How much more could it be? :)

Added 2 minutes later
IronAxe;60176
Yes, strong players sometimes fell there, but they would resurface in the next tournament. But again, the strength of the tournaments was taken into account in the calculations. That is, they went where there were strong players and tried to win - if they succeeded, the status quo was restored :) If they didn't succeed, they slowly climbed back up in the following tournaments :)
How many tournaments were there in a year? Roughly?
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#101
Auto-translated
AmberSoler;60179
:D :D :D
Essentially, your position is understandable, and therefore acceptable. I was simply responding to your nostalgia for the 1/3 aging rate... That's a lot, and I explained why I think so... 1/10 is closer to reality. And your calculation of a gradual (but not rapid) decrease in the ranking is correct... Yes, an unlucky player will halve their ranking in five tournaments, but even that is a disaster for a normal player! And you want to lose half your ranking in two tournaments?
Five tournaments with a tight schedule is two months of play. How much more could it be? :)
Added 2 minutes later
How many tournaments were there in a year? Roughly?
:) There were hosts who ran games 2-3 times a week. The game could last from 35 to 50 or more turns. Well, depending on the players participating.
Roughly speaking, 2-3 months per game.
If I PLAY and score fewer points than are deducted from me in two tournaments, why not? No, I certainly don't want to, but ask Vasya Pupkin - does he care what I lose? The more I lose, the better it is for him :) But if I am forced to skip and don't meet the deadline, then of course it's a pity to lose rating just like that. But how will we find out who and why didn't play? And Vasya Pupkin cares about this the least! :) He always has his tires in order - he has a good time :)
0
⚖️
0
In reply to IronAxe
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#102
Auto-translated
IronAxe;60181
But how will we know who didn't play and why? And it's the least of Vasya Pupkin's concerns! :) He always has good tires - he has a skating rink!
It won't bother me either. Or you. Or the host... Don't overcomplicate things, no advantages. The count will be done on equal terms - whether you played or not... it doesn't matter. Otherwise, we'll just get bogged down. The question is this - a person only plays the second game, for example... What will happen to his rating? Will it decrease after tournaments in the first, third, or fourth games? I think that would be fundamentally wrong... It's even worse than comparing players from different versions...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#103
Auto-translated
AmberSoler;60182
It won't bother me either. Nor you. Nor the host... Don't overcomplicate things, no advantages. The scoring will be done on equal terms - whether you played or not... it doesn't matter. Otherwise, we'll just get bogged down. The question is this - if a person only plays the second game, for example... What will happen to their rating? Will it decrease after tournaments for the first, second, third, or fourth games? I think that would be fundamentally wrong... It's even worse than comparing players from different versions...
That's exactly why we need clear accounting on the resource. And my answer is this - if a person only plays the second game and there were no tournaments for the second game on this resource this month, then the rating should not decrease - the player is NOT to blame.
And this organizational issue is one of the most difficult if we want to have a GENERAL CURRENT RATING for heroes on this resource.

Added 5 minutes ago
AmberSoler;60182
It won't bother me either. Nor you. Nor the host... Don't overcomplicate things, no advantages. The scoring will be done on equal terms - whether you played or not... it doesn't matter. Otherwise, we'll just get bogged down.
But what if a person is on vacation and away - I think we should definitely give a month for various unforeseen circumstances? But if we simplify it and reduce it by 1/10, maybe we can live with that, but even if Vasya Pupkin, even if he works the levers like the wolf in "Wait for Me!", will be working hard, it will take him a long time to overtake cool players :) even if he doesn't go on vacation.
0
⚖️
0
User Avatar
Offline AmberSoler
Registered:
Posts: 1,394
#104
Auto-translated
That's true, and the opposite is also true... For example, in the third game, a team wins tournament after tournament, while in the second game, they don't – for various reasons... Players in the third game are gradually moving up. The rest are standing still... In my opinion, the GENERAL CURRENT RATING is a utopia.

Added 4 minutes later
IronAxe;60183
But if you simplify it and reduce it by 1/10, maybe it can be tolerated, but even if Vasya Pupkin, like the wolf in "Wait for Me!", works hard on the battlefield, it will take him a long time to overtake cool, but inactive players :) even if he doesn't take vacations.
That's what I'm talking about... I don't think it's necessary to collect vacation requests. This is an unjustified burden on the administrator. Why? That's why we reduced the aging penalty, so that a couple of missed games don't undermine the foundations of the game and the player's desire to continue... Agree, nothing kills initiative like setbacks... A couple of failures will simply stop a player's enthusiasm because they are sure that failures will periodically repeat... And they are right – that's how it is...

Therefore, the less the element of randomness affects the rating, the better it is for everyone...
Сначала было слово...
FizMiG v.2.0 *** Очередное обновление справочника! *** http://vk.com/fizmig

0
⚖️
0
In reply to AmberSoler
User Avatar
Offline IronAxe
Registered:
Posts: 1,495
#105
Auto-translated
AmberSoler;60185
That's true, and the opposite is also true... For example, in the third game, players consistently perform well from tournament to tournament, while in the second game, they don't - for various reasons... Players in the third game are gradually moving up. The others are standing still... In my opinion, a GENERAL CURRENT RATING is a utopia.
As the rating ages by 1/3, it's still unclear whether it's better to stand still or move forward. Those who play for the rating understand this perfectly; that's what makes them players. They crave the game to gain more rating, but they also understand that if they perform significantly worse compared to the previous tournament, their rating might decrease. It's a kind of hunt for luck.
If the rating ages by 1/10, then even with a very poor performance compared to the previous one, the rating will likely not decrease. You yourself pointed out one of the reasons why VGAP used a 1/3 reduction. And in general, it seems to be taken from real-life rating calculation formulas, perhaps from Elo or some other system; I'm not sure.

Added 4 minutes later
AmberSoler;60185

That's what I'm talking about... I don't think it's necessary to collect leave requests. This is an unjustified burden on the organizer...
That's why they gave a one-month grace period without requiring reasons for absence.
AmberSoler;60185

And that's why, the less randomness affects the rating, the better it is for everyone...
Tell that to Vasya Pupkin. Right now, you're advocating for the interests of consistently strong players. So, whether it's better for everyone is debatable.
And the point of the rating is lost if it's too stable and doesn't depend on randomness. The current rating is the current rating. And no one can take away the glory of victories - they all go into the Hall of Fame
0
⚖️
0