Website founded by
Milan Velimirović
in 2006

9:28 UTC
ISC 2023



Remember me

Forgot your
Click here!
to create your account if you don't already have one.

Rating lists


MatPlus.Net Forum Threemovers Peter Gvozdják, Štefan Sovík (Slovakia), 11th WCCT, 1st place
You can only view this page!
Page: [Previous] [Next] 1 2
(21) Posted by Neal Turner [Thursday, Jan 19, 2023 09:40]

This has all been very interesting, the general consensus being that it's a great problem, with one or two dissenting voices making what many would consider minor quibbles.
So I'm surprised that nobody has mentioned the claim of 'Cyclic dual avoidance'.

After the key we have a situation with three knight checks, none of which work.
Each of the pawn moves enables one of the checks, but - importantly - does not prevent the others because they were never possible in the first place.
No duals are being avoided here.

Before you write this off as the rantings of an idiot, I give the following quote:
'I do not believe that we can use the term “dual avoidance” if it is achieved merely by means of departure effects.'
Friedrich Chlubna in his commentary to problem 257 in Chris Feather's 'Black To Play'.
(Read Only)pid=24318
(22) Posted by Hauke Reddmann [Thursday, Jan 19, 2023 10:44]

Can't resist to quote another famous problem composer:

"‘When I use a word,’ Humpty Dumpty said in rather a scornful tone,
‘it means just what I choose it to mean–neither more nor less.’
‘The question is,’ said Alice, ‘whether you can make words mean different things–that’s all.’
‘The question is,’ said Humpty Dumpty, ‘which is to be master–that’s all.’"

My Bayesian estimate we can define all problem chess mathematically exact
such that even a computer can output annotate a problem with a tag list: 98%
My Bayesian estimate we can get the problem chess community to follow
these definitions without talking back: 2% :-)
(Read Only)pid=24319
(23) Posted by Joose Norri [Thursday, Jan 19, 2023 13:03]

As for dual avoidance: claiming it for this problem is contrary to all accepted definitions more or less for whole of problem history. But increasingly many terms are used in a lax way. (If John Rice is reading this, he might like to quote from Chess Wizardly. This being MatPlus, Milan is also quotable.)

As a layman I am also interested in the take of the experts on the constructional principle: since there is not much Black initiative to provide for, the White moves seem to need to have less effects, making it that much 'easier'.

Whether I'm right or wrong there, it does make the problem less exciting to my eyes and mind.
(Read Only)pid=24320
(24) Posted by Neal Turner [Thursday, Jan 19, 2023 14:03]

You imply that it's all just a matter of opinion, but if we're going to use these terms they should at least be well-defined, and in the case of Dual Avoidance it is!

Taking the advice of a 'layman' and looking at John Rice's 'Chess Wizardry' and the Encyclopaedia they both start in a similar vein:
1) that Black's move makes an error that seemingly allows White multiple continuations.
And then they go on to say that:
2) a secondary, compensatory effect of the black move eliminates all the dualistic lines except one.

So it's not just a matter of a position where one line succeeds and others fail, but of how we arrived at that position.
As we see above, to be classed as Dual Avoidance the moves leading to the position must meet specific criteria - which none of the pawn moves do in this problem.

So I would say that anybody continuing to support the cause of Dual Avoidance in this example needs to come up with a new definition that will take in what is happening here.
(Read Only)pid=24321
(25) Posted by Viktoras Paliulionis [Thursday, Jan 19, 2023 23:21]

The Encyclopedia also has a definition for another type of dual avoidance:

Passive Dual Avoidance. White second move carries a harmful effect which eliminates it as a mate.

I think there's Passive Dual Avoidance in this problem:

1.b8=S! waiting
1…b4 2.Se6+ Ke3/Kd5 3.Sc4/Bc4#
2.Sxb5+? (withdrawal) Ke3! 3.Sc4??
2.Sc6+? (unguard of c6) Kd5! 3.Bc4??

1…g5 2.Sxb5+ Ke3/Kd5 3.S×f5/R×f5#
2.Sc6+? (unguard of c6) Kd5! 3.R×f5??
2.Se6+? (unguard of e6) Kd5! 3.R×f5??

1…f×g4 2.Sc6+ Ke3/Kd5 3.R×e4/B×e4#
2.Se6+? (unguard of e6) Kd5! 3.B×e4??
2.Sb5+? (unguard of c4) Kd5! 3.B×e4??
(Read Only)pid=24322
(26) Posted by Hauke Reddmann [Friday, Jan 20, 2023 10:01]

@Neal: Fact aside that I am my usual trolling self :-) - no, I'm NOT
the opinion of Humpty Dumpty that words *shall* mean what you want
them to mean, but alas, they *will* mean what you want them to mean,
since you don't get three problemists to agree on one opinion, and
even worse, there is no way from the mere facts, even if agreed upon
("this problem shows/shows no dual avoidance) to aesthetics
("I like this problem").
(Read Only)pid=24324
(27) Posted by Neal Turner [Friday, Jan 20, 2023 12:41]

Yes, it's true that criteria 2) above refers to what is known as 'active' dual avoidance, but it's not correct to invoke passive dual avoidance here because criteria 1) hasn't been met.
The essential point of dual avoidance is that the dualistic possibilities are created dynamically by the black moves, we do not have that in this problem.
(Read Only)pid=24325
(28) Posted by Viktoras Paliulionis [Friday, Jan 20, 2023 13:23]

The definition in the Encyclopedia is very limited, applying only to two-move problems, and it is unclear how to apply it to three or more-move problems or to helpmates.

I think that the first criteria is also satisfied, since Black's defense has a weakening - unguard, and this enables three possible continuations.
(Read Only)pid=24326
(29) Posted by Neal Turner [Friday, Jan 20, 2023 14:59]

"this enables three possible continuations"

Yes, but only one at a time - there's no duals!
(Read Only)pid=24327
(30) Posted by Viktoras Paliulionis [Friday, Jan 20, 2023 21:00]

1…f×g4 unguards square e4 and allows 3.R×e4/B×e4# after Ke3/Kd5. We have to attack bK to force it to move. We can do this with three possible checks by wS's (possible duals), but two of them have a harmful effect, so only one works. The duals are avoided. It is the same in two other variations.
(Read Only)pid=24328
(31) Posted by Neal Turner [Saturday, Jan 21, 2023 11:59]

Let's look at a problem which definately does show dual avoidance.

René Millour
1st HM Stella Polaris 1967
(= 12+7 )

1.Se2! thr. 2.Qd4#

1.- Qf6 2.c4# (2.Sc3?, 2.Sef4?)
1.- Re4 2.Sc3# (2.Sef4?, 2.c4?)
1.- Sb3 2.Sef4# (2.c4?, 2.Sc3?)

1.- Se6 2.Rf5#
1.- Kd6 2.d8Q#

After the key, the moves 2.c4, 2.Sc3, 2.Sef4 fail only because of the flight on d6.
So when the black moves leave white lines opened on to the flight square we have three mates - duals - oh no!
But on completion of Black's move we find that two of the possible mating squares have become guarded - duals avoided - phew!

We compare this with Peter's and Štefan's example.
The first thing to say is that it doesn't matter whether it's a 2 or 3 mover - the logic is the same.

(= 11+6 )

Peter Gvozdják, Štefan Sovík (Slovakia). 11th WCCT, 1st place. Mate in 3

1.b8=S! waiting
1…Ke3 2.Be5 b4/g5/f×g4/e×d3 3.Sc4/S×f5/R×e4/Bd4‡
1…Kd5 2.e3 b4/g5/f×g4/e×d3 3.Bc4/R×f5/B×e4/Rd4‡
1…b4 2.Se6+ (2.Sb5+?, 2.Sc6+?) Ke3/Kd5 3.Sc4/Bc4‡
1…g5 2.S×b5+ (2.Sc6+?, Se6+?) Ke3/Kd5 3.S×f5/R×f5‡
1…f×g4 2.Sc6+ (2.Se6+?, 2.Sb5+?) Ke3/Kd5 3.R×e4/B×e4‡

After the key we have the three knight checks which fail because of the black king flights.
The black moves each unguard a square which will allow the king to be mated on these flights - but this is where the problems part company.
Where in René's problem Black's error allows in principle all three white moves, here it only allows one.
The reason the failing moves fail hasn't changed, the black move has no effect on them - nothing has been avoided.
Put them down as 2nd move tries, but not duals.

It would be interesting to hear Peter's take on this.
I would go with his opinion (as long as it's the same as mine).
(Read Only)pid=24330
(32) Posted by Viktoras Paliulionis [Saturday, Jan 21, 2023 20:05]

I'm not an expert on direct mate themes, I'm just trying to apply the definition of a dual avoidance to a three-mover. With two-movers everything is clearer, it is better to compare with three move examples.

In the FIDE Album, I found a three-mover with the keyword "cyclic dual avoidance" which is somewhat similar to Peter's and Štefan's example.

Sygurov, Aleksandr; Fomichiov, Evgenij
I. Yarmonov-50 JT
Spec. Prize
(= 14+9 )

FEN: F5B1b/2NRPp2/1p1N1k1P/1qP4B/Q2P2pK/1bnP2P1/7P/4r3

1.h3! [2.Sde8+ Kf5 3.hxg4#]
2.e8=S+? Rxe8! 2.Sce8+? Ke6!
1...Bd1 2.e8=S+ Rxe8 3.Rxf7#
2.Sce8+? Ke6! 2.Sde8+? Kf5!
1...Qxd3 2.Sce8+ Ke6 3.Bxf7#
2.Sde8+? Kf5! 2.e8=S+? Rxe8!

1...gxh3 2.g4 Re4 3.Sde8#
1...Qxd7 2.Qxd7 Re6 3.Sce8#
1...Re4 2.dxe4 Sxe4 3.e8=S#

If there is a cyclic dual avoidance in this problem (I hope the FA judges are right), why is it not in the Peter's and Štefan's example?
(Read Only)pid=24331
(33) Posted by Neal Turner [Saturday, Jan 21, 2023 23:04]

What a joke.
There's no semblence of dual avoidance here.
(Read Only)pid=24332
(34) Posted by Miodrag Mladenović [Sunday, Jan 22, 2023 08:37]

I agree with Neal. There is not at all "dual avoidance" in the problem published in the post (32). The black defenses Bd1 & Qxd3 do not contain common black errors at all. The Bd1 unguards f7 and Qd3 open lines a4-d7. In order to achieve "dual avoidance" there should be common black error created by the black defense. Most likely it's unintentional error in FA.
(Read Only)pid=24333
(35) Posted by Jacques Rotenberg [Sunday, Jan 22, 2023 10:11]

Let's have a look at what happens in the Peter's and Štefan's problem

1…b4 unguards c4 and allows white to mate on c4 in both variations 2.Se6+!
2...Ke3 3.Sc4‡
2...Kd5 3.Bc4‡
2.Sb5+? removes the mating piece needed after 2...Ke3 and removes the control of c4 needed after 2..Kd5
2.Sc6+? gives a flight on c6 and prevents the mate after 2...Kd5
2.Se6+! also gives a flight but, fortunately 3.Bc4‡ controls e6

1…f×g4 unguards e4 and allows white to mate on e4 in both variations 2.Sc6+!
2...Ke3 3.R×e4‡
2...Kd5 3.B×e4‡
2.Sb5+? unguards c4 so that 3.Bxe4+ is no more a mate after 2...Kd5
2.Se6+? gives a flight on e6 and prevents the mate after 2...Kd5
2.Sc6+! also gives a flight but, fortunately 3.Be4‡ controls c6

1…g5 unguards f5 and allows white to mate on f5 in both variations 2.S×b5+!
2...Ke3 3.S×f5‡
2...Kd5 3.R×f5‡
2.Sc6+? gives a flight on c6 and prevents the mate after 2...Kd5
2.Se6+? removes the mating piece needed after 2...Ke3 and removes the control of e6 needed after 2..Kd5
2.S×b5+! has no negative effect

sorry, it's a bit boring, but it helps for clarity.

So... how can we explain the content ?
I would say :

Were white be able to give a "neutral" check (for example, after all these 3 black moves will be born a white knight a1) this neutral check would solve (with a Sa1, 2.Sb3+! would work each time)
this underlines that after 1…b4, 1…f×g4, 1…g5 white only needs to give check, so the fact that only one works each time, is because of the negative effects of the others

And so : there is here an anti-triple.

Perhaps because the effects are on the one hand not so clean, and on the other hand rather simple, the authors did not write the theme, but only the tries.
(Read Only)pid=24334
(36) Posted by Jacques Rotenberg [Sunday, Jan 22, 2023 10:29]

@ neal : I would not use the word "cyclic" here.
(Read Only)pid=24335
(37) Posted by Viktoras Paliulionis [Sunday, Jan 22, 2023 12:47]

I think that Jacques explained in more details the same that I tried, and he showed that formally the Peter's and Štefan's problem satisfies the definition of "dual avoidance".

@Miodrag: I read the definition of "dual avoidance" in the Encyclopedia once more and didn't find that it required black defenses to contain common black errors.
(Read Only)pid=24336
(38) Posted by Neal Turner [Sunday, Jan 22, 2023 14:56]

As we'd expect, a very good analysis from Jacques and I'm glad of it because I see it as proving my point!
What's demonstrated is a cycle of second move tries, but not cyclic dual avoidance!

One last go.

We can see when it's written down (from the Millour example):
1.- Qf6 2.c4# (2.Sc3?, 2.Sef4?)
1.- Re4 2.Sc3# (2.Sef4?, 2.c4?)
1.- Sb3 2.Sef4# (2.c4?, 2.Sc3?)
where the confusion might be coming from, as it really does look as though it's all about the white moves and how they're cycling through the variations.
But this is completely misleading, because (as Misha points out) it's actually ALL ABOUT THE BLACK MOVES.
I attempted to get the ideas across in post(24) here I'll try to do a better job.

For Dual Avoidance the effects of the black moves are very specific - they can be put in a hierarchy:
- Primary effect: the defensive motif.
- Secondary effect: the error - which enables multiple white replies. It's the ENABLING which is important here.
- Tertiary effect: the duals are eliminated, leaving just a single refutation.

All this content from a single move, it's no wonder dual avoidance has attracted the attention of composers down the years.
Many famous themes address dual avoidance, differentiated by the mechanisms by which the effects are produced.
The secondary effect features in the Herpai with its own way of producing the duals, while for the tertiary effect we have the Java, Mari, Fröberg, Barulin each with their individual mechanisms for achieving the avoidance.
And there are lots of others to find for those who enjoy browsing the Encyclopaedia.

BTW -the definition of 'passive' dual avoidance in the Encyclopaedia is very loose, other authorities take a stricter view.
Encyclopaedia: 'White second move carries a harmful effect which eliminates it as a mate.' but John Rice in Chess Wizardry points to a particular scenario: White must avoid annulling a specific weakness that resulted from Black's move. So for instance in the so-called Mansfield-Schiffman theme Black actively self-pins a piece and White must avoid unpinning it.

After all this many must be wondering: So what does this guy do with dual avoidance in his own stuff?
So here you are:

neal turner
Springaren 161 mars 2022
(= 10+7 )
S#2 SAT royalgrasshoppers

1.d7! (> 2.Be2+ rGc2# )
1... Se7 2.S4g6+ f6# (2.S4e6+? f6+ 3.rGf7)
1... Sd6 2.Sd5+ Se4# (2.Sg2+? Se4+ 3.rGa5)

White ponders the idea of forcing the black king to c2, but sees that the check on b1 is easily handled with blocks on both b1 and d3.
With the key a new front is opened on c8 so now Black's arrival on c2 will result in double-check and mate.
However the knight, which was pinned, is now free and by vacating c8 it leaves the square guarded, thwarting White's plan.
In the resulting sequences we get Dual Avoidance arising from moves of the black knight guarding squares in the white king’s extended field – characteristic effects of SAT.
(Read Only)pid=24337

No more posts
Page: [Previous] [Next] 1 2

MatPlus.Net Forum Threemovers Peter Gvozdják, Štefan Sovík (Slovakia), 11th WCCT, 1st place