Komodo 2.03 release is imminent

[Mike S.]

And i want to show that SMK and his shredder have started the race to a complete chessplayer in silicon.

I can support that impression 100%, emphasizing the word "complete". Shredder is not on top anymore, or currently but that is most probably because it is an "old" engine which AFAIK does not use bitboards etc., so it may be difficult to keep up with current developements in terms of search speed. But in terms of knowledge and understanding of a position, in 9 from 10 cases where another engine fails, you can start Shredder and it will solve sooner or later. It is what I call a "quality engine". I put Fritz in the same category. There are (in general!) stronger engines than these, but they have holes, in the early opening and especially in the endgame - if have seen unbelievable ugly things at TCEC - which commercial engines just cannot afford to have.

Major example blind bishop knowledge. As long as an engine doesn't have it, it's an just amateur engine disregarding it's ranking on a rating list.

I have my demanding day today. Again, many thanks for so much which is provided for free anyway. Maybe chess players are not easy to satisfy.

[stevenaaus]

There is lot of work to do, because computers are not complete chessplayers.

Again I have to say you are rambling on here - tell us something we don't already know. You act like you are just now figuring this all out but everything you are saying is common knowledge that has been known for years and your insights are not original. They are correct, but not original or even interesting.

O-O-O by Carlsen ~is~ an aggressive move, and inspirational to see him playing chess like this. But i tend to agree with Don, sadly. Computers are correct. Magnus couldn't beat the top computers anymore than Kasparov could.

Of the six best engines i could test, none of them selected O-O-O in their top 5 moves. Shredder Classic alone initialy chose it, but after more time changed it's mind.

[Don]

My response came off stronger than I intended and I apologize. Your thoughts are good ones but as I said I've been hearing it for almost 40 years. It's generally presented as if it's some kind of revelation so it does tend to be annoying when presented as if it's a new idea that would be of great benefit "if only" the dumb programmers would just listen.

There was nothing you said that was wrong or not a good idea in general.

First i want to show that it has to be done further work to play chess in all options to reach the olymp of perfect chess.

Why don't you tell us something we don't know

I have posted many times here that I believe there are still hundreds of ELO left before perfect play. As a computer chess developer don't you think I have any perspective on this?

It´s not a question of style, it´s a question of ability to play the move that is neccessary to make progress in the position.

We don't strive to make it play like humans because we think that makes it play stronger. We just feel that is more pleasing and more fun to play against.

In fact I think making it play "like" humans probably is not best for the ultimate strength.

A perfect player will not play that much like a human in my opinion.

Secondly with my example of how Shredder it handles i wanted to show that someone has made the first steps in this area and the flag is on the moon.

By showing us 1 example? I think Junior is a better example for this type of thing.

I honestly do not know if we play more like a human that other programs as I have never spent any serious time trying to evaluate the "humanness" of any give program. All I really know is that Larry and I prefer it to play like a human and have put some effort into this. We are not claiming we have succeeded or that we do it well.

A project Larry has been pestering me to start on at some point is a serious attempt to tune our program to play more like humans. The trick is to do this without sacrificing ELO. So we are far from our ultimate goal and I don't know if we will ever fully be satisfied that we have the most human playing program. Maybe we never will - but it's not something we lose any sleep over.

There is lot of work to do, because computers are not complete chessplayers.

Programs have the same flaws they did 40 years ago, they are just hidden better. The horizon effect is minimized and disguised by much greater depths but it's still there. And programs still do not reason about positions so they do approach the game differently that humans.

Again I have to say you are rambling on here - tell us something we don't already know. You act like you are just now figuring this all out but everything you are saying is common knowledge that has been known for years and your insights are not original. They are correct, but not original or even interesting.

They won they games mainly by search and tactics.

Now the aliens of chess are stronger than humans. But they have enormous weaknesses and will never play perfect chess if they didn´t learn from human chess.

And the first lesson is to get a better feeling for attackingchess.

Only that was my mission.

And i want to show that SMK and his shredder have started the race to a complete chessplayer in silicon.

[Ron Langeveld]

[d]r3k2r/1p2bppb/1qp2n1p/p4P2/P1BP4/1PN5/4Q1PP/R1B2RK1 b kq -

Komodo insists on playing Kf8 here, a move that humans would reject almost instantly. Maybe a good position to start looking for further improvements.

[mclane]

First i want to show that it has to be done further work to play chess in all options to reach the olymp of perfect chess.

Hi Thomas.
IMO the olymp of perfection, that would be the MACHEIDE lasker dreamed about in his book: die philosophie des unvollendbar, 1919

It´s not a question of style, it´s a question of ability to play the move that is neccessary to make progress in the position.

do you believe that there is in any position only ONE move that
makes progress ?

My theory about it is, that chess players / chess programs
play a thesis. and sometimes the thesis works, sometimes its refuted.

Secondly with my example of how Shredder it handles i wanted to show that someone has made the first steps in this area and the flag is on the moon.

shredder won the championship in japan. but has not sold this engine
to the public. right. i wonder why.
if i would win a championship, i would sell the engine.

There is lot of work to do, because computers are not complete chessplayers.

They won they games mainly by search and tactics.

Now the aliens of chess are stronger than humans. But they have enormous weaknesses and will never play perfect chess if they didn´t learn from human chess.

And the first lesson is to get a better feeling for attackingchess.

Only that was my mission.

And i want to show that SMK and his shredder have started the race to a complete chessplayer in silicon.

aha.
which version ? shredder 12 ?

[Don]

[d]r3k2r/1p2bppb/1qp2n1p/p4P2/P1BP4/1PN5/4Q1PP/R1B2RK1 b kq -

Komodo insists on playing Kf8 here, a move that humans would reject almost instantly. Maybe a good position to start looking for further improvements.

It's easy to find positions to improve Komodo or any other program, I'm not sure whether this is one of them or not.

You picked a position here where a bishop is pinned in front of the king and that castling loses. I tried this position in Houdini and it also wants to play Kf8. After a time it switches to another move, but Kf8 is still in the PV. Komodo also switches to another move later with Kf8 still in the PV.

One problem here is the center is open, the king is on e8 and it cannot castle. So you find a position where the king is between a rock and a hard place and then criticize it's choice as being non-human like?

Kf8 gets the rook trapped but I assume Houdini as well as Komodo feel that a more serious concern is the king on an open file in the center of the board with a lot of activity in the center.

It's easy to see this is a king safety issue because if you turn King Safety Aggressiveness down to 3, the program will not play Kf8 but instead will find a way to eventually castle.

So I forced a sequence in that allowed it to castle and let the computer play a few moves and the white bishop went to g2 and saced itself on g7 with a king side attack.

I don't know if Kf8 should ever be played in this position due the rook, and maybe there is a way to deal with the attack in the center without having play that or castle right away, but I'm having a hard time thinking Kf8 now or in a couple of moves is totally ridiculous. I do know that Komodo hates having to play Kf8 so there must be a pretty good reason it does. I assume Houdini also hates Kf8 even though it also plays that move.

I think you need a much better example to show moves where a human would never play what the computer does. You example needs to one where the move the computer plays is wrong. I know many such positions exist, but they are a lot harder to find than they used to be and I don't think this is one of them.

Part of the reason computers are now stronger than humans is that they find moves the computer wouldn't play, but unfortunately for the human player the move it finds is GOOD.

[FWCC]

It seems to me that 0-0-0 is just a simple preference in this matter.

FWCC

[Don]

It seems to me that 0-0-0 is just a simple preference in this matter.

FWCC

I'm not a chess master, but O-O-O looks horrible to me. It cannot be played now because it's an outright blunder and I think it takes several moves of preparation to even be able to play it all. And even then there is not reasonable safety there.

Are you telling me that a human would think this is a ok?

[Ron Langeveld]

It is true that black is already in a tough spot. The moves leading up to this position, starting from :

[d]r3kb1r/1p1n1ppb/1qp4p/p4P2/P1BP4/2N5/1P4PP/R1BQ1RK1 b kq - 0 17

were 17...Nf6? (Be7!) 18. Qe2 Be7 19.b3 Kf8?

Komodo's 17...Nf6 surely deserves a question mark, but with 19...Kf8 black opts to play with 1 rook less for quite some time. Given the fact that white already had a lead in development means that 1-0 is just a matter of time. Komodo may count a penalty for this trapped rook but I doubt that the penalty is high enough. Humans see the trapped rook on the spot and start too look for better moves like 19...Qxd4!. Houdini finds this rather soon @ 23 plies I think, without Kf8 in the PV. 20.Kh1 0-0 21.Qxe7 Qxc3 22.Ba3 b5 after which I believe black should be able to hold. In short: my assessment of the position after 19...Kf8 is somewhat different than yours or Komodo's. I am wondering how Larry would assess the position.

[Don]

It is true that black is already in a tough spot. The moves leading up to this position, starting from :

[d]r3kb1r/1p1n1ppb/1qp4p/p4P2/P1BP4/2N5/1P4PP/R1BQ1RK1 b kq - 0 17

were 17...Nf6? (Be7!) 18. Qe2 Be7 19.b3 Kf8?

Komodo's 17...Nf6 surely deserves a question mark, but with 19...Kf8 black opts to play with 1 rook less for quite some time. Given the fact that white already had a lead in development means that 1-0 is just a matter of time. Komodo may count a penalty for this trapped rook but I doubt that the penalty is high enough. Humans see the trapped rook on the spot and start too look for better moves like 19...Qxd4!. Houdini finds this rather soon @ 23 plies I think, without Kf8 in the PV. 20.Kh1 0-0 21.Qxe7 Qxc3 22.Ba3 b5 after which I believe black should be able to hold. In short: my assessment of the position after 19...Kf8 is somewhat different than yours or Komodo's. I am wondering how Larry would assess the position.

I looked at similar lines yesterday where black castles and tries to hold off the attack. My point was not whether he could survive it, but what motives Houdini and Komodo to prefer Kf8 at any reasonable depth below 20 ply or more instead of castling. Any program with good king safety will think twice about castling. As I mentioned I can get Komodo to play O-O early by turning king safety down.