Human versus Machine

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Harvey Williamson, bob

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Fri Oct 13, 2017 12:23 am

APassionForCriminalJustic wrote:
Lyudmil Tsvetkov wrote:Comparing 'The Secret of Chess' and 'Human versus Machine: How to beat Stockfish and Komodo', I wrote the latter much quicker, the former took whole 4 months, but the interesting thing is how notions presented in 'The Secret of Chess' are visible in the games showcased in 'Human versus Machine'.

For example, the games exhibit patterns and notions like:

- twice backward shelter pawn on f7
- pointed chains
- white and black KID structures
- fully closed sides of the board, etc., etc.

all of which could be found in 'The Secret of Chess'.

Of course, it is actually the other way round: the many thousands of games(over 50 000, to be clear) I have played against engines and top engines and the knowledge I derived from them are reflected in the knowledge presented on the pages of 'The Secret of Chess'.

That is how I verified that knowledge: by playing an infinite number of games against the very top, and it seems to work.

If anyone would like to consider the games in 'Human versus Machine' as fake ones, well, you simply don't have a point, looking at the specific positions, you will not find even a single one that even distantly resembles any human or engine game you could find in any database.

There are simply no such games and positions, so who came up with the concept and system?
Also, checking evaluations, you will easily see the games are for real. Current Stockfish development version still does not understand most of them.

Again, why would beating Stockfish and Komodo be less interesting than reproducing a routine game from a public database?
Stockfish doesn't need to understand anything. It plays chess at a level that you could not even comprehend... who cares about fluke positions. Sure humans can understand when they've been mated and stuff like that. But chess is nothing more than brute force and computation. Computers have proven that no doubt. There's nothings left anymore for human interferences. Maybe you should consider playing a super GM. Not sure what there is to prove about playing chess versus code that can do things that the human brain simply cannot. The idea of understanding things is completely overrated which is pretty hilarious if you ask me.
and yet, with every single new version, the evaluation of Stockfish only seems to improve. once you try to simplify it or introduce terms that don't reflect objective knowledge, and you see a regression.

it is quite the other way round, dear Adam, the human brain is a much more powerful tool than any existing machine, you just have to discover it.

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Fri Oct 13, 2017 12:32 am

MikeB wrote:
Lyudmil Tsvetkov wrote:Anyone interested in human-computer matches might check my new book:
https://www.amazon.com/dp/1549916785/re ... 1250226011 (paperback)

https://www.amazon.com/dp/B0768G8R2C/re ... 1250226011 (ebook)

also available on amazon.uk(search by author and title), amazon.de, etc.

Amply commented and diagrammed games.

Seems like the first book with extensive coverage of a large number of winning games against the top engines.

Kasparov, Carlsen and Nakamura still have not written one. :)
Good luck with the books Lyudmil. Nicely done. Getting it for free is pretty cool ( by using my Amazon points 👍 ). Should have it by Saturday.

Image

Image

Will be releasing McBrain 3.0 this Saturday.
thanks Mike.

I don't know if I'll be getting any of that.
What do you mean by Amazon points? Sure you did not use a prime student account?

I am glad you will read the book. Hope you will find something to implement
in MB.

Amazon does not stop to surprise me, I am new there(just 2 months),
but it does not stop to surprise me.

especially those ranks...

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Fri Oct 13, 2017 12:42 am

btw.(I hope Harm will not find this too comm, it is just an interesting real-world observation), it is incredibly difficult for an independently published author, like myself, to compete with books published by mainstream publishers: they have much more exposure, maybe exponentially so, their books are as a rule offered discounts on, while indies don't quite have that feature available, once you get a competitive advantage, Amazon boosts that advantage further, so it is even more difficult to come along, etc., etc.

of course, thinking that a large number of chess books published on Amazon don't have even a single purchase for their entire lifetime is a bit of a morale-booster, but the point is when will a good indie chess book rival Fischer's 'My 60 memorable games' or similar books?

Maybe never?

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Fri Oct 13, 2017 12:53 am

Nay Lin Tun wrote:Saw the 10 preview pages and you wrote about winning against stockfish 4 with stonewall .Well, Stonewall is notoriously drawish opening (since dinosaur age )against computers, of course at least 99% good players already knew and might have tried against computer already. (similar to a player nicked name Grandfather who frequently posted his drawish games with stone wall here). I can tell the reason behind the loss of computer, it was because black did not develop queen bishop on c8. However I highly doubt that the same opening will work against current stockfish or Komodo of 2017 version. If you play stonewall in either 1. e3, d4, bd3 , f4 , Nf3,Ne5 in whatever order, black will respond with d5, nf6 bf5 or bg4 in either sequence. Black will not badly play like e6 and stop developing c8 bishop in that opening. Of course you may still get a draw 1 in 100 against modern computers and it should be according to your rating difference vs computers. ( your 2100+ FIDE rating vs 3100+ computers)
I don't know when people on this forum and elsewhere will finally learn that my objective current chess strength, especially under quiet conditions, is at least 2800?

that 2100+ estimate(actually 2200+) dates back from 2004-2006, and during that period I had a couple of 2400 performances.
I simply played too few rated games during that period, mostly at the weekends or in the evening, after a strenuous working day. Lack of concentration and fatigue play an enormous role in deciding a chess game.

in the course of 12 years, I have added too much chess strength, I am even afraid to think of how large that quantity is nowadays. Matter of fact is that in the last 5 years I have played and analysed at least 3 times more games than during my entire chess training span before that.

I don't know what are you talking me drawing 1 out of 100 against Stockfish. Out of 100, I am usually drawing at least 67.

I acknowledge you are right, though, concerning Bf5 in the commented Stonewall game. It is the much better choice, Komodo does play that move, but not Stockfish.
Just check, current Stockfish will again choose e6 instead.

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Fri Oct 13, 2017 12:57 am

APassionForCriminalJustic wrote:
Nay Lin Tun wrote:Saw the 10 preview pages and you wrote about winning against stockfish 4 with stonewall .Well, Stonewall is notoriously drawish opening (since dinosaur age )against computers, of course at least 99% good players already knew and might have tried against computer already. (similar to a player nicked name Grandfather who frequently posted his drawish games with stone wall here). I can tell the reason behind the loss of computer, it was because black did not develop queen bishop on c8. However I highly doubt that the same opening will work against current stockfish or Komodo of 2017 version. If you play stonewall in either 1. e3, d4, bd3 , f4 , Nf3,Ne5 in whatever order, black will respond with d5, nf6 bf5 or bg4 in either sequence. Black will not badly play like e6 and stop developing c8 bishop in that opening. Of course you may still get a draw 1 in 100 against modern computers and it should be according to your rating difference vs computers. ( your 2100+ FIDE rating vs 3100+ computers)
They are higher than 3100. Humans are nothing to a computer. Lol. Of course you know that - but some do not. Stockfish 4 is a fossil anyways.
Should be scheduling an official match against a top engine.

Maybe then the hype will start... :)

User avatar
MikeB
Posts: 2505
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Human versus Machine

Post by MikeB » Fri Oct 13, 2017 1:55 am

Lyudmil Tsvetkov wrote:
MikeB wrote:
Lyudmil Tsvetkov wrote:Anyone interested in human-computer matches might check my new book:
https://www.amazon.com/dp/1549916785/re ... 1250226011 (paperback)

https://www.amazon.com/dp/B0768G8R2C/re ... 1250226011 (ebook)

also available on amazon.uk(search by author and title), amazon.de, etc.

Amply commented and diagrammed games.

Seems like the first book with extensive coverage of a large number of winning games against the top engines.

Kasparov, Carlsen and Nakamura still have not written one. :)
Good luck with the books Lyudmil. Nicely done. Getting it for free is pretty cool ( by using my Amazon points 👍 ). Should have it by Saturday.

Image

Image

Will be releasing McBrain 3.0 this Saturday.
thanks Mike.

I don't know if I'll be getting any of that.
What do you mean by Amazon points? Sure you did not use a prime student account?

I am glad you will read the book. Hope you will find something to implement
in MB.

Amazon does not stop to surprise me, I am new there(just 2 months),
but it does not stop to surprise me.

especially those ranks...
You should be - I get points from using my "amazon" credit card ( it's actually a chase bank credit card with amazon branding. In the US a prime account gets you free two day shipping and amazon video streaming - about $10/month. Not a student account for sure..

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Sat Oct 14, 2017 6:42 am

MikeB wrote:
Lyudmil Tsvetkov wrote:
MikeB wrote:
Lyudmil Tsvetkov wrote:Anyone interested in human-computer matches might check my new book:
https://www.amazon.com/dp/1549916785/re ... 1250226011 (paperback)

https://www.amazon.com/dp/B0768G8R2C/re ... 1250226011 (ebook)

also available on amazon.uk(search by author and title), amazon.de, etc.

Amply commented and diagrammed games.

Seems like the first book with extensive coverage of a large number of winning games against the top engines.

Kasparov, Carlsen and Nakamura still have not written one. :)
Good luck with the books Lyudmil. Nicely done. Getting it for free is pretty cool ( by using my Amazon points 👍 ). Should have it by Saturday.

Image

Image

Will be releasing McBrain 3.0 this Saturday.
thanks Mike.

I don't know if I'll be getting any of that.
What do you mean by Amazon points? Sure you did not use a prime student account?

I am glad you will read the book. Hope you will find something to implement
in MB.

Amazon does not stop to surprise me, I am new there(just 2 months),
but it does not stop to surprise me.

especially those ranks...
You should be - I get points from using my "amazon" credit card ( it's actually a chase bank credit card with amazon branding. In the US a prime account gets you free two day shipping and amazon video streaming - about $10/month. Not a student account for sure..
maybe so.
I am still extremely puzzled by the way they handle orders and purchases.

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Human versus Machine

Post by Lyudmil Tsvetkov » Sat Oct 14, 2017 6:46 am

btw., to add some humour to the thread(Milos would say more humour), one big danger I recognise in people reading this and similar books is that, in case people get immersed and learn the appropriate opening and middlegame patterns to beat Stockfish, too many people might suddenly start winning against the top engines, and that would certainly mean poor Stockfish will need at least another 400 elo to compete successfully. :) :)

Anyone seeing Stockfish adding another 400 elo soon?

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Part 2

Post by Lyudmil Tsvetkov » Fri Oct 27, 2017 6:01 am

Part II of the book is out:

https://www.amazon.com/Human-Versus-Mac ... 087&sr=1-1

(both paperbacks and ebooks included, just switch between versions)

Also available on other amazons, uk, de, etc.(search by author and title)

In this edition, games against Stockfish DD, Stockfish 5, 6 and Komodo 10 are represented.

7 or 8 different openings featured, basically boiling down to 4 main winning pawn structures:

- Stonewall Attack(Stonewall Defence): pawns on d4-e3-f4, d5-e6-f5 for black
- King's Indian Attack(King's Indian Defence): pawns on d3-e4-f5, d6-e5-f4 in the standard KID for black
- Central chain structure, arising out of the Queen's Pawn Game, Torre Attack(or out of the possible Slav for black): pawns on c3-d4-e5, c6-d5-e4 for black
- Central bind structure, arising out of the English Opening(Sicilian Defence for black): pawns on c4-d3-e4, c5-d6-e5 for black

It is not easy to beat the top engines, so take a look at the games and explanations.
Maybe, you will find that interesting.

Lyudmil Tsvetkov
Posts: 6031
Joined: Tue Jun 12, 2012 10:41 am

Re: Part 2

Post by Lyudmil Tsvetkov » Fri Oct 27, 2017 6:05 am

Page count has risen from 165 to 238, so I guess it has been a bit more difficult overpowering the engines, but that also supposes higher quality of the games compared to the first part.

Post Reply