OFB-12 — Tournament History online

No ratings yet — be the first to rate!

Tournament Ended

Overview Participants Reports

Announcement

Sanin

7/13/2018 11:15 AM UTC

Auto-translated

«OFB - 12»

Registration is open for the qualifying tournament for the most prestigious online HoMM 3 tournament on heroesworld.ru – «Final Battle»

Organizational Matters:

1. Tournament host – Sanin (ICQ - 423164784); judges – LuckyF (ICQ - 681350680), Dave_HUN (ICQ - 635192642), Dirty Player (ICQ - 408722873).
2. The tournament is official, regular, and conducted according to the Rules for organizing, conducting, and participating in online HOMM3 tournaments.
3. All tournament games are included in the HoMM3 HW Player Rating.
4. After the start of tournament games, the tournament regulations will not change.
5. Disputed matters, if any arise, are discussed outside the forum. To resolve, contact the host or one of the tournament judges on ICQ and provide daily save files, ICQ logs, and other necessary information.
6. Tournament games must be played within the established deadline. Violation of game deadlines may, at the discretion of the host and tournament judges, result in penalties up to and including disqualification from the tournament.

Requirements for Participants:

- Players with unconfirmed reports will not be accepted into the tournament.
- Applications from players recently registered on the forum will be reviewed on a case-by-case basis.
- The organizers reserve the right to refuse registration to players who have repeatedly withdrawn or been disqualified from other HeroesWorld tournaments.

Regulations

Rules:

1. Game Version: The tournament is played on the SoD version + HD mod.
2. Tournament games use the HW_rulez 1.40 and HW_rulez 1.38.2 patches.
HW_rulez 1.38.2 patch is used for playing on the h3dm1 template and for games in "Battle of Mages" mode.
HW_rulez 1.40 patch is used for all other templates and in all rounds except the second.
3. The tournament is conducted according to the rules of online HoMM 3 games.
Clarification to the rules: on all templates, the "Mage Guild Restructure" option must be enabled, while when playing with the HW_rulez 1.40 patch, the "Library Restructure" option is disabled. When playing with the HW_rulez 1.38.2 patch, there are no restrictions on Library Restructure.
4. Restarts total 3 days (111+111+111 or 111+112 or 113), for the h3dm1 template – one day (111).
5. Templates:
Panic, h3dm1, Diamond, Mini-Nostalgia, Jebus Cross, 6lm10a, Spider.
The template is chosen by standard drawing of lots, the order of the draw is determined by a coin toss.
6. The tournament game regulations will change depending on the round.
First Round: standard bidding, standard rules.
Second Round: standard bidding, games are played with the "Battle of Mages" option enabled (more about the option).
Third Round: standard bidding, standard rules.
Fourth Round: standard bidding, but opponents choose each other's starting heroes (see the 2010 tournament for an example).
Details. For all templates except h3dm1: after generating a pair of towns in the lobby, standard bidding takes place. The player who wins the bidding chooses their town first, after which opponents assign each other starting heroes.
For the h3dm1 template: a random town is generated in the lobby, after which opponents bid for the right to assign the opponent's hero first.
Fifth Round: standard bidding, standard rules.
Sixth Round: 200% fullrandom on all templates. Color is determined by a coin toss and changes after each subjective restart.
Seventh Round: standard bidding, standard rules.
Eighth Round: standard bidding, games are played without rules on all templates.
Ninth Round: standard bidding, standard rules.
7. Color selection: bidding is conducted for the right to choose color (exception - sixth round).

Tournament Format:

1. Registration from July 13 to July 21 (dates may be adjusted, follow updates).
2. Start of tournament games – July 23, 2018. End of tournament – November 26, 2018.
3. The tournament is conducted using the Swiss system in 9 rounds. Each round lasts 14 days.
4. Within the round deadline, a report of the completed round game must be submitted in the tournament thread to record the tournament result. Failure to comply with this rule requirement will result in either a technical result or a 0-0 score being assigned for the match.
5. Round deadlines:
Round 1: July 23 – August 5
Round 2: August 6 – August 19
Round 3: August 20 – September 2
Round 4: September 3 – September 16
Round 5: September 17 – September 30
Round 6: October 1 – October 14
Round 7: October 15 – October 28
Round 8: October 29 – November 11
Round 9: November 12 – November 25
6. A win gives 3 points, a draw gives 1 point, a loss gives 0 points.
Places are determined based on points earned.
In case of a tie in points, the following tiebreakers apply in order of priority:
1) Buchholz coefficient;
2) Number of wins;
3) Head-to-head result;
If the tiebreakers do not allow for the distribution of places among participants who may qualify for the final stage of the tournament, tie-breaks will be held.
Tie-break rules: drawing lots for templates, random pair of towns + bidding for town choice and color choice.
7. 16 players participate in the «Final Battle–12»:
FB-11 Winner LuckyF,
FB-11 2nd place Dave_HUN,
FB-11 3rd place FireToSky,
Winner or finalist of the «Only XL-5» tournament,
Winner or finalist of the «Favorites, Into Battle-10» tournament,
Winner or finalist of the «Wowa Memorial» tournament,
Winner of the «Новичок - 2018. Winter» tournament, provided they have a rating of 1750+,
The top 9 places (guaranteed) – from the OFB-12 tournament.

Additions:
If the winner of the «Новичок - 2018. Winter» tournament refuses to participate in FB-12, their slot will be taken by the finalist of that tournament;
If the winner (finalist) of the «Новичок - 2018. Winter» tournament does not have a rating of 1750+ at the start of the FB-12 tournament, their slot will be taken by the player who finished 10th or lower in OFB-12;
If LuckyF, Dave_HUN or FireToSky refuse to participate in FB-12, their slot will be taken by a player from the OFB-12 tournament;
If LuckyF, Dave_HUN or FireToSky win the «Only XL-5», «Favorites, Into Battle-10» or «Wowa Memorial» tournaments, the right to participate in FB-12 passes to the finalists of those tournaments.
If the winners of the «Only XL-5», «Favorites, Into Battle-10» or «Wowa Memorial» tournaments are among the players who take a qualifying place in OFB-12 or refuse to participate in the FB-12 tournament, the right to participate in FB-12 passes to the finalists of those tournaments.
If the finalists of the «Only XL-5», «Favorites, Into Battle-10» or «Wowa Memorial» tournaments are also among the players who take a qualifying place in OFB-12, or refuse to participate in the FB-12 tournament, their slot will be taken by the player who finished 10th or lower in OFB-12.

ICQ 423164784

In reply to Xocenk

Dirty_Player

12/2/2018 1:32 PM UTC

Auto-translated

Xocenk

But why bother figuring out how the program calculates it (it would be good for the future, of course)? According to the regulations, there can't be any 0.5:

You can just calculate the total score, but the program doesn't give these bonuses for no reason. And why is that? For example, in the first round, Amilu played against an opponent with the nickname Johnny the bony, who lost two games and left the tournament without finishing it. Therefore, his coefficient is 0! Which automatically eliminates him from the candidates to advance, in almost any distribution. Is he to blame for encountering such an opponent in the first round? I don't think so. It is precisely because of such situations that the system gives additional points for unplayed games. However, other questions arise. Why +0.5? And why do different versions of the program calculate differently for Tvarina? Who to believe? :smile39:

Xocenk

and the coefficient provided by the program is decisive.

I recommend including exactly this wording in the next official rules. What the program shows should be considered the truth. What to do now - we will think about it.

Added 3 minutes ago

Georg_

Perhaps the confusion arose due to this nuance:

"Appendix to the FIDE Tournament Rules
Rules for the application of additional indicators
Approved by the General Assembly in 1998
1. Accounting for unplayed games
To eliminate the undesirable influence of unplayed games on the distribution of places, these games should be counted as follows.
Regardless of the result of the unplayed game (win in the case when the player receives a point due to being "exempt from play," win or loss by default, or the game was not played because the player withdrew or was absent for one or more rounds), this result should be considered when determining the additional indicator as a draw of the player against himself."

Yes, that's exactly what I was talking about. The problem is that I completely forgot about the phrase highlighted in bold. Then I will try to figure out Tvarina's coefficient again.

Georg_

That is, when calculating the Buchholz score, the program treats technical wins or unplayed games of withdrawn players as draws with oneself. However, it is still not clear where these 0.5 come from...

I only have one hypothesis about +0.5. The program was originally designed for a 1-0.5-0 point system, and a draw gives +0.5. In theory, with a 3-1-0 system, it should give +1, but it still gives +0.5. Which looks like a clear bug.

In reply to Dirty_Player

kamukag3e

12/2/2018 1:38 PM UTC

Auto-translated

Xocenk

Otherwise, it contradicts the tournament regulations: the program-generated coefficient, not the Buchholz coefficient, should be the determining factor.

Dirty_Player

That's exactly the wording I recommend including in the next official rules. Whatever the program shows should be considered the truth.

If the tournament rules are made opaque, and the reference to the program in the rules is also an opaque rule, then many will think so as well;) :

Dirty_Player

Perhaps, it's just another conspiracy against Twaryna in the official rules :D

艰难获胜 herolwy

In reply to kamukag3e

Grando

12/2/2018 2:04 PM UTC

Auto-translated

the idea of blindly believing in a "program" is terribly bad. Programs are only needed to the extent that they do exactly what a person needs them to do. As for the next tournament, it would be good to abolish Buchholz as a deciding criterion and introduce play-off matches.

https://www.twitch.tv/grando__

In reply to Grando

Dirty_Player

12/2/2018 2:19 PM UTC

Auto-translated

Al Feyn

The idea of blindly believing in a "program" is terribly bad.

Then why, at all levels of chess tournaments, do they use this "terribly bad idea"?

In reply to Dirty_Player

kamukag3e

12/2/2018 2:33 PM UTC

Auto-translated

Dirty_Player

Then why is this "terrible idea" used at all levels of chess tournaments?

Using a program in itself is not a bad idea; the main thing is that everyone understands how this program should calculate the Buchholz score, and this should be specified in the rules. And then everything will be fine.

艰难获胜 herolwy

In reply to kamukag3e

Dirty_Player

12/2/2018 2:43 PM UTC

Auto-translated

kamukag3e

Using the program itself isn't a bad idea; the main thing is that everyone understands how the program should calculate the ratings, and this needs to be specified in the rules. And then everything will be fine.

I'll repeat once more. The regulations for chess tournaments state: "The draw and calculation of coefficients are performed in the Swiss Master x.x program," and that's it! That's all! Those who want to can find out how this program calculates things. If you come to participate in a tournament, it means everything suits you. If it doesn't suit you, don't come. No one coddles anyone.

In reply to Dirty_Player

Ihromant

12/2/2018 3:33 PM UTC

Auto-translated

There used to be a chess player named Robert Fischer. At first, he complained about Soviet chess players arranging draws and the way points were calculated. And then he buckled down and became the world champion. :smile54:

Мой любимьій размер - тройка. Будь то конса, утопа или сиськи
ICQ:5676388

In reply to Dirty_Player

Grando

12/2/2018 3:48 PM UTC

Auto-translated

Dirty_Player

Then why is this "terrible idea" used at all levels of chess tournaments?

I see two possible answers to this question:
1. Because everything has been tested, they use 1.0-0.5-0.0 scoring, and they know that the program calculates correctly for this case?
2. Because the previous organizer wrote such rules, and this time they simply copied them.

The second option is more likely because no one cares.

https://www.twitch.tv/grando__

In reply to Ihromant

kamukag3e

12/2/2018 4:38 PM UTC

Auto-translated

Ihromant

There used to be a chess player named Robert Fischer. At first, he complained about Soviet chess players arranging draws and the way points were calculated. And then he buckled down and became world champion. :smile54:

Well, since we're still arguing, it means the time to buckle down hasn't come yet:D

艰难获胜 herolwy

In reply to kamukag3e

MicroElf

12/2/2018 7:24 PM UTC

Auto-translated

The thing is, the program isn't that simple, and it has its own view on free wins. It considers all unplayed games (whether free wins or free losses) as +0.5 (possibly a bug, as +0.5 fits well into the 1-0.5-0 point system, but not so well into the 3-1-0 system). Let's try to estimate for our situation. We subtract the points for free wins from the total and add the points for unplayed games.

I haven't dealt with chess or the accountant's calculations for a long time. But as far as I remember, a free win or free loss is important. In the case of a free win, the coefficient is considered to be 0.5 (compared to 1 for a win and 0.5 for a draw), and in the case of a free loss, it is 0.

With a 3-1-0 point system, logically, a free win should be given 1 point (like a draw), and not 0.5. But a free loss should still be 0. I haven't dealt with this system, so this is just based on logic; I don't know what the correct approach is.

I can't calculate my own theory because there is no complete table of results (why?), only the top part.

In general, I would like to ask the organizer to publish the full tournament results. It might be important for me to know what place I took :D

ICQ 289320583

In reply to Dirty_Player

kamukag3e

12/3/2018 12:43 PM UTC

Auto-translated

Dirty_Player

Let me repeat myself. The regulations for chess tournaments state: "The draw and calculation of coefficients are performed in the Swiss Master x.x program," and that's it! That's all! If someone wants to know how this program calculates things, let them figure it out. If you come to participate in a tournament, it means you agree with everything. If you don't agree, don't come. No one is going to coddle anyone.

I'm not arguing with this approach; just include it in the regulations, and everything will be fine. No one will have any questions.
Currently, the regulations only state that in case of a tie in points, additional criteria apply, and the first criterion is the Buchholz coefficient. How it is calculated and which program is used for this tournament is not specified in the regulations, so we need to Google it.
For example, Wikipedia provides the following definition: "The Buchholz coefficient of a particular participant is the sum of the scores of all opponents with whom the participant played, regardless of the results of the matches between them. The idea is that a participant who played against stronger opponents (opponents who scored more points in total) is awarded a higher final ranking. All unplayed matches (+/-) are conditionally considered draws."
In this definition, there is no mention of half a point. For example, with twaryna's Buchholz coefficient, it doesn't matter whether he actually played against Dust_HL or not; the result of the game is recorded – it's a technical loss for twaryna, so Dust_HL's points should be included in the calculation of twaryna's Buchholz coefficient.
In my opinion, it's all simple, and there's no need to overthink how the program calculates half a point when it can be calculated manually.

艰难获胜 herolwy

In reply to kamukag3e

Dirty_Player

12/3/2018 12:56 PM UTC

Auto-translated

kamukag3e

The rules don't specify how it's calculated or which program is used for this tournament, so we'll have to look it up.

If you open not just the first link on Wikipedia, but at least the second, third, or fourth, you'll read something like this:

Note: Regarding the calculation of coefficients when a round is skipped or an opponent is absent.

As a rule, the following approach is used.

All unplayed matches (skipped round, absence, or lack of an opponent, "plus") are treated as draws for the purpose of calculating coefficients:

in case of an opponent's absence, the match is treated as a draw
in case of no opponent (in a Swiss system with an odd number of participants), the match is considered to have been played to a draw against oneself
in case of skipping a round at the beginning of the tournament (due to being late) or at the end of the tournament (due to withdrawal), the matches are considered to have been played to a draw against oneself

Regardless, we will follow this. I don't see any reason to think otherwise. Why? I already explained it above.

In general, it took a long time to figure out these programs; I spent almost half of Sunday on it, but we finally figured out the coefficients. Soon, Sanin will post the results.

In reply to Dirty_Player

Sanin

12/3/2018 4:01 PM UTC

Auto-translated

Let's start with the most important thing that everyone is interested in: Redwhait and twaryna will compete for the right to replace the winner of the Novice tournament, kolianxxxxxx.

Second: Dima Surmay and Yostairs advance to the Final Battle, as announced at the beginning.

Unfortunately, Georg_ and Amiloo will not advance to the Final Battle (unless someone withdraws, but that is unlikely).

Now, let's go into detail about what happened, what the possible solutions were, and why we settled on this one.

Initially, the results generated by my program were published. Then it was noticed that the program calculated the scores in a peculiar and not very obvious way. What's worse, it turned out that Dirty_Player's program gives different results for twaryna. At this point, we had to either figure out the reason or simply schedule tiebreaker matches for all six players. This option seemed unfair to the players who had already qualified, so we immediately abandoned the idea of tiebreaker matches for all six.

What did we have? We had four ways to calculate the Buchholz score: the simplest one (from Kamikadze) - which was not seriously considered. Two program calculations, in which 0.5 is awarded for an unplayed game in the Buchholz calculation - the calculation provided by my program and the calculation provided by Dirty's program. And finally, a manual calculation, in which 1 point is awarded for a draw, not 0.5. The latter option was also quickly rejected because it gave the same top three players as Dirty's program, or the same top three as my program - the whole question is how to calculate the Buchholz score for twaryna.

So, my program calculated twaryna's Buchholz score as 125, while Dirty's program gave him 132.5. This would not have been a big deal, but the fact that the Buchholz score is calculated differently affects who ultimately takes kolianxxxxxx's place. As some have already noticed, the whole controversy arose from the first-round match between Dust_HL and twaryna, in which I recorded a loss for twaryna due to a no-show. In general, the logic is as follows: for a 0-0 draw, I always recorded a loss for the player who didn't show up; for technical wins/losses, I recorded a loss for the player who didn't show up if the game had not started. If the game had started, I recorded a loss by default, and the program considered this a game result with all the consequences for the Buchholz score. So: my program considered that the match took place, and the result of Dust was used for the Buchholz calculation; Dirty's program broke the match, and the result of twaryna himself was used for the Buchholz calculation. Dirty_Player can explain this issue and the program logic in more detail.

And then we decided to compare the program versions, and they, of course, turned out to be different. That is, at some point, the program logic changed, and we ended up with these problems in our tournament. Some will definitely ask, so I'll say it right away: the tournament was held on a newer version, and it is the newer version that puts Redwhait ahead, while the older version puts twaryna ahead. The places of Dima Surmay and Yostairs do not change depending on the program versions: they are ahead of both Redwhait and twaryna.

And then the choice is very interesting: on the one hand, it is more correct to take the newer version - and Redwhait wins; on the other hand, not to get a prize because the program was updated - that's just ridiculous. Even a pre-arranged draw is more sporting than this nonsense. And we decided that it is better to give the players the opportunity to compete for a place in the Final Battle in a face-to-face match than to decide which version is more correct.

Since this game is part of the tournament, the rules will be the same as in the tournament. The same pool of templates, library randomization disabled by default (note: this no longer requires rule set 1.40), standard rules, and timers. By default, games are played in a bo1 format; by agreement, you can play in a bo3 format. Post the games in this thread; the deadline is preferably by the end of the week.

ICQ 423164784

In reply to Sanin

Dirty_Player

12/3/2018 4:55 PM UTC

Auto-translated

Let me explain what the fuss was about. The screenshot shows the difference between the program versions. In the old version, the program would sometimes disconnect and, when calculating the coefficient, would declare a draw against itself. At the same time, players could meet each other again. In the latest version, the program doesn't disconnect, but instead declares a conditional draw for the players and calculates based on that. As a result, the score is noticeably lower. Most likely, this is correct, based on the algorithm:

if an opponent doesn't show up, the match is considered a conditional draw
in the absence of an opponent (with an odd number of participants in the Swiss system), the match is considered to be played to a draw against oneself
if a round is missed at the beginning of the tournament (due to being late) or at the end of the tournament (due to withdrawal), the matches are considered to be played to a draw against oneself

But still, we decided to give Tvarina a chance to prove himself and disprove the conspiracy theory))))

P.S.: +0.5 or +1 difference doesn't matter, the order of ranking is the same, I checked. The overall table is also attached.

Attachments 2

2 files attached • Total size: 120.3 KB

In reply to Dirty_Player

Dima Surmay

12/3/2018 5:07 PM UTC

Auto-translated

When you say that the program assigns a conditional draw and awards points, are these points only for the Buchholz coefficient? Or does it also add points for the draw to the main points earned?