Greengard's ChessNinja.com

MTel 08 Starts Hot

| Permalink | 55 comments

Actually it started a little slowly on all three boards in the first round of the 4th edition of the MTel Masters in Sofia, Bulgaria. But eventually all three games heated up and finished decisively. An exciting start to this category 20 (2737) double round-robin. Four of the six players currently inhabit the top ten: Topalov, Aronian, Ivanchuk, and Radjabov. Relative outsiders Cheparinov and Bu Xiangzhi round out the field and faced each other in the first round, a win for the Bulgarian. Cheparinov and Radjabov hustled to this event directly from the FIDE Grand Prix in Baku.

Radjabov lost rather horribly with white against Ivanchuk in the first round. An inspired defensive exchange sac by Ivanchuk looked good enough to draw until Radjabov went hard for the win by giving back the exchange. This quickly led to a winning endgame for Black. If you count his last few games from Baku, Radjabov has now lost three whites in a row. Topalov took a big step toward winning his fourth MTel tournament by beating 2nd seed Aronian with the black pieces. It was classic Topalov, finding dynamic attacking chances from a calm position. He got a little excited and nearly flubbed the finish, but by then Aronian was in terrible time trouble -- 32 seconds for six moves -- and the Armenian had no time to find 37.Qxd5! Topalov could have avoided that miracle save with 36..Qf4+ 37.Kh1 Ne3 38.Qe2 Nxd1 39.Qxe8+ Kh7 with a winning queen and pawn endgame.

The other hometown player, Topalov's second Cheparinov, also started out with a win. On Chess.FM, my co-pilot GM Joel Benjamin thought that Bu Xiangzhi came out of the Najdorf quite well. But after exchanges the Chinese player drifted into a totally passive endgame and Cheparinov ground him down convincingly. Going for counterplay with 32..Rg8 is one of the last chances to keep at least one piece active. The horrible black bishop stayed locked behind the "Najdorf Pawn" on d6 for all 55 moves of the game despite shuffling between d8 and f8.

This year's time control is 40/90' + g/60', no increment. The players are inside a soundproof double-paned glass cage on the stage. The 1995 Kasparov-Anand WCh match used something similar with mediocre results in both the sound and temperature departments. According to Macauley Peterson, there on the scene for the ICC, the players are sitting on 16th century chairs (oddly not visible in the photo), which will be auctioned off later. This reminded me of my favorite chess chair ever, Karpov's in the 1997 FIDE WCh mini-match against Anand. Vishy was in a regular office chair and Karpov shows up with something from the Wizard of Oz meets Captain Kirk. I couldn't find a pic online but apparently I scanned that one from an Argentine paper 11 years ago. My archives are scary.

Round 2 (8am EDT, 1200 UTC; 1500 local): Topalov-Ivanchuk, Bu Xiangzhi-Radjabov, Aronian-Cheparinov.


What's with that "90 min for 40 moves + 1 hour to the end of the game" time control?
Does this nonsense apply to all "Grand Slam" tournaments or only to this one?

Good question. I thought only FIDE went in for this crap. Although I have noticed that Eastern Europe seems keener on it than Western events.

Just a correction: From the M-Tel press release, it's only TOPALOV's chair that will be auctioned. Not all the chairs.

"The starting price for the antiquarian chair will be BGN 1500. The gathered money will be used for the award fund of the circuit of children’s chess tournaments M-Tel Masters Junior starting this year."

I remember seeing that photo as a kid. Thought that world champions, even FIDE ones, get to have special "thrones".

Soikins wrote:

"What's with that "90 min for 40 moves + 1 hour to the end of the game" time control?"

That's 2.5 hours for the whole game. Not too bad. In fact, I think 2 hours per game for each player is more than enough to strike a nice balance beteen good chess and spectator endurance. After all, it is just a game and neither spectators nor players should have to spend more than 4 hours on something that is supposed to be trivial and fun...

Ivanchuk immediately rejected his chair and was provided a backup sans arms. This manoeuvre is visible on the website video. Perhaps he read the postings here likening him to an antique.

I know it's not important, but I very much doubt that those chairs are from the 16th Century...
I know Americans are not very interested in "old stuff", but still...

Irv wrote:
"After all, it is just a game and neither spectators nor players should have to spend more than 4 hours on something that is supposed to be trivial and fun..."

If chess for you is just a game, then I guess you really have no reason to mind that Topalov has 11 minutes for 20 moves and Ivanchuk has 14 minutes. But I consider chess to be something more - a kind of art. There are chess games that were played hundred years ago, that provide aesthetic pleasure even today. I doubt we will see such games in this tournament.

agreed strongly with soikins


Your purist nature is showing, if somewhat outdated. a 5-hour contest is plenty long enough.

Compare to other sports/past-times/endeavors:

Red Sox - Yankees (arguably the longest of all games): ~ 3.5 hours
American Football: 3 hours
Soccer: 2 hours
Basketball: 2+ hours
Hockey: ~ 3 hours
Curling: who cares
Go: see above
Checkers: 5 minutes

Get the point?

And 'Topalov having 11 minutes..' left is irrelevent. He knows the time control. If it's a problem for him, he needs to manage his clock better. If you gave him 4 hours for 40 moves, he's probably use all of it up just as surely.

I don't consider chess an art, but I do consider some games very artisitc. The same can be said for just about every sporting endeavor regarding certain plays/contests.

Quit your crying and climb out of the past. You'll find plenty of good chess here if open your eyes.

To Mark Goodwin

First of all I have no idea what your assumptions about my "purist nature" has to do with the subject of this discussion.

Second, I don't see what time controls of physical sports (american football, soccer, hockey and the like) have to do with mind "sports". I think the differences are too great to use them as some sort of example.

Third, about mind game time controls you wrote:
"Go: see above [ed. - ?]
Checkers: 5 minutes"

First about Go:
"A typical time control is "60 minutes + 30 seconds byo-yomi", which means that each player may make as many or as few moves as he chooses during his first 60 minutes of thinking time, but after the hour is exhausted, he must make each move in thirty seconds or less."
Wikipedia: "Byoyomi"

International checker tournaments are mainly played under the time control of 2 hours for first 50 moves + 1 hour for next 15 moves.

Regarding chess tournaments - if there would be a World Blitz Chess Championships organized in the same manner as Snooker World Championships (knockout matches), I would follow this tournament very closely and with great interest. I undoubtedly would see a lot of entertainment there - petite combinations, fast movement of players hands, arguments with the arbiters etc. It would be a lot of fun. Would a I care to browse through the game texts the next day? Would I care to read GM analysis of these games? I wouldn't. The games would simply lack in depth, the ideas would be too shallow to be of great interest.

I might be purist, but the quality of chess games played under the standard time control, IMHO, are better than those played under FIDE control or this MTel time control. For example, today it could be a great battle of minds in Topalov - Ivanchuk, but instead I saw a cheap blitz-style bluff that failed. Fun for a minute, but nothing of interest a day after.

Mark Goodwin, I fear your nomination of "Red Sox - Yankees" as the longest baseball game of all time is not arguable at all.
If only the major leagues are considered, the longest game was between the Milwaukee Brewers and the Chicago White Sox that lasted for twenty-five innings with total time of eight hours and six minutes. This was on May 9, 1984.

There were many other major league baseball games that lasted far, far longer than 3 hours. Goes to show Mark Goodwin's whole philosophy of life - just as he views chess as a trivial game not worth much effort, he seems to feel the same way about his own posts. (Baseball game-length records being readily available on the Internet, I'm sure.)

Irv, who started this theme of chess being no more than a trivial game, has been an articulate and consistent advocate of his view here.

Why he chooses to devote what must be significant portions of his personal leisure time playing chess, may spend more time studying the game (he is a formidable player, I can personally attest), and regularly show up here, a chess-fan site, to push that truncated vision of what he thinks chess should be (yeah Irv I know you're gonna insist "Not should be, but IS")...that's all a mystery to me.

Cricket, highly popular around the world (although not in the US), can literally last for days. A Test match (game) is commonly four to five days. Three two hour sessions per day, with lunch and tea breaks in between.

And yes, it can end in a draw.


In Mark Goodwin's defence, he may be referring to the average time a typical Sox-Yankees game usually lasts. And yes, 3.5 hrs usually is par for the course.

"Cricket, highly popular around the world (although not in the US), can literally last for days. A Test match (game) is commonly four to five days. Three two hour sessions per day, with lunch and tea breaks in between."

I'm not sure cricket is the best example though, as the test version seems to be dying out in favour of one day or twenty-twenty cricket, which is much shorter. Hopefully chess won't go down that same road, but on the other hand these shorter versions do seem to be far more popular, and lucrative.

Personally, I agree with what Svidler said during the grand prix. There is nothing wrong with shorter time controls, just as long as it doesn't replace the classical events. If people think that blitz tournaments will get on TV, or attract more sponsorship, or whatever, then fine, organize as many blitz tournaments as you like. I just don't see why classical chess also has to die out.

I fully agree with KinG (and Svidler, of course) with the fact that theres no problem in having lots of blitz tournaments, should it be on tv, internet, etc. The point may be in doing something like tennis, for example. They have quick surfaces where a point lasts for just a few minutes, and slow ones, like Roland Garros, where a match lasts for hours, especially when going to tie-breaks. And those various ways coexist without harming each other, everyone finds his pleasure. I think FIDE (or someone, GlobalChess, whatever) should take this into account and include blitz and even lightning into rating and developing full serious tournaments, the same way tennis does. I think it would help everyone, players, fans, and chess itself.

Just wanna add on the time control for Go.

For some of the biggest Go tournaments, each player is given 8 hours thinking time, and each game could take 2 days.

Another interesting thing to notice: the rounds might be separated by weeks. so the entire tournament takes days.

Ivanchuk - Bu - Chucky is on a roll. Nice to see blunders in the opening at such a high level.

Nice? Well yes, if by nice you mean horrible. (And that's no matter what level...blunders are always horrible!)

Correction on my comment on Go :

... in the same tournament, one round and the next might be separated by weeks. So one entire tournament could takes MONTHS.

Interesting games we're getting eh. Poor Bu. I was watching it live and felt his pain...do i resign on move 9? ugh. or drag it out for an excruciating hour
without the ghost of a chance against a top GM? ugh again. He must`ve forgotten the queen was on c2 not d1 coz of Ivanchuck´s slightly unusual move order. Anyone got any insights on where exactly Cheparinov went wrong? single move or general plan? (no transmission interruption jokes pls :))

Another question! Ivanchuk is on 3/3, a "+800" rating perf. How is it possible to calculate the performance of either 0/3 or 3/3? Theoretically the hypothetical player could be rated 10,000?! Or 0?
Thanks for any info

According to crafty, he went from += to +- in a single, horrible Bh5?? allowing Bxb5+!. Not a mistake a player of his level would oridinarily be expected to make (possibly excepting Ivanchuck, who is often somewhat insane).

levski_sofia wrote:
Today is Topalov - Cheparinov, quick draw is expected.

IChep as White: Kh1, Re1, Qf2, Bg2-d4, Pa2-b2-c4-h2.
VTopa as Black (to move): Kg8, Rc8-d8, Be8, Qe7, Pa5-c6-d5-f7-g6-h7.

2rrb1k1/4qp1p/2p3p1/p2p4/2PB4/8/PP3QBP/4R2K b - - 0 28

White is down two pawns and the B:R exchange; Shredder_11 evals as -.8/12(depth).

Next from Topalov came 28... Qe7g5, the right move (protecting Qe7 from Re1, and defending f6 against Qf2-f6 threatening mate by Qf6-g7).

But then from Cheparinov came 29. Re1-e5? (-3.2/12) instead of forking with Bd4-f6 (which would have forked Black's Qg5 and Rd8).

Duif wrote:
Cricket, highly popular around the world (although not in the US), can literally last for days. ...

And yes, it can end in a draw.

Ties and Draws

Sometimes there are lots of ties in soccer. But in soccer not all ties are equal.
There is a very different spectator experience from a 4-4 tie versus a 0-0 tie (where the rules seem to create an imbalance that gives too much power to defense over offense).

My guess is that...
Soccer fans in Europe did not switch to the semi-unjust Bilbao scoring system (3/1/0) because there were too many 4-4 ties, or for any reason having to do with ties.
Rather it was the extremely low rate of scoring that prompted Bilbao. It just happens that by its very nature, extremely low scoring tends to produce a lot of 1-1 and 0-0 ties.

In chess, most hard fought draws/ties are loosely equivalent to the 0-0 form of a soccer tie. There is no ongoing scoring during a chess game. Nor do most high quality draws contain major shifts in advantage (often some minor shifts):

In soccer a team can be losing 4-1 and come back for an exciting tie 4-4. In contrast, in grandmaster chess (and even most class level chess), when you fall behind by say a knight, you might as well resign. Chess is not a sport that facilitates comebacks like most other sports can.

That is a funny looking chair. Look how unevenly matched the chairs are! That says a lot about player privileges. This could be distracting at times. Luckily it was against Anand otherwise we would have seen a chairgate. Imagine the champions asking for and playing with a diffent size and type of pieces for their color. Or half the board and pieces from a different set.

Chesshire Cat,

Of course it is not possible, it is 'undetermined'.

'How is it possible to calculate the performance of either 0/3 or 3/3?'

Could not Cheaparinov have thrown that game without being so obvious as GeneM pointed out. He will go 0-2 to his master and hope to pocket the Bulgarian Bonus that Topalov will win, as I'm sure he would graciously give that to his grasshopper for his assistance. Something seems wrong when a bonus is anounced just when the tournament starts and excludes 67% of the players. Like the comments by everyone...too bad there are some crooks in chess that spoil the game(s) for all......

Chesshire Cat,

I would imagine that they simply calculate the rating starting from which you would expect a 3/3 or a 0/3, and give that as the rating performance.

i.e. if a 3500 rated player would be expected to go 3/3 against Topalov, Bu, and Radjabov, then that would be Ivanchuk's rating performance. The fact a 10000 rated player would be expected to do the same is irrelevant. He hasn't shown that performance, he's shown a 3500 performance.

Anyway, I am just guessing but it makes sense to me.

Buahahaha, that is a RIGHTEOUS chair. Total Alice in Wonderland or Roald Dahl extravagance. They should have played in giant peach.

lol Righteous chair. A judgement throne?

One easy way to calculate rating performance for a perfect score is to subtract 0.25 from the score and calculate performance for that. Say for 3/3, we calculate performance for 2.75/3. People with that rating are expected to get 2.75, which will be rounded off to 3, according to the actual score. But people below that rating are expected to get less than 2.75, which will be rounded to 2.5, lower than the actual score. So that's the minimum rating expected to get 3/3.

The result is the same as following mungeono's suggestion.

Some things are easy but not correct.

Most tournament reports do not list a TPR for players with 100% or 0% score.

The FIDE handbook '10.0. The working of the FIDE Rating System' indicates that 'For a zero or 100% score dp is necessarily indeterminate'. dp is the rating difference between your performance and the average rating of your opponents.

Nothing extraordinary about that. Perfect score in a small sample is also why sometimes it is impossible to measure someone's IQ.


I don't think I am not correct, if you follow my explanation. Notice that I wasn't trying to calculate the precise rating for a perfect score.

Of course, the formula is asymptotic, hence strictly speaking a perfect score will produce infinite difference.

However, what my formula gives is the smallest rating difference which will produce expected score which when ROUNDED off produces the perfect score.

For example, to get a PRECISE 10/10 score, the rating difference must be infinite, hence the rating performance also infinite. That's what the FIDE statement meant. What my formula calculates is the rating that is expected to get not exactly 10/10, but 9.75/10. And in practice, of course you cannot get a score of 9.75, so the score is rounded off to 10.

In that practical sense, I believe my explanation is correct :-).


I don't think the issue with the rating formula is "small sample". The formula itself is asymptotic. No matter how many games (sample) you have, if it is a perfect score, the rating difference is infinite.

Yes, that's right. My previous post mixed up two different things. First, that any perfect score will yield an unmeasurable, infinite rating. Second, that in chess such perfect scores will probably only turn out over a small number of games.

So if someone goes to their first tournament and wins every game, do they not get a rating until they draw or lose?

Very funny, IWillWorkHarder.

Anyway, with Chucky winning again by beating Chepa, time to calculate 3.75/4

No seriously, what happens in that case?

3.75/4 (about 94%) against 2731 average, you add 444 by http://www.fide.com/component/handbook/?view=article&id=75
to get 3175.

So Ivanchuk would have a 3175 rating performance if he had 3.75/4- but he has 4/4.
There you go..


Googling for 5 minutes gives me the following from "http://fide.com/official/handbook.asp?level=B0210"

Ru = rate of unrated player
Rc = average rating of competition

"10.23 If he scores more than 50%, then Ru = Rc + 12.5 for each half point scored over 50%. (GA `94)"

So if I (definitely an unrated player) played as Chucky's replacement and score 10/10, I will be rated 2731 + (12.5 * 10) = 2851, I think.

2851? The all-time highest record hold by GK, not bad. Oops, time to wake up from my dream :)

We have "Audi" for 0-0-0-0. Anything for Chuky's 1-1-1-1?

A strings of wins has been called a picket fence.

Ivanchuk might be heading towards his best result in a elite tournament in a relatively long while.

Ivanchuk produced the greatest results early 1990s, but afterwards his slowed down (compared to his early standards). He hasn't won any elite tournament since at least 2000. He wasn't even invited that often anymore to elite tournaments, until this year. Good to see him back. Hopefully he won't have accidents in the rest of MTel.

By the way, a great result for Ivanchuk here will keep Carlsen from reaching a higher rank towards the top for now.

I don't have time to calculate, but 3.75/4 against opposition that Ivanchuk has will definitely produce rating performance at least 3100. And that's only the lower bound.

Another common problem with unrated players is that often in the tournament they play in, there are other unrated players as well. It's not realistic to calculate rating performance in such cases. That's why it is more practical to give them fixed additional rating points, as in the FIDE system.

I thought that performance rating for a tournament was calculated like this:
"A performance rating for an event is calculated by taking (1) the rating of each player beaten and adding 400, (2) the rating of each player lost to and subtracting 400, (3) the rating of each player drawn, and (4) summing these figures and dividing by the number of games played."

So Ivanchuk's performance after four rounds is 3130,5.

Soikins, the very primitive formula you just described WAS the official system used by the USCF (which was the prototype for FIDE's version of Elo ratings), until 15 or 20 years ago.

Today both FIDE and USCF use somewhat more complicated algorithms. But I the formula you sketched still gives a close approximation. I've been using it regularly to estimate my own rating gain or loss from each tournament I play in lately, and it's always coming up + or - 1 or 2 rating points from the actual result that shows up in my MSA record.

There is a big difference, however, from the way FIDE assigns a performance (and hence an initial rating) to an unrated player, and the way the TPR is computed for those with established ratings.

For previously unrated players, FIDE applies the rule given in the comment by "tsn" above. You can see that it greatly downplays the player's RESULT (if positive), and instead emphasizes the average rating of opposition - so the performance rating and initial FIDE rating assigned to an unrated player won't differ much from the average opposition rating, even if they go 10 / 10. (However, if the unrated player has a minus score, I think a different formula applies.)

I learned this the hard way when I got my initial FIDE rating some 2-1/2 years ago. I scored 5.5 / 9 against opponents whose FIDE rating averaged 2285. In fact I'd had 5.5 / 8, and coming into my final round game (which was against GM Yudasin), I was salivating over the thought that if I could manage to beat him, my initial FIDE rating would come in close to 2450. Turns out it would have been less than 2350, because of that formula that awards a measly 12.5 extra points for each +1 in your final score. (In the event, I lost, and earned an initial FIDE rating of 2309.)


The formula you gave is not the one used by FIDE. However it is a useful approximation, because in the actual formula a 400 rating difference means the expected score is more than 90%. So it makes sense to add 400 to the rating of the opponent you beat.

Sorry to go off topic, but with all this rating talk there probably isn't a better time to ask these questions. And yes, I'm being lazy by not rooting through the FIDE site for the answers.

First, does FIDE update the ratings after each game (like on a chess server), each tournament, or only for each rating list (every 3 months isn't it?)? I believe it's only for each rating list. If so, why don't they do it after each tournament? Is it just too much hassle to do it in a reliable and timely manner considering the results must be submitted, verified, entered into the database, etc? I ask because it seems the resulting rating discepancy could be quite large if, say, during a series of tournaments one player achieved much higher than his expected result against a small number of opposing players. Game by game updates would downplay his rating gain because eventually the wins would be rewarded with smaller gains and the losses with more severe punishments. Taking it to an extreme, picture two similarly rated players competing in a very long match where one guy is able to maintain close to an 80% score for the duration. His rating would eventually 'settle' with game-by-game updates whereas it would continue to rise indefinitely with no updates. Anyway, I do realize this might not be much of a problem in practice since this sort of thing is unlikely to happen. Just wondering....

I visited this page first time to get info on people search and found it Very Good Job of acknowledgment and a marvelous source of info......... Thanks Admin! http://www.reverse-phone-look-up.net

That was a more than healthy advice. Thanks for your article and the discussion.

Twitter Updates

    Follow me on Twitter



    About this Entry

    This page contains a single entry by Mig published on May 9, 2008 1:06 AM.

    Underdogs Bite in Baku was the previous entry in this blog.

    Incredible Ivanchuk at MTel is the next entry in this blog.

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