Knight Relay Chess

by George Jellis
From Variant Chess, Volume 3, Issue 28, Summer 1998, pages 156-157.
Back to: Home Page

The original form of Relay Chess was invented by the US problemist Mannis Charosh and first published in the last full year of Fairy Chess Review , April 1957. It permits the powers of any piece to be relayed to any other piece it guards. This proved to be a difficult stipulation to get sound problems from, since the number of possibilties were so great.

From the start pawns were not allowed to use a relayed move to reach the first or eighth ranks, since “without this rule there would be many cooks by long-distance promotion”.

For playing purposes (in 1972 according to ECV ) Charosh developed the simplified form of Knight-Relay Chess in which only the knights relayed their powers. This has proved a popular variant ever since, though through further playing experience, in NOST and AISE, the rules have been further refined. The rules described here are those currently played in AISE.

The refinements are that
  1. The knights can no longer capture directly, but only by relaying their powers
  2. The knights cannot be captured, and can thus act as blocks to opposing pieces
  3. The knights do not relay knight moves,
  4. The kings cannot relay knight moves, this is mainly to prevent kings escaping mate too readily,
  5. There is no e.p. capture.

Here is a short game ending in mate to illustrate some of the rules. A relayed move is shown by ®.

George Jelliss v Aldo Andreotti
AISE Grand Prix 1996

1.Nf3 Nc6 2.h2-g4® f6 3.d2-e4® f6×g4® 4.Nc3 Nf6 5.Bg5 (this B can be taken by h7×g5® but this would permit R×Rh8) 5...e6 (missing the double attack on h7) 6.R×h7 R×h7 7.Bg5×h7® g4-h2® 8.Bh7-g6† Ke7 9.e4-d6®† c7×d6 10.Cf3-e5‡ (The knight's check is relayed via Bg5. Black cannot escape by Ke7×Bg5® because Kings do not relay.)

MATE!

The alternative 8...Qd8-f7® is interesting—: after 9.B×Q, K×B 10.e4-g5®† Kg8 11.Ne4 h1=Q is prevented since 12.f2×h1®, and 11...Ng4 does not threaten 12...h2×f1®=Q† since pawns are only allowed to promote by a normal pawn move.

Here is a well-played game between seasoned campaigners:

Alessandro Castelli v Paul Yearout
AISE International Knight-Relay Championship (1993).

Italic notes by Paul Yearout

1.h3 e5 2.h3-f4® c6 3.f4×e5 c6×e5® ( White spends three tempi to exchange an outside pawn for a centre pawn, leaving Black ahead in development. 3...d7×e5® could have been better for Black, opening up his second bishop. ) 4.d4 Nh6 5.d4×e5 f7×e5® 6.e4 d7-f6® 7.Q×d8† K×d8 8.c3 Ng4 9.Bd2 Be6 10.Bd2-b3® B×b3 11.a2×b3 f6×e4® 12.c3×e4® Na6 13.Nd2 b6 14.f3 Bb4 15.Bf1-g3® e5×f3® 16.g2×f3 Bb4-c2® ( Threatening Na6-b4 †) 17.Rc1 B×b3 18.Ne2 Be6 19.f3-d4® ( White's two centralised and knighted pawns are very powerful and adequate compensation for a lost pawn. )

19...Rc8 20.R×c8† B×c8 21.Be5 Nf6 22.Bb8 Bb7 ( This begins a plan to trap White's bishop. ) 23.d5 Nd7 24.Rh5 b6×d5® 25.e4×d5 Nc7 26.B×a7 Nb6 ( Ah, ha! Did it!. ) 27.Nd4 Kd7 28.Nc6 Ba6 ( Perhaps the price is to high. Black's bishop and both knights are tied up in restraining White's bishop. ) 29.Ne4 g6 30. Rg5 Kd6 31.Nd8 ( Good move. I thought my rook would get his bishop. ) 31...Rf8 ( I was thinking of the pawn exchange, but hoping also he might think of ‘gaining' the rook because of my supposed carelessness. But I gain a pawn. )

32.Rg5-e6®† ( Rg5×h7® would likely have led to a draw. ) 32...K×d5 33.Re6×f8® Ba6-b4®† ( Black's prospects have bright- ened considerably. White's inactive bishop will trouble him, though he uses his resources to give an admirable defence against the pair of Black pawns. ) 34.Ke2 B×f8 35.b3 Kc6 36.Nb7 g5 37.Ne4-c5 ( Aiming to get his pawn and king in front of Black's two pawns. ) 37...h5 38.b3-d2® h4 39.d3 Bd6 40.Kf1 Ne8 41.Kg2 Kd5 42.d3-f2® Ke4 43.Nd3† (from Pf2) Kf4 44.Nb7-c5 Ng7 45.Ne6 Nf5† (from Ph4) 46.Kg1 Kf3 47.Nd4 h4-g2® 48.Nb5 ( White finally gets his bishop free, but has concentrated on that to the exclusion of Black's attacking potential. 48.Nf4 would have prolonged the game, which still could be a draw. ) 48...Bh2† 49.Resigns (0-1) . (49.K×h2 K×f2 50.Ba7-c6® g1=Q† 51.Kh3 g4‡)

The next game shows the power of pawns relayed towards the promotion rank and stonewall defence by the white knights.

Paul Yearout v George Jelliss
AISE Grand Prix 1996

1.d4 f6 (with the idea of building up an attack on f2 via g4 by relay.) 2.f3 h5 3.c3 f6-g4® 4.Qc2 d7-f6® (5.Qg6† is guarded against by N-relay from Pe7) 5.e4 c6 6.Be2 e5 7.h3 e5×d4 8.h3×g4 f6×g4® 9.f3×d4®? (The Black g-pawn remains a thorn in White's side. White seems to have over- estimated the danger from the d-pawn. 9.f3×g4 looks good for White.) 9...Nh6 10.Qd3 g6 11.e5 Bf5 12.Qd1 Be7 13.Nf3 Bh4† 14.Kf1 Bf5-g3® 15.Be3 g4-h2®! 16.e5-f7®† Kf8 17.R×h2 B×R 18.Bd3 Nf5 19.Ke2 Bh4×g2® 20.Qd2 Bg3† (by relay) 21.Kd1 Qh4 22.c3-e4® Qh1† 23.Ne1 B×e4 24.B×e4 Bg3×e4® 25.b3 K×f7 26.Nc3 h4 27.Ne2 Nd6 28.Qb2 Be4-f6® 29.Rc1 a6

30.Kc2 Qe4† 31.Nd3? (I was expecting Rc1-d3® guarding Be3) 31...Q×e3 32.Resigns (0-1) .

Now some amusing pattern play.

Michele DeGiglio v George Jelliss
AISE Grand Prix 1996

1.Nf3 d7-e5® (occupying a cell in an opposing knight's field reduces its scope for mischief) 2.e4 Na6 3.d2-c4® Bd7 4. Be3 e7-c6® 5.Nbd2 Ne7 6.c4×e5® c6×e5® 7.c4! (reloading the knight's slingshot) 7...Nc5 8.c4×e5® Be6 (Didn't like the look of 8...Bd7×e5® 9.Qa4† or Bb5†) 9.e4-g5® f5 (no e.p. capture!) 10.g5-f7®† B×f7 11.e5×f7®† K×f7 12.Bc4† (recharging the slingshot again!) 12...Ke8 13.Bg5 Ne6 14.Bc4-e5® (once more unto the breach...) c6 15.Bg5-f7®† Qd8×f7®† 16.Be5×f7®† K×f7 (completing the repeat perform- ance) 17.00 g5 18.Ne4 Bg7 19.Nd4 h5 20.f3 g5×f3® 21.Q×f3 Resigns (1-0) . The threat of Qf3-g5®† seems decisive.

George Jelliss v Michele DeGiglio
AISE Grand Prix 1996

1.c3 Nc6 2.c3-e4® Nd4 3.Nc3 e5 4.Qb3 Be7 5.a2-b4® c6 6.b5 d5 7.e4-d6® B×d6 (the threat is 7...B else 8.b5-c7® with connected passed pawns.) 8.b5×d6® Q×d6 9.Nf3 b5 10.e4 (threat e4×d6®)

10...d5×e4 11.d2×e4® Be6 12.Q× b5!? (this gives back B for 2Ps again, but after 12.e4×d6 B×b3 the Ra1 and Bc1 are forked and d6 difficult to support, while retreat allows Black to build Q-side attack) 12...c6×b5 13.B×b5† Qd7 (13...Bd7 14.Bb5× Qd6®) 14.B× d7† K×d7 (equal). The rest of play is anticlimax: 15.00 Ne7 16.Nd5 Rhc8 17.b4 Nec6 18.Ng5 a7-b5® 19.b4-a6®? (R×R) Rcb8 20.e4- c5® Rb8×a6® 21. Resigns (0-1) .

Andrea Mori v George Jelliss
AISE Grand Prix 1996

1.e4 d7-f6® 2.c3 f6-g4® 3.Nf3 f6 4.Bc4 c6 5.Qb3 h6 (relays guard to f7) 6.h×g4® f×g4® 7.d4 Bd7 8.e5 c×e5® 9.d4-e6® Bd7-e5® 10.Nf3-d4 (Ns cannot be captured) Nb8-d7 11.e×Bc5® (backward captures by Pawns are easily missed) e5×Bc4® 12.Q×c4 b6 13.c5×b6 Q×b6 14.Qe2 Bf8-g6® 15.b4 Nd7-e5 16.Bd2 g4×f2®† (premature, Qf6 perhaps) 17.K×f2 h6-g4®† 18.Kg1 R× Rh1† 19.K×Rh1 Qf6 20.g3 (if 20.Qb5† Nc6 for a7-b5®) Qf6-h5®† 21.Kg1 Nf6 22.Bd2 -e4® B×Be4 23.c3×Be4® Nd3 24.e5 e6 25.Nb1-d2 Qh3? (premature again) 26.Qg2! Qh5 27.Q×a8† Kf7 28.Qb7† Nd7 29.Qe4 Nf2 (threat Qh3‡) 30.Qf4† Ke7 31.Ne2 Qh2† 32.Kf1 Qh1† 33.Ng1 Qh3† 34.Ke2 g4-h2® 35.Ng1-f3 Qh3×Qf4® 36.g×Qf4 Nd137.Rc1 h1=Q 38.e5-c6® Qh2† White resigns (0-1)


Back to: Home Page