The Kibitzer 30 
by Tim Harding

Bring Back Free Castling!

THE GAME OF CHESS as we play it today is essentially
unchanged since the late 15th century. However, pressures are
building up that may make it necessary to have a revolution in the
rules. Maybe not in the lifetime of anybody reading this column in
the final month of 1998, but quite conceivably within the next 100
years. The main reason, of course, is computers.

Even without computers, there have been times when some players
(such as Capablanca) felt the game was becoming played out with
too many draws, and required a change of rules. 

The most recent well-publicised proposal was Fischer-Random in
which former world champion Bobby Fischer suggested a variant
on the old idea of randomised chess where the game can start with
various arrangements of pieces along the back rank (Black's and
White's to match). At a stroke this gets rid of all present opening
theory and makes players think for themselves, but to some people
(including myself) the idea is too radical. For example, if I cannot
a fianchetto a bishop then I think I will play some other game and
not chess at all.

Chess as it is at present is probably satisfactory for the majority of
casual players, who are unlikely to want a radically different game.
It is the stronger and more active players (not necessarily the same
thing) who may want a change - especially in correspondence
chess where many feel that the game is of less interest if the
opponent's moves are being decided with the aid of a computer. In
this article, I have one interesting suggestion to make which I think
is interesting and not as radical as Fischer-Random.

Changes to the rules can be of two kinds - alterations to the actual
laws of chess or just changes in tournament conditions. Changes in
tournament conditions (e.g. to do with clocks, time-limits,
eliminating adjournments, possibly even allowing computer
"sidekicks") have all been tried in recent years; most of these don't
affect the ordinary social and family player who does not join a
club or enter a tournament or other competition.

The famous book by Jonny von Neumann, "The Theory of Games
and Economic Behaviour", classifies games into various
categories. For example, chess and checkers are zero-sum two-
person games with perfect knowledge whereas poker is zero-sum
(what someone loses, somebody else must win) but normally
involves more than two players and less than perfect knowledge
(only some, and sometimes no, cards are revealed during betting).

Changing chess can involve tampering with each of these
parameters. If you eliminate "perfect knowledge" you get
something like kriegspiel, in which players get limited information
about the opponent's moves, transmitted through an umpire. It's
fun, but not a serious alternative to chess for serious sporting
competition or artistic exploration.

Changing the "two-person" parameter is an option, but again I
think not a serious one; chess appeals to individuals and is not an
ideal partnership game. Of course, chess may have evolved from a
game in India called chaturanga which was a four-handed game
played with dice. Other four-handed versions have been suggested.
We had a board at school and used to have a lot of fun with it; but
this is not a serious alternative to "proper" chess.

Eliminating "zero-sum" is definitely a possibility as a change in
tournament conditions that would not affect the laws of the game
as such. Instead of the present one point for a win, half-point for a
draw, you can have 3 points for win, 1 point for draw as is done for
example in soccer. By discouraging draws, this would be a change
to tournament conditions that would have a knock-on effect to how
game itself would be played. However, in chess there is rarely a
paying audience to entertain. Would this would be a beneficial
change or would it just lead to surreptitious point-selling in the
washrooms of tournament venues as players saw prizes eluding
them in a game heading for a drawn ending?

Proposals to change the rules of chess tend to involve either a
change in the board, or changes to the move sequence or capturing
arrangements, or most often to a change in the pieces. The number
of possible chess variants is limited only by the human imagination
and a vast number of variants have been tried by the members of
the correspondence chess club, kNights Of the Square Table
(NOST) whom you can contact via their website at
http//www.accelnet.com/NOST/index.html.

A change in the board can involve an extra rank between White
and Black, or three-dimensional chess (3 boards arranged
vertically) or cylindrical chess in which the a-file and h-file are
adjacent.

Changes to the move sequence can be illustrated by progressive
chess in which White makes one move, then Black makes two,
White makes three etc. until the game is (quite rapidly) over. This
is quite a challenging and intellectually fascinating chess variant,
definitely one of the best.

Variant chess pieces (sometimes known as "fairy chess") are quite
popular with problemists but the only variant chess piece I ever
played with much was one called the Transcendental Prelate,
invented in the 1960s by George Botterill who later won the
British Championship twice. A transcendental prelate's move is
calculated like that of an ordinary bishop EXCEPT that, like a
knight, it always changes the colour of the square that it lands on.
It is allowed to go to any square of the opposite colour adjacent to
any of the squares a normal bishop can go to. 

This may sound awkward but you soon get the feel of how to use
the piece after 1 e2 e4 White's prelate on f1 has the following
squares available - e3 (via e2), c3 and d4 (via d3/c4), c5 and b4
(via c4), a5, c5 and b6 (via b5) and it can also capture the black
pawn on a7. Some of these moves would of course be blunders, but
2 TPf1-c5 is OK since the black prelate on f8 cannot capture it!

At Oxford University we had a lot of fun with these prelates; in
fact we used to enjoy late-night transcendental prelate kriegspiel
sessions to unwind after serious matches. My favourite memory of
these involves a game in which my opponent (White) had his king
on c2 and a knight on d2. The position of his other pieces,
somewhere over on the kingside, are irrelevant.

As Black, I had moved a transcendental prelate to a2, controlling
the squares b2, c1 and c3 near the enemy king. My other prelate
was on somewhere like e6 and I played it (via b3) to b2 whereupon
the umpire to my surprise announced checkmate!

Diagram
White: Kc2, Nd2
Black: Kg8, Ba2, Bb2

All squares in the king's field are attacked by prelates except for d2
which was blocked by his own knight!

However, I do not seriously propose replacing present-day chess
("ortho-chess" as they call it at NOST) with transcendental prelate
chess. 

When considering possible reforms of chess, it is worth
remembering that the game we play now itself evolved from earlier
forms of the game by rule changes which largely involved a
change in the powers of several pieces.

Modern chess developed in the second half of the 15th century
from the slower form prevalent 700 years ago in the Middle Ages
and the Arab world as a result of a few linked reforms

a) The double advance option for pawns on their first move, with
the "en passant" rule that arose as a concomitant of it to avoid too
many blocked positions;

b) The transformation of the bishop from a puny diagonal-mover
that could only hop two squares at a time to today's wide-ranging
piece;

c) The modern queen (or "firs" or "vizier") combining the moves of
the rook and new bishop;

d) Castling, to protect the king in the new high-speed, high-
firepower environment that the chessboard had now become.

If you want a detailed history of how modern chess developed, I
can recommend the early chapters (especially chapter 3) of "Chess,
The History of a Game" by Richard Eales (Batsford 1985). Eales is
a university history lecturer who was a member of the same
Cambridge University team with Keene, Hartston and other
leading English players of that generation. 

Eales makes the point that the double-pawn move had been
experimented with from about the year 1200, along with other
innovations to speed up the game that did not make it into modern
chess, but the first description of the modern rules is to be found in
Lucena's book "Repeticion de Amores E arte de axedrez" printed
in Spain in 1496. 

The new form of the game rapidly spread throughout Europe in the
early 16th century, supplanting the mediaeval game, but was
slower to become known in the Arab world and Asia. 

The printing press was obviously important for the spread of the
new rules. In Italy, however, the new form of the game was
adopted with local modifications that were still in use in that
country as late as 1880 (according to Professor Carlo Pagni of
Turin, the great historian of chess in Italy). One of these rules,
Passar Battaglia ("pass over the battle"), was in effect a negation of
en passant; the pawn could not be captured if this rule was
invoked. 

Free Castling is the rule that interests me. This meant that the king
and rook could end up on various squares, e.g. the rook on e1 and
the king on h1 for White in kingside castling, or in the case of
queenside castling, the King could go to a1 (a8 for Black) or b1
(b8) and the rook to e1 (e8). 

I certainly do not advocate Passar Battaglia but I would seriously
suggest that Free Castling is a reform that would be well worth
adopting to reduce the impact of theory and computers on the game
as it is played now. This would send publishers in a scurry to find
books of old Italian games as 99.9% of current opening theory
would immediately become redundant. 

Here is an example, found in one of Professor Pagni books.
Warning - you won't be able to play this game into your database
program so get out a proper chess set to look at it! Played by
correspondence between the clubs of two Italian cities in 1877, it
was (he remarks) a good example of the level of play in Italy in
that period".

White: Ferrara Black: Modena

1 e4 e5 2 Nf3 Nc6 3 Bc4 Bc5 4 c3 Nf6 5 d3 d6 6 Qe2 h6 7 Be3
Bb6 8 Nbd2 Kg8-Re8 

Until this point it was a "normal" game, but now Black makes use
of Free Castling to get his rook to the central file. The players of
Modena could have put their king on h8 but maybe thought it was
safer on g8. Incidentally, Modena is the home town of the famous
tenor Pavarotti and if you travel by train from Milan to Bologna (or
onward to Florence and Rome) you will probably stop there. The
Italian Rules seem to have been particularly popular there.

9 h3 Ne7 10 g4 Bxe3 11 Qxe3 Ng6 12 g5 Nf4 [DIAGRAM?]

13 Ka1?-Rd1!

Free Castling again. The rook does not go to e1 because of the
threatened knight fork on g2, but I add the question mark because
the king would have been better placed on b1 (see move 26). In
"normal chess" of course, the king would go to c1. Now the game
proceeds "normally" for a while...

13...hxg5 14 Nxg5 Be6 15 Bxe6 Nxe6 16 Nxe6 Rxe6 17 Rdg1
Nh5 18 Nf1 Qf6 19 Ng3 Nxg3 20 Rxg3 Qf4 21 Rhg1 g6 22 Qxf4
exf4 23 Rf3 Rd8! 24 Rxf4 d5 25 exd5 Rxd5 26 d4 Rg5!

A back rank trick - if 27 Rxg1 Rd1 mate. See my note to White's
13th move!

27 Rf1 Kg7 28 a3 Rg2 29 d5 Rf6 30 Rb4 b6 31 Rc4 c5!

This move was only playable because of the "Passar Battaglia"
rule; in ortho-chess White would now win by 32 dxc6 en passant.

32 b4 cxb4 33 axb4 Rfxf2 34 Rxf2 Rxf2 35 Rd4 Kf8 36 c4 Ke7 37
Kb1 Rf3 38 Kc2?

According to contemporary sources, 38 c5 would have led to a
draw. Even later White might have been able to salvage a draw but
the game ended

38...Rxh3 39 c5 Rd7 40 d6 Rh8 41 Kc3? Kc6 42 Kc4 bxc5 43 b5+
Kb6 44 Re4 Rh1 45 Re3 Rd1 46 Rd3 Rxd3 47 Kxd3 Kb7 and
White resigned.

Looking through Professor Pagni's book made me think some
more about the implications of transforming chess by the
introduction of Free Castling. I was struck by how radically it
could affect modern theory yet how quietly White played in the
opening of the above game and several others. 

I thought first of the Marshall Gambit in the Spanish (Ruy Lopez).
After 1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 Be7 6 Re1 b5 7
Bb3 0-0 8 c3 d5 9 exd5 Nxd5 10 Nxe5 Nxe5 11 Rxe5 c6 12 d4
Bd6 13 Re1 Qh4 14 g3 Qh3 15 Be3 Bg4 etc. the main problem
Black encounters when trying to develop his attack is that ...f5 will
expose his king (on g8) to a pin by the bishop on b3 against the d5-
knight. If Black could free-castle with 7...Kh8 (and maybe ...Re8
too) he would have a much more dangerous attack.

However, I soon realised that sharp openings would be more
radically affected at an early stage. In that variation, why should
White play 5 0-0 and 6 Re1 if he could move the rook to e1 at the
same time as he moved his king to g1? 

Therefore a Free Castling Spanish should look like this1 e4 e5 2
Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 Kg1,Re1 Be7 (The Open Variation
becomes a thing of the past!) 6 d4! when 6...exd4 would be very
bad because of 7 e5, probably forcing 6...d6. Also the Exchange
Variation (1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Bxc6 dxc6 looks very
awkward if White brings his rook to e1 when he castles on move 5.

In other openings, the freedom of the king to go directly to the
corner could be more important than the rook's chance to go to e1.
For example, an Evans Gambit between Bologna and Modena in
1877 began 1 e4 e5 2 Nf3 Nc6 3 Bc4 Bc5 4 b4 Bxb4 5 c3 Ba5 6 d4
exd4 and now White free-castled with K-h1, R-f1. This avoided
pins or checks on the a7-g1 diagonal when White played f4 later.
Also some variations of the King's Gambit could become more
dangerous with the white king going immediately to h1 while the
rook went as normal to f1 to enjoy the opened f-file.

Why did the Italians abandon Free Castling? Partly, it may have
been because with the increasing international communications
they found themselves out of step with the rest of the chess world
and partly because they felt Free Castling gave White too great an
advantage?

According to Prof. Pagni, the last correspondence game played
with the Italian rules may have been in the intercity match Livorno
(i.e. Leghorn) and Modena in 1877-78 which was annotated by
Steinitz in his column in the English publication, The Field, in
1879. normal rules were used in the game with Livorno as White
but with Modena playing White, the Italian rules were used. Both
games were drawn.

Modena-Livorno began1 e4 c5 2 Nf3 e6 3 c3 Nc6 4 d4 d5 (unwise
in a Free Castling game, if not in any game) 5 exd5 exd5 6 Bb5
Qd6 (Contemporary theory preferred 6...cxd4.) 7 Qe2+ Be7 8 dxc5
Qxc5 and now 9 Re1/Kg1.

Steinitz now wrote"The possibility of putting the Rook in the e1-
square gives, according to most of the players, too great an
advantage to White in many openings".

Does that mean, therefore, that the Free Castling idea should be
rejected? I think no such thing. What it probably does mean is that
after 1 e4 the reply 1...e5 is dubious in games played with Free
Castling and that Black should strive to keep the play more closed,
so that after both players have castled (freely or otherwise) an
orthodox game results.

For example, in the game Modena-Padua 1877, Black adopted the
French Defence (1 e4 e6) about which the leading Italian player of
the time, Serafino Dubois (1817-1899) commented, "This is the
most monotonous and annoying play you can imagine - rarely it
gives rise to combinations of some interest". This is a backhanded
way of complimenting Padua on choosing a good defence - with
the e-file closed there is much less advantage to be obtained by
castling your rook directly to e1, so Modena adopted the Exchange
Variation (1 e4 e6 2 d4 d5 3 exd5 etc.) in order to open the file
quickly but Black blocked the file and the game was drawn.

I suspect that top players today would adopt either the Sicilian or
Caro-Kann Defences against 1 e4, to keep the e-file closed as long
as possible.

I think it would be very interesting to have some new experiments
with Free Castling to see how the players like it and whether White
really does have a significantly increased advantage as a result.
Even if it is so, that is not necessarily bad if it reduces the
percentage of draws in high-level play, so long as White does not
have a forced win. 

What do readers think? Maybe NOST, specialists in this kind of
thing, should hold a Free Castling tournament to test the idea.