Website founded by
Milan Velimirović
in 2006

6:41 UTC
ISC 2024



Remember me

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

Rating lists


MatPlus.Net Forum General Direct Retro Analysis (a retro variant)
You can only view this page!
(1) Posted by Kevin Begley [Friday, Apr 1, 2011 23:00]

Direct Retro Analysis (a retro variant)

April 1st is a poor day for an honest post... but, why put it off until tomorrow...

In the latest issue of Best Problems (, #58, page 17), Enzo Minerva describes a new chess variant, which he calls Direct Retro Analysis (a poor name, imho), where the goal of each player is to stay on a unique Proofgame path, or lose. (if a unique game ends in #/=, the result is a draw).

I'm not certain whether he is the first to publish rules for this game, but it's certainly not a new chess variant.
I was first introduced to this variant about 10 years ago, by Kostas Prentos.
If memory serves, Kostas had invented it (he may have mentioned a co-inventor), and was calling it "Proofgame Chess."

There are some alternative ideas possible.
I'm not entirely certain, but Proofgame Chess may have had alternate rules to cope with unique games ending in mate (perhaps checkmate was still checkmate, providing it was unique).
It also required that a player claim victory, subsequent to the opponent going wrong, by demonstrating an alternative path (in the same # of moves, or fewer w/ same player on the move).
I'm not certain whether it was theoretically possible to "slide" onto a separate unique path (where the position holds a unique PG, but in fewer moves than the actual game).

Despite having lost every game to Kostas, I can attest to this being a thrilling game (and I learned by playing it-- especially to avoid falling into check!).
The longer a game goes, the more time you require -- so, it might be ideal for correspondence.

I remember encouraging Kostas to publish something in this variant... but, I'm not sure whether he did.
(Read Only)pid=6829
(2) Posted by Kostas Prentos [Saturday, Apr 2, 2011 18:34]

Together with a Greek friend George Tsolakos, we used to play this hybrid back in 1987 or 1988. We never explored it in depth, and there is a possibility that it can be fully analyzed. Back then, it was very difficult to check the soundness of the proof games. Most of the times, the games we played did not produce proof games of any aesthetic value and the whole idea was soon forgotten.

Many years later, I showed this game to Kevin, who remembers the details much better than me. Some years after, I remember discussing this game with Andrey Frolkin, with whom I composed a great deal of joint problems. He told me he already knew this game and he had played it in the 80's with his retro colleagues. He wrote something about it in his book with Gerd Wilts about proofgames. As the book is out of print and I was also out of the composing business at the time it was published, I am not able to provide any details. Perhaps, Andrey or Gerd can give more information.
(Read Only)pid=6830
(3) Posted by Geoff Foster [Sunday, Apr 3, 2011 00:11]

Here is Chapter XI of "Shortest Proof Games", by Gerd Wilts and Andrey Frolkin, 1991.

SPG-Chess: Will It Become Popular?

Andrey Frolkin and Dmitry Pronkin were the first (and probably still the only ones) to play a match of a new kind of chess they invented: SPG-chess. The rules of it are easily defined: there has to exist just one single-line SPG [with account for the turn to move; otherwise tempo-losing SPGs like 1.a3 Sf6 2.a4 etc would be banned] leading to the position formed after each move of the players; the player who cannot prolong a unique SPG is the loser. The match between the two inventors was stopped abruptly after the 4th game (the score was 2:2), when the partners had realized how easy it was for white to win. All that was needed was to prepare a sacrificing check leaving black helpless (a couple of examples: 1.e4 Sc6 2.Qh5 Sa5 3.Sf3 c6 4.Qxf7+, since the position obtained after the forced reply 4…Kxf7 could have been alternatively arrived at by 1.e4 Sc6 2.Qh5 Sa5 3.Qxf7+ Kxf7 4.Sf3 c6; or 1.c3 c5 2.Qa4 Qc7 3.Kd1 Qxh2 4.Qxd7+). The two published a SPG-chess problem in ‘Die Schwalbe’ in 1985. The stipulation was as follows: 'SPG-chess: Position after the 31st move of black. White to play and win'.

Dmitry Pronkin & Andrey Frolkin, Die Schwalbe, 1985
(= 14+9 )

White has several moves with their Bd8 to continue a single-line SPG, but only 32.Bc7+ lead to their victory, since after 32.Ba5 (Bb6 etc) black replied with 32...Bxf4+, capturing the fifth promoted bishop and winning. Surprisingly, this reasoning was omitted by the publication of the solution; the problem was considered to be a presentation of the capture of just 4 bishops by the judge who evidently didn't think too much of it.

It really does look like SPG-chess as a game will hardly ever reach even the not too high level of popularity of CIRCE or three-dimensional chess. But at the same time the use of SPG-chess (as well as its anti-form, anti-SPG-chess: the partner who is the first not to be able to continue an SPG sequence is the winner; this variety has so far not been looked into at all by its inventors) in composition may produce some very interesting results and lead to the creation of some very difficult problems (maybe even SPG-chess 'endgame studies', or, rather, 'debut studies'!).
(Read Only)pid=6831
(4) Posted by Sarah Hornecker [Sunday, Apr 3, 2011 06:51]

While this might or might not be appropriate here, in regards to popular chess variants, we have great fun playing KöKo in our chess club (that is, me and another crazy guy :-) ) because Circe ends too often in a draw with a typical circe-circle of:
wBg5 bQd8 and 1.-Qxg5 2.Bxg5 Qxg5 3.Bxg5 ad infinitum
wQd1 bQd7 and 1.Qxd7+ Qxd7 2.Qxd7+ Qxd7 ad infinitum
(Read Only)pid=6832
(5) Posted by Hauke Reddmann [Monday, Apr 4, 2011 18:23]

Just for the record, among the Hamburg youth
"Hoppelpoppel chess" was a bit en vogue. B/S move like B/S
but capture like S/B. The Hamburg girls master
(from my club) always beat me to pulp, although
my "normal" ELO is roughly double of hers :-)

(Read Only)pid=6833

No more posts

MatPlus.Net Forum General Direct Retro Analysis (a retro variant)