Website founded by
Milan Velimirović
in 2006

7:43 UTC
ISC 2021



Remember me

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

Rating lists


MatPlus.Net Forum General Forced mutual perpetual
You can only view this page!
Page: [Previous] [Next] 1 2
(1) Posted by Hauke Reddmann [Sunday, Sep 26, 2021 21:50]

Forced mutual perpetual

I know there was an article about fairy mutual perpetual check
(where again was the proof that orthodox isn't possible? :-),
but what about the forced version?

A known position, somewhat adapted:
(= 3+3 )

Circe, 1.Rh3+ and so on

I now would be interested if this can be done with (a,b)-riders
and -knights only. Note that the well known camel solution doesn't
work here anymore!
(Read Only)pid=21598
(2) Posted by James Malcom (Rewan Demontay) [Monday, Sep 27, 2021 16:36]

Another classic is easily drained for points. Nightriders, trapped in hell!

(= 5+5 )

(Read Only)pid=21601
(3) Posted by Joost de Heer [Monday, Sep 27, 2021 17:10]

1. Kc3 Rd3. Or even simpler: 1. Nd5
(Read Only)pid=21602
(4) Posted by Hauke Reddmann [Monday, Sep 27, 2021 19:07]

...which is exactly the point ;-)
With riders-only, the checking riders have to
be pinned, and the pinners also pinned,
probably requiring a large board.
(Read Only)pid=21604
(5) Posted by Joost de Heer [Wednesday, Sep 29, 2021 13:32]
(Read Only)pid=21607
(6) Posted by seetharaman kalyan [Wednesday, Sep 29, 2021 23:37]

I could not understand (in post 2) why the black rook cannot interpose?
(Read Only)pid=21608
(7) Posted by Andrew Buchanan [Thursday, Sep 30, 2021 08:09]

Seetheraman wrote:
I could not understand (in post 2) why the black rook cannot interpose?

Yes as others posters noted, that one is cooked. A rare inaccuracy by Rewan.

Meanwhile, I have found a more economical forced perpetual check than post 1, in Circe. It just has 5 pieces, and I believe the loop is unique. (The only variable is what point in the loop one chooses to show in the diagram.) I think the most effective way to test such a claim is to present its construction as a challenge for other folk here. If no-one gets it, I will post my idea in a couple of days.
(Read Only)pid=21610
(8) Posted by Jacques Rotenberg [Thursday, Sep 30, 2021 20:27]

Yes Andrew, it seems rather simple in 5 pieces, and nice.

It there a possibility to get a longer loop ?
This question led me to this :
(= 9+8 )

1.Sh3+ B×a3(Sg1)+ 2.S×a3(Bf8)+ B×h3(Sb1)+ 3.S×h3(Bc8)+
but still... is there a longer loop ?
There should be!
(note that the same knight g1 plays again to h3 only at move 7 (after 6 moves!))
(Read Only)pid=21611
(9) Posted by Siegfried Hornecker [Thursday, Sep 30, 2021 20:38]

I composed an endgame study a while ago where a mutual perpetual is forced by check. It was something like White having cannons (Pao in fairy terms) on the entire right side, king on the g-file. Black has rooks on the entire a-file (should have made that d-file, now that I think about it), king on the e-file.

So the kings can wander up and down, but there is no way to escape the perpetual check.

And because I'm a moron, White could checkmate in 1 move in the initial position instead of promoting to a cannon, so the study was incorrect, and I narrowly avoided being banned from feenschach, for being too stupid, by finding it myself and apologizing. But the idea of it fits here.

So here is a corrected version:

(= 9+9 )

= Cannon (Pao)

And if I knew how to make a 5x5 board, that would be the perfect setting probably.
(Read Only)pid=21612
(10) Posted by seetharaman kalyan [Thursday, Sep 30, 2021 21:10]

While your idea is fine your setting has a few #1 like Sf3, Sc3/d2. A few wP plugs necessary -- unless we consider a help perpetual !!
(Read Only)pid=21613
(11) Posted by Jacques Rotenberg [Thursday, Sep 30, 2021 21:50]

I just wanted to show that such a loop may have a "normal" beginning, put at the beginning the diagram obtained after any of the moves of the loop if you prefer...
(Read Only)pid=21614
(12) Posted by Jacques Rotenberg [Thursday, Sep 30, 2021 22:40]

if fact, the stipulation I had in mind was something like "get a loop", but if you say that the stipulation may be "white plays and draw", it seems that you can even get a better position with less pieces.
(= 8+5 )

White plays and draw

1.Sa3+ B×h3(Sb1)+ 2.S×h3(Bc8)+ B×a3(Sg1)+ 3.S×a3(Bf8)+
it seems to work
(Read Only)pid=21615
(13) Posted by Joost de Heer [Friday, Oct 1, 2021 08:00]

With a fairy piece, 4 pieces can be achieved:
(= 2+2 )

= Fers+Alfil ((1,1)+(2,2) leaper)
1. Rh1+ Qd1+ 2. Rd1(+Qd8)+ Qd1(+Rh1)+ etc
(Read Only)pid=21619
(14) Posted by Andrew Buchanan [Friday, Oct 1, 2021 09:23]

(= 2+3 )

Here is what Hauke & I'm pretty sure Jacques spotted. Thanks for confirmation!

What is the most economical "Type B" Circe inexorable mutual perpetual check? "Type B" means that the position does not contain a check, but we are told who has the move? That's quite easy - I have found one with 6 units. And what about "Type A", where there is no check, but we can deduce who has the move? Or "Type D" = Duplex, where the position heads inexorably towards mutual perpetual check whoever is on move?

I like inexorability, so I prefer Jacques' original and beautiful SSSBB loop. I don't have a problem with the couple of extra pawns this implies, and it's nice that the only non-standard material is the thematic S.
(Read Only)pid=21620
(15) Posted by Hauke Reddmann [Friday, Oct 1, 2021 10:21]

@Andrew: Forget the PM, hadn't read the preceding post then. And I have the 6-piecer. Really tricky, a Bc4 protecting itself :-)

(= 4+5 )

An alternate possibility to force White's entry into the loop.

(= 7+5 )

A mixture of the above and Jacques idea. (WTM again) Maybe this could be
used for Andrew's task - this *almost* it:

(= 10+11 )

WTM is stalemated, BTM must enter the loop.

Finally, by the fairy mantra, "If in doubt, cheat" :P,

(= 7+13 )

Either Diagram Circe or Fischer starting BQ...any.
(Read Only)pid=21621
(16) Posted by Andrew Buchanan [Friday, Oct 1, 2021 18:22]

(= 2+4 )
Circe. White to move.

Here's the 6-piece Type B I had in mind (and I think Hauke guessed a colour-flipped form). It's not unique, the bBc5 could equally, but less interestingly, be on e3, f2 or g1.
(Read Only)pid=21622
(17) Posted by Hauke Reddmann [Saturday, Oct 2, 2021 10:33]

Yup. (Had a Qa2 instead of Ra4, in that case the B must be there.)
(Read Only)pid=21623
(18) Posted by Andrew Buchanan [Sunday, Oct 3, 2021 11:13]

Here is the "Type D" case:
(= 12+4 )
Circe. Result? Duplex.

Could alternatively have wSf8, but then the *non-checking* Sf8xd7[+Bc8] entry to the true mutual perpetual check loop can be postponed indefinitely, which doesn't seem in the spirit of the "forced mutual perpetual check". Or could have wBf8, but I now feel that standard material is more attractive than strict economy, for this particular challenge.
(Read Only)pid=21628
(19) Posted by Jacques Rotenberg [Sunday, Oct 3, 2021 19:49]

@ Andrew

I had not your very nice 5 pieces B/Q.
I thought I had a R/Q feature, but it was a mistake!
(Read Only)pid=21630
(20) Posted by Hauke Reddmann [Monday, Oct 4, 2021 09:40]

@Jacques: Had QR also first, also overlooked something...I know I'm telepathic but I didn't know my powers reach THAT far :-)

@Andrew: I see you already edited the post, saving me the one or other nagging. :-)
But one comment: Since the stipulation is "Results", one could troll the solver
a bit: e.g. add wPc7,wSg8,wPa5,bPa6 and a bP tempo. Now White could desperately try to free
the Q before the inevitable happens...
(Read Only)pid=21631

Page: [Previous] [Next] 1 2

MatPlus.Net Forum General Forced mutual perpetual