Go back
Question about engine choice in opening

Question about engine choice in opening

Only Chess

S

Joined
14 Jul 06
Moves
20541
Clock
24 Apr 08
Vote Up
Vote Down

Just out of interest, what does your copy of Fritz, Shredder or whatever suggest is the best play in the opening for both sides if you leave it running for half an hour or more?
Is there a general concensus or are they all different?
Also, I'm guessing that since they're tactics-driven they'd all go 1.e4.
My betting is on some form of Ruy Lopez.
Interested to see your results.
🙂

coquette
Already mated

Omaha, Nebraska, USA

Joined
04 Jul 06
Moves
1121334
Clock
24 Apr 08
Vote Up
Vote Down

1. f3 has been shown by my research with engines to be the move with the highest probability of winning.

JoL
Curb Your Enthusiasm

London

Joined
04 Nov 07
Moves
4259
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by Squelchbelch
Just out of interest, what does your copy of Fritz, Shredder or whatever suggest is the best play in the opening for both sides if you leave it running for half an hour or more?
Maybe some of the time but far from reliable and I'd guess most of the time the answer will be no.

Depends how exactly you're defining "opening" I suppose and the nature of the position.

J

g

Joined
29 Jul 01
Moves
8818
Clock
24 Apr 08
1 edit
Vote Up
Vote Down

ChessMaster plays the first move in a milli sec and does not start thinking until it is out of the book. Without the book Fritz plays the Petrov Defense and Rybka likes to play the Knights first. P.S. ChessMaster plays whatever book opening its selects by chance?

JoL
Curb Your Enthusiasm

London

Joined
04 Nov 07
Moves
4259
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by gambit3
Without the book Fritz plays the Petrov Defense ....
That depends on what version of Fritz you have, what its running on and how long you leave it.

I've just let Fritz 8 have a 30 minute think on my HP 510 laptop...

I gave it 1. e4 to consider and after half an hour its top line was:-

1. ... d5, 2. exd5 Qxd5 3. d4 e5 etc etc

Now, you'd be hard pressed to prove that 1. ... d5 loses but equally I don't think there are many people who would say it is Black's objectively best choice against 1. e4. 3. d4 must also be inaccurate btw.

Fritz's second line after half an hour was the Petroff

1. ... e5, 2. Nf3 Nf6, 3. Nxe5 d6 4. Nf3 Nxe4, 5. Nc3 Nf6 etc

Not too long ago 5. Nc3 would have been considered harmless but it gets a outings at the very highest level these days. Not at all sure about ... nf6 as a response however.


Curiously, after about 20 minutes Fritz had this Petroff line as its number once choice. After about 10 minutes it liked the Petroff but with the rather unlikely variation

1. ... e5, 2. Nf3 Nf6, 3. Nxe5 Qe7

Z

Joined
18 Apr 08
Moves
382
Clock
24 Apr 08
1 edit
Vote Up
Vote Down

Originally posted by JonathanB of London
That depends on what version of Fritz you have, what its running on and how long you leave it.

I've just let Fritz 8 have a 30 minute think on my HP 510 laptop...

I gave it 1. e4 to consider and after half an hour its top line was:-

1. ... d5, 2. exd5 Qxd5 3. d4 e5 etc etc

Now, you'd be hard pressed to prove that 1. ... d5 loses but e ed the Petroff but with the rather unlikely variation

1. ... e5, 2. Nf3 Nf6, 3. Nxe5 Qe7
d5 is equal with e5 and c5. Otherwise it would not be a famous opening that GMs play.

EDIT: And where's the flaw after Qe7?

g

Joined
29 Jul 01
Moves
8818
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by JonathanB of London
That depends on what version of Fritz you have, what its running on and how long you leave it.

I've just let Fritz 8 have a 30 minute think on my HP 510 laptop...

I gave it 1. e4 to consider and after half an hour its top line was:-

1. ... d5, 2. exd5 Qxd5 3. d4 e5 etc etc

Now, you'd be hard pressed to prove that 1. ... d5 loses but e ...[text shortened]... ed the Petroff but with the rather unlikely variation

1. ... e5, 2. Nf3 Nf6, 3. Nxe5 Qe7
The copy I have prefers the Petrov Defense. I have the free 5.32 released with no book. The Rybka demo is released with no book as well.

JoL
Curb Your Enthusiasm

London

Joined
04 Nov 07
Moves
4259
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by Zweite
d5 is equal with e5 and c5. Otherwise it would not be a famous opening that GMs play.

EDIT: And where's the flaw after Qe7?
You think ... d5 is as good as ... e5 and ... c5? Well that's up to you I suppose. As for whether it's a 'famous' opening that GMs play ... I think if you count the games in these defences that are played by, say, 2500+ elo you'll see ... d5 is somewhat less popular than the Sicilian or the open games.

That said, Vishy got a good position with it against Kasparov in the 1995 match. doesn't prove it's a top opening but certainly is playable. Probably less so at GM level 13 years on though.

w
If Theres Hell Below

We're All Gonna Go!

Joined
10 Sep 05
Moves
10228
Clock
24 Apr 08
Vote Up
Vote Down

it depends on what kind of 'positional' values has been hard coded into the engine. there are big differences on how different engines will evaluate moving the pawns in front of the castled king for example. some will storm them far easier than others (which is visible as big differences in the evaluation of such positions), but it's not because the 'saw something' ahead, as they can't really see far enough. they'll just add up the gain in space, mobility and decrease in king safety, and spit out a number. they don't really understand whether that king became weak or not in reality, they're just adding some preset value which the coder in his infinite wisdom saw fit to accompany 'h3' with.

so, what I'm getting at is this: if you think engines will give some knowledge on which opening is objectively the best, well, it won't happen. they'll just reflect the preconceptions the coder had about basic opening principles.

same applies to any later position as well btw, excluding endgame tablebases, which really are the objective truth (but at the same time largely meaningless to unassisted human play).

JoL
Curb Your Enthusiasm

London

Joined
04 Nov 07
Moves
4259
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by Zweite

EDIT: And where's the flaw after Qe7?
Didn't your daddy ever tell you about not moving your queen in the opening? ;-)

For specific reason as to why 3. ... qe7 is inferior rather than that general principle, well I'll leave you to research that one. You might also want to take a look at how often Kramnik has played it compared to the number of times he's played 3. ... d6 when you take a decision on which is best.

Z

Joined
18 Apr 08
Moves
382
Clock
24 Apr 08
Vote Up
Vote Down

Moving Qe7 in the Petroff has won me some strong scalps. I think that the Scandinavian is rarely seen, but then so is the King's Gambit.

JoL
Curb Your Enthusiasm

London

Joined
04 Nov 07
Moves
4259
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by Zweite
Moving Qe7 in the Petroff has won me some strong scalps. I think that the Scandinavian is rarely seen, but then so is the King's Gambit.
King's gambit is rarely seen at GM level for the same reason (which again doesn't mean it's not playable lower down the ranks).

Well done on your scalps with ... Qe7. Care to supply any games or contexts (when/where played under what conditions etc)?

S

Joined
14 Jul 06
Moves
20541
Clock
24 Apr 08
Vote Up
Vote Down

As White they prefer 1.e4 though?

I'm surprised that the Ruy (or Italian Game) hasn't featured though.
The Scandinavian is a surprise choice though.

Z

Joined
18 Apr 08
Moves
382
Clock
24 Apr 08
2 edits
Vote Up
Vote Down

Originally posted by JonathanB of London
King's gambit is rarely seen at GM level for the same reason (which again doesn't mean it's not playable lower down the ranks).

Well done on your scalps with ... Qe7. Care to supply any games or contexts (when/where played under what conditions etc)?
Fischer played the King's Gambit.
And indeed I would supply the games if I had the time to write down 50 moves for both sides. 😕
The context was in fairly strong tournaments and from quite recently, I started playing the Petroff a few months ago as Black.

Z

Joined
18 Apr 08
Moves
382
Clock
24 Apr 08
Vote Up
Vote Down

Originally posted by JonathanB of London
You think ... d5 is as good as ... e5 and ... c5? Well that's up to you I suppose. As for whether it's a 'famous' opening that GMs play ... I think if you count the games in these defences that are played by, say, 2500+ elo you'll see ... d5 is somewhat less popular than the Sicilian or the open games.

That said, Vishy got a good position w ...[text shortened]... top opening but certainly is playable. Probably less so at GM level 13 years on though.
To be honest, I think that e5, d5 and c5 are all adequate responses to e4. Seeing as the current world champion Anand plays it, that makes it fine by me, particularly if Kasparov could not gain advantage. 🙂

Cookies help us deliver our Services. By using our Services or clicking I agree, you agree to our use of cookies. Learn More.