Thoughts on Komodo 11, Free engines and the future

Ras · Post by **Ras** » Mon May 29, 2017 12:07 pm

Lyudmil Tsvetkov wrote:seemingly, the French is bad, because already on the 3rd move, after 1.e4 e6 2.d4 d5 3.e5, white can achieve the aforementioned central pawn structure.

Alternatively, this pawn structure is irrelevant. What is the actual evidence anyway?

Lyudmil Tsvetkov · Post by **Lyudmil Tsvetkov** » Mon May 29, 2017 2:29 pm

Ras wrote:
Lyudmil Tsvetkov wrote:seemingly, the French is bad, because already on the 3rd move, after 1.e4 e6 2.d4 d5 3.e5, white can achieve the aforementioned central pawn structure.
Alternatively, this pawn structure is irrelevant. What is the actual evidence anyway?

try running 10 000 games with echange French, which SF likes, 3.ed5 ed5, and another 10 000 with e5 advance, either immediately 3.e5, or, even better, after 3.Nc3 with transpositions going into a similar structure, to measure the difference.
I bet the advance line will score much better, even if only using weak SF for the test.

this is the secondary evidence, however.

primary evidence is that with that particular structure the black king shelter, whether the king is castled or not, becomes significantly inflexible because of the f7 semi-backward pawn.
sometimes, you need to move your shelter pawns and, even if they are only partially backward, that starts being a big problem. the shelter is compromised, either black will stay with its pawn on f7, which is not tolerable for a longer time, or attempt f6/f5 push, after which ef6 Nf6(or something similar) will leave black with a central backward e6 pawn, while ef6 gf6 will leave black with an isolated shelter h pawn. not very promising in both lines.

Lyudmil Tsvetkov · Post by **Lyudmil Tsvetkov** » Mon May 29, 2017 2:31 pm

also, I guess the advanced French(including Winaver, Classical, etc.) scores much better in human tournaments than the exchange variation.

another major evidence: my gut feeling

Ras · Post by **Ras** » Mon May 29, 2017 2:50 pm

Lyudmil Tsvetkov wrote:try running 10 000 games with echange French, which SF likes, 3.ed5 ed5, and another 10 000 with e5 advance, either immediately 3.e5, or, even better, after 3.Nc3 with transpositions going into a similar structure, to measure the difference.

It is well-known that the exchange variation is very drawish because it doesn't put pressure on Black, and because it conveniently solves Black's ugly queenside bisphop problem.

This doesn't prove that e5 is dangerous, it just proves that the exchange variation is not the best tool for White - unless White is heading for a draw because of the tournament or match situation.

primary evidence is that with that particular structure the black king shelter, whether the king is castled or not, becomes significantly inflexible because of the f7 semi-backward pawn.

That is not evidence, that is rewording the claim.

sometimes, you need to move your shelter pawns and, even if they are only partially backward, that starts being a big problem.

Obviously with the fixed centre, White will play on the kingside since this is where his "pawn finger" points to. Black will try to get counterplay on the queenside.

the shelter is compromised, either black will stay with its pawn on f7, which is not tolerable for a longer time

Of course it is.

or attempt f6/f5 push

That would be the suicide in two ways:
1) you don't attack a pawn chain at the head (e5), but at the root. That's also where the Black e6 problem would come from.
2) Black has to seek counterplay on the queenside, not on the kingside.

The pawn fingers cleary indicate where the chances are, and if Black decides to counter on the kingside, the game is lost because White has much more space to maneuvre there. It is obvious that Black will crack under the pressure. And that's not because of e5, but because Black counterplay on the kingside amounts to passive defence.

Lyudmil Tsvetkov · Post by **Lyudmil Tsvetkov** » Mon May 29, 2017 3:18 pm

Ras wrote:
Lyudmil Tsvetkov wrote:try running 10 000 games with echange French, which SF likes, 3.ed5 ed5, and another 10 000 with e5 advance, either immediately 3.e5, or, even better, after 3.Nc3 with transpositions going into a similar structure, to measure the difference.
It is well-known that the exchange variation is very drawish because it doesn't put pressure on Black, and because it conveniently solves Black's ugly queenside bisphop problem.

This doesn't prove that e5 is dangerous, it just proves that the exchange variation is not the best tool for White - unless White is heading for a draw because of the tournament or match situation.

primary evidence is that with that particular structure the black king shelter, whether the king is castled or not, becomes significantly inflexible because of the f7 semi-backward pawn.
That is not evidence, that is rewording the claim.

sometimes, you need to move your shelter pawns and, even if they are only partially backward, that starts being a big problem.
Obviously with the fixed centre, White will play on the kingside since this is where his "pawn finger" points to. Black will try to get counterplay on the queenside.

the shelter is compromised, either black will stay with its pawn on f7, which is not tolerable for a longer time
Of course it is.

or attempt f6/f5 push
That would be the suicide in two ways:
1) you don't attack a pawn chain at the head (e5), but at the root. That's also where the Black e6 problem would come from.
2) Black has to seek counterplay on the queenside, not on the kingside.

The pawn fingers cleary indicate where the chances are, and if Black decides to counter on the kingside, the game is lost because White has much more space to maneuvre there. It is obvious that Black will crack under the pressure. And that's not because of e5, but because Black counterplay on the kingside amounts to passive defence.

well, I can not agree with even a single of your conclusions.
they might be learned, somehow agree with current (book) theory, but are completely untrue.

f7-f6 is a standard freeing attempt in the French. Usually, if black does not play that on time, simply loses the game, and that is seen in countless instances.
f7-f6 does precisely that: attack the spearhead of the pawn chain at its 'head'. if you destroy a strong advanced pawn, you gain in advantage.

please look at the first couple of your sentences, they do not quite make sense, I do not know what rewording means, there is logic behind each and every eval term, and that must be sought first.

what better evidence than e5 advance lines score much better than the Exchange, preferred by SF? there are simply not other possibilities for white, the e4 pawn should either advance, after which we get to the mentioned structure, or trade on d5. (de4 should be even weaker for black)

concerning playing on the queen/king side, I would say that is more or less an outdated approach, typical of Steinitz. After Lasker, almost everyone has been playing modern, multi-faceted and unprejudiced chess.

I would say the state-of-the-art maxim concerning game play would be: play there where terms would give you the biggest advantage.

is not that what engines have actually been doing?

Leo · Post by **Leo** » Mon May 29, 2017 5:17 pm

This is a great thread. This is a Golden Age for chess. Also, there are some powerful new processors on the horizon.

Kirk · Post by **Kirk** » Mon May 29, 2017 5:27 pm

leavenfish wrote:If you ask me (and no one has of course....) in furtherance of greater and greater elo, 'speedup's' and search technique seems to be about all any programmer/developer of the top engines are really going for at the moment. I mean...K11 is released with a 4 or 5 'elo improvement' over K10.4....perhaps though that was just because it was time to get a fully 'new' version to Chessbase (?). But it seems pretty clear we may be at an impass when it comes to any notable leap in improvement. True, we cannot see the future, but can we not at least acknowledge that this possible truth may just be staring us in the face?

I'm a Komodo subscriber and as such appreciate ANY overall improvements being made available to me, but I find this something interesting to speculate on....not just for Komodo but other top engines.

I know the trend for Team K seems to be to have less emphasis on knowledge and more on these 'speed-ups' and refined searches and the various tweaks made in furthering those. But, if in the end computer chess is only about as deep a search as one can obtain as fast as one can with as little 'knowledge' as need be...then are we not at a crossroads where developers are gravitating to that mean and we may only ever seen incremental improvments? Heck, we can always get 'more' simply by using faster processors and more cores, etc.

If that is likely to be so, then I feel those who own the (commercial) programs need to offer...something more. I am not sure what that is, but that is the point of this tread - to discuss what a commercial program could do to make itself more (dare I say) 'relevant' when free engines (Stockfish and its derivatives) continue to offer essentially the same and even outpace the commercial engines a bit.

One could argue that it is simply the programs which make use of these engines where one should look to make them more useful in non-game play areas. Lets say how they are used WITHIN programs like Chessbase or Chess Assistant (or Aquarium) or the nice training programs like Chess Position Trainer or Chess Openings Wizard - both VERY similar? Okay, maybe so...but a free engine is essentially as good as a commercial engine there, right? I mean, I'm using the 'Deep Analysis' feature in Chessbase as I type this, but even there, isn't a current development version of Stockfish every bit as good as Komodo 11.1?

Be that as it may, what about relevance at the engine level? I know Team Komodo has made attempts by offering interesting ideas one can tweak oneself. Items like 'White Contempt', but is that about all we will ever get when it comes to engines alone? Is that all we should expect for our $$? Just wondering...

If the future is AI, I still see programming far away from being able to annotate a game like a classic book in chess. Calculations get the ELO, but "understanding" a position and providing that to a Class A player (In a verbal /verbose format) has a way to go.

cdani · Post by **cdani** » Mon May 29, 2017 9:39 pm

Lyudmil Tsvetkov wrote: white pawn on e5, black pawns on e6,f7,g7, black king on e,f,g or h file, or, alternatively, on the queen side, white pawn on d5, black pawns on d6,c7,b7, black king on d,c,b or a file, some bonus for white(who knows how big, I can only estimate:) )
...
when will engines be able to implement it?
...
so that, unless Daniel succeeds in implementing it, obviously no one is going to be able to do so...

Made a long time ago

Was already in Andscacs 0.85. Is pending of more serious test to better tune it, but was a win.

Lyudmil Tsvetkov · Post by **Lyudmil Tsvetkov** » Tue May 30, 2017 8:39 am

cdani wrote:
Lyudmil Tsvetkov wrote: white pawn on e5, black pawns on e6,f7,g7, black king on e,f,g or h file, or, alternatively, on the queen side, white pawn on d5, black pawns on d6,c7,b7, black king on d,c,b or a file, some bonus for white(who knows how big, I can only estimate:) )
...
when will engines be able to implement it?
...
so that, unless Daniel succeeds in implementing it, obviously no one is going to be able to do so...
Made a long time ago Was already in Andscacs 0.85. Is pending of more serious test to better tune it, but was a win.

why do I then see Andscacs 0.86 pick the French on 1.e4, and then, after 2.d4 d5, wants to change ed5 ed5?

Code: Select all

o - Andscacs 0.86134, Blitz 1m Microsoft
rnbqkbnr/pppppppp/8/8/4P3/8/PPPP1PPP/RNBQKBNR b KQkq e3 0 1

Analysis by Andscacs 0.86134&#58;

2.d4 d5 3.exd5 exd5 4.Nc3 Nc6 5.Qe2+ Be7 6.Nf3 
  =  &#40;0.19&#41;   Depth&#58; 6/13   00&#58;00&#58;00  2kN
2.d4 d5 3.exd5 exd5 4.Nc3 Nc6 5.Qe2+ Be7 6.Nf3 Nf6 
  =  &#40;0.19&#41;   Depth&#58; 7/13   00&#58;00&#58;00  2kN
2.d4 d5 3.exd5 exd5 4.Nc3 Nc6 5.Nf3 Nf6 6.Bb5 Bd6 
  =  &#40;0.21&#41;   Depth&#58; 8/15   00&#58;00&#58;00  3kN
2.d4 d5 3.exd5 exd5 4.Nc3 Nc6 5.Nf3 Nf6 6.Bb5 Bd6 7.Qe2+ Qe7 
  =  &#40;0.21&#41;   Depth&#58; 9/18   00&#58;00&#58;00  5kN
2.d4 d5 3.exd5 exd5 4.Bd3 Bd6 5.Qe2+ Ne7 6.Nf3 0-0 7.0-0 Nbc6 8.Re1 
  =  &#40;0.19&#41;   Depth&#58; 10/21   00&#58;00&#58;00  13kN
2.d4 d5 3.exd5 exd5 4.Bd3 Bd6 5.Qe2+ Ne7 6.Nf3 0-0 7.0-0 Nbc6 8.Re1 Nf5 
  =  &#40;0.15&#41;   Depth&#58; 11/21   00&#58;00&#58;00  18kN
2.d4 d5 3.exd5 exd5 4.Bd3 Bd6 5.Qe2+ Ne7 6.Nf3 0-0 7.0-0 Nbc6 8.Re1 Re8 9.Nc3 
  +/=  &#40;0.27&#41;   Depth&#58; 12/21   00&#58;00&#58;00  54kN
2.d4 d5 3.exd5 exd5 4.Nf3 Bd6 5.c4 dxc4 6.Bxc4 Qe7+ 7.Be3 Nf6 8.0-0 0-0 9.Re1 Nc6 10.Bf4 
  =  &#40;0.22&#41;   Depth&#58; 13/23   00&#58;00&#58;00  77kN
2.d4 d5 3.exd5 exd5 4.Nf3 Nf6 5.Bd3 Qe7+ 6.Be2 Nc6 7.0-0 Bf5 8.Nc3 0-0-0 9.Bf4 Kb8 10.a3 
  =  &#40;0.24&#41;   Depth&#58; 14/28   00&#58;00&#58;00  126kN
2.d4 d5 3.exd5 exd5 4.Nf3 Bd6 5.Bd3 Nf6 6.Qe2+ Be6 7.Ng5 Qd7 8.0-0 0-0 9.Nc3 Nc6 10.Nxe6 Qxe6 11.Nb5 Rad8 
  =  &#40;0.17&#41;   Depth&#58; 15/28   00&#58;00&#58;00  173kN
2.Nf3 d5 3.exd5 exd5 4.d4 Bd6 5.Bd3 Nf6 6.Qe2+ Be6 7.Ng5 Qd7 8.0-0 0-0 9.Nc3 Nc6 10.Nxe6 
  =  &#40;0.21&#41;   Depth&#58; 15/28   00&#58;00&#58;00  240kN
2.Nf3 d5 3.exd5 exd5 4.d4 Nf6 5.Bd3 Bd6 6.Qe2+ Be6 7.Ng5 Qd7 8.Nc3 0-0 9.Nxe6 Qxe6 10.Nb5 Re8 11.Qxe6 Rxe6+ 12.Be3 Ng4 
  =  &#40;0.18&#41;   Depth&#58; 16/28   00&#58;00&#58;00  306kN
2.Nf3 d5 3.exd5 exd5 4.d4 Nf6 5.Bd3 Bd6 6.Qe2+ Be6 7.Ng5 Qd7 8.Nc3 0-0 9.Nxe6 Qxe6 10.Nb5 Re8 11.Qxe6 Rxe6+ 12.Be3 Ng4 13.Nxd6 
  =  &#40;0.18&#41;   Depth&#58; 17/28   00&#58;00&#58;00  462kN
2.Nf3 d5 3.exd5 exd5 4.d4 Nf6 5.Bd3 Bd6 6.Qe2+ Be6 7.Ng5 Qd7 8.Nxe6 Qxe6 9.Nc3 0-0 10.Nb5 Re8 11.Qxe6 Rxe6+ 12.Be3 
  =  &#40;0.18&#41;   Depth&#58; 18/31   00&#58;00&#58;00  657kN
2.Nf3 d5 3.exd5 exd5 4.d4 Nf6 5.Bd3 Bd6 6.Qe2+ Be6 7.Ng5 Qd7 8.Nxe6 Qxe6 9.Nc3 0-0 10.Qxe6 fxe6 11.Nb5 Nc6 12.0-0 a6 13.Nxd6 cxd6 14.Be3 
  =  &#40;0.19&#41;   Depth&#58; 19/31   00&#58;00&#58;00  929kN
2.Nf3 d5 3.exd5 exd5 4.d4 Nf6 5.Bd3 Bd6 6.Qe2+ Be6 7.Ng5 Qd7 8.Nxe6 Qxe6 9.Nc3 Nc6 10.Nb5 Qxe2+ 11.Kxe2 0-0-0 12.Nxd6+ Rxd6 13.Be3 Re8 14.c3 Kb8 15.Bc2 
  =  &#40;0.15&#41;   Depth&#58; 20/32   00&#58;00&#58;01  1419kN

(, Microsoft 30.05.2017&#41;

Lyudmil Tsvetkov · Post by **Lyudmil Tsvetkov** » Tue May 30, 2017 9:10 am

[d]rnbqk1nr/ppp2ppp/4p3/3p4/1b1PP3/2N5/PPP2PPP/R1BQKBNR w KQkq - 0 4

and in this MacCutcheon line, it also prefers 4.ed5 to the much better 4.e5

Code: Select all


o - Andscacs 0.86134, Blitz 1m Microsoft
rnbqk1nr/ppp2ppp/4p3/3p4/1b1PP3/2N5/PPP2PPP/R1BQKBNR w KQkq - 0 1

Analysis by Andscacs 0.86134&#58;

4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 6/6   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 7/7   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 8/8   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 9/9   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 10/10   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 11/11   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 12/12   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 13/13   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 14/14   00&#58;00&#58;00  0kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 15/15   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 16/16   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 17/17   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 18/18   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 19/19   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 20/20   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 21/21   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 22/22   00&#58;00&#58;00  1kN
4.exd5 
  +/=  &#40;0.35&#41;   Depth&#58; 23/23   00&#58;00&#58;00  1kN

(, Microsoft 30.05.2017&#41;

Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future

Re: Thoughts on Komodo 11, Free engines and the future