Greengard's ChessNinja.com

Dresden Olympiad r4

| Permalink | 36 comments

Podcast with animated diagrams and round wrap-up here.

There are only two perfect scores left in the Dresden Olympiad. One is no surprise, the top-seeded Russians. Kramnik has looked rocky on board one and they've won a few matches by the minimum, but they keep winning. The other is the home team, Germany 1, with some serious home mojo working so far. That mojo will be severely tested in tomorrow's fifth round when they face Russia. Today Russia squeaked past India 2.5-1.5 and Kramnik was in trouble with white against Sasikiran, who decided to force a repetition draw in time trouble instead of playing for a win by grabbing the b-pawn. Harikrishna exploded Svidler's Grunfeld in just the way the Grunfeld gets exploded occasionally. The spectacular rook sac 25.Rxe6!! wins instantly. Black gets mated if he takes the rook and in the game f7 is just too weak. Svidler resigned on move 27. Unfortunately for India, losses on the lower two boards cost them the match. Tough to match up when Morozevich is lurking on board four, yeesh.

Germany stayed perfect thanks to Naiditsch's third win. He beat Shirov on the black side of the Berlin defense. The heavyweight matchup of the round was Ukraine-Armenia, a battle between the last two gold medal winners. Ivanchuk netted an exchange against Aronian but couldn't convert. The other three boards also ended in draws. So two more perfect scores disappeared. Can I mention again how much I dislike the change to match scoring from board scoring? It's so conservative it's just painful. There's so little movement in the standings and if you fall off the pace it takes forever to get into the mix. We're used to seeing the cream rise (and the curds fall) fairly quickly. Here it is the fifth round and we're still seeing 4-0 wipeouts against teams with one GM, if that, by the top teams. Silly.

Cinderella team Norway continued to impress, this time drawing with England. Magnus Carlsen beat Mickey Adams on board one with a little help. White had pressure, but he would have a lot of work to do after 43..Rb4 or 43..Rc7. Howell saved the match for England with a win. They join the big pack of teams with 7/8, including Hungary, Azerbaijan, Ukraine, Netherlands, and Armenia. Radjabov and Mamedyarov beat Kamsky and Nakamura, respectively, to put a big dent in the Americans' hopes to repeat as medal winners. Onischuk was finally held to a draw after starting with three wins. China bounced back by stomping FYROM 4-0.

Fabiano Caruana beat Korchnoi in rather brutal fashion. There was already no love lost by Viktor the Terrible toward Caruana after the teenager blitzed him from an inferior position at the "Rising Stars vs Veterans" event this year. Of course Korchnoi isn't shy about sharing such feelings after a loss and being cursed out by him is almost a rite of passage. Veselin Topalov and Bulgaria are trying to work their way back up after a slow start by beating Mongolia.

In the Women's event they are also down to two perfect score, Poland and China. They will meet in the fifth round. India surprised host Germany with a 3-1 win. A shout-out to Rusa Goletiani of the US, who has a 4/4 score.

Apart from the key Germany 1 - Russia matchup, some round five highlights: Armenia-Netherlands, Hungary-Ukraine, Norway-Azerbaijan (a Dragon on board 1?). The unbalanced pairings continue elsewhere with China-Scotland, England-Italy, and India-Austria. Something is just wrong when we have teams that have barely faced any GMs, let alone defeated any, on the top boards in the fifth round. Not blaming the teams and players, of course. Austria lost to Norway in round one and then beat Jordan, Sri Lanka, and Turkmenistan, who have no GMs among them.

Podcast with animated diagrams and a few game comments here on ICC Chess.FM. Don't miss Macauley Peterson's Olympiad blog with video and more direct from Dresden. Live games at the official site. Don't bother downloading the PGN archive from the official site, btw. So far they have been full of completely wrong scores. Stick with TWIC, where poor Mark is battling to get them from the live broadcast and verify them.


Mig, I just don't understand your negative comments with respect to match scoring. Using game pts means the determing factor is how well you beat the weaker clubs. Furthermore, game pt scoring dimishes the importance of each game within a match. When it is 1.5-1.5, that last game is critical, and the players know it. Furthermore, it is much more difficult to create an insurmountable lead in the match score system. Having played in regular leagues under both scoring systems, I find match point does much more to foster team unity and interaction between the players. Game scoring is more individual in nature.

I think the error is in using the game scores as the fourth tiebreak rather than the primary TB [or second, using their nomenclature]. Thus when two teams meet in later rounds, if one has a significant edge in game score, the other knows it must win the match, and the games should be that much more exciting. Even being down 4,5,6, game pts., knowing that you can win the thing by beating the other team, well.

Finally, I suspect the pairings you are complaining about are due to the accelerated pairing system rather than the scoring system.

But enough of that, we shall see as we move into the second week of the matches. And dang, there have been some exciting moments and games already.

Match point format generates too little variance, too little spread. You get a "more efficient market", so to speak, with a game point system. (I.e., the team with the best rating performance is more likely to place first in game point format.)

As I understand it, the bizarre pairings that we're seeing now are a result of how teams are ranked within each score group. Instead of the top half playing the bottom half as determined by rating (or even game points), they're using S-B points.

I know this isn't the right place but there may be Russians around and today is TWIC day so I'm on a deadline. I'm looking for the Russian Cup games which should be downloadable from http://russiachess.org/online/2008/08/ but, at least in a western browser not set up for cyrillic, the link to the PGN fails. Can anyone download these games?

On the subject of the scoring system, yes its different, but I'm used to it playing a lot of team chess, I'm not necessarily against, lets see how it works out.

I'm not surprised Norway are doing fine. Its like the US with Bobby Fischer in the 1960s, a great top board really gives wings to the rest of the team.

How about the following format:

Winning 2.5 - 1.5 or 3-1 gives you 2 match points.
Winning 3.5 - 0.5 or 4-0 gives you 3 match points.
Draw gives you 1 match point.
Any defeat gives you 0 points.


I think the obvious negative to match scoring is the conservative strategy that it rewards.

For instance, if your team has scored a win on one of it's boards, the rest of the team should immediately switch gears from winning to avoiding losing their games.

Game points means it is necessary to win lots of games, and 'beating up on the weakies and drawing the big guns' is hardly a problem in my opinion, as it is the normal strategy for all but the most elite and exclusive tournaments.

The beauty of the game point system is that in last rounds leaders can change if one of the teams pulls a big win. Like Russia won the gold medal in championship when they smashed China in the last round 3.5:0.5. Even 3:1 win for Russia would mean that China would get the gold. With match points that would be impossible.

Then, match points remove incentive to fight in each game. When one team is winning 3:0, the remaining player has (almost) no incentive to fight and pull out a 4:0. Which makes it boring-look at wins after Russia-2.5:1.5 against really weaker teams.

The match point system, IHMO, works especially bad with less rounds-we have 11 instead of the regular 13-14.

Overall, there will be much more draws and stupid pairings in this Olympiad. In the last rounds, there will be less tension than usual. They know how to spoil things that worked sooooo well before.

The bridge system of Victory Point conversion tries to combine the best aspects of game and match scoring.
It probably works better than this binary conversion because it gives losers some credit in a hard-fought match, while still allowing a big winner some extra credit.

Thanks to the guy that sent the Russian Cup Games.

The disadvantage of the board count system is that it rewards huge wins in the final rounds, which can often be the result of a favourable draw.

Oh if you get chance try the official free tv, really good picture. Not sure about Tom Jones, "Its not unusual" in the background (it is however before the start of play in case you're wondering).

its Macedonia, or Republic of Macedonia, not Former Yougoslav Repiblic of Macedonia

Details on Caruana's win over Kortchnoi are on the Chess.FM blog:

Caruana leads Italy to third straight win



I agree, if a team is already secure of a match win, there's almost no incentive for the remaining players to win.

But that doesn't quite explain Russia's 2.5-1.5 wins. From checking some of the games, they are actually having some problems. Besides, Kramnik would draw while they were still equal. Like today.

Is it my biased perception or we are witnessing a shift in fashion from Slav to Grunfeld ?
Svidler aside, Lenier, Kamsky, Mamedyarov, Morozevich, Topalov have been played it these days.

At first I was also annoyed by their using game points as the last tiebreak and not the first, but alas, it's necessary. If you pair with match points, board points can be very easy to come by if you played the Swiss gambit. Just look at the standings right now and see what they would be if you used board points. They're all over the place. Teams that haven't beaten a single GM have more than some of the teams with 7 match points. Annoying. That will probably flatten out by the end though.

I think everyone else has covered my other objections. The main two are less dynamic standings, which means it takes longer for the good teams to rise and the bad ones to fall in the pairings and removes much of the drama from the final rounds; and the conservative effect it can have on the games themselves. As someone mentioned, it can increase the drama if you have a head-to-head match between two teams tied for first in the final round, but the chance of that is quite small. By the 11th round most of the favorites should have met.

Zigomar: I know the various naming arguments about Macedonia and I also call it that on occasion. But when FIDE is still listing it as FYROM it can be confusing to call it something else and I don't want to have to explain it every time.

Yah, I looked up the bridge system a few years ago during one of these discussions. Certainly more fun than this. It's nice to have something simple and easy to figure out though. Something like what Duncan mentions would be all right, though why not distinguish between 2.5 and 3 or 3.5 and 4? A point is a point.

And any system that has the USA playing a team with a 1600 on it in the fifth round is a total joke. What a waste of time. I know that's the pairing system as much as anything, but jeez.

>And any system that has the USA playing a team >with a 1600 on it in the fifth round is a total >joke

It is a non-elo-racist system, it promotes fairness, "social-justice" and afirmative action.
It is meant to promote poor people chess not the elo greed of the few brain privileged.
"Gens una sumus regardless of Elo."

India doing surprisingly well so far both men and women. Absence of Vishy and Humpy a blessing in disguise perhaps for promoting the tier-2 players.

Is there a no draw under 30 move rule? If so then it does not appear to be enforced as in round 5 there were several games quickly agreed drawn ....

What else is it due to?

--I know that's the pairing system as much as anything, but jeez.--

Great defense by Gustaffson!

@gg: The moral deficiencies of the HKG team. Getting paired with the USA when they can only field a lowly 1600 on board 4? Sheer presumption!

Winning big not only increases your tiebreak (which puts you ahead of other countries which finish on the same match point total), but I think it also gives you easier pairings. Game points still count.

"And any system that has the USA playing a team with a 1600 on it in the fifth round is a total joke".
I am not really impressed by this argument. Maybe the USA did not play well in the first four rounds? I play in open tournaments in the summer and occasionally I see a grandmaster playing against a much weaker player somewhere in the lower tables, far away from the other players of his kind. Whose fault is it? It is his fault, because he performed below his real strength in the previous rounds, and if he has 2.5/4 he might well play with a 1600 player who just beat someone rated 2000. This is the Swiss system.
The reason I like the system used in this Olympiad is because it makes chess a slightly more "team" game.

I'm not ignorant of the vagaries of the Swiss pairing system. Had the US suffered a serious upset or had Hong Kong surprised a strong team, great. That's the way it usually works. But this year we are seeing many strange pairings relatively late in the event without any surprising results to lead up to them. The US went from playing the #4 seed Azerbaijan to facing Hong Kong the next day. The Hong Kong team, unsurprisingly, has not defeated a single titled player. ALL of the games they have won have come against UNRATED PLAYERS. And yet they faced the US, the 10th seed, today. Ridiculous, and it's happening up and down the table.

I disagree that using match points instead of game points is a bad idea. In fact, I like this new system. It allows scrappy tough teams that are not superGM powerhouses (e.g. Norway) to compete for medals. Under traditional system, a team like Russia would string at some point a couple of 4:0 victories and that would be the end of the fight. Norway wouldn't be able to reproduce such results and thus would never catch up. Under the new system Norway has a chance to hang in there by winning 2.5:1.5.

The worry that match scoring would lead to less combative chess is baseless. This is one of the hardest fought Olympiads in a long time precisely because the margins are small, so every team still has something to fight for.

Also, under the new system the importance of each individual game may be higher than under the old one. Take a look at the Leko-Ivanchuk game today. Ivanchuk had a R+B+3p v Leko's R+N+2p in a dead drawn position. Normally they would just shake hands and go home. However the score in the match was 0.5-0.5 and Ukraine had advantage on 1 of the remaining boards and a disadvantage on the other. So Ivanchuk kept playing. The score became 1.5-1.5 and the game became decisive. Ivanchuk kept playing this dead drawn position for more than 100 moves and actually won in B+R v B endgame. Bloody priceless. Would this game happen under the old system? I don't think so.

What is really screwed though is the accelerated pairing system COMBINED with match scoring. It's grossly unfair (e.g Armenia-Farer Islands and China-Norway in the same round) and it creates all kinds of weird matches even late in the tournament. In first rounds weak teams play each, so some of them are bound to score points. Then, a winning weak team plays another winning weak team. With a system like that, there will be some weak teams near the top of the table even in the middle of the Olympiad.

Accelerated parings itself is not a bad idea. However, see here how they should have done it properly: http://en.wikipedia.org/wiki/McMahon_system_tournament

Making the occasional 4th game to finish harder fought and/or more relevant comes at the cost of making performance less relevant overall. Once a team has a win in the bag, the other players can, and should, play it safe. How many times do you think the fourth game to finish got an extra boost compared to how many times it made more sense to stop taking risks and to go for draws? Not nearly enough to compensate, I'd say, not that we can quantify these things easily.

At least not with all the info we discard in the chess world. If everything were databased properly, including game times, we could easily figure out the duration and nature (decisive or drawn) of the fourth game to finish in every match and compare it to the same info in the last Olympiads. That is, how long after the third game to finish finished did the fourth game finish, and was it drawn? (Especially for the Finnish team, har har.) Would be cool, if not necessarily causative.

It's not just the fourth game, Mig. You never know which game is going to finish last, so until your team locked up 2.5 points, you've got to fight it. Even if your team stays OK at the moment, someone might blunder, someone might get outplayed, you never know. So if there is a slightest spark of life left in the position, the guys on top teams keep the game going. That's my observation. Maybe it's a bit different in the middle of the pack, I wouldn't know since I mostly follow the top teams.

I also don't buy that match point system encourages more conservative play. I would argue that it's exactly the opposite. With the game point system, every point counts and every half point counts too. So if you have a choice between a move that gives you a minimal edge with no risk and a sharper move that gives better winning chances with some risk, one would normally chose the minimal edge. With the match point system, your decision depends on the match situation. If your team is down, you should probably go for a sharper line.

I think the only way to resolve this debate is to do some real statistics of how often players keep going when their team already clinched the match. Compare it with previous olympiads.

There are too many factors to resolve it by just making arguments (on both sides) that perhaps make sense, but haven't really been supported by data.

Earlier today, I was watching the Bulgarian games. When the match win was already secure, the last guy (I think it was Iotov) kept pushing, and added another winning point. Also in Israel vs Denmark, Postny kept playing when even a loss wouldn't change the match result, and he actually won. Not that these examples constitute definite proof of an argument. Just examples I remember from today.

Interesting theory, Osbender, but it's just not the way it's happening in practice. They use match points in pro leagues all the time and they still have plenty of short draws. (Pro leagues are very imbalanced and a top few teams would totally dominate with board scoring.)

Same goes with the decision to play more sharply. Losing has an even higher risk when by losing you are risking 100% of the team's total score instead of 25%. So 50% is going to be a safer bet. That's why solid, no-loss players are so prized on teams. Not that the players sit there cogitating all this all the time. (The coaches on the other hand...) Of course most of the time they just play the position. If someone is better they'll play. They are competitors. There's rating to think about as well.

The difference between pro-leagues and Olympiad is the number of boards. With just 4 boards in Olympiad, you can afford a short draw only when your team already has at least 2 points on the scoreboard. In practice that means no more than one relatively boring draw per match. In a typical pro-league you have 6-8 games per match, many more chances to score a team win, hence less urgency to win your own game, hence more draws.

The proof of the pudding is in the eating. So far we have an excellent fighting chess at the top of the standings with virtually zero non-games. For example, Ukraine scores 2.5 point per match not because they take it easy, but because opponents give them hard time.

Also, match scoring in Olympiads puts the fight where it really belongs, e.g. into matches like Ukraine-Hungary, Armenia-Azerbajdzhan, etc, instead of the matches like England-Italy where the goal is to slaughter a weaker opponent with the maximal score.

I like the match scoring. It makes it more like other sports. The important thing becomes winning the match rather than the "goal differential" as it were. Also, unlike Mig I strongly disliked the big (and IMO unfair) last round swings that could happen from the randomish 4-0 blowouts that could take place with a lucky pairing. Having the Olympiad be a Swiss is random enough already. We'll see how the tiebreaks work in practice, but so far they seem reasonably fair about reflecting both strength of schedule and performance.

I also enjoy the fighting chess, especially on the top boards.

But these fighting chess doesn't prove anything either way. They fight because both sides play well, and the win is not secured. There's no option but to fight on.

The cases we need to see in the context of the argument (whether the match point system reduce fights or otherwise) are those where the match win is secured before all games are finished. Now count what percentage of the remaining games are still fought, how many quickly end in peaceful handshakes after the match is secured, etc. Of course this is not as easy as it sounds. Perhaps the remaining game is a win, obviously the guy still wants to continue despite the settled match result. Etc.

What happened in most cases so far, at least on the top boards I was watching, is that that type of case happens rarely. So we don't have enough data to argue on the merits of the match point system as far as fighting chess is concerned.

From watching some games, it seems that the match point rule could at least sometimes have a good effect. When the match is secured, the last player can play as he like. He could draw if he wants, but he could also play wild and take all the risks if he's in the mood for fighting. It doesn't matter for the team anymore, so it's his own game.

Did you know about freelance writing service? I just want to say that your data referring to this good post is smart! Thank you so much for doing it!

Twitter Updates

    Follow me on Twitter



    About this Entry

    This page contains a single entry by Mig published on November 16, 2008 10:58 PM.

    Dresden Olympiad r3: So Much for German Efficiency was the previous entry in this blog.

    Dresden Olympiad r5 is the next entry in this blog.

    Find recent content on the main index or look in the archives to find all content.