﻿﻿ MatPlus.Net

Website founded by
Milan Velimirović
in 2006

5:45 UTC
 ISC 2021

Remember me

 CHESS SOLVINGTournamentsRating lists1-Oct-2021
 B P C F

MatPlus.Net Forum General Definiton of Republican Chess and its source needed

### Definiton of Republican Chess and its source needed

Can anybody provide the original definition (and its source) of Republican Chess?

Cf. http://sourceforge.net/tracker/index.php?func=detail&aid=2007222&group_id=200122&atid=972237 to find out why I'm asking.

Thanks
--
Thomas

Here you can find a wording :

http://christian.poisson.free.fr/problemesis/condra.html

you'll find in French :

"Il n'y a pas de Rois : si le camp qui vient de jouer peut placer le Roi adverse sur une case où il serait légalement mat, alors le camp adverse est mat."

and in English :

"There are no Kings : if the side which has played can put the opposite King on a square where it would be legally mate, then the opposite side is mate."

you'll find there also another fairy condition "Republican Chess II" :

French :

"Il n'y a pas de Rois : si le camp qui vient de jouer peut placer le Roi adverse sur une case où il serait légalement mat, alors il le fait. Le camp adverse peut alors lui-même placer l'autre Roi sur une case où il est maté."

English :

"There are no Kings : if the side which has played can put the opposite King on a square where it would be legally mate, then the opposite King is put on such a square. The opposite side can then put itself the other King on a square where it is mated."

This last sentence seems not enough : what then ? pat ? or the play goes on ?

@ Jacques : The second King is mate, of course.
@ Thomas : Maybe someone could find the original definition of Republican Chess in an old 'feenschach' issue. Nowadays, we use Christian Poisson's definition(s).

I see that Christian Poisson's definition is "if the side which has played can put the opposite King on a square where it would be legally mate, then the opposite King is put on such a square". In other words, it is mandatory - the side which has played must do it. In Popeye 4.37 it is optional. I have an original help-selfmate which is a very good example, but if I post it on the forum then it will be regarded as published.

Milan, if I post the problem here, will it be able to compete in the Mat Plus tourney?

(= 8+8 )

Juraj Lörinc
Klüver MT 1990-1993, 1st commendation
Mate in 1

1.Bxe5 mate

I think, later he said, that also would fit as a republican chess thing, which he didn't know then, but I doubt so. But maybe I remember wrongly.

Hopefully that helps in some way.

(PS: Yes, the reply is correct, he said Kriegspiel!)

Well, there are many more #1 in Republican chess in this position, e.g. 1.Bd6(+Kd8)#

It would qualify as a Kriegspiel #1, however.

Republican Chess was invented by the late J.C. (Joop) van Gool, who wrote an article about it in the march/april 1989 issue of Probleemblad, with some examples.

Literal definition from that article: "Republican Chess: both kings fail in the diagram position. Purpose of the game is to put the hostile King on the board in a legal mate position. Both sides are allowed to put the hostile King on an empty square on the board, but only after a move of the own side and only if the hostile King is immediately mate. The mate position has to be legal".

I vaguely remember there being some discrepancy about the interpretation of the definition, but browsing through subsequent issues could not find it anymore, just a couple of variations on RC with increasingly complex definitions (and consequences).

In hindsight I'd say the definition is somewhat shaky, especially the "mate postition must be legal" bit. Legal in what sense? As in "must be reachable from the initial game array"? Methinks there's not a chance of that happening with the other king missing in the mate position.

So it was an article in Probleemblad that I have seen sometimes in the past after all...

I have written more personal recollections on Popeye SourceForge page:
http://sourceforge.net/tracker/index.php?func=detail&aid=2007222&group_id=200122&atid=972237

Thanks to the many contributions here, on the sourceforge site (Joost de Heer, Juraj Lörinc) and "private" E-Mail (Hans Gruber, Bernd Schwarzkopf)!

Here is my synthesis:

1. Original definition (of Type 1)
==================================
The original definition of Republican Chess was published in Probleemblad III-IV/1989 (p.38-39) by Johan Christoffel van Gool.

The interesting points are:

- inserting the opposite king is *not* mandatory (so Juraj's problem http://jurajlorinc.tripod.com/chess/m99_104.htm#uloha2 is correct, even if White could, were he stupid enough, add the black king on h8 after the try 1.Qh7?); this is the main difference to Christian Poisson's definition, which states that mate is already reached if the king *can be inserted* [this difference is similar to that between selfmate (mate has to be forced) and reflexmate (mate has to be made possible)]

- the mate resulting from inserting the opposite king has to be "legal" (apart from a missing king); retro-analysis may be necessary to determine the legality, for instance in the following problem:

Gerard Smits

(= 3+3 )

#2
RepublicanChess Type1

1.Ba8-d5 ! zugzwang.
1...d7-d6
2.Rf8-f7[+bKe6] #
2.Rf8-f7[+bKa8] #
1...Bb8-a7
2.Rf8-h8[+bKa8] #
1...Bb8-d6
2.Rf8-f7[+bKe6] #
1...Bb8-c7
2.Rf8-a8[+bKb8] #

Note that after 1.-Ba7, 6 white moves allow the insertion of the black king for a mate, but only 2.Rh8 creates a legal mate, with Bg8-d5 as last move; bPd7 prevents 2.Bc6[+bKa8] with last move Be8-c6.

This obviously means that computer programs will probably never be able to fully check Republican Chess problems.

van Gool writes that a dualistic insertion square for the black king (as after 1.-d6) is "less serious" ("minder ernstig").

Probleemblad, IX-X/1990, S. 113-114 and Probleemblad, IX-X/1991, S. 130 contain definitions of variants.

2. Definition of Type 2
=======================
I haven't found the original definition of Type 2.

But the main difference to Type 1 is that RepublicanChess is reactivated (after having been temporarily suspended for determining whether the inserted king is mate) after the insertion of a king; the side that has just been "legally mated" can therefore attempt to defend against that "mate" with a Republican move.

[In Type 1, on the other hand, play is "non-Republican" (should I write "democratic"? :-) ) after the insertion of the king, which means that play ends at the move inserting the king.]

This definition of Type 2 is more exact from Christian's, which is quite vague about when the second king can be inserted.

The difference can be illustrated in the following positions:

(= 3+2 )

RepublicanChess Type1

a) h#1

1.Rg2-h2 Be6-d5[+bKh1] #

b) #1

1.Ra1-h1[+bKh3] # !

No surprise so far. Let's swap colors for Type 2:

(= 2+3 )

RepublicanChess Type2
h#1

1.Ra1-h1[+wKh3] + Kh3-h4[+bKh2] #

Now 1.Rh1 still allows Kh3 to be inserted, but this no longer is a mate, since White can defend with the Republican move Kh4[+bKh2]#.

Support by Popeye
=================
Popeye has supported Type 2 (under the name "RepublicanChess", without indication of a type) for quite some time, obviously without the retro-analysis.

Starting at the next release, Popeye will also support Type 1. Type 2 is the default, if no type is indicated in the problem description.

Popeye does not currently support the variations published later in Probleemblad, nor RepublicanChess according to Poisson's definition.

Support by other programs
=========================

Maybe it's just me, but where is the legalisation for a bKh1 in the h#1 in the previous to last example (the first of the color-swap twin). To me the position looks illegal then.

Same for the double check below, the bK or wK doesn't look insertable on h3.

Years ago, I had the idea to make a Republican Chess Proofgame.
But I aborted, when I wasn't certain about whether dropping for mates/stalemates was mandatory.
(The plan was to base the Proofgame on avoidance of mates and stalemates.)

So, I hope Christian Poisson's interpretation is correct (mate drops mandatory),but it raises questions...

How else can I make a h=n w/ this condition?

And, if stalemate drops are mandatory too, what if I have both cases (a checkmate drop, and a stalemate drop)?
Don't try calling this a new aim (say "chalemate" and award 0.75 points) because the problem only multiplies!

Maybe one should take precedence over the other (give priority to stalemate, and create tries w/ this?).

As for somebody's question about "legal justification" -- the answer is obvious: you need a Consequent form of Republican Chess, if your goal is to avoid illegal mating positions. Otherwise, it's a fairy condition -- I wouldn't try to make "sense" of it in terms of the FIDE game.

I'd be interested to hear if somebody has rules that make these questions clear.
Because I really like this condition, but have avoided it because of such uncertainties.

I suppose I could instead exploit the uncertainties, and prompt discussion through problems...
Until somebody makes a TT for such problems, I'll do my best to avoid fairy conditions with uncertainties.

But, sometimes I just can't help myself...

QUOTE
Maybe it's just me, but where is the legalisation for a bKh1 in the h#1 in the previous to last example (the first of the color-swap twin). To me the position looks illegal then.

Same for the double check below, the bK or wK doesn't look insertable on h3.

Agreed on both accounts.

I posted the output of Popeye's current (unreleased) state, which doesn't check for legality at all. I assume that other programs that support Republican Chess behave the same way.

I have somewhere used the word "nightmare" when speaking about coding the Republican Chess with legality checks. That is just like with any fairy condition taking into account legality of position: Circe Alsacien, Sentinelles Angevines etc.

Virtually impossible to code as it would mean that the program would have to have ability to solve retro problems and the routine would have to be run after every move done.

I think the legality check is mainly to avoid impossible checkmates, such as a triple-check or impossible double-check (although I quite like seeing this kind of thing!). This should be fairly easy to test in a program. For example, if White has just checkmated Black then the mate is allowed if White could play a move (which may be an uncapture), after which Black is no longer in check.