#2 FBR enhancements for "Pocket Mutation Chess"

[Trevor Davis]

http://www.chessvariants.com/large.dir/pocketmutation.html

Three main changes to FBR to allow specification of "Pocket Mutation Chess".

1. Expand range operator to allow specifying a players "holding" area, at least with respect with the pieces FBR rule in order to restrict the number of pieces that can be in each players holding area to one.
2. Like in previous comment add a attribute to holding FBR rule to restrict holding area only to pieces that have "warped" in (and not pieces that have been captured).
3. Add new attribute to promotion FBR rule to allow pieces to promote immediately after warping out of the specified range.

Example usage:

:Pocket-Mutation

8x8 -- 18

rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - - 0 1

a1=R b1=N c1=B d1=Q e1=K a2=P

royal holding=pocket warp pieces=({^},"holding",1) promotion*=[1-7],warp promotion+=[8],warp drop={^},[1-7]

K
P=cefF+mfW+[2]mfW02 +P==BN
B *B==BN +B==R:NR::SB:
N *N==BN +N==R:NR::SB:
R *R==R:NR::SB: +R==C:SR:
:NR:|Nightrider|==N0 *:NR:==R:NR::SB: +:NR:==C:SR:
:SB:|SuperBishop|==B+W *:SB:==R:NR::SB: +:SB:==C:SR:
C|Cardinal|==B+N *C==C:SR: +C==QH:CR::SC:
:SR:|SuperRook|==R+F *:SR:==C:SR: +:SR:==QH:CR::SC:
Q *Q==QH:CR::SC: +Q==:CR::SH::XC:
H|Chancellor|=R+N *H==QH:CR::SC: +H==:CR::SH:XC:
:CR:|CardinalRider|=B+N0 *:CR:==QH:CR::SC: +:CR:==:CR::SH::XC:
:SC:|SuperCardinal|=B+N+W *:SC:==QH:CR::SC: +:SC:==:CR::SH::XC:
:HR:|ChancellorRider=R+N0 *:HR:==:CR::SH::XC: +:HR:==A:XH:
:SH:|SuperChancellor|=R+N+F *:SH:==:CR::SH::XC: +:SH:==A:XH:
:XC:|SuperCardinalRider|=B+N0+W *:XC:==:CR::SH::XC: +:XC:==A:XH:
A|Amazon|=Q+N *A==A:XH: +A==:AR:
:XH:|SuperChancellorRider|=R+N0+F *:XH:==A:XH: +:XH:==:AR:
:AR:|AmazonRider|=Q+N0

p.s. Technically the "fifty-moves" rule for pocket mutation chess is fifty moves without capture or promotion instead of fifty moves without capture or pawn movement. For this purpose "mutation" within a value class does not count as a "promotion" (i.e. '*' promotion doesn't but '+' promotion does).

[Gregor Cramer]

A few notes:

1) Due to specification only pieces from eight rank will promote (immediately), so FBR has to be:

royal holding=pocket warp pieces=({^},"holding",1)
promotion*=[1-7] promotion+=[8],warp drop={^},[1-7]

Do you agree, or have I overseen something?

2) Do you know recorded games for this variant?

3) "fifty-moves" rule : the "move clock count" should not belong to the definition of a chess variant, because such rules normally will be defined by organizations (e.g. FIDE), and may change from time to time. Furthermore any application may have his own rules how to count/reset the move clock for a specific variant. So this problem with move clock counter belongs to an application, but not to C/CIF specification.

[Trevor Davis]

1) Well pieces only "mutate" (represented by promotion rule '*') if they warp from ranks [1-7] to the pocket. So why shouldn't it be promotion*=[1-7],warp assuming the "warp" attribute is defined as promotion upon warping away from the specified range)? Without the "warp" attribute i.e. just promotion*=[1-7] wouldn't the promotion rule imply that a piece "mutates" if it moves to ranks [1-7] (either normally or via a drop into that square). Also pieces only "promote" (represented by promotion rule '+') if they warp from the last rank, in particular pawns do not automatically promote upon merely moving to the last rank, they only promote upon warping from the last rank. So the other promotion rule should be promotion+=[8],warp.

I realize now that C/CIF probably also needs a new promotion attribute or notation to cleanly allow a single piece to "mutate" or "promote" an arbitrary number of times which can happen in Pocket Mutation Chess. Maybe something like "replace" (piece is replaced/captured by a new 'unpromoted' piece) or "multiple" (piece may promote multiple times, promoting twice doesn't cancel initial promotion). So the Pocket Mutation Chess promotion rules should maybe be

promotion*=[1-7],warp,replace promotion+=[8],warp,replace

2) There are dozens of "Game Courier" logs of Pocket Mutation games:

http://play.chessvariants.com/pbmlogs/index.php?age=0&stat=any&gamewcp=Pocket+Mutation+Chess

In their log notation the white pocket is represented by "p1", the black pocket by "p8" and "mutation/promotion" is represented by that player doing an immediate second move on their turn "capturing" their piece in the pocket with its new "mutated/promoted" replacement.

3) Sounds reasonable

[Trevor Davis]

For example here are two implicitly defined promotion chain rules for Pocket-Mutation I am not sure is respected by the default C/CIF promotion notation.

**B==BN i.e. the following promotion chain is possible B -> B -> N
**+**+P==R:NR::SB: i.e. the following promotion chain is possible P -> N -> B -> N -> R -> :NR: -> :SB:

[Trevor Davis]

I'm not entirely sure how the "Control position" field should be set but I guess all the possible piece upgrades should also have been included assigning each of them an arbitrary point within the board prepended with an '@'?

Likely additions to be added the standard chess "Control position":

@c1=:NR: @c2=:SB: @c3=C @c4=:SR: @c5=H @c6=:CR: @c7=:HR: @c7=:SH: @c8=:XC: @d1=A @d2=:XH: @d3=:AR:

[Gregor Cramer]

promotion*=[1-7],warp,replace promotion+=[8],warp,replace

This "replace" is not required. See this specification:

promotion#=[1-8],move
Q #Q ##Q

The following happens:
1. Moving a queen (Q) is promoting to #Q
2. Moving promoted queen (#Q) is promoting to Q

Here we have the promotion chain Q->#Q->Q->#Q->Q->...
And here you can see the difference in syntax, why default definition is "#Q=Q" (defining move capability for #Q), but default for ##Q is "##Q==Q" (defining promotion from #Q to Q).

Now see this specification:

promotion#=[1-8],move
P #P==N N #N==B B #B==R R #R==Q Q

1. Moving a pawn (P) is promoting to N
2. Moving a knight (N) is promoting to B (any knight, even if promoted from pawn)
3. Moving a bishop (B) is promoting to R (any rook, even if promoted from bishop)
4. and so on

Here we have the promotion chain P->N->B->R->Q.

So your original MBN for Pocket-Mutation-Chess works.

I'm not entirely sure how the "Control position" field should be set but I guess all the
possible piece upgrades should also have been included assigning each of them an
arbitrary point within the board prepended with an '@'?

Likely additions to be added the standard chess "Control position":

@c1=:NR: @c2=:SB: @c3=C @c4=:SR: @c5=H @c6=:CR: @c7=:HR: @c7=:SH:
@c8=:XC: @d1=A @d2=:XH: @d3=:AR:

I see, the control position is not yet documented. The character '@' is denoting the holding, so your right. This control position helps to identify piece letters, in most variants this works perfectly, but not in all. For example in Pocket-Mutation additonal information is required, in this case the "standard" order used at chessvariants.com seems to be appropriate.

http://play.chessvariants.com/pbmlogs/index.php?age=0&stat=any&gamewcp=Pocket+Mutation+Chess

Thanks for the link. I see that these games are recorded with a single-letter set. I prefer to use established letter sets, so I replaced your letter set with the letter set of play.chessvariants.com. Example: ':NR:' now is 'I'.

Please check the specification of Pocket-Mutation at C/CIF - Chess Variants.

[Trevor Davis]

1) Slight tweaks to promotion rules required: should be +Q==FWX and +M==FWX
2) I must have miscounted number of pieces, board should be 8x8 -- 19 since there are 19 pieces
3) I don't know if it matters but control position has two different pieces at @a7 i.e. @a7=V @a7=F

[Gregor Cramer]

1) Slight tweaks to promotion rules required: should be +Q==FWX and +M==FWX
2) I must have miscounted number of pieces, board should be 8x8 -- 19 since there
are 19 pieces
3) I don't know if it matters but control position has two different pieces at @a7 i.e
@a7=V @a7=F

I have done the appropriate changes, see C/CIF - Chess Variants.

One open issue:

If a player's pocket is empty, the player may remove any of his
pieces (except his King) from the board and put it in his pocket
as a move. White may not use the pocket on the first move.

What does the last sentence mean? It's clear that White cannot drop in first move, but Black also cannot drop in first move. So does this sentence mean: White cannot warp in first move?

[Trevor Davis]

Sorry I missed that. That is the logical interpretation, that white can't
warp a piece to the pocket in their first move, I guess the "warp" FBR
would need an additional attribute specifying first turn players can warp
ie. warp=(1,0) or less generally a pocketmutation attribute i.e.
warp=pocketmutation.

[Trevor Davis]

Sorry if you get this twice. I replied to the e-mail but it hasn't shown up on Sourceforge (like it would in a Github issue tracker).

Sorry I missed that rule. That is the only logical interpretation, that white can't warp a piece to the pocket in their first move, I guess maybe the "warp" FBR would need an additional attribute specifying first turn players can warp ie. warp=(1,0) or less generally a pocketmutation attribute i.e. warp=pocketmutation.

[Trevor Davis]

text deleted (accidental cross-post)

[Gregor Cramer]

Sorry if you get this twice. I replied to the e-mail but it hasn't shown up on Sourceforge
(like it would in a Github issue tracker).

Often it takes a while on Sourceforge.

That is the only logical interpretation

I was unsure because I cannot see the indentation behind this rule. I added "warp=start(1)". Meaning: Warping is not allowed in first Nth half moves (here N=1).

I am closing this request, it is fulfilled. I have set the status of Pocket-Mutation to "verified", because a specialist has done the specification, and proof-reading.

Many thanks for your voluntary work for C/CIF.

[Gregor Cramer]

Additional note:

The specification now is using a macro for the Nightrider, this makes it possible to use the Nightrider for the definition of I|Nightrider|, H|SuperChancellorRider| and Z|AmazonRider|,
see Macros.

[Trevor Davis]

No need to re-open this issue. It would seem that the intention of that Pocket Mutation Chess rule is since white usually has an a first-move advantage in chess then letting black use the pocket sooner should reduce white's advantage making the game more evenly balanced.

https://en.wikipedia.org/wiki/First-move_advantage_in_chess