Website founded by
Milan Velimirović
in 2006

2:09 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 General The Official Dummy Pawn Thread
You can only view this page!
(1) Posted by Rewan Demontay (Real Name: James Malcom) [Friday, May 8, 2020 17:42]

The Official Dummy Pawn Thread

This is a thread for all to use to post problems involving the famous dummy pawns. All types are welcome-directmates, helpmates, selfmates, reflexmates, studies, retros, PGs, fairies, etc. Have fun composing!
(Read Only)pid=19013
(2) Posted by Andrew Buchanan [Friday, May 8, 2020 19:02]; edited by Andrew Buchanan [20-05-14]

Thanks Rewan, here is a very simple proof game. The dummy pawn theorist & part-time world champion, Wilhelm Steinitz, might have annoyed Howard Staunton with this kind of composition in the 1860s :-)

(= 13+14 )
(13+14) PG in 8.0 (-:
(Read Only)pid=19014
(3) Posted by Kostas Prentos [Saturday, May 9, 2020 12:41]

Dummy Schnoebelen on f8! A promotion to any officer would interfere with the bK's path.
(Read Only)pid=19016
(4) Posted by Olaf Jenkner [Saturday, May 9, 2020 19:36]

Some people claim that this is not Schnoebelen because the pawn wasn't promoted. 5. exf8=P is only a pawn move but not a promotion. If you consider each pawn move onto the 8th rank as a promotion then we have the Schnoebelen.
(Read Only)pid=19017
(5) Posted by ichai [Sunday, May 10, 2020 03:33]
Some examples (mine in post (8))
(Read Only)pid=19019
(6) Posted by Rewan Demontay (Real Name: James Malcom) [Sunday, May 10, 2020 03:40]; edited by Rewan Demontay (Real Name: James Malcom) [20-05-10]

Well of course that thread exists ichai, but that thread is “clogged” IMO, hence why I made a new ome

And your problem is in the database, along with all others from that thread. See here:
(Read Only)pid=19020
(7) Posted by Peter Wong [Sunday, May 10, 2020 04:46]; edited by Peter Wong [20-05-10]

(= 2+2 )

Series-helpstalemate in 5 (-:

Just found this on the PDB:
(Read Only)pid=19022
(8) Posted by Andrew Buchanan [Sunday, May 10, 2020 06:29]

Hi ichai. I like your problem. What author name do you want to have associated with it in PDB? Feel free to send me a message in matplus, if you are shy.

There's also a version posted in the stackexchange dummy pawn saga (soon to be made into a Netflix mini-series).

The stackexchange version replaces bBa2 for bP for minimality. Obviously, your original twinning version is more interesting as a composition, however it is not the minimal example showing Dummy Pawn in a study. The honorable term I use occasionally for a composition which expands beyond the narrow remit of a task is "director's cut". Both versions should be in PDB, I think.
(Read Only)pid=19023
(9) Posted by Peter Wong [Wednesday, May 13, 2020 03:17]

Here's another 6-unit study!

(= 2+4 )


1.h8=Q? Sf4! (threats: 2…Bf3 and 2…g2+ 3.Kh2 g1=Q) 2.Qa8 g2+ 3.Qxg2+ Sxg2, or 2.Qh3+ g2+ 3.Kh2 Sxh3.
(Not 1…Bf3? 2.Qa8! Sh4+ 3.Qxf3+ Sxf3 =, or 2…Kf2 3.Qa2+ =.)

1.h8=P! Ke2/Ke1 2.Kxg2 ~ 3.Kxg3 =

Tablebases links:
(Read Only)pid=19059
(10) Posted by Rewan Demontay (Real Name: James Malcom) [Wednesday, May 13, 2020 04:58]

Very nice Peter! This is surely minimal in terms of piece economy! This is also more valuable because it contains content beyond the promotion.
(Read Only)pid=19060
(11) Posted by Peter Wong [Wednesday, May 13, 2020 08:57]

Thanks, Rewan! Yeah, I suspect this is the most economical... With this matrix, you generally can't replace the BB with a BS because then W can draw by sacrificing the Q for the BP to leave a S+S ending. I also tried to replace the BB with a BP, but if this BP is well-placed to beat the WQ (, then the intention P=P fails because Black can release stalemate and still win with the remaining P.

That's incredible research by you and Andrew in the StackExchange thread!
(Read Only)pid=19062
(12) Posted by Andrew Buchanan [Wednesday, May 13, 2020 09:31]; edited by Andrew Buchanan [20-05-14]

Peter that's really beautiful. Here's my effort:

(= 2+4 )
(2+4) Draw

1. g8=Q? g1=Q!
2. Qxg1 Rf8+! 3. Qg8 Rxg8+ 4. Kxg8 Kg6! & #13
2. Qxf7 Qa1(d4)+ 3. Qf6+ Qxf6+! (3. Qg7+ Qxg7#) 4. Kg8 Qg7#
3. Kg8 Qa8(g1)+ 4. Qf8+ Qxf8+! & #13
1. g8=! (here the "=" denotes the draw rather than promotion! :-)

This is the only arrangement of these pieces which will work, I think. If bRf7 is further to the left e.g e7, then there is 1. g8=Q! g1=Q 2. Qe6+! If bPg2 is further to the left, it is useless, until it reaches a2 when it becomes too powerful.

The key can't be shown on Syzygy, because here the dummy pawn does affect the play.
(Read Only)pid=19063
(13) Posted by Peter Wong [Wednesday, May 13, 2020 15:13]

Hey Andrew, That's such an original scheme and clever to use the dummy P to help with the stalemate! There's a big contrast between own versions – I aimed to prevent WQ checks but you were unfazed by them and allowed 5 such checks, all defeated by Black in different ways. And you don't even consider them as the main variations, as the captures of Black's two major pieces are indeed great.
(Read Only)pid=19065
(14) Posted by Rewan Demontay (Real Name: James Malcom) [Wednesday, May 13, 2020 16:35]; edited by Rewan Demontay (Real Name: James Malcom) [20-05-13]

Well the idea of using the idea of using the dummy as a self-block is very old of course-Andrew has found the minimum form for it. The earliest use of it in a study that I know of is from 1879-P1372578

Henry John Clinton Andrews, Chess Player's Chronicle 1879
White plays and draws
(= 5+6 )

The earliest occurrence that I know of though is from 1873-P1340366

James White, Recreationist 01/1873
s#5 via triple check
(= 11+8 )

(Read Only)pid=19066
(15) Posted by Andrew Buchanan [Thursday, May 14, 2020 07:58]; edited by Andrew Buchanan [20-05-14]

Hi Peter, thanks for your kind words. I think your composition contains more subtle chess, but it is curious in my composition that wQ moving from a square so near to bK is so weak. An unsupported queen needs to operate from a distance for best effect, except in "desperado" mode (i.e. to force its own capture to result in immediate stalemate) which turns out not to be possible here. The location of sR on f7 severely cramps White, forcing him to lose the initiative, which is really fatal in this kind of position. This argument has nothing to do with dummy pawn, and could be a factor in conventional end-game analysis. However my main creative goal was to make something that syzygy tablebase can't evaluate correctly! :-)

As James rightly observes, the idea of using dummy pawn as a blocker has been around since people first started exploring the implications of this rule. There is even (I now see) another (less economical) anonymous 6-unit study, but the play after 1.b8=Q? is archi-simple, see here:

(= 2+4 )
(2+4) Draw
(Read Only)pid=19068
(16) Posted by Siegfried Hornecker [Thursday, May 14, 2020 16:33]

The source given in the PDB (" 2011") is a bit wrong, however. Apparently the example was added by the user KnightMove in 2006.
(Read Only)pid=19069
(17) Posted by Peter Wong [Wednesday, May 27, 2020 03:47]

Here's a new blog on the subject on my site:

It quotes two of the studies above and has two originals, one diagrammed below.

(= 3+5 )

(Read Only)pid=19133

No more posts

MatPlus.Net Forum General The Official Dummy Pawn Thread