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

I agree. But there is no consideration of the tournament's popularity either. Everyone will always receive the same amount. I would like the number of grail rewards to depend on the number of participants. However, if the majority votes for the option proposed by the respected VDV_forever, then it will be adopted. I am not insisting on my own.

I agree. But there is also no accounting for the popularity of the tournament. Everyone will always receive the same amount. Personally, I would like the number of grails to depend on the number of participants.
But if the majority votes for the option proposed by the esteemed VDV_forever, then that's the one that will be adopted. I'm not insisting on my own.

It seems to me that this problem should be solved more simply. Well, how many people can actually participate and make it to the end of a tournament? We already have experience with this. Just create a scale for awarding grails based on how many participants submit their final save.
up to 5 - award option 1
from 5 to 10 - award option 2
from 10 to 15 - award option 3
and so on.
Why bother with recalculations and complicate the calculation systems? Who is going to do all that? Moreover, many people don't even try to understand these calculations; at first glance, the system seems confusing and complicated to them. Consequently, there is no enthusiasm. It needs to be simpler.

This system could also work. The main thing is that it functions. Otherwise, we have less than a month until the tournament starts, and we still haven't decided how we're going to evaluate the submissions. It's time to start discussing the maps and the rules.

This system could work too. The main thing is that it functions. Otherwise, we have less than a month until the tournament starts, and we still haven't decided how to evaluate it. And it's time to start discussing the maps and the regulations.

That's what I'm suggesting. Take our purely gameplay-based rating system. Calculate the Grail points using the VDV system, introducing a scale for awarding points based on the number of participants and the seed. Perhaps simply continue calculating the rating for players who participate in your tournament and who have already participated in VDV's Best Of Best and already have a rating. Players without a rating will start with zero.
By the way,

To avoid getting bogged down in details and causing unnecessary arguments, I will try to make only general comments:

1. The rating is purely a game-related metric and has no bearing on contributions to the portal.

2. In a single tournament, there should be only one evaluation criterion for all participants, chosen from the following list (which can be expanded):

- speedrun,
- score-based game,
- game based on the amount of resources collected within a specified and uniform time limit for all,
- game based on the amount of experience gained by a specified hero within a specified and uniform time limit for all,
- game based on the number of skeletons, demons, earth elementals, etc. - you can specify a time limit, or the number; i.e., whoever gathers the most first wins, etc.

For variety, you can hold a new tournament each time with a pre-agreed condition to determine the winner.

3. The scoring will be standard: all participants are ranked in a row, depending on their position, based on the winning criteria, and a simple mathematical calculation is performed:

The player who finishes last receives 20 rating points, the player who finishes second to last receives N/2 points more (where N is the number of participants, 2*N seems too much), and so on, each subsequent player in the list has an advantage of N/2 points up to the first place... This is how it should be - the more competitors you beat, the more significant your victory... The player who finishes first (or 1st to 3rd place) receives an additional N points on top of the calculated score (a special prize for winning).

4. At the end of the year, a balance is calculated separately for each game version - a small cup for Heroes of Might and Magic II, a small cup for Heroes of Might and Magic III, etc.

The sum of absolutely equivalent rating points across all versions will determine the overall winner in the combined standings.

It's simple and clear. The fewer complications, the better for everyone.

5. All stages should be taken into account, and it doesn't matter how many each player has completed. What is the purpose of limiting them?

6. There is no need to submit preliminary applications to participate. I won't give an example; everyone has seen it. Fewer than half of the participants reached the finish line. A third of the posts in the thread were simply empty statements of participation and thanks for such a bold decision. Why bother the moderator? The end date of the tournament is set and will not be changed; the information about the number of players and the calculation of results is all that is needed. That's it.

7. I would like to comment separately on the concept of aging the rating by 1/3... Have you tried to calculate what will happen to your 50 rating points earned in the first stage after 10 stages? Only 0.86 points will remain... :D Why play the first stage for such a small amount? You can just play the last two...

Please don't criticize me too harshly; how long can we discuss a simple question? And what does it have to do with those who will come to the portal in six months? Are you starting tournaments and calculating ratings for them, or for yourselves? Those who miss out, miss out. Play better, and everything will work out, instead of aging the rating to make it easier for newcomers... Yes, the rating should be reset at the end of the year, but it is not wise to age it during the year. IMHO, as always. :)

Very positive. But then, a general ranking should definitely be held for those who have played 3/5 of the total number of tournaments. Otherwise, with the addition of N/2, the winner of a tournament with 30-40 participants will easily outperform those who have participated in 3-4 tournaments with fewer than 10 participants. Or, instead of counting from the last place, we could count from, for example, 15-20, creating a single point balance for all tournaments (this would also give participants in less prestigious tournaments who finished in last place more points than those who played in large tournaments but didn't make it to the top. This would increase attendance!).

That's a very positive suggestion. But in that case, we definitely need to have a General ranking for those who have played 3/5 of the total number of tournaments. Otherwise, with the N/2 bonus, the winner of a tournament with 30-40 participants will easily outperform those who have participated in 3-4 tournaments with fewer than 10 participants. Or, instead of counting from the last place, we could count from, say, 15-20, creating a unified point balance for all tournaments (this would also give participants in less prestigious tournaments who finished in last place more points than those who played in larger tournaments but didn't make it to the top. This would increase participation by 8)).

So, if the issue remains only in balancing the points you mentioned, the discussion will be more dynamic, and we'll sort everything out before the tournament starts. These are details – they certainly need to be adjusted, and there's nothing complicated about it... I'm trying to propose a concept, not a final calculation... We can increase the points not by N/2, but by a fixed number – for example, by 5. But definitely from the bottom up, to reflect the number of participants. The essence won't change much...

The goal is to eliminate the very sharp discrepancies that will become apparent later.

If the proposed concept doesn't raise any objections in principle, then everything else will be developed very quickly. :)

I don't really like the idea of tying points to participants... For example, I had 10 people playing Heroes of Might and Magic II, while the VDV (airborne troops) had 20+ people playing Heroes of Might and Magic III... I think this is a perfectly reasonable reflection of the games' popularity, and it's not worth burdening the organizers with this problem, as they already have enough on their plate.

I don't really like the idea of tying points to participants... For example, I had 10 people playing Heroes 2, and VDV had 20+ people playing Heroes 3... I think this is a perfectly reasonable reflection of the games' popularity, and it's not worth burdening the organizers with this problem, as they already have enough on their plate.

Participants in Heroes 2 compete for their own small cup and are equal among themselves... The winner in Heroes 2 will still have the highest number of points among all participants in their version, even if that number is several times less than what the winner in Heroes 3 scores...

Please don't compare them... These are not comparable things...

And when summarizing the results in the overall standings, this can be taken into account differently. For example, a coefficient can be introduced for the calculation. This will only need to be done once and will not be difficult.

Or maybe I'm not seeing the ultimate goal of the tournament yet? Do you want to see during the tournament who is stronger - a player who plays Heroes 1 + Heroes 3 compared to a player who only plays Heroes 4??? But that's impossible... I don't think you'll argue with that... You probably wouldn't be interested in seeing yourself in the tournament table among participants who haven't played your map and don't even know your version of the game...

Or, for example, if we consider the calculation mechanism from top to bottom, then the last place in any tournament in Heroes 2 will be awarded more points than the last place in Heroes 3... Again, due to the different number of participants. How do you compare?

Thank you to everyone who continues to discuss this topic. I ask that everyone who proposes their own rating systems present them in the form of a table so that it is easy to understand how many points a participant can score according to each rating.

I think that each installment of Heroes should have its own separate rating, distinct from the other installments. Comparing players across different Heroes games within the same tournament will inevitably be inaccurate. The only exception could be a separate tournament specifically designed to identify the most versatile player, much like the "Van Caningham" tournament, which was conceived with that goal in mind.

2 AmberSoler, I ask you to please pay attention to my posts №8 and 12. You repeated them word for word (I am also against the General rating as an indicator of a player's strength). That is why a limit of 3/5 of the number of Tournaments was proposed. This, as already mentioned, will reveal not so much the strongest in any single version, but rather well-rounded players (this is essentially the goal of the "Van Canigem" tournament, as I understand it).

If we limit the number of points to a certain place, say 10th, then the 12th participant in 2 and the 38th in 3 will have an equal number of points, and the winners will also have an equal number of points, unlike your proposal.

My proposal is post №12 (/topic/at/58925/)
I will provide the point calculation table a bit later at AlexeyPank`s request.

Added after 5 minutes

Place<>	Number of points<>
1<>	25<>
2<>	20<>
3<>	16<>
4<>	13<>
5<>	11<>
6<>	10<>
7<>	9<>
8<>	8<>
9<>	7<>
10<>	6<>
11<>	5<>
12<>	4<>
13<>	3<>
14<>	2<>
15<>	1<>

7. Separately, I want to comment on the concept of rating decay by 1/3... Have you tried calculating what will happen to your 50 rating points earned in the first stage after 10 stages? Only 0.86 points will remain... :D Why play the first stage for such a pittance? You could just play the last two and that's it...

To be honest, even I didn't get what you calculated here :) Do you mean that after playing one stage, the player will no longer participate in any of the stages?
If a player only plays HoM1 and doesn't play any others, then for the same tournament, their rating doesn't need to decay in the stages where they don't participate. However, if they skip one of the subsequent HoM1 stages, then the General rating must definitely decay. That is why I wrote that the rating decay can be sorted out as the tournament progresses.

To be honest, even I didn't understand what you calculated here :) Do you mean that after playing one stage, a player will no longer participate in any of the other stages?
If a player only plays Heroes 1 and doesn't play any other versions, then on the King's Bounty tournament, you don't need to calculate their rating in the stages they don't participate in. However, if they skip one of the following stages in Heroes 1, then you definitely need to calculate their Overall rating. That's why I wrote that we can figure out the rating calculation as the tournament progresses.

Or maybe we shouldn't rate anyone at all, especially if the 3/5 rule is adopted. Because a person might play for time, but not play the arena... and then you tell them, "No, we're going to lower your rating now" 8)). Let each person decide which stages to skip and which to play. We don't need to scare off participants with unnecessary complications.

To be honest, even I didn't quite understand what you calculated here. Do you mean that after playing one stage, a player will no longer participate in any of the other stages?

Quoting:

4. We obtain a new rating for each player, which we will call R.new (new rating).

R.new = (2/3)*R.old + R.tur

This formula indicates that the old rating R.old of each player decreases by one-third, even for those who did not play in this tournament but who previously participated in the tournament. For new players, R.old=0.
Therefore, for new players, the new rating will consist solely of R.tur.
For old players and those who played in this tournament, the new rating will consist of R.tur and R.old, reduced by one-third.

From this, I understand that the rating decay occurs regardless of whether the player participated in the current tournament or not. After each tournament, you simply reduce the old rating by one-third and add the new one. Logically, this means that my result in the first tournament (let's say 50 points) will be subject to the decay procedure 10 times by the end of the eleventh tournament. This means that only 0.86 will remain of the 50. From the result of the second tournament (let's say 50 again), 1.3 will remain. From the result of the third tournament, 1.95 points will remain... I won't continue, the point should be clear.

Let's say I'm the strongest player and played 12 tournaments in a year, finishing them with a victory - 50 points each. The total is 600. After the decay, it will be 148.84. You started playing in the middle of the year and also received 50 points for six tournaments, i.e., you shared the prize with me. Your total is 300, after applying the decay procedure, it is 136.8. (although you played twice as few). Let's say I missed the last tournament. As a result, I immediately lost 50 points - 98.8 points (!!!)... And if I had missed the first tournament, my result at the end of the year would hardly have changed and would be 148.26. Is that clear?

I am forced to repeat my words:
"Why play the first stage for such a small gain? You can just play the last two and that's it..."

And I can only hope not for winning the tournament, but for you to miss it... :)

Let's still use numbers for the example, and not theory... The "weight" of the result in the final formula should not depend on the time that has passed since it was achieved... Please, test this mechanism in Excel, and you will understand that by the end of the year, the "weight" of a victory achieved many months ago will dissolve even against the background of weak results from newly joined players... Imbalance.

If you want to see only the player's strength at a specific current point in time, and it doesn't matter how he played a few months ago, then what is the point of calculating an annual rating?

Do you really need to play several tournaments to be sure of this flaw?
Or maybe I'm not calculating it correctly?