Website founded by
Milan Velimirović
in 2006

13:31 UTC
ISC 2020



Remember me

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

Rating lists


MatPlus.Net Forum Misc An interesting retraction problem
You can only view this page!
(1) Posted by Bob Baker [Sunday, Aug 18, 2019 23:07]

An interesting retraction problem

Here is a problem I found interesting. It was composed and posted yesterday by 'Arisktotle' in one of the forums.

Find a legal sequence of moves leading to the position below such that the player who made the last move can retract it and replace it with a winning move.

(= 4+2 )

(Read Only)pid=17770
(2) Posted by Rewan Demontay (Real Name: James Malcom) [Monday, Aug 19, 2019 05:19]; edited by Rewan Demontay (Real Name: James Malcom) [19-08-19]

It's not the fastest possibile PG by far, but it's still something-

1. g4 f5 2. gxf5 e5 3. d4 exd4 4. Qxd4 Bc5 5. Qxc5 Nf6 6. Qxa7 O-O 7. Qxa8 h6 8. Bxh6 Nh7 9. Qxb7 Rf6 10. Qxb8 Rg6 11. fxg6 Qg5 12. Qxc7 Qxh6 13. Qxd7 Qh4 14. Kd2 Qxf2 15. Kd3 Qxg1 16. Ke4 Qxh1+ 17. Ke5 Qg2 18. Ke6 Qxe2+ 19. Kd6 Qxc2 20. Ke7 Qxb1 21. Ke8 Qxa2 22. Qxc8 Qxa1 23. Bc4+ Kh8 24. Bg8 Qxb2 25. Qf5 Qf6 26. Qxf6 Nxf6+ 27. Kf8 Nh7+ 28. Bxh7 1/2-1/2

The retraction move is easy: gxh7 instead of Bxh7.

I have interacted with Arisktotle myself on the fourms before. They are a prominent problemist over there.
(Read Only)pid=17774
(3) Posted by Siegfried Hornecker [Monday, Aug 19, 2019 05:49]

Sadly I can't find it in the forum, as I am really curious as to what the solution is. I tried everything, I think, and nothing works. Of course Rowan's idea still is a draw, as the only way to get Ph2 to the g-file is per stalemate.
(Read Only)pid=17775
(4) Posted by Geir Sune Tallaksen Østmoe [Monday, Aug 19, 2019 10:00]

Retract Ph5xPg6 and play 1.Bg8. Then Black has one pawn too many to get stalemated after 1...gxh5 2.h4 or 1...g5.

Pf5xPg6 does not work, I think, as after 1.Bg8 gxf5 White does not manage to stalemate Black while blocking the f-pawn.
(Read Only)pid=17776
(5) Posted by Bob Baker [Monday, Aug 19, 2019 11:25]

Geir has found the solution. When I posted my reply on, I gave the clearly legal position below with the moves 1. Bh7 Rg6 2. Nxg6+ fxg6 and noted that instead of hxg6, White has a mate in 14 starting with Bg8. Arisktotle commented that, as a composer, he prefers the mate-in-5 variation after ...Pgxh5 since it forces the unique winning move h4! And it's true, once White finds the first move, it's Black's only chance to give White a unique winning move that's not obvious.

Geir is correct that with the h5 pawn on f5, the result is a draw. Siegfried is correct that Rewan has reached the final position, but has given a retraction that produces a drawn position. As I imagine Siegfried did, I failed with every possible king capture and bishop capture before hitting on the correct pawn capture for White's last move.

(= 5+4 )

(Read Only)pid=17777
(6) Posted by Guus Rol [Saturday, Nov 28, 2020 17:01]

Thank you for posting that. I forgot all about that problem and now take vitamins to recover my memory. One day I will remember!
(Read Only)pid=20105
(7) Posted by Andrew Buchanan [Saturday, Nov 28, 2020 18:25]

Amiright that White has 3 winning retractions: Ph5xPg6, Pg5-g6 & Bg8xPh7, but only the first is derived from a legal position?
(Read Only)pid=20106
(8) Posted by Guus Rol [Saturday, Nov 28, 2020 19:32]

That looks about correct. I'm somewhat serious about my memory though it is returning now that I am tackling the position again. SF confirms what you suggest and its analysis appears correct.

I am somewhat sad about posts disappearing mysteriously from (haven't searched for this one) as it destroys compositions I can't even recall having made anymore. Not all my stuff is in my own topics, I often place them as comments in different threads. By the way, I started to occasionally scan the last months and add the best content into a document. Most not retro though.
(Read Only)pid=20107
(9) Posted by Andrew Buchanan [Sunday, Nov 29, 2020 04:46]

Hi Guus,
In in a browser on my PC, I discovered yesterday that for message chats, I can get the older messages to appear by dragging the slider on the right to the top, and then repeatedly dragging it down and up. It's quite time-consuming, but with this undocumented feature I can go all the way back to the beginning of our conversation.
(Read Only)pid=20108
(10) Posted by Guus Rol [Sunday, Nov 29, 2020 09:43]

Thx! I'll check that! Btw, most of my lost material is (was) in the forums. I commonly communicate in the public places; private exchanges are for special purposes and persons :-)
(Read Only)pid=20112

No more posts

MatPlus.Net Forum Misc An interesting retraction problem