Skip to content

Posts from Rules for organizing, conducting, and participating in online Heroes of Might and Magic III tournaments.

No ratings yet — be the first to rate!
In reply to Oelm
User Avatar
#76
Auto-translated
Oelm
Well, what reasonable suggestions could there be? It's a classic situation. One player won against everyone. And three players behind him have the same number of points – each won against the next one. What to do? Either a random draw, a rematch, or a tiebreaker. + of course, you can look at their rating or who won in how many days, but this doesn't seem entirely objective.

And it certainly wouldn't hurt anyone to learn how to play tiebreakers better.

The table should compare the discipline in which this tournament is being played!!!

The first comparison should definitely be based on head-to-head matches between the opponents, but if there are three players and each has won against each other, something needs to be figured out, but it shouldn't involve placing positions in the table based on the seasonal rating, as the season and tournaments are different things, like different disciplines.

There is a way to compare these three: who finished the match earlier in real time – this is also an incentive to play on time and not drag out the tournament... but, as I was told, there might be opponents who one player wouldn't be interested in playing against, and it would be as if they were deliberately delaying the match, and this can be avoided (probably) by providing logs about the request to play or about their refusal (anyone who wants to play in the tournament and registers for it should approach this responsibly)
User Avatar
#77
Auto-translated
Here's an idea that came to me. Why not create a Player Tournament Rating? Devise an algorithm that translates tournament achievements into numbers and compile a table.

It would be convenient for tournament organizers: 1) to conduct seeding (after all, we usually try to roughly "sort into baskets" and do it subjectively), 2) to resolve tiebreaker issues when points are equal. And besides, it's simply very interesting.

We divide tournaments by "weight" – significance (for example, Best, PL on the Portal, RP at the Bay – tournaments among the top players), by the number of players...
For participation in any tournament (we only consider НОММ3 online), a base number of points is awarded (if at least one game was played). And bonuses for the number of players, tournament quality, placement... We'll come up with a formula, and I can write a little program.

Over the entire existence of ХВ, there haven't been many tournaments – all are in the History, so it won't take long to calculate. Those who have played successfully on ХВ automatically get into this rating. And any player who has participated in tournaments on other resources leaves the info – resource, tournament, link, number of participants, their achievements, etc. And the rating organizer evaluates it and adds it to the table.
In reply to Pit for LM
User Avatar
#78
Auto-translated
Pit for LM
Here's an idea I had. Shouldn't we create a Tournament Player Rating?

Eureka! Here's the solution to the problem!
But maybe we should only start calculating it for this year? Otherwise, the question will arise as to how far back we should dig on other resources?
РП (Реальное папство!)
*Автосейвы не беру и не даю.

Скачать WT (Чат, античит, одновременный ход) http://www.heroeswt.net/download.php

Награда Лучшему игроку сезона Осень 2009 за Оплот
In reply to Forrest
User Avatar
#79
Auto-translated
Will the tournament ranking fairly determine the placement? Is it better than the seasonal one for many?
The first comparison should definitely be based on head-to-head matches between opponents, but if there are three players and each has won against each other, something needs to be devised.
It's clear that a head-to-head record is the most normal criterion. But when there are three players, something else needs to be considered. And the criterion should be unambiguous, not based on 100 points.

Tournament Committee 3 is also not an option for two reasons.
1. Tournament Committee 3 also takes time. And the problem is often scheduling. A match can last for several hours.
2. A similar situation may arise where a player has one win and one loss. What then? Play again?
In reply to Pit for LM
User Avatar
#80
Auto-translated
Pit for LM
Here's an idea that occurred to me. Shouldn't we create a Player Rating for Tournaments? Find an algorithm that translates tournament results into numbers and create a table.

It would be convenient for tournament organizers to 1) conduct the draw (we usually try to "sort players into tiers" and do it subjectively), and 2) resolve the issue of qualification when the number of points is equal. And it would also be very interesting.

We divide tournaments by "weight" - their importance (for example, Best, PL on the Portal, RP on Bay - tournaments among the best), by the number of players...
For participating in any tournament (we only consider HoMM3 online), a certain basic number of points is awarded (if at least one game was played). And bonuses are added for the number of players, the quality of the tournament, and the place taken... We'll come up with a formula; I can write a program.

Over the time that Hv tournaments have existed, there haven't been many - everything is in the History, so it won't take too long to calculate. Those who have successfully played in Hv tournaments will automatically be included in this rating. And any player who has participated in tournaments on other resources can provide information - the resource, the tournament, a link, the number of participants, their results, etc. And the rating organizer will evaluate and add it to the table.


I'm all for it. All that's left is to find a volunteer with a calculator;)
Даже если вас съели - у вас два выхода:smile19:

Зима 2011-12.
In reply to Dirty_Player
User Avatar
#81
Auto-translated
Dirty_Player
Will the tournament rating provide a fair determination of rankings? Is it better than the seasonal one?
There's nothing to compare with the seasonal rating. It will take into account the results of tournaments and tournament activity, and on all resources. In the seasonal rating, even our tournament players from World of Xeen don't all participate, not to mention those who have just come from other resources...
Of course, such a rating cannot be an absolute measure of skill. But it will be more objective. And choosing one out of several with an equal number of points is unfair in any case. But in the end, we're not playing for a cow, and it's not just one tournament...

Chronologically, it can be limited, but I think it's better to give a longer period. For example, starting from the founding of HeroesWorld.
In reply to Forrest
User Avatar
#82
Auto-translated
Forrest
Is it only possible to start counting from this year? Otherwise, the question arises as to how deeply to delve into other resources?
Not that many tournaments have taken place. Many have, but not as many as Best, PL, and RP. On HW, it's the first, on HP, it's the second, and there were 4 on Bay. So, a total of 7.
Although, it seems to me that each resource should primarily value its own "bests" and then look at the other two.
You could take the 7 "bests," sum up the points.
Discord  5884
In reply to Pit for LM
User Avatar
#83
Auto-translated
Pit for LM
Here's an idea that came to me. Shouldn't we create a Tournament Player Rating?

An enticing idea... :D

It would be great to find some objectivity in the formulas....

But it's clear that this rating will be much more highly regarded than the seasonal one!
Сейчас балуюсь этим https://vcmi.eu/download/
За то что нейтралы не бегают за единичкой, для меня это уже стоит того
In reply to Lider1
User Avatar
#84
Auto-translated
So what? Does that mean a player with a higher rating should have an advantage? For example, there could be a situation where two players in a group only need to win 3 games (in a group of 5), while one player is obligated to win all 4. Furthermore, there could be a situation where two leading players, by agreement, play a match, and both leave the group. What would prevent them from making a deal?
Земля нуля.
Все в ней нулево и кукольно.
In reply to Oelm
User Avatar
#85
Auto-translated
Oelm
Additionally, a situation is possible where two rating leaders play a match by arrangement, with both advancing from the group. So what would prevent them from colluding?
Such a situation is possible and is already happening perfectly without any rating system.

The main questions are - 1) the formula. It should take into account the tournament difficulty + number of participants + final placement + ....?
2) I can write a program based on the formula.
3) we need to find someone who will maintain this rating.
4) I can calculate it for HeroesWorld. For HeroesPortal, Neo probably has tournament information. For Bay, maybe someone will volunteer? Or, if anyone from other resources wants to be included in the rating, let them provide the info themselves. One can more or less calculate their own victories.
In reply to Pit for LM
User Avatar
#86
Auto-translated
Such a situation is possible and is already being implemented effectively without any ranking system.

That's not true! Let's look at three possible scenarios: 1) determining places based on ranking, 2) determining places by drawing lots (in case of a tie), 3) determining places using tiebreakers.

For example: There are 5 people in a group. 2 obvious underdogs. Now, what will happen if, in the first match, one of the three contenders for advancing from the group with the lowest ranking loses to one of the other two?

In the second and third ways of resolving the problem, everyone remains motivated to play, and the winning player even has an incentive to win all remaining matches.

And only in the first case (and I can't think of a better name for it than a blatant ranking system) does the winning player have a motivation to lose to another contender for advancement so that both can advance from the group. Also, the player with the lowest ranking loses almost all chances of continuing the competition. And there are still 3 rounds ahead! Even in an Olympic system with best-of-two matches, there is always a chance for a mistake. How else can such a system be described?
Земля нуля.
Все в ней нулево и кукольно.
In reply to Pit for LM
User Avatar
#87
Auto-translated
Pit for LM
Here's an idea that came to me. Shouldn't we create a Player Rating for Tournaments? Find an algorithm that translates tournament successes into numbers and create a table.
Great idea.
we only consider HOMM3 online
What did club tournaments do wrong? They are the same qualifying tournaments that are currently running.
Participating in any tournament gives a certain basic number of points.
It shouldn't. Because it encourages point farming. Like, you sign up for all the tournaments, lose in the first round, and do the points still get added?

In my opinion, it's worth giving points starting from the quarterfinals (and even then, only in major tournaments), and not many, with the main bonus going to the winners.
Maybe it can only start counting from this year? Otherwise, the question arises of how far back to dig on other resources?
I think it's better to go back three years, with coefficients for the time elapsed.

For example, 0.97^N, where N is the number of months since the end of the tournament. Tournaments that are one year old will have a coefficient of 0.7, and tournaments that are three years old will have a coefficient of 0.33.
We divide tournaments by "weight" - importance.
I suggest the following formula:

Tournament weight (N) - the number of points the winner receives:
1) N=100 for top tournaments, which you qualify for through qualifying rounds (RP, PL, Best, future CIS championship),
2) N=75 for regular tournaments with a strong lineup (for example, seasonal cups, St. Petersburg championship, qualifying rounds for top tournaments, etc.),
3) N=50 for fan tournaments, various irregular tournaments, etc.

Points for performance are calculated as "tournament weight (N) * performance success coefficient (K)".

1. Winner, K=1

2. For round-robin tournaments like RP, K = 1 - 0.9 - 0.8 - 0.7 - 0.6 - 0.5 (exactly 6 places that retain their spot for the next season)

3. For cup tournaments with an Olympic system (or group stage + Olympic system)
K = 1 - 0.75 - 0.5 - 0.5 (that is, points are awarded starting from the semi-finals)
In reply to Pit for LM
User Avatar
#88
Auto-translated
Pit for LM
Here's an idea I had. Shouldn't we create a Tournament Player Rating? Find an algorithm that translates tournament successes into numbers and create a table.
I'm against this idea for several reasons:
1. Why should only a player's tournament experience be considered when determining their skill? Many newcomers don't have any. Besides, it's not a fact that a person will play better in a tournament than, for example, in a seasonal game. I, for example, have the same mindset. Often, the more you prepare, the worse it turns out. For me, tournaments are just games with specific rules, devised by the organizer.
2. How can tournaments be divided by "weight" based on classes? This has nothing to do with skill, and some organizers might be offended when their tournament is considered insignificant.
3. The place taken in a tournament is not an indicator. It may be the result of glitches or depend on the level of opponents with whom one was lucky enough to participate in the group.
В дикім полі
User Avatar
#89
Auto-translated
Andrey, it seems like you just want to argue. I disagree with every point. Shit happens :D
VIK suggested an interesting scheme. It's been a long time coming to create a general rating. Dream's attempt was good, but the PR was weak.
In reply to KobRa_RR
User Avatar
#90
Auto-translated
Rankings in general are a creation of the devil, designed to oppress unrecognized geniuses within the old guard. It's a caste system; one wrong step and you're out.

Statistics

Welcome our newest member: Emil