All-in-Chess: The Art and Amusement of Shared Control!

Let’s Imagine a World Where…

...the chessboard is a puppet theatre. Normally, each player controls only their own marionettes. But what if the strings got all mixed up, tangled into one great snarl, so that both puppeteers could yank at every figure in sight, pawns and pieces of either colour doing the bidding of whichever hand gave the pull? That, in essence, is All-in-Chess. There’s no such thing as ‘my pawn, my piece’ or ‘your pawn, your piece’ in this hall of whimsy. It’s like a messy divorce settlement: the children — pawns and pieces alike — don’t stay with one parent. They’re up for grabs every turn — all of them (hence, All-in!) — with custody contested one move at a time!

I’ll skip the metaphors and state it plainly: in All-in-Chess, each side takes turns just like in normal chess; however, they may move any piece or pawn on the board, regardless of colour. In other words, if I can move a piece or pawn, my opponent can move it too — and vice versa. But… but wait… I hear you, the astute reader, protest: doesn’t that mean my opponent can simply undo the moves I make? If that’s true, how could the game ever progress? I move a piece from a to b, and my opponent whisks it straight back from b to a, trapping us in an inescapable loop. Bravo, reader! Well, precisely to prevent such futile back-and-forth, there’s one key restriction: a player may move any unit of either colour, provided the board doesn’t revert to the position from two plies ago. Beyond that, the usual laws of chess remain in force: pawns must move in the correct direction for their colour, and no side may put itself in check.

Now, to see this in action, we turn to a problem by the late Indian composer S. N. Ravishankar, who had a particular fondness for the oddities of shared control. The mate-in-2 that follows is, in its simplicity, as perfect an introduction to this fairy type as one could hope for.

Problem 01

S. N. Ravishankar, The Problemist, 2016

If this were normal chess, 1.a8=Q+ Kxh2 2.Qg2# would simply solve the problem. Or would it? Well… if this were normal chess, Black couldn’t have moved last, and the position itself would be illegal with White to play! In All-in-Chess, however, the position is perfectly legal — Black could have moved last with either the pawn on a7 or the white king — and the sequence no longer works: the moment White queens the a-pawn, Black can slide the newly crowned queen off the long diagonal, rescuing their monarch from check. This is one of the quirks of this variant...frequently, a check isn't mate, as the other side can just move away the checking unit!

Then how do we deliver mate here? The trick is to underpromote the a7 pawn to a bishop, since a bishop on a8 can't be moved off the diagonal on the very next turn. And after Kxh2, we simply pull the black king back to h1, delivering mate: 1.a8=B! Kxh2 2.Kh1#. But why is this mate? Couldn't Black just shift the king to h2 again? No — this is where the crucial anti-repetition rule comes into play, making 2...Kh2?? illegal!

It’s important to note that the problem wouldn’t work if the h2-pawn weren’t present, since then, after 1.a8=B Kh2, 2.Kh1?? would be illegal, repeating the previous position. In the actual solution, White could tug the king back to h1 on the second move precisely because the previous move by Black was a capture. Consequently, by placing the king on h1 again, White wasn’t really undoing Black’s move.

I hope this problem has given you a solid grasp of the nuances of AIC. Perhaps you might try your hand at the next one yourself, a creation by Chris Tylor, the inventor of this fairy condition himself.

Problem 02

Christopher M.B. Tylor, Chessics, 1976

This time the bK sits in a different corner, h8, lined up against the bishop on b2. However, if you followed the logic of the previous problem, you’ll know the wB can serve as an effective mating piece only when it stands on the corner opposite the bK. Therefore, the key move should be obvious: 1.Ba1!, putting Black in zugzwang. Any push of wPc3 or wPd4 now brings a battery mate: 1…d5 2.c4# or 1…c4 2.d5#. Black could also try shifting wKf8, 1…Ke7 or 1…Ke8, but that walks into 2.Kg7#, since 2…Kh8?? is clearly illegal under the anti-repetition rule. Quiz: is 2.Kg8 also mate?

Having explored these two basic examples and gotten a feel for how AIC works, let’s now move on to something more serious. Shall we? The following is the brainchild of India's first International Master of chess composition, Narayan Shankar Ram.

Problem 03

N Shankar Ram, feenschach, 1986

If this were normal chess, 1.Qa8 or 1.Bxf6 would be mate in one. However, under the AIC condition, these are mere checks, as Black can simply whizz the attacking piece away on the very next turn. The key move is 1.e7+! — once again, in normal chess this would be mate; here, though, Black has …exNf8, and depending on which piece is promoted on f8, White delivers a unique mate on the second move: 1…exf8=Q 2.Qxf6#, 1…exf8=R 2.Bf7#, 1…exf8=B 2.Be7#, and 1…exf8=N 2.Kd7#.

Can you figure out why, after each promotion, White's second move is the only way to deliver mate? For instance, why isn’t 2.Qe7 mate in the first variation? Why would 2.Bg6 be wrong in the second variation? Think!

By the way, this is likely the only AIC problem to accomplish the well-known Allumwandlung task, which requires promotions to all four pieces in the solution.

Next, let’s move on to tackling compositions that are longer than two moves.

Problem 04

S. N. Ravishankar, N Shankar Ram; Chessics, 1986

It’s clear that the solution should begin with 1.h8=R+! (by now, you know why 1.h8=Q+ wouldn’t work). Black can interpose the check on the h-file with both Bd6 and Rd7, yielding two perfectly harmonious variations that mirror each other along orthogonal and diagonal lines:

• 1…Bh2 2.Bb8+! Rdh7 3.Rc7# (Bb8 can’t interpose on the h-file, as it’s blocked by Rc7, and playing 3…Rdh7?? again would be illegal due to repetition)

• 1…Rdh7 2.Ra7+! Bh2 3.Bc7# (In this case, bB and bR swap roles. Consequently, Ra7 can’t interpose on the h-file, as it’s blocked by Bc7, and playing 3…Bh2?? again would be illegal due to repetition)

White’s second move in each variation above is a critical move. A critical move arises when a line-piece crosses a square where it could potentially be obstructed by another unit. For instance, 2.Bh2–b8 is critical because it passes over c7, a square the bR subsequently occupies. By contrast, 2.Rh7–a7 is critical for crossing that very same square, this time blocked by the bB, creating an interference.

Also notice that the uniqueness of White’s second move in the second variation is secured by the wKb1 — since 2.Rb7?? would obviously be illegal, putting White in self-check!

The penultimate problem, a more-mover, comes with a twin. First, solve the given diagram; then, for part (b), shift the bQ to a3 and tackle it again! This may well be the toughest problem of the set, but if you’ve taken in the themes from the preceding mate-in-3s, you’ll find you already have the tools to solve it.

Problem 05

N Shankar Ram, Die Schwalbe 1986, 4th Hon. Mention

Of course, the play must begin with the advance of the wPc2. But the question is — do we push it one square or two? In the diagram’s solution, all the action unfolds along the a2–f7 diagonal, so the correct start is 1.c3+!, shielding bRf3 from barging in. The main line runs 1...Beb3 2.Bf7+! Qb3 3.Qa3+! Bfb3 4.Ba2+! Qb3 5.Qf7+! Bab3 6.Be6#, while 1...Qb3 leads to mate one move sooner: 2.Qa3+ Beb3 3.Ba2+ Qb3 4.Qf7+ Bb3 5.Be6#.

The observant reader will notice that the bB crossing over b3 and the bQ crossing over e6 are both critical moves, each underpinned by a distinct line-clearance motif. In the first instance, the bishop makes way for the queen to traverse in the opposite direction, known as Voidance; in the second, the queen clears the path for the bishop to proceed along the same course, the so-called Bristol manoeuvre. Meanwhile, the effect around the b3 square — with the bQ giving way as the bB crosses it and then executing a switchback — constitutes the Klasinc theme. These terms are defined formally below:

Part (b) shows the same combination of ideas, except the action now unfolds along g3–a3. Consequently, the key must screen the bBe6: 1.c4+!. The analogous main line runs 1...Rb3 2.Rg3+! Qb3 3.Qa2+! Rb3 4.Ra3+! Qb3 5.Qg3+! Rb3 6.Rf3#, while 1...Qb3 results in mate one move sooner: 2.Qa2+ Rb3 3.Ra3+ Qb3 4.Qg3+ Rb3 5.Rf3#. Can you clearly make out Voidance, Bristol, and Klasinc from the definitions provided above?

Our final problem is yet another fun creation by S. N. Ravishankar. This one leans more on humour than sophistication, wrapping up the article with a smile!

Problem 06

S. N. Ravishankar, KoBulChess, 2017

Surely, the coup de grâce has to come from the b6 pawn. Needless to say, 1.b7+ doesn’t work immediately, as Black can reply with 1…b8=B or b8=N. White must therefore first block the b8 square, and the only unit which can serve that purpose is the wPh2, which must then promote to a wB and make its way to b8. Once you see this logic, the solution almost plays itself.

1.h4! h5 2.h6 h7 3.h8=B Bg7 4.Bh6! Bf8 5.Be7 Bd8 6.Bc7 Bb8 7.b7#

Note that, on account of anti-repetition, every move by Black is forced, making the whole sequence feel almost like a helpmate. After all, White couldn’t possibly get the h2-pawn all the way to b8 in just seven moves without Black lending a hand!

And with that, we’ve reached the end of our journey. We’ll leave you with one last challenge to mull over — until next time!

Acknowledgement

1. The author of this article is immensely grateful to N. Shankar Ram, International Master of Chess Composition, for his inspiring guidance and encouragement.

2. The author also acknowledges the use of Viktoras Paliulionis' Glossary of themes and terms.

Contact

If you have any feedback on this article, or if you’d like to share your solutions to the two bonus problems, you may write to chessbaseindiasocial@gmail.com

Related news:

Gone Too Soon: Exploring the Unfinished Chess Legacy of Dr S. Manikumar @ 10/09/2025 by Satanick Mukhuty (en)