# No.1346 (HG)

 No.1346  Hubert Gockel  (Germany) Original Fairy problems JF-2018/II: July – December’2018

Definitions: (click to show/hide)

 No.1346 Hubert Gockel Germany original – 22.11.2018 Solutions: (click to show/hide) white Ka4 Qa5 black Pa6 Kh1 neutral Se4 Pb5f2 h#2             2 solutions         (2+2+3) SuperGuards Neutral pieces e4, b5, f2 1.nSe4-c3 Qa5-b6 2.nPf2-f1=nR Qb6-g1 # 1.nSe4-c5 Qa5-d2 2.nPf2-f1=nB Qd2-g2 # { (C+ by Popeye 4.79)} The curiosity is that only neutral Pawns (in contrast to S,B,R,Q) can capture and check in Superguards. Proof of concept: replace nPb5 with a nBb5 and you get 10 solutions! (Author)

### 15 Responses to No.1346 (HG)

1. Jacques Rotenberg says:

It seems that something goes wrong here.
After 1.nSc5, Qd2 should not be legal because the neutral knight threatens the king and nothing « superguards » it

• Georgy Evseev says:

It is unfortunately current logic of neutral pieces. Today they are considered as “both white and black”. So, black part attacks white king, while white part superguards it. Similar logic is used in Madrasi, Back-to-Back, etc.

I do not like it, but have to live with it.

2. Jacques Rotenberg says:

at least, such a strange way of understanding things should be written with the stipulation.

3. Jacques Rotenberg says:

by the way, now, I understand the comment of the author : among the neutral pieces (and according to this way to associate neutral pieces and superguards) only the pawn can give check (or take any piece)

4. seetharaman says:

I never intended it to be like this. I omitted to define superguards clearly for Neutral pieces. This is a grave omission on my part I realised later.

I assumed that neutral pieces also can capture as each side has the option to consider each neutral piece as either white or black. So I assumed they do not observe.

Unfortunately both the programmers in Popeye and Winchloe independently thought differently.

5. seetharaman says:

Considering Neutral pieces as “half white-half black” was not the intention of its inventor T.R.Dawson and neither it is logical. If that is taken as definition, then no one can capture neutral pieces, as capturing one’s own pieces is not allowed without added conditions. He said while introducing Neutral pieces, “either side can consider them as black or white”.

6. Geoff Foster says:

A neutral pawn moves up the board when played by White and down the board when played by Black. The “white” part of a neutral pawn on b5 guards a white piece on a6 or c6, while the “black” part attacks a white piece on a4 or c4. So if the nPb5 is considered to guard the wK, then the guard is by the “black” part of a “both white and black” unit. I can’t see how the “black part attacks white king, while white part superguards it”.

• Kjell Widlert says:

The “parts” referred to are parts of the nS, not parts of the nP!

7. Nikola Predrag says:

-“either side can consider them as black or white”-

So, in a helpmate, both White & Black will consider a Neutral piece in a way that helps fulfilling the stipulation?!

8. Frankly speaking, I am quite surprised that the question of the neutral pieces mechanics is an issue.

It is already quite well established how they work together with Madrasi, Back-to-Back, Face-to-Face, Annan… and other conditions. There are at least 143 problems using these conditions in combination with neutral pieces in big WinChloe database, by 57 authors (! – not just a very small group of prolific advertisers).

I also see no wonder in fact that both WinChloe and Popeye programmers have implemented these interactions more-or-less in same way (I leave some percent for peculiarities), precisely because the mechanics is standardized. By the way, in the ancient times, Popeye was not programmed to work with this combination, Madrasi + neutrals resulted in error message.

It would be rather confusing to have the interaction between neutral pieces and fairy conditions defined differently. Of course, anyone can do it in it own way, but then the confusion would be probably growing.

9. Kjell Widlert says:

This is a very odd effect, that the same piece which would like to capture the wK also superguards it to make it immune to capture. As Georgy points out, this is really a consequence of the interpretation that neutral pieces are white and black at the same time – and as Juraj points out, this convention is well established in several fairy forms.
With the other possible convention (the side to move may decide for each neutral whether to treat it as white or black at this particular moment) other effects would arise: for example, two neutrals (nRa1 + nRb1, say) would not paralyse each other in Madrasi: the side to move could choose to treat both pieces as his own and move one of them. This is also quite odd.

I really prefer the established convention of having the neutrals white and black simultaneously.

10. Georgy Evseev says:

Historically, at least as I believe, this kind of apporoach was proposed by Petko Petkov in his article concerning interaction of neutral pieces with Madrasi condition. Sorry, I do not remember the year and source (in the nineties, I think).
It was then soon programmed in Popeye according to that interpretation.
While there are some intersting effects/motives which become possible with such approach, other equally interesting possibilities specific for another approach are now probably lost forever.

11. Thomas Maeder says:

> With the other possible convention (the side to move may decide for each neutral whether to treat it as white or black at this particular moment)

This is not “other” (i.e. “different”).

It’s Black’s move when we are finding out if the black king is super-guarded.

12. Geoff Foster says:

Apologies for my mistake.

Another unusual possibility with neutral units that are both white and black is CnPP (Complete neutral Pawn Promotion): Neutral pawns always promote whenever they reach either promotion rank (first or eighth), regardless of how they come to arrive there. This was actually the case in some early versions of Popeye (e.g. version 4.65), but the “bug” was later corrected. CnPP was used in problem no.4 in Fairings 57.