A 30 month tournament

by Martin Fischer
2/18/2016 – Despite computers, engines, databases and critical voices - correspondence chess is very much alive. Of course computers have changed correspondence chess but to play it with success handling engines well is not enough. Leonardo Ljubicic knows what modern correspondence chess is about. He just won the 28th Correspondence World Championship with 10.0/16.

30 months of competition

Nowadays, a match for the world championship in chess lasts about three weeks. For correspondence players this is not much - even though they do no longer exchange moves by sending postcards around the world. But the  final of current world championship began 10th June 2013 and the tournament is not yet over.  But the winner is already certain. The 49-year old engineer Leonardo Ljubicic from Omis, a small village near Split, is the new king of the correspondence players. For him, playing in the final of the world championship meant to spend about two to three hours per day analysing chess. At weekends twice as much.

Over-the-board chess is a kind of mental wrestling match between two human beings but todays correspondence chess is more like a kind of scientific duel, fought on a virtual chessboard. One thing is particularly important: using the possibilities of modern computer technology efficiently. For once, the human factor is not a source of error but helps to get better results.

A common prejudice about correspondence chess has it that it does not make much sense any more in our time of strong engines and easily available good hardware. Humans would basically do not much more than executing the moves of the engines without contributing much of their own.

Well, of course, correspondence chess has undergone drastic changes since the times of the Australian Cecil Purdy, winner of the first official correspondence chess championship 1958. But this kind of chess still has its supporters and friends as the extensive tournament schedule of the International Correspondence Chess Federation, the ICCF, and the many national correspondence chess organisations prove.

Leonardo Ljubicic, winner of the 28th World Championship in Correspondence Chess

Correspondence chess is more or less played without without an audience. After all, the players do not face each other directly but are sitting in the private of their study and exchange moves on a virtual board. But the final of the 28th World Championship in Correspondence Chess was played on the ICCF webserver and this allowed spectators and fans to follow the tournament on the ICCF and to get information about the players.

With eight draws and four wins the new World Champion scored 10.0/16, won more games than any other player and achieved a smooth victory. You see a lot of "correct chess" in correspondence chess and thus only about 13% percent of all games were decided. But a look at the games reveals that there were only a few short, friendly draws.

After winning games after eight and thirteen months the Czech player Boukal took an "early" lead in the tournament. But at the end of 2014 Ljubicic scored his second and third wins and took the lead. On 30th September 2015 he then managed to bag his fourth win.

The new World Champion of Correspondence Chess
indulging in his hobby

Among the participants were a number of remarkable players. For instance, Mark Noble, a Correspondence Grandmaster from New Zealand who also is an excellent blitz player who has won a number of titles in this discipline. On top of that he also plays Boule for New Zealand - a sport that is quite popular in New Zealand and other countries of the former British Empire.

Not to forget the 76-year old Italian Fabio Finocchario, who won the 25th World Championship in Correspondence Chess. For these players - and for all the participants in the tournament - chess is a hobby. A hobby that takes a lot of time but that still remains a hobby. But their continuing success proves that correspondence chess is more than copying moves from engines.



Das Final of the 28. World Championship of Correspondence Chess

Games of the final


Martin Fischer has been tournament director on the playchess.com server for many years. He plays in Germany for the club of Johanneum Eppendorf, in the season 2013/14 in Germany's second league.
Feedback and mail to our news service Please use this account if you want to contribute to or comment on our news page service


Rules for reader comments


Not registered yet? Register

Bendigo Bendigo 2/18/2016 12:12
I played correspondence chess in the pre-computer age and loved it but gave it away once I knew I would be playing against a computer. What this article doesn't say is how the humans still have an input. I would be fascinated to hear how GM Ljubicic imposes his skill and judgement on that of the remorseless computer analysis.
ashperov ashperov 2/18/2016 12:31
I second what Bendigo says. Want more info
Mr TambourineMan Mr TambourineMan 2/18/2016 02:15
Sometimes I think that correspondence chess is so far above our heads that the outcome of the games are more or less by chance. And also Bendigo that no one could explain how the humans can have an input. And if someone could I think we can talk as much of an human output but thats a negative approach but still to be taken seriously. I dont know if someone had won these tournament and said "The computer made every move" and "My job was a mechanic only to see the computer run properly". But the intencive to say this if you win are of course non as its degenarative agianst one self and is much better to say som bla bla bla about how good one are at bla bla bla.
Hawkman Hawkman 2/18/2016 02:40
Bendigo "What this article doesn't say is how the humans still have an input."

Agreed. By not doing so, the author is admitting it's not true.
Exclam Exclam 2/18/2016 03:14
I play correspondence chess and you guys could enter your engine against these guys and have them wipe the floor with you. The difference is, most of these guys are already elite level chess players so when they watch a computer evaluation and they have 10 to choose from all with roughly the same evaluation they will move that analysis forward 15 moves for example and then redo their analysis to see the resulting positions, they will look for theoretical won endgames. They also can tell when a weak player is using an engine and by using their chess knowledge help the engine choose the strongest continuation when it has to decide between multiple moves of the same evaluation. Spouting that it isn't true because the author didn't describe how it is done is just asinine.
hansj hansj 2/18/2016 03:35
Did you notice the vast number of draws?
Mr TambourineMan Mr TambourineMan 2/18/2016 04:04
Look at the games. You only have to look at the end position that comes after aprox. 40 move (many times as soon as move 25) and we are in middlegame position or early ending and it are much to play for but theyll take draw anyway. I has never been a friend of anti draw rules but her I could definitely be! Why so many Draws. I guess one fact is that players with computers start to think like this "I cant outsmart this guy/computer" best to take a draw and focus my energy on the games I feel I can do progress. Also. This years Winner waited with taking Draws. Get my Point?
genem genem 2/18/2016 06:27
This article desperately needed to show a few games and moves where the winner overruled the judgment of his computer fritz (little 'f'), to support the premise of the article. Unfortunately no such moves are shown.
More likely, the winner uses several fritzes, like Komodo, Rybka, Houdini etc, and sees that the engines often enough slightly disagree about which move is best. The human likely does not "overrule" the fritzes, but rather just picks his favorite move among the 2-3 moves the fritzes assess as best.
I suspect that these games are little different from plain fritz versus fritz games. I could be wrong about that, but unfortunately this article gives no insight into such questions.
chessmatt chessmatt 2/18/2016 07:26
To compare CC to OTB makes little sense today. It is a completely different animal using different skills, tools and a different kind of perseverance and needs to be appreciated in that context. The existence of CC and CC titles does not diminish OTB titles or accomplishments in any way. The two really should not be compared although they are obviously complementary. Think of ICCF as the academia of chess, and OTB as applied practice. There can be no doubt that the efforts of strong ICCF players push the limits of theory. It might not appeal to you, and that is fine. I think the mistake some ICCF critics make is to assume strong ICCF players want you to think they are strong OTB players or that an ICCF rating should have any correlation to OTB ratings. That really isn't the point.
chessmatt chessmatt 2/18/2016 07:59
This may be of interest as it touches on some of the comments

adp adp 2/18/2016 11:05
Are you kidding? Engine-vs. Engine. Who cares? The first 15 games I went through were drawn. This is a joke.
Queenslander Queenslander 2/18/2016 11:20
New Zealand connection: Bowls, not 'boule'. FM Mark Noble (FIDE 2224) is OK at blitz (FIDE blitz 2227) but nothing remarkable.
Karbuncle Karbuncle 2/19/2016 01:21
I'm also on ICCF, having just got my SIM title and now rated over 2500. I started at the bottom with an 1800 rating and have faced computers the entire way up to 2500. I also don't have good hardware, yet I made USA CC champion along the way to my master titles, and have won numerous tournaments. So how do I succeed when I have an 8-year-old computer? Answer: Opening preparation. Very rarely can you win a CC game these days if you don't get a middlegame with chances. That's why so many games are drawn now on ICCF, and only those special games of opening prep get the win. That's how I approach the game, and I spend a lot of time reviewing my opponent's games, looking for weaknesses in their opening preferences. It pays off just enough to keep me climbing the ranking ladder. I should be ranked #8 in the USA by the next rating period in April. Again, I had to climb up from the very bottom, and it was a long and difficult climb in modern cc era.

T J Bozat T J Bozat 2/19/2016 03:46
According to How Purdy Won by Frank Hutchings and Kevin Harrison, the First World Correspondence Chess Championship ended in June 1953, not 1958 as stated above.
JonWH JonWH 2/19/2016 10:49
It should be called "The World Championship Management of Several Engines Tournament". There is no resemblance to normal chess. If managing computer output is of interest to you, then fine. But IMO it's boring and of no interest to a typical OTB chess-player.
Karbuncle Karbuncle 2/20/2016 03:38
@JonWH, that's a typical ignorant view. If it was just simply an engine contest, people like me would have no hope of advancing up the ranks. Yet I am now a Senior IM with a 2500+ ranking on a computer that is 8 years old, sporting a Q6600. It's not just about the hardware, it's about how you USE everything at your disposal. This means lots of research and analysis techniques the computer would otherwise be quite weak without. That's why I always invite those that trash-talk to try and do what I did: Start at the bottom with an 1800 rating, use only a Q6600 CPU, and become a National Champion in addition to making the Senior IM title. I'm not done yet though. I'm going form the GM title next. To put it quite simply: You get out of centaur cc exactly what you put into it. So if you just thinks it's letting the engines play, you're not going to have much fun and you're going to a hit a wall on your rating.
Edward Labate Edward Labate 2/20/2016 09:48
Our 'new world champion' is OTB...UNDER 2200!!!!!!!!!!!!!!!!!!!!!!!!
I don't even want to hear that he's anything BUT very proficient at using engines. The old champions like Purdy, Berliner, Estrin, etc. were all OTB IMs or better. Todays PRETENDERS aren't even national masters. Give me a bleeping break.

Correspondence chess has ceased to be relevant as a human achievement!
Mendheim Mendheim 2/20/2016 06:02
Modern correspondence chess is not just executing computer moves, but working on strategic and tactical ideas by use of computers, which is a big difference. Computers may fail in understanding specific positions and plans, and often you cannot trust their evaluations. The human factor is still the last resort.
Yet, what cannot be denied, is the tremendous rate of draws even though players are looking for ways to achieve a decisive advantage. By chance, Ljubicic had White against the 4 last players, and he won all those 4 games, but he did not beat any of the players No. 2-13. The rate of draws among the first 14 players was 94%. When does ICCF start to think about its draw problem?
sava_ sava_ 2/21/2016 01:56
Unfortunately, most of the people who comparing ICCF with OTB chess never had an experience of "advance" chess and do not understand what they are talking about. They are trying to compare two completely different disciplines from the point of view of one of them. It is like comparing a runner with formula - 1 pilot. Yes, it is completely different set of skills. In fact, some very strong titled OTB players performed badly in ICCF tournaments. And I do not think that the only reason is that they did not know how to use a computer. The problem is that positions which looks promising for the OTB chess can be not suitable for ICCF and vice verse. It requires a long practice in correspondence chess, before a correct positional intuition (suitable for ICCF) will be developed. So, for people who are trashing achievements of the winning the ICCF World Champion title, you do not have a slightest idea how much time and effort it takes. So, if you want your opinion to have any weight at all try to win any serious ICCF tournament yourself, may be than you will be qualified to express your opinion on this topic.
Dragon Mist Dragon Mist 2/21/2016 04:14
@bendigo, @ashperov, @Hawkman, @genem: I've given the ChessBase an extensive interview and 2 fully annotated games around Christmas. They were a bit slow publishing this article, and promise to have games and interview published "very soon". You'll then be able to judge for yourself.
@hansj, @Mr TambourineMan: draw rate at top CC nowadays is almost 90%. I see no solution to this, but you can read more about it and one proposal in the above article "How many points should you get for a draw?"
@adp: engine vs engine is on Playchess, not on ICCF.
@T J Bozat: there are several mistakes in the above text, also to note is that Omiš is not a village, but a town of 6,000, for instance.
@Karbuncle: I admire your efforts so far Wolf, let's see how you fare in the league with big boys now :)
@Mendheim: let's wait till the tournament is over to see if the games that I won were against the bottom four. Also, in a tournament with 85%+ draw rate, if you lose one game you should expect to be at the bottom. :)
@sava_: exactly! One of the most underrated things is that, compared to top CC, human play is simply erratic. Most of the times you see OTB game with commentary "white is slightly better" it is just a dead draw from CC perspective. Enormous effort is being put in achieveing ANY kind of advantage, let alone win the game in the end.
SushiDefense SushiDefense 2/22/2016 06:50
I have to agree with Edward Labate. These matches are basically (tweaked engine + user) vs (tweaked engine + user). Convince me the engine doesn't do 90% of the work.