Rybka Chess Community Forum
hello, which one is the "state of the art" public available book (free or not)?
Is there any information about his author? Is it downloadable book from any site?
It's a good tournament book but it's fairly limited when it comes to general use. By that I mean that many lines are one sided. Any deviation away from the lines the author was attempting to play leads to variations that are only sparsely covered and fairly unreliable.
I'll go and suggest the Vesely book:http://www.ulozto.cz/7610611/vesely-1-0-7z
Sure, it's got some really weird marks in there, like marking good moves red, or bad moves green, and adding ? marks in there for no reason, but what matters to me is the great width and depth that it has, and the useful refutations and ideas that it contains. It has surprised me that many times it suggest best moves that all my engines were missing.
I have to thank TheHug that sent me the book since long time ago, and I think it was a Private Book until Mark Mason made it public
Thanks for the info. I'll check it out. I'm a huge fan of all-purpose opening books.
The download took almost seven hours on the second try! It timed out the first time and the download speed can drop to about 50 kB per second. Big relief that I got it the second time! So be prepared not to be hasty... I have not yet checked if I got it without errors, hopefully it will unpack and I don't get that "draw counter overflow" error message that Pia talked about. Not sure if that is a Chessbase message? I think the file is probably too big too post on the TDDB opening forum for easier download and it is also a bit against etiquette I suppose. It is also not clear to me how much of Jeroen's book is in there
If it is I will not put it on the Opening forum on TDDB and Shaun would probably not want that either. Still thanks for mentioning it Uly! I don't play much myself but I think it is hard to resist checking your own opening repertoire with the best sources that are available. Assuming this Vesely 1.0 from R.Vesely is not totally pirated from the Rybka 3 book
Sadly WinRAR and 7-Zip both could not unpack the resulting 1.318.636 kB large file, which, presumably, got corrupted during the long download. So probably it's not advisable to try this link.
> It is also not clear to me how much of Jeroen's book is in there
TheHug assured me the that the book is fine, if it was a pirate copy of another book he wouldn't let people talk about it on the forum (and wouldn't be recommending it).
>So probably it's not advisable to try this link.
That was the link I used to download it, there were also links in where it could be downloaded in several parts and another in mediafire I had trouble with, so unless another link has been created since then, the one at ulozto.cz is the best.
I have actually meet this R.Vesely over the net and I can say I'm pretty sure its not. I have seen lines that are not in the R3 book.
A 1.3GB book! How do you guys manage?!
Not sure were you got that number. My book will be close to 30GB when I'm done. Its around 10GB right now. It takes a while for the sources to pull together.
I know you can't give a lot of info on your book, but just for my curiosity how long have you been collecting games?
Since chessbase 1 for me (my m8 first got it and it was run on an Atart ST comp with a tiny black and white monitor
) although it is only in the last dedcade that access and the amount of games being played has exploded.
The book is only 1.3GB when zipped, after un-compression it goes up to 5GB.
> could not unpack the resulting 1.318.636 kB large file
The file I downloaded is 1 350 291 861 Bytes long (or 1.318.644 kB) and I have no problem unpacking it with 7-Zip
Your file seems a bit short...
You are right Ernest! I tried the download again and now it worked! Now my copy is 1.318.645 KB
1.350.291.861 bytes. I added a b to its name to distinguish the downloads and the size in "properties" is exactly the number you have, so I guess this is the right size. It opens in my Chessbase Light 2007 just fine. Thanks guys!
It does seem R.Vesely has some unorthodox opening ideas. Have you looked at all the first moves, for which there were games in the database Where did he get this material? I think he included all Steve B's or Tim Krabbé's games against the Chess Challenger 1? I remember Tim writing about the Chess Challenger in the magazine Kijk, somewhere in 1976 or thereabouts. 1. Rxh8! A killermove
Is this just a precaution against bookcrashes if the opponent plays illegal moves, or if only the book is tested but no engine or something? Does everybody make their books like this nowadays?
There are obviously problems with the database that built this book.
1. Illegal moves - a sign that the author included fragmentary games in his database.
2. e3 - the game-count and success rate are both preposterous.
3. h3 - the success rate does not remotely match what I have seen, plus the average ELO is off the charts.
4. b3 - success rate well off what I have seen
...and there are other variances that are more trivial. I have to wonder how rigorously this guy eliminated duplicates and what kind of filtering procedure he used.
As for me, I only see what moves are in the book and that's enough for me, I don't care about statistics and it has worked for me. It doesn't matter if a position has been played tons of times and has a high success rate if I can find a refutation for it that hasn't been played, so the statistics besides the moves don't matter much, but the moves themselves can be critical to know.
Yes, but from my perspective the internal numbers need to add up otherwise the whole database just becomes a move-list, which is almost tragic underutilization. Statistics provide useful information so long as the underlying methodologies employed to compile the database are generally understood and have credibility. With this book it isn't merely a matter of doubting the statistical success rates (e3 73%?), but the game-counts are in some cases totally out of synch relative to the actual number of blitz-or-longer games that have been available to the public over the past several years.
Now, the game-count phenomenon can be explained easily enough if someone has a massive database of private games in obscure opening lines, or has incorporated a fairly large number of cherry-picked non-Playchess server games, or has thrown in a large number of hyperbullet games. But these possibilities all strike me as unlikely for different reasons. Instead, I suspect there are just an incredible number of duplicate games in there.
> otherwise the whole database just becomes a move-list
For correspondence games that's the value I find in books. A move list, in where it's an exercise for the user to know what moves should be investigated.
And I would say the same about engines, I ignore their PVs, and only use the scores for comparison but what really matters is their move choices, that I just put on a list, and my analysis file is just a lot of these move-lists in where I point to myself what are the most favorable.
After all, the moves that are in the book are being played at time controls much faster than at corr time controls, so what matters is the moves that the book has in there to be played against e3, not the % Performance that it has about them.
It could suggest a line with many moves and 0% performance, that is best and engine don't like to play, and as long as the user sees it and finds it best, that's what matters. I'd take a book with such a line over a book that has perfect statistics for all moves but omits this line, any day.
Its was one reason why I offered you the book. I knew you would like the vastness of it.
It has been very useful to get Vesely out of book, that was the first thing I had to do against Schachmatt
I'm getting to the point now that I'm close to doing my manual book out of all my reference stuff. Right now I have a big reference book that is just a mass of different books. But what I would like to do is when I start on the manual. I'll add all the moves I have for now and delete the reference books. Then I'll go back and take my time and start marking the moves and once I get to a point I feel good with were I'm at with it. I'll keep look for new books and after downloading them I'll manually add these as a difference color say probably black. Telling me these are lines I need to explore. Sound like a good plan to me. What do you think? I know I have my own ideas for this. Its mostly about eval, but I'll use all my main engines to tell me if its worth a green move or add a !? and say its interesting and worth taking a try in a game, even if its not a forced winning line (I think you get my meaning here)
Yeah, sounds like a great plan when dealing with huge books, I may like the vastness of Vesely very much, but probably the useful variations that it has could be manually condensed into 200MB or something, manually, though it'd be a lot of work, I'd rather keep the 4.8GB of bloat than go for such a project
But what matters is that you enjoy doing it and that it's useful to you.
What part of it have you consider the most useful so far?
The moves it has that other engines don't suggest and that aren't in the other opening books that I have.
Very good idea Uly! I will implement this as well
> Instead, I suspect there are just an incredible number of duplicate games in there.
Am I overstating your suspicion, Nelson, by saying that this book is (unfortunately) utter crap?
I wouldn't dismiss it out of hand. There are obviously a lot of games in there and a lot of work went into it. But I have serious doubts about quality control and statistical integrity. I base this only on the things I am seeing in the opening position, mentioned above, which instigate these concerns. If there are problems in the opening position that usually means there are bigger and more obvious problems further into the book. The first place I would look is the 1.e3 line. How did he run up such a game count, success rate and average ELO? Are the games really unique? Another thing to check is when the file was last modified. That would establish the cut-off date for games in the database.
Beyond expressing doubts I will not go. I haven't downloaded the book myself; I don't have time to serve on an expert panel like some people with the intent of running down the work of others. I would only caution users that some things don't appear to add up. If you want to use it in the limited way Uly and Jimmy are using it, then go right ahead.
> If you want to use it in the limited way Uly and Jimmy are using it, then go right ahead.
Can you tell what other uses does a book have than giving ideas of lines to analyze for the correspondence chess player?
Your question answers itself. A correspondence chess player only uses it one way. But people who aren't correspondence players may find a number of alternative uses.
In my defense, I'll say that the OP probably meant to ask 'which one is the "state of the art" public available book (free or not) for correspondence chess?', and then Vesely is a good answer because its "problems" aren't so for this "limited" usage.
Any book that gives the green light to 1.e3 and shows a success rate of 73% after 280k+ games is immediately suspect in my mind. No one could possibly be that successful with it in meaningful games over a prolonged period of time.
Is there any chance, this book can be shared in small bits - like break it up in 10 pieces and then uploaded?
If you go to the original thread where the book was mentioned, Uly's second link where he says "Mark Mason made it public" there is the same book but now in seven parts. Each of these can be downloaded in about ten minutes or so with ADSL, but then you have to recombine them. I have not actually tried if it works. Downloading the whole book went faster the second time well actually the third time, but it depends a bit on how busy the server is I think, and the download speed is not so great.
si quieres un bun libro,que ha demostrado ser unos 50-80 puntos elo superior a otros libros de aperturas(enfrentando match de libros,no los engines),compra el libro de aperturas de HIARCS,se actualiza cada 3 meses con las innovaciones que van sacando los GM's en sus partidas,y esta tambien 'tuneado'`para las partidas por correspondencia y las que se usan los engines...
ademas,con la suscripcion,te dan un libro de aperturas sin 'tunear' para el uso de engines y correspondencia,que es perfecto para el uso contra jugadores humanos
yo lo compre,he hecho matchs contra el libro de aperturas del rybka,y le gana de largo
Does someone here want to suggest a book that is FREE and doesn't take up gigabytes of space? I'm downloading at a slow-poke speed, and suggest JapanTopTeam won't help me at all, since I don't have 30 dollars to spend to buy the book. The only book I can think of off the top of my head is Morpheus... Seekwanda is problematic as the guy earlier said...
Have you tried building your own book? There are lots of games in public domain now, all kinds. If you're selective you can come up with a pretty decent book from the first try.
Problem is, I'd need a fat guide and lots of help to make a decent book. All I've been doing is plugging in moves from Houdini into Morphius 11.6 ABK... It has helped, but it's very limited in that it takes forever to put up lots of variations. I did succeed in making a good Anti-Rybka book, though. :-)
Actually right now I have the cpu cycles to put up quite a bit of analysis for quite a lot of moves, just not sure how to start... Maybe only take openings from 2700 rating players and up, and engine games with rating 3000 and over...
But uh... how do I do that?
> But uh... how do I do that?
I'd like to help, but I am too busy to write several step-by-step paragraphs on this. There are a lot of guys on this site can offer sound advice. They may not all agree, but that's the nature of things: there are multiple ways to get the job done. The first thing that needs to be defined is what kind of a book you want. Some of the possible choices:
1. A book to play engine games on Playchess or some other server
2. A quasi-archival book to include every game you can find
3. A book to test engines under fair conditions
4. A book to enter in competitive engine/book tournaments like WCCC used to be
5. A book to help you improve your opening play in human games
6. A book to help you in correspondence matches
Each of these would involve a different approach, though there might be some overlaps. Mostly it is about the depth and breadth you are aiming for and to a lesser extent the time you have for customization.
My number one priority is to make a strong book. A strong book for engine vs engine play. For the sake of saving time, I'd like to specialize the book for White, but I'm fine doing both.
I am definitely not going to include every game I can find. I'm going to use my book for testing against other books, and to play against other people's engines/book configurations, or maybe to use against high ranked players.
Yeah sure, if my book somehow becomes good enough to compete (and I doubt that, considering how professional those other guys look with their books...), I'll send it to compete.
EDIT: If I really had to pick a second option, I'd say my second priority for building this book would be to learn a few openings, but at 1050 rating-ish, I don't think I need to break out the top level books just yet.
For some reason, I've been obsessed with building the best computer chess setup, and that includes a good opening book...
You probably want to take a look at HS-Masterbook 3.0:http://www.megaupload.com/?d=CQGD2TUW
It's 360+ MB.
For a long while I used the 2.0 version, and was very happy with it, I have no reason to believe that this one is worse, though I've never tested it.
I would second Uly's recommendation. Also you might look at Hiarcs subscription-based books if you don't mind paying a little. (Comparing what they ask for in relation to the many, many hours your would spend to create something comparable yourself, it's a bargain.)
See, that's the whole point: I don't have money to spend. Else I'd buy JapanTopTeamII.
Well, then, you're just like me. You haven't got capital, but you have a lot of labor. So start working!
Well, what I was hoping for is a book tournament... for free books. :) (Or include many free books.) For ctg or ABK... I have no clue how to get polyglot working.
What I can do is get the best free book, then edit it to help make it a tiny bit stronger.
>>What I can do is get the best free book, then edit it to help make it a tiny bit stronger.
YES... share result in PGN file of Your book tournament on this forum, I know that many members use it :-)
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill