ChessPub Forum - Chess Book Review blog

proustiskeen YaBB Moderator Offline Hello from Omaha! Posts: 678 Joined: 08/11/08	Re: Chess Book Review blog Reply #447 - 10/27/19 at 15:38:36	Post Tools	Print Post
	My October review of Sasha Chapin's All the Wrong Moves: A Memoir about Chess, Love, and Ruining Everything. https://chessbookreviews.wordpress.com/2019/10/27/quest-or-obsession/


	WWW IP Logged

proustiskeen YaBB Moderator Offline Hello from Omaha! Posts: 678 Joined: 08/11/08	Re: Chess Book Review blog Reply #446 - 09/30/19 at 21:15:39	Post Tools	Print Post
	My September review of two new works by Romain Edouard. French players will want to pay special attention to this column. https://chessbookreviews.wordpress.com/2019/09/30/multi-tasking-to-the-max/


	WWW IP Logged

proustiskeen YaBB Moderator Offline Hello from Omaha! Posts: 678 Joined: 08/11/08	Re: Chess Book Review blog Reply #445 - 09/01/19 at 01:08:50	Post Tools	Print Post
	My August review, posted just in the nick of time! https://chessbookreviews.wordpress.com/2019/08/31/data-driven-chess/


	WWW IP Logged

proustiskeen YaBB Moderator Offline Hello from Omaha! Posts: 678 Joined: 08/11/08	Re: Chess Book Review blog Reply #444 - 07/16/19 at 14:45:58	Post Tools	Print Post
	Here's my July review of Vladimir Tukmakov's new book, Coaching the Chess Stars. https://chessbookreviews.wordpress.com/2019/07/16/trainer-to-the-stars/


	WWW IP Logged

IMDavidCummings YaBB Newbies Offline I Love ChessPublishing! Posts: 12 Joined: 11/20/16	Re: Chess Book Review blog Reply #443 - 07/13/19 at 13:40:22	Post Tools	Print Post
	Nickajack wrote on 07/13/19 at 05:05:21: kylemeister wrote on 06/26/19 at 17:08:49: Pantu wrote on 06/26/19 at 11:21:14: The example given there is 1.c4 e6 2.g3 d5 3.Bg2 Nf6 4.Nf3 d4 5.d3 c5 and now apparently 6.e4! is the move - if 6...dxe3 white is just better and otherwise 7.e5 followed by Qe2/Nbd2 with a good version of the KIA. I wonder if that is really just good for White ...in the introduction/overview of the latest update, IMDavidCummings didn't appear to think so. I am quite surprised that IM Cummings would not like 6.e4!, since it is a thematic move in this Benoni Wall (reversed) type of structures. It's an idea that comes up in similar lines of the English/Reti. Here it looks like the perfect time to play it To set the record straight, I do like this move! With hindsight, my conclusion “Food for thought for Reti Players!” was a bit too modestly worded, but I didn’t think it would be controversial My analysis in the PGN gives the nod to White, so the onus is on Black to find improvements. First of all, 6 e4 was a great find by GM Donchenko, apparently uncovered using LC0 (indeed it is Leela’s first line). When researching games for the June ChessPub Update, his move really stuck out for its originality. In this opening, White, including recent games from the likes of Aronian, Nakamura, Wei Yi, Ding Liren etc., had almost always gone for a reverse Modern Benoni structure (most often with 5 0-0 and 6 e3). In the database, there are hundreds of games with e2-e3 and only five games with 5 d3 and 6 e4. The point is not to play the reversed KID/Benoni wall, but the King’s Indian Attack structure after 6 e4 Nc6 7 e5, since Black can’t prevent 7 e5 in this line. Now, if Black continues with standard development, White does indeed get a good version of the KIA. See Donchenko-Villegas – a convincing win for White. In Donchenko-Oleksienko, Black varied with an early ...Ne7-g6 followed by trading White’s important f4-bishop. Black was actually better after 15 moves, so analyze this game before declaring victory! As mentioned in my article, White has earlier improvements such as 11 b4. I’ve since done some more analysis here and will update this game in the PGN Archives next month. Analyzing this game using LC0 and SF10 side-by-side is fascinating, they have clearly diverging evaluations in many places. Coming back to the earlier comments in the thread, I agree that we will see many more LC0-inspired novelties in the coming months!


	IP Logged

Pawnpusher Senior Member Offline I Love ChessPublishing! Posts: 274 Joined: 01/04/18 Gender:	Re: Chess Book Review blog Reply #442 - 07/13/19 at 11:24:32	Post Tools	Print Post
	I am entirely in agreement with Descartes, I have played some ganes where the other fellow simply understood the ideas in the position and I suffered. Not that there was a huge difference in strength, he just had a better idea about where the pieces should go.


	IP Logged

Nickajack Junior Member Offline Yabba Dabba Doo! Posts: 57 Location: USA Joined: 02/21/17 Gender:	Re: Chess Book Review blog Reply #441 - 07/13/19 at 05:05:21	Post Tools	Print Post
	kylemeister wrote on 06/26/19 at 16:08:49: Quote: The example given there is 1.c4 e6 2.g3 d5 3.Bg2 Nf6 4.Nf3 d4 5.d3 c5 and now apparently 6.e4! is the move - if 6...dxe3 white is just better and otherwise 7.e5 followed by Qe2/Nbd2 with a good version of the KIA. I wonder if that is really just good for White ...in the introduction/overview of the latest update, IMDavidCummings didn't appear to think so. I am quite surprised that IM Cummings would not like 6.e4!, since it is a thematic move in this Benoni Wall (reversed) type of structures. It's an idea that comes up in similar lines of the English/Reti. Here it looks like the perfect time to play it.

	Dubious, therefore playable
	IP Logged

brabo God Member Offline Welcome chessfriend Posts: 1071 Joined: 02/02/07	Re: Chess Book Review blog Reply #440 - 06/26/19 at 17:39:05	Post Tools	Print Post
	ReneDescartes wrote on 06/26/19 at 15:22:11: this is a limitation of GM-Rep-style books, and perhaps of theory, as such! GM-Rep-style books are written to give the reader a solid base for a repertoire which can be used for many years at a high level (even 2500 elo). So modest lines will be disregarded for more critical ones. Also these books must be pushing the theory further in many lines or at least give an up to date view of the theory (gigantic research is critical). Players should of course first try to master the books covering their repertoire before making any analysis themselves. As the books are the base for any further work, it is very important that the quality of them is as high as possible.


	WWW IP Logged

kylemeister God Member Offline Posts: 4946 Location: USA Joined: 10/24/05	Re: Chess Book Review blog Reply #439 - 06/26/19 at 16:08:49	Post Tools	Print Post
	Quote: The example given there is 1.c4 e6 2.g3 d5 3.Bg2 Nf6 4.Nf3 d4 5.d3 c5 and now apparently 6.e4! is the move - if 6...dxe3 white is just better and otherwise 7.e5 followed by Qe2/Nbd2 with a good version of the KIA. I wonder if that is really just good for White ...in the introduction/overview of the latest update, IMDavidCummings didn't appear to think so.


	IP Logged

ReneDescartes God Member Offline Qu'est-ce donc que je suis? Une chose qui pense. Posts: 1240 Joined: 05/17/10 Gender:	Re: Chess Book Review blog Reply #438 - 06/26/19 at 15:22:11	Post Tools	Print Post
	That is a good illustration and supports your point well regarding ability. I would only add that since one has to spend the time to understand how the plan works, learning such a line is not nearly as simple as learning an engine-generated tactical refutation. Faced with a GM-Rep book with many, many such lines, it would be hard in practice to learn the repertoire. Some time ago, I myself did research into how to play a not-particularly-impressive line in the French which I found interesting. Beyond a point, there were not many games played with it, so I used an engine (not a new linear-algebra engine, but a regular one) to investigate and acquire a strategic understanding of this modest line. As a result, I have had success with it. The point is that investing the time to understand an equal line strategically will pay off even when that line is not shocking, doesn't have to be discovered by an engine, and is somewhat rare, but not new. And one can't expect this kind of education from a detailed repertoire book--this is a limitation of GM-Rep-style books, and perhaps of theory, as such!
	« Last Edit: 06/26/19 at 18:41:58 by ReneDescartes »

	IP Logged

brabo God Member Offline Welcome chessfriend Posts: 1071 Joined: 02/02/07	Re: Chess Book Review blog Reply #437 - 06/26/19 at 11:02:22	Post Tools	Print Post
	Quote: Heh, I cheated slightly, but 0-0!? looks fun. Or is there something else? Dorfman makes the same argument as brabo in this interview, specifically with regards to AlphaZero/Leela: https://chess24.com/en/read/news/dorfman-the-real-talent-is-the-ability-to-work-... The example given there is 1.c4 e6 2.g3 d5 3.Bg2 Nf6 4.Nf3 d4 5.d3 c5 and now apparently 6.e4! is the move - if 6...dxe3 white is just better and otherwise 7.e5 followed by Qe2/Nbd2 with a good version of the KIA. Consider this: if you have the resources, and are interested in what to do in a variation, let's say the 7.Nd5 Sveshnikov that Carlsen has been playing. You then set Leela0/AlphaZero to play against itself continuously, learning and improving, for a week from that starting position. You then check the games and analysis and you find a completely new approach along with a couple of near perfectly played games. You then prepare this before your next game against a Sveshnikov player and get to display some cutting edge stuff. The advantage of this approach is that it's fairly effortless - the machine generates ideas. It also tells you if they are worthwhile or not, and demonstrates how to play. A bunch of human grandmasters could generate the same but horribly inefficiently (and much more expensively). For the top-level we could see a return to the early computer era where how efficiently you work with the engine (and how big your team working behind it) is a big factor. Games like Kasimdhzanov-Kasparov, Linares 2005, where 17...0-0!! basically put a line out of commision for white, that was suggested by the engine to Kasparov. Nowadays the top players have all learned how to use normal engines to find/avoid this stuff, but the next wave is showing another level. P.S. to Brabo - I appreciate your English language blog and do follow it. 0-0 is indeed the right answer. My opponent almost fell of his chair when he saw me playing that ridiculous move. I already published the game on my Dutch blog and will later translate it. The last decades we saw engines becoming continously stronger tactically. Many openinglines were refuted as tactical flaws were discovered. Today we see indeed a second wave which starts to question also the openings strategically. Till only a couple of years ago it was impossible to get good answers from engines for many strategic problems. You don't need to be a super strong grandmaster to benefit from the second wave.


	WWW IP Logged

Pantu Ex Member	Re: Chess Book Review blog Reply #436 - 06/26/19 at 10:21:14	Post Tools	Print Post
	Heh, I cheated slightly, but 0-0!? looks fun. Or is there something else? Dorfman makes the same argument as brabo in this interview, specifically with regards to AlphaZero/Leela: https://chess24.com/en/read/news/dorfman-the-real-talent-is-the-ability-to-work-... The example given there is 1.c4 e6 2.g3 d5 3.Bg2 Nf6 4.Nf3 d4 5.d3 c5 and now apparently 6.e4! is the move - if 6...dxe3 white is just better and otherwise 7.e5 followed by Qe2/Nbd2 with a good version of the KIA. Consider this: if you have the resources, and are interested in what to do in a variation, let's say the 7.Nd5 Sveshnikov that Carlsen has been playing. You then set Leela0/AlphaZero to play against itself continuously, learning and improving, for a week from that starting position. You then check the games and analysis and you find a completely new approach along with a couple of near perfectly played games. You then prepare this before your next game against a Sveshnikov player and get to display some cutting edge stuff. The advantage of this approach is that it's fairly effortless - the machine generates ideas. It also tells you if they are worthwhile or not, and demonstrates how to play. A bunch of human grandmasters could generate the same but horribly inefficiently (and much more expensively). For the top-level we could see a return to the early computer era where how efficiently you work with the engine (and how big your team working behind it) is a big factor. Games like Kasimdhzanov-Kasparov, Linares 2005, where 17...0-0!! basically put a line out of commision for white, that was suggested by the engine to Kasparov. Nowadays the top players have all learned how to use normal engines to find/avoid this stuff, but the next wave is showing another level. P.S. to Brabo - I appreciate your English language blog and do follow it.


	IP Logged

brabo

God Member

Offline

Welcome chessfriend

Posts: 1071
Joined: 02/02/07

Re: Chess Book Review blog
Reply #435 - 06/26/19 at 05:24:37

Post Tools

Print Post

ReneDescartes wrote on 06/25/19 at 18:03:34:

Perhaps Carlsen and Anand have the time and ability to learn how to follow up these engine ideas the way an engine would, in order to cash in the supposed advantage. But is there any evidence that, for an ordinary mortal who might read an opening book and memorize lots of lines, getting the position from such a new move actually gives any practical advantage--that is, in the case which matters most for your argument, the case where a human wouldn't independently produce the same evaluation?

Try to guess which move I played in the position below which I got on the board a couple of months ago in a standard game.

My opponent a FM told me that he played hundreds of games with this opening but never he encountered that idea. He thought for a very long time during the game that black was at least ok. Afterwards he was amazed that I had seen this idea from an engine and an engine evaluates it as strategically close to winning.

Now before I decided to play the idea, I had spent of course some time with the engine to understand why and how it works. This allowed me in the game to capitalize on it. I am rated about 500 points lower than Carlsen and Anand so even at my mediocre level looking and studying games of engines will more and more impact the results.

WWW

IP Logged

ReneDescartes God Member Offline Qu'est-ce donc que je suis? Une chose qui pense. Posts: 1240 Joined: 05/17/10 Gender:	Re: Chess Book Review blog Reply #434 - 06/25/19 at 18:03:34	Post Tools	Print Post
	brabo wrote on 06/25/19 at 07:10:30: proustiskeen wrote on 06/24/19 at 19:21:45: No doubt SF8 is still plenty strong for Avrukh and others to blundercheck their work. We are entering a new era in which engines will not be used for only blunderchecking. Till recently it were humans developing the latest trends in the openings. That starts to change which we can already see at the very highest levels. Topplayers start to copy openings played by engines in their matches. I expect we will soon see the rise of new databases with millions maybe even billions of engine games all played at a level way beyond the level of a world-champion and containing a wealth of new ideas. Does this sound ridiculous to you? Well just check the https://lczero.org/ in which the counter is already at quarter of a billion games. Also one of the top novelties used in the latest worldchampionship by Carlsen was directly coming from an engine-match or what about the same 15 moves happening at the same day in a human and engine-match see https://www.chess.com/news/view/fide-grand-prix-moscow-semifinal-chess Thinking or working with an engine still as only a blunderchecking tool is old school. I predict we will see in the next years a shift to a dominance of engine-games which will largely define the new trends in the openings. Perhaps Carlsen and Anand have the time and ability to learn how to follow up these engine ideas the way an engine would, in order to cash in the supposed advantage. But is there any evidence that, for an ordinary mortal who might read an opening book and memorize lots of lines, getting the position from such a new move actually gives any practical advantage--that is, in the case which matters most for your argument, the case where a human wouldn't independently produce the same evaluation?
	« Last Edit: 06/25/19 at 19:31:11 by ReneDescartes »

	IP Logged

brabo God Member Offline Welcome chessfriend Posts: 1071 Joined: 02/02/07	Re: Chess Book Review blog Reply #433 - 06/25/19 at 16:39:16	Post Tools	Print Post
	proustiskeen wrote on 06/25/19 at 14:59:24: Chesspub has not been what it once was for a long time, and Avrukh quotes the same correspondence game that Brabo does. (No offence was meant to you, Brabo - thus the emoticon.) Perhaps look at the book before dismissing it because he doesn't cite this forum or use your preferred engine? A correspondence game can be already irrelevant after 1 year which is here the case. Avrukh has always been regarded as one of the very best authors. Is it wrong then to expect that he uses the best tools and is aware of all the most recent publications even if it happened on chesspub? I don't doubt the book is still very good but I was expecting more than that. The standard before was brilliant and after reading your review and interview I only got convinced this isn't the case anymore.


	WWW IP Logged