Carry on up the Armageddon – reader feedback

Feedback by our readers

Peter Rajcsanyi, Press Director of WWCC Nalchik
Congratulations – it has been the best ten minute of the day for me so far, reading John Saunder's emotions and thoughts of the chess Armageddon or Sudden Death game.

Beyond the Zatonskih-Krush event, no doubt that the recent Socko-Foisor tie break game in the Women's World Chess Championship brought the issue into the forefront among chess lovers. There are some who liked this type of solution for deciding the outcome of the match of two players and there are some who hated it from the very beginning of its introduction. Anyhow, it is in the rules applied for many years in different tournaments, including World Cup, World Championship, USA Championship, etc. So far it has been applied without any big trouble.

Now, in the Women`s World Chess Championship, the arbiters made the decision based on the general rules of chess (that was also included among the paragraphs of the WWCC rules, i.e. Article 9.6.) and called the game draw, maintaining the view that one cannot achieve victory if there are only kings and knights (one for each player) on the board, only in case of help-mate. Then the Appeal`s Committee overruled the decision by the same paragraph of the same WWCC Regulations which says that a very unskilled player can drive him/herself into a mate position thereby a mate can be achieved. No doubt that these players are professionals and the probability that they make such a mistake is by far less than 0.1 per cent. Still, theoretically a mate is possible. Irrespective of the final position of the pieces on the board, which have shown that to reach the mating position may require another half hour playing (if it comes at all), theoretically a mate is possible.

This is the moment at which everybody agrees that the Regulations are not appropriate anymore and they must be changed. How to change them and what kind of solution would emerge, that is another issue. You can use tennis match rules, you can let allow the players to play until the first victory in blitz games, etc. But, please, no Armageddon, no nuclear war, no Abaddon, no Apocalypse or any kind of destruction or self-destruction of the reputation of FIDE.

Duif Calvin, San Rafael, California
In your recent Chessbase article on the Armageddon game, you ask who first coined the term. I may be mistaken, but I believe I was the first to do when I was the Editor and webmaster for the US Chess Federation's website, and wrote the weekly news column for them. The concept of a one-game-playoff was introduced in the 1997 FIDE world championship tournament, and we needed a quick descriptive phrase to distinguish it from "sudden death." Hence "Armageddon" – the end of the conflict. (FIDE had referred to it in the official rules as "one game sudden death," and we'd gotten a number of questions about it, so felt a new term would be useful.) You can see the use of the term in our December 1997 coverage via the Internet Archive:

The term was introduced during the earlier event (that Anand won) using the same phrasing, but that page wasn't archived. Some of the commentators on various servers picked up the term, and it filtered into general usage that way. However, if someone else has proof of using it earlier, I wouldn't dispute it. But at least this gives you one reference.

The interesting thing is that in the original event (the 1997 FIDE World Championship Tournament) there were two types of "sudden death" time odds games used as tiebreakers.

The first, called sudden death games, gave White four minutes and Black five minutes. (Yes, White had less time.) There was an increment of ten seconds per move. These games were not played in pairs, but just as one game after another. The first one to score a full point advanced to the next round. In this phase, the extra time given to Black was to make up for White's benefit of first move. The problem was that draws didn't stop the conflict in this system. If there was a draw, they just reversed the colours and played another game under the same time control.

The Armageddon game (which FIDE called "one game sudden death") switched things up. Suddenly it was Black who had less time, not White. However, Black was also, for the first time, given Draw odds. This meant that the entire conflict not just could, but would be completed in a single game. White had to win to advance. Otherwise, Black would advance.

So the original sudden death games did have an increment. The Armageddon game, the one without the increment, was only introduced at the Arbiter's discretion if the players had been unable to break the tie during the series of previous "sudden death" time odds games. And the conflict did have to come to an end because of the fact that the next round in the knockout tournament started the very next day.

Nowadays, though, the Armageddon format without increment seems to be used as the first tiebreaker, rather than the tiebreaker of last resort. I think introducing the draw odds game so early in the process is part of what has led to the complaints about it. However, I do know that some of the elite players complained about having to play two different types of time odds games, one right after the other, as they felt that was a very difficult playing rhythm to manage.

Personally, I think that once you introduce an increment the draw odds will be an overwhelming advantage. That's why the original FIDE system introduced draw odds and took away the increment at the same time. I myself see nothing wrong with saying "On this day, these two players have played to a dead draw" and then either declaring them Co-Champions or flipping a coin. I also see nothing wrong wtih giving a defending Champion "draw odds" for the match, which is another way of settling the issue for matches. However, that's just speaking as a fan. I know that most grandmasters prefer to have event titles settled through some form of chess combat, however distorted it may be.

Eduardo Moura, Vancouver, Canada
I have just watched a video of the Women's World Chess Championship in your web site. It was an Armagedon game, which finished in an ugly time scramble with the players banging the pieces around without any intention other than to win on time. Recently at the final of the Women's USA Chess Championship we witnessed the same situation. Very sad to see such a thing at a World Chess Championship.

It has become evident, at least to me, that Armagedon blitz is a very poor method to decide a chess match at this level of competition. It is of course better than a coin toss but it suffers from a serious fault, which is the incentive it creates for the players to play only to win on time irrespective of the position on the board (ie to try to flag the other player). This is perfectly legimate way to win a chess game given the rules of armagedon but it is not an elegant solution to the problem at hand. There is however a simple solution to this problem: to add an increment of 2 seconds per move for both players. Say White starts with four minutes, Black with two minutes, both players haveing increment of two seconds per move. A draw favors Black.

Picture the following situation in say Anand-Kramnik match. The match is tied on classic time control, rapid and blitz chess. They play the Argamedon with no increment to decide the WCC. White has to win but is very short of time and end up in a situation of king and queen vs king. It is mate in a few moves but Black flags White and thus becomes World Champion! What a bizarre situation! Of course the same could happen with two second increment, but it is unlikely as the increment gives the players enough time to win a completely winning position.

Thus I think tha FIDE should seriously reconsider the time control for Argamedon games otherwise we face the risk of havig a World Chess Championship being decided in a scenario similar to the one described above. It would be a pity were that to happen to the Game of Kings (and Queens!). Better not to take the chance and change the rules of armagedon blitz asap.

Jim Monaghan, Cold Lake, AB, Canada
It's ludicrous that Armageddon rules are applied to determine chess victors in these prestigous events of the last few months. I think that tennis is a good model to emulate. Why not have sets of two extra blitz games (mini-match) played at a TC of 5min + 1sec, with the victory going to the first player to win a mini-match. If a player wins the first game, a second game is still played as the opponent could tie and then another mini-match is played. Games would proceed at a pace of about fove games each hour. You would most likely have a champion in about an hour and a half. Theoretically, it could go on forever, but in practice someone will tire and the other will win in an hour or two. It simply takes one decisive game and a draw (or a draw and then a decisive game) and it's over ... At least "chess" would still be played. Otherwise, why not toss a coin, and everyone can stay home?

Royce Campbell, Tulsa, Oklahoma, USA
Mr. Saunders article was entertaining. We have played a variation of speed chess in our day that is extremely fun, and tends to even out differences in ratings at the lower classes (USCF class A, myself), but may be worthy of consideration for replacing the Armageddon game at tournaments. We called it 'cutthroat,' an equally as graphic visual connotation as 'Armageddon,' and it goes like this:

Each player starts with five minutes (I've never played with an increment, but one second or so does not seem improper). Draw for colors and play the game (change colors each game). Each game is worth one minute. If you win, you get one less minute in the next game. If you draw, you each get 1/2 minute less in the following game. The outcome is simple ... the first person to achieve no time at the beginning of the game wins the match.

For example, when you win the first game as White, you must now play another game at 4:5 time odds as Black. Win again, and you play as White at a 3:5 disadvantage. A draw now means you would play Black at 2.5:4.5, and so on until someone reaches 0, who is then declared the winner.

This game has the effect of dampening the advantage of having White in the first game, and can take no longer than (10+9+8+7, etc, minutes) an hour or so to complete. The only drawback is if the final game at .5:.5 is drawn (an unlikely scenario), but the answer is to simply play another game with colors reversed, thus giving the Black player in that last game the draw advantage of having White in the next game.

This method of playing was a little more difficult with the old analog clocks, but we never had a controversy, and with the advent of digital clocks, there should, it seems, be less chance of one. Play a few matches of cutthroat, and you'll see the appeal of the game for fun. It is up to others to decide if it is viable for serious competition.

Robin Lindsay, Montreal, Canada
I have been following with some interest the debate regarding Armageddon blitz finishes. The debate seems not so much to revolve around whether or not major titles should be resolved by blitz finishes, but the problems, other than the quality of the games themselves, pose. Controversy seems to erupt because the players knock the pieces over, or misplace the pieces, make illegal moves, move before their opponent has moved, etc. Would all of this not be solved simply and surely with a couple of laptops?

Philip Feeley, Vancouver, Canada
My views on an Armageddon finish in short: I agree with Woody Allen. As for your theory about why time controls are chosen, didn't the Zatonskih - Krush armageddon game refute that? I thought they played well past midnight because their earlier games went long into the evening (wasn't one of them over 100 moves?) So much for getting off for dinner...

Lonnie Kwartler, Chester, NY, USA
In the Monika Socko-Sabina Foisor game with K+N v K+N, could the rule requiring the players to try to win by normal means have been applied to stop the game and declare it drawn?

Frits Fritschy, Leiden, The Netherlands
When chess clocks were introduced in the nineteenth century by tournament organisers who didn't like cold dinners, there may have been quite some grumbling about the loss of quality in chess. Nowadays time is even more scarce and stakes are higher. Circumstances you can't beat, so you've got to get the rules right.

The problem with Armageddon finishes is that they use rules that were devised for other circumstances.
In a blitz tournament a single game is much less important.You won't make much fuss over either giving a draw or about the odd person going on with knight against knight. There are twenty or thirty other games where you can put things straight. Moreover, it's simply impossible to put an arbiter at every board where someone claims a draw. So that's why rule 10.2 (draw claims) is not in force in blitz games.

With Armageddon finishes in knock-out tournaments there's a lot more at stake. It's simply not fair to make just the players responsible for a fair outcome. In the Socko-Foisor game, people have expected too much of Socko. (she had only one second left herself!) In knock-out tournaments there are only a few games to watch, contrary to blitz tournaments. Why not change the rules and apply rule 10.2 (in a way) to Armageddon games? Then Foisor could simply have stopped the clocks and summoned the arbiter, who, even without seeing any more moves, would have granted her a draw.

Michael Allard, Bowie, Maryland, United States
The "public" desires Armageddon?! The BCM editor strikes the nail true with one swift blow. One additional point: and that is if a chess opponent is physically impaired to preclude participation in such a travesty (witness at least one opponent in the last U.S. Championship). Not only is Armageddon discriminatory and absurd, it's just as inane as the addiction to chess software engines for OTB preparation and postal play. We are witnessing the de-evolution and de-construction of human expertise. It's time to return to the classical time controls.

Kung-Ming Tiong, Universiti Malaysia Sabah, Kota Kinabalu, Malaysia

If you must have Armageddon, have a double. John Saunders's article (Sept. 7, 2008) wonderfully revisits the issue of Armageddon chess with some satirical twist. With reference of two examples (the 2008 US Women's Chess Championship and a game from the current ongoing 2008 Women's World Championship), Saunders made a valid and clear point: a single Armageddon game is not a satisfactory way to reach a conclusive result.

However, other parts of the article's suggestion and comments did not prove to be sound. In the article, it was pointed out that:

1. it was not a satisfactory conclusion for the Socko-Foisor game where the win was awarded to Socko after an appeal to the initial draw result;
2. adding time increments to Armageddon seems a plausible improvement to Armageddon games

Regarding point 1, Mr Saunders implied that the final decided result (win for Socko) was attributed to the unacceptability and illogicality of FIDE Laws of Chess Article 9.6. He appears to favour a draw as a result simply because with the expected level of play of the two players, it would not be likely for Socko to force a win in a K+N vs K+N ending. Herein lays the faulty statement. I am of the opinion that the appeals committee made the correct and acceptable decision through FIDE law. Possibility and likelihood are two different things.

In the game, although the possibility of Socko to force a win is practically nil, the other factor, time, is to be considered. When a player's flag falls, that player loses. Period. Otherwise, what is the use of clocks and flagging if we were to look at, evaluate and give a result to chess endgames based on possibility and likelihood of checkmating/draw? Take a K+N+B vs K ending for example. It is possible for the player with K+N+B to checkmate a player with K, but the likelihood of achieving it differs according to the level of the player. It is also possible for a player with K to draw against a player with K+P. Again, the likelihood differs according to the level of the player. Granted, players in high level tournaments/match will definitely be able to convert those possibilities to a high degree of likelihood (in the first case, to certain win for K+N+B and for the second, to certain draw, depending on the positions of K and P of course).

But likelihood is not a good measure at all. Mr. Saunders misses the point that players do and will make mistakes regardless of their playing strength. Referring to the reconstructed game between Socko and Foisor, he commented that it was amazing for Foisor to miss out on a draw in one with her "failure" to capture the white knight after white's blunder. That's the whole point. She was likely to have captured the knight but did not. Reason? Time pressure where in a race against time, sometimes players overlook moves.

Therefore, just because it was not likely for Socko to force a win in that situation, it does not mean that the game should be drawn. The flag should NOT be disregarded. Losing on time can be painful, but that is the dramatic and interesting part of chess.

Regarding point 2, adding time increments does not work as the concept of a single Armageddon game itself does not present equality to both players. A tie-breaker should always be equal to both players. Odds should not be present; there should be nothing left to chance (coin-toss to decide who to choose color and time distribution). Armageddon rules are skewed and do not have inherent equality. Yes, players get to choose... but the choice is partially influenced by chance. An element of chance should never be present in a skill-based game/sport.

If Armageddon rules have to be used, then the number of games played using that rule MUST be an even number (Double Armageddon) where both players get to play with both colors. That way, both players will have equal odds unlike the present single game Armageddon.

Think: if Armageddon rules are acceptable, then how about playing a whole tournament/match consisting totally of games with Armageddon rules?

Joaquin Font, Ponce, Puerto Rico
One solution to the Armageddon conundrum is to have a chess clock in such a way that the arbiter can intervene and press a button (or better yet, through a remote or computer signal) to change the timing of a game to a time increment if a position is reached of even material between the parties, no pawns. Or, to change the time after x number of moves (for example, 10) in such a position. Technically speaking,this could also be designed in an automatic fashion (a board with sensors) so that the arbiter is not involved at all--that would be even better.

As to the number of moves which could be played after a materially even position without pawns arises before a draw is declared, perhaps one can say, 10 moves per every piece that each side has. (If one knight, 10 moves, if two, twenty, if two rooks and a queen, thirty). This would allow real fighting chances for the side with a positional advantage in materially even positions.

Vitaly S. NJ, USA
This letter is about the N vs N ending. First of all, it looks extremely unethical for the "stronger" side to keep playing, whatever are the formal rules. However, in general there exists a formal procedure to deal with this: the "defending" side should have urged the referee to watch the game before it ended. Upon observing a train of obviosly senseless moves (but before one side ran out of time!) the referee would have declared a draw, at least this was a rule of normal blitz tournaments.

Earlier ChessBase articles on the subject