Attached is a database of 17485 LSS corr games finished between Jan-September 2009. I'm sure some of you already have these games but I'm just as sure that there are others who will find them useful. Hint, hint, nudge, nudge...there are a lot of interesting games here that aren't the Najdorf!
Attachment: LSS-2009-09.zip (1570k)
Thanks ;)
![[hr]](/mwf/v12/flags/hr.png "Croatia")
Thank you Rowlando. What is LSS anyway?
> What is LSS anyway?
It's the site where IECG server games are played. http://lss.chess-server.net/
Earlier this month IECG made the following announcement:
"Due to the fact that email play declines quickly in popularity making it impossible to build tournaments in a reasonable time, and due to the practical issues of email (viruses, spam filters) making this playing mode difficult, IECG has decided to stop its operations in December 2010 and to transfer its activities to our partner the Lechenicher SchachServer Server (LSS) in the meantime."
![[eg]](/mwf/v12/flags/eg.png "Egypt")
Thanks but my first impression is that most of the games are of very low quality....
Dr.D
P.S.I digged deep into the database for about 3-4 hours so my conclusion is not based on a quick glance....
Dr.D
P.S.I digged deep into the database for about 3-4 hours so my conclusion is not based on a quick glance....
![[it]](/mwf/v12/flags/it.png "Italy")
true about the very low quality, but I am happy anyway as I am a games collector and I missed all those games! (except one)
Anyone else knows somewhere where to download some more corr games? (except the well known FICGS and ICCF)
Eros.
Anyone else knows somewhere where to download some more corr games? (except the well known FICGS and ICCF)
Eros.
> Anyone else knows somewhere where to download some more corr games?
If you have UltraCorr2 you'll probably have many of these games but I figured it wouldn't hurt to post it anyway. Here's a db of 44539 LSS games up to the end of 2008.
Attachment: LSS-Full-2008-12.zip (3886k)
![[se]](/mwf/v12/flags/se.png "Sweden")
is there a download-page at ficgs or do you have to be a member to download?
If you don't supply IECG with a FIDE or national rating you start at 800 so you can't judge the quality of the game by ratings alone because that 800 player might be a very strong centaur who's on his way up the rating charts or a patzer that learned the game yesterday. The only way to tell is by taking a closer look at the games. There are a number of Rybka Forum regulars that play on the LSS server, all of whom are quite skilled at correspondence, so I can assure you there are many high quality games in this database. I can personally vouch for a couple dozen games, one of which features a refutation of a sideline in the Breyer that book authors would appreciate. ;-)
This database is just like any other -- there are great games, crap games and everything in between.
This database is just like any other -- there are great games, crap games and everything in between.
![[us]](/mwf/v12/flags/us.png "United States")
In my estimation about 40-50% of these games are keepers. That's not bad at all.
thank you again! another 13.000 good games :-P
Eros.
Eros.
Careful there, Eros. I'm getting a high percentage of wrongful outcomes. It turns out this is very low quality even after filtering out the worst stuff. If you're using this for booking it has a pretty high sabotage rate!
> If you're using this for booking it has a pretty high sabotage rate!
Are you joking? Or referring to games that were decided by forfeit when one side violates the time control? There are no sabotaged games here, the files came directly from the LSS site. To make these games more suitable to book making, games that were decided by forfeit should be removed. You can do this in Chessbase by doing a search for the word "forfeited" (without quotes)on the Annotation tab of the Filter window or simply sorting by the VCS column -- games that have been forfeited will have annotations, good games won't. In theory, this should remove most if not all of the games with wrongful outcomes.
When someone violates the time control, the side that won by forfeit must claim the win to make the result count (i.e. effect ratings etc) and to claim the win you have to show the Arbiter that you would have won the game had play continued. If the claim is unsuccessful or no claim is filed, then the result is null and void and the word "forfeited" is written into the notation.
I was talking to Eros, and he would know what I was referring to. I am not suggesting that the motive of the players is sabotage. I am referring to the effect on an opening book which is based on games with a high percentage of wrong outcomes. (I speak with some authority on this.) A good source of games generally has 1-2% errors. This batch is around 11-12%.
> This batch is around 11-12%.
I presume you have some sort of automated method for gathering these statistics? Is the method documented somewhere or is it a trade secret? (for lack of a better description) Just curious.
Yep, and even I don't have documentation.
thank you for letting us know, Nelson, anyway I don't really care... it's a long time since I have stopped filtering the games I get. Not completely stopped, but I don't waste anymore as much time as before. What I do now is simple: I get new games? I "throw" them all in my database and click on the chessbase command "search doubles", without wasting time to check manually or with other complicated search commands for other games to delete or fix, as I know I will never be able to check every single one; some wrong games will always survive anyway. Moreover, I am sure that some games have been "created" by someone, they in reality have never been played. I have to add that my skills of finding doubles and bad games from a large database are quite bad, and I am jealous of your cleaned database(s) and I admire your continuous "cleaning work".
Fortunately I don't import games to make books ;-) I search games usually when I work manually on book, so usually I realize when a game is bad or not.
Fortunately I don't import games to make books ;-) I search games usually when I work manually on book, so usually I realize when a game is bad or not.
The Americans spent a fortune trying create a pen that would work in zero-gravity. The Russians had the same problem. They solved it by using a regular wooden pencil.
Your solution to this problem is Russian in spirit. By doing it the way you do it you have freed an enormous part of your finite lifetime for better things. You envy my database; I envy you.
Your solution to this problem is Russian in spirit. By doing it the way you do it you have freed an enormous part of your finite lifetime for better things. You envy my database; I envy you.
Hi,
Thanks for the DB, one of the problems is that there are so many games with the same prob, below is an example where white is totally winning but!!
Hunt,Dylan (800) - Van Den Berg,Eddy (1597) [D07]
OP-2006-0-00030 LSS, 29.04.2006
1.d4 d5 2.c4 e6 3.Nf3 Nc6 4.Nc3 h6 5.Bf4 Bd6 6.cxd5 Bxf4 7.dxc6 bxc6 8.e3 Bd6 9.Bd3 Nf6 10.0-0 0-0 11.Rc1 Re8 12.Na4 e5 13.dxe5 Bxe5 14.Nxe5 Bb7 15.Nxc6 Qd7 16.Bb5 Qe6 17.Nd4 Qxa2 18.Bc4 Qxc4 19.Rxc4 Ba6 20.Qc1 Bxc4 21.Qxc4 Rac8 22.Ra1 Kf8 23.Qc5+ Kg8 24.Nc3 Nd7 25.Qxa7 Ne5 26.Qc5 Nd3 27.Qb5 Nxf2 28.Kxf2 Rb8 29.Qc5 Rxb2+ 30.Kf3 Rb7 31.Ra7 Rbb8 32.Rxc7 g5 forfeited 0-1
Regards
Thanks for the DB, one of the problems is that there are so many games with the same prob, below is an example where white is totally winning but!!
Hunt,Dylan (800) - Van Den Berg,Eddy (1597) [D07]
OP-2006-0-00030 LSS, 29.04.2006
1.d4 d5 2.c4 e6 3.Nf3 Nc6 4.Nc3 h6 5.Bf4 Bd6 6.cxd5 Bxf4 7.dxc6 bxc6 8.e3 Bd6 9.Bd3 Nf6 10.0-0 0-0 11.Rc1 Re8 12.Na4 e5 13.dxe5 Bxe5 14.Nxe5 Bb7 15.Nxc6 Qd7 16.Bb5 Qe6 17.Nd4 Qxa2 18.Bc4 Qxc4 19.Rxc4 Ba6 20.Qc1 Bxc4 21.Qxc4 Rac8 22.Ra1 Kf8 23.Qc5+ Kg8 24.Nc3 Nd7 25.Qxa7 Ne5 26.Qc5 Nd3 27.Qb5 Nxf2 28.Kxf2 Rb8 29.Qc5 Rxb2+ 30.Kf3 Rb7 31.Ra7 Rbb8 32.Rxc7 g5 forfeited 0-1
Regards
my suggestion is, like for playchess games you may type "disconnected" to find such games, here you can try "forfeited" ;-)
But unless you check those games manually anyway, together with the real wrong games you will also delete correct games, where one player who forfeited was also really losing. As I said before, to avoid such a boring work (like loading engine at the end of every game to check eval) since some time I prefer to keep everything as it is...
But unless you check those games manually anyway, together with the real wrong games you will also delete correct games, where one player who forfeited was also really losing. As I said before, to avoid such a boring work (like loading engine at the end of every game to check eval) since some time I prefer to keep everything as it is...
The Americans spent a fortune trying create a pen that would work in zero-gravity. The Russians had the same problem. They solved it by using a regular wooden pencil.
:-D Thanks Nelson! I will be retelling this story a lot in the future. Spending valuable resources designing new, but not necessarily better approaches to solving problems is rampant in my business...
:-D Thanks Nelson! I will be retelling this story a lot in the future. Spending valuable resources designing new, but not necessarily better approaches to solving problems is rampant in my business...
![[ch]](/mwf/v12/flags/ch.png "Switzerland")
It's a fascinating story, but unfortunately, false. :-) Snopes has the story as always.
/* Steinar */
/* Steinar */
![[mx]](/mwf/v12/flags/mx.png "Mexico")
Well slap me with a wet fish. I was going to launch into a tongue-in-cheek diatribe about how the story may not be true, but makes a nice point anyway. But no, strike that idea. In your shoes, knowing that someone had made a provably inaccurate statement, I surely would have offered the same correction and I might not have been as gentle doing it. Nicely done.
![[nl]](/mwf/v12/flags/nl.png "Netherlands")
You judge the quality of a collection of games based on the correct outcome instead on the sequence of (good) moves. Obviously this has got to do with the automatic generation of books where percentages determine which moves is played. Forfeited games where there was actually an advantage for the player that forfeited is of course undesirable for a bookmaker, but I would like to add the nuance 'lazy' bookmaker. In general the quality of the moves in any correspondence database is better than the these engine room games. The percentage of incorrect outcomes may be lower but so is the quality of play, or are you suggesting that engine blitz games with just a few seconds per move are to be preferred above typical correspondence reflection time where both high quality GM games (theory) are consulted as well as chess engines with considerable more calculation time than just a few seconds. Depending on how you look at it your remarks about the quality of the database is totally missing the mark. Using large quantities of 'correct' engine blitz games may be a statistical recipe for 'success' in the engine room, this does not say a whole lot about the quality of the games. Compared to the LSS database the engine room archives are of very low quality.
Well, you are entitled to your opinion. I would submit that "quality" is a subjective word because all of us lack the thing that could allow us to judge objectively--absolutely correct answers to every position. A chess engine with a poor reputation, playing blitz, can sometimes unwittingly come up with a superior move in a given position than a correspondence player pondering for days. Perhaps that is not the way to bet, but it is the truth. It isn't for me to sort out these vagaries; no one could. I just collect games and try to weed out those that contain verifiable empirical misinformation. For example, if white simplifies the board to six pieces, why would I rely on the actual outcome of the game rather than endgame tablebases? Likewise, if the final position in a game shows one side with a decisive advantage, yet the game result shows the opponent the winner, how can that be helpful empirically speaking? The fact is that human games invariably result in a much higher percentage of misinformation than engine games, and one must draw certain conclusions from that. As for correspondence games, I grant you that some of them approach perfection. Yet taken as a whole, there are a number of turkeys in the data set, too.
The fact that an engine can play a brilliant move in just a few seconds by accident doesn't mean shit in a statistical sense and you know it.
I think I will reply some more.
I am entitled to my opinion yes, that is the standard reply of almost anybody that is short of real arguments. You are no exception in this case. All I hear is some rhetoric about *nobody* can tell the truth in any chess position but yet you completely fail to address my point about the quality of moves when comparing engine blitz games to correspondence games. A weak engine can 'find' a move that is not discovered in a correspondence game after days of analysis ? I doubt whether this shows your real opinion of correspondence play. My guess is that you have discovered the value of correspondence games a long time ago and just try to discourage people from discovering the same. Keeps you ahead in the book race, right ? You can take offence all you want because I cannot prove any of this, but neither can you, so mutual ignorance is probably the best decision for the both of us.
I am entitled to my opinion yes, that is the standard reply of almost anybody that is short of real arguments. You are no exception in this case. All I hear is some rhetoric about *nobody* can tell the truth in any chess position but yet you completely fail to address my point about the quality of moves when comparing engine blitz games to correspondence games. A weak engine can 'find' a move that is not discovered in a correspondence game after days of analysis ? I doubt whether this shows your real opinion of correspondence play. My guess is that you have discovered the value of correspondence games a long time ago and just try to discourage people from discovering the same. Keeps you ahead in the book race, right ? You can take offence all you want because I cannot prove any of this, but neither can you, so mutual ignorance is probably the best decision for the both of us.
I am baffled by your angry reply to my meek answer. Can you calm down please? Or are you the type of person who does not brook any difference of opinion or perspective?
I am in no way whatsoever diminishing the value of correspondence games. However I do look at them as a whole, as an aggregated value. What I see is that a good number of them provide the most accurate reading of a position that is currently available. But you also will find a fair number between less-skilled players that appear to have not used an engine at all because clearly sub-optimal moves were played. My larger point is that quality is not uniform and I personally have no way of differentiating or even defining what is good and bad on a game by game basis without an unproductive (and probably speculative) diversion of finite resources.
Conversely I have the same issue with engine games, whatever their time control. Even long time-control games are at the mercy of the opening books used, engines, settings, etc. It is very possible for a three-minute blitz game to be of great quality and an hour-plus-increment game to be silly. How? Well, the former may involve two players with strong machines and very strong opening books. Thus you wind up in book-exit positions that are quite cutting-edge and from there interesting things can happen. Conversely longer games can follow already-refuted lines that add little to overall knowledge.
Given the limitations of knowledge and resources a game collector has to make practical choices, and in doing this he needs to establish some kind of philosophy about games and positional truth. Mine is quite simple. Collect everything you can and let sheer numbers and time sort out things out. What I know from years of doing this is that this approach actually works remarkably well in most positions. I grant you a good correspondence player going against my book will likely have an advantage, but that is hardly surprising.
I am in no way whatsoever diminishing the value of correspondence games. However I do look at them as a whole, as an aggregated value. What I see is that a good number of them provide the most accurate reading of a position that is currently available. But you also will find a fair number between less-skilled players that appear to have not used an engine at all because clearly sub-optimal moves were played. My larger point is that quality is not uniform and I personally have no way of differentiating or even defining what is good and bad on a game by game basis without an unproductive (and probably speculative) diversion of finite resources.
Conversely I have the same issue with engine games, whatever their time control. Even long time-control games are at the mercy of the opening books used, engines, settings, etc. It is very possible for a three-minute blitz game to be of great quality and an hour-plus-increment game to be silly. How? Well, the former may involve two players with strong machines and very strong opening books. Thus you wind up in book-exit positions that are quite cutting-edge and from there interesting things can happen. Conversely longer games can follow already-refuted lines that add little to overall knowledge.
Given the limitations of knowledge and resources a game collector has to make practical choices, and in doing this he needs to establish some kind of philosophy about games and positional truth. Mine is quite simple. Collect everything you can and let sheer numbers and time sort out things out. What I know from years of doing this is that this approach actually works remarkably well in most positions. I grant you a good correspondence player going against my book will likely have an advantage, but that is hardly surprising.
...compared to what .... ?
I have also dug deep and I have to agree. Especially the drawn games...It will take a good amount of patience to filter this batch. There are also some really good games--after some time analyzing, you see some clever lines. Better to have than not to have...Thanks for the collection.
![[de]](/mwf/v12/flags/de.png "Germany")
Good starting point. Needs just a little more work...Here are a few games where the result does not reflect the possible outcome.
[Event "BI-2008-F-00008"]
[Site "Lechenicher SchachServer"]
[Date "2009.03.02"]
[Round "?"]
[White "Yarkov, Evgeny"]
[Black "Digalakis, George"]
[Result "0-1"]
[WhiteElo "2066"]
[BlackElo "1726"]
[PlyCount "56"]
[EventDate "2009.03.03"]
1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Bc5 5. Be3 Qf6 6. c3 Nge7 7. Bc4 Ne5 8.
Be2 Qg6 9. O-O d6 10. f3 O-O 11. Kh1 Bb6 12. a4 N5c6 13. Na3 f5 14. Nab5 fxe4
15. fxe4 Rxf1+ 16. Qxf1 Bd7 17. Bc4+ Kh8 18. Nxc6 bxc6 19. Bxb6 cxb5 20. Bf7
Qh6 21. Bg1 Rf8 22. Re1 Be8 23. Be3 Qh4 24. Bf2 Qf4 25. Bxe8 bxa4 26. Bxa4 Qxf2
27. Qxf2 Rxf2 28. e5 dxe5 0-1
[Event "SA-2008-0-00034"]
[Site "Lechenicher SchachServer"]
[Date "2008.10.29"]
[Round "?"]
[White "Zuccotti Bozzano, Marianella"]
[Black "Shishkin, Victor"]
[Result "0-1"]
[WhiteElo "2013"]
[BlackElo "2071"]
[PlyCount "56"]
[EventDate "2008.10.26"]
1. c4 e6 2. d4 d5 3. Nf3 Nf6 4. g3 Be7 5. Bg2 O-O 6. O-O dxc4 7. Qc2 a6 8. Qxc4
b5 9. Qc2 Bb7 10. Bd2 Be4 11. Qc1 b4 12. Bg5 h6 13. Bxf6 Bxf6 14. Nbd2 Bd5 15.
Qc2 Nd7 16. e4 Bb7 17. Nc4 Rb8 18. Rfd1 Ba8 19. Rac1 c5 20. dxc5 Qc7 21. Nd6
Rfd8 22. c6 Nb6 23. e5 Be7 24. Qd3 Nc8 25. Qxa6 Nxd6 26. exd6 Rxd6 27. Rxd6
Bxd6 28. Nd4 Bb7 0-1
[Event "ZH-2009-0-00256"]
[Site "Lechenicher SchachServer"]
[Date "2009.02.16"]
[Round "?"]
[White "Aalderink, Gerwin"]
[Black "Denayer, William"]
[Result "1-0"]
[WhiteElo "2117"]
[BlackElo "2337"]
[PlyCount "66"]
[EventDate "2009.02.09"]
1. Nf3 d5 2. d4 Nf6 3. c4 c6 4. Nc3 a6 5. e3 b5 6. b3 Bg4 7. Be2 e6 8. O-O Nbd7
9. h3 Bh5 10. Bb2 Bd6 11. c5 Bb8 12. Ne5 Bxe2 13. Nxe2 Qc7 14. Qc2 a5 15. a4 b4
16. Rad1 O-O 17. f3 Ba7 18. Nf4 Nxe5 19. dxe5 Nd7 20. Nd3 Qd8 21. Bd4 Qe7 22.
e4 Rac8 23. Qc1 Rfd8 24. Kh1 h6 25. Rd2 dxe4 26. fxe4 Qg5 27. Rf4 Nf6 28. Qb2
Bb8 29. Rdf2 Nh5 30. Rg4 Qe7 31. Nf4 Nxf4 32. Rgxf4 Rf8 33. R2f3 Rcd8 1-0
[Event "BI-2008-F-00008"]
[Site "Lechenicher SchachServer"]
[Date "2009.03.02"]
[Round "?"]
[White "Yarkov, Evgeny"]
[Black "Digalakis, George"]
[Result "0-1"]
[WhiteElo "2066"]
[BlackElo "1726"]
[PlyCount "56"]
[EventDate "2009.03.03"]
1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Bc5 5. Be3 Qf6 6. c3 Nge7 7. Bc4 Ne5 8.
Be2 Qg6 9. O-O d6 10. f3 O-O 11. Kh1 Bb6 12. a4 N5c6 13. Na3 f5 14. Nab5 fxe4
15. fxe4 Rxf1+ 16. Qxf1 Bd7 17. Bc4+ Kh8 18. Nxc6 bxc6 19. Bxb6 cxb5 20. Bf7
Qh6 21. Bg1 Rf8 22. Re1 Be8 23. Be3 Qh4 24. Bf2 Qf4 25. Bxe8 bxa4 26. Bxa4 Qxf2
27. Qxf2 Rxf2 28. e5 dxe5 0-1
[Event "SA-2008-0-00034"]
[Site "Lechenicher SchachServer"]
[Date "2008.10.29"]
[Round "?"]
[White "Zuccotti Bozzano, Marianella"]
[Black "Shishkin, Victor"]
[Result "0-1"]
[WhiteElo "2013"]
[BlackElo "2071"]
[PlyCount "56"]
[EventDate "2008.10.26"]
1. c4 e6 2. d4 d5 3. Nf3 Nf6 4. g3 Be7 5. Bg2 O-O 6. O-O dxc4 7. Qc2 a6 8. Qxc4
b5 9. Qc2 Bb7 10. Bd2 Be4 11. Qc1 b4 12. Bg5 h6 13. Bxf6 Bxf6 14. Nbd2 Bd5 15.
Qc2 Nd7 16. e4 Bb7 17. Nc4 Rb8 18. Rfd1 Ba8 19. Rac1 c5 20. dxc5 Qc7 21. Nd6
Rfd8 22. c6 Nb6 23. e5 Be7 24. Qd3 Nc8 25. Qxa6 Nxd6 26. exd6 Rxd6 27. Rxd6
Bxd6 28. Nd4 Bb7 0-1
[Event "ZH-2009-0-00256"]
[Site "Lechenicher SchachServer"]
[Date "2009.02.16"]
[Round "?"]
[White "Aalderink, Gerwin"]
[Black "Denayer, William"]
[Result "1-0"]
[WhiteElo "2117"]
[BlackElo "2337"]
[PlyCount "66"]
[EventDate "2009.02.09"]
1. Nf3 d5 2. d4 Nf6 3. c4 c6 4. Nc3 a6 5. e3 b5 6. b3 Bg4 7. Be2 e6 8. O-O Nbd7
9. h3 Bh5 10. Bb2 Bd6 11. c5 Bb8 12. Ne5 Bxe2 13. Nxe2 Qc7 14. Qc2 a5 15. a4 b4
16. Rad1 O-O 17. f3 Ba7 18. Nf4 Nxe5 19. dxe5 Nd7 20. Nd3 Qd8 21. Bd4 Qe7 22.
e4 Rac8 23. Qc1 Rfd8 24. Kh1 h6 25. Rd2 dxe4 26. fxe4 Qg5 27. Rf4 Nf6 28. Qb2
Bb8 29. Rdf2 Nh5 30. Rg4 Qe7 31. Nf4 Nxf4 32. Rgxf4 Rf8 33. R2f3 Rcd8 1-0
![[ru]](/mwf/v12/flags/ru.png "Russian Federation")
I have removed games less than 20 courses and game with a low rating. To reconsider all 17000 games rather inconveniently. It is possible to find bad games in any base. Nevertheless, now the share of bad games is less.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill