thoughts on Big Chess and tips


Back to forum

Normajean Yates    (2009-05-11)
thoughts on Big Chess... and tips...

I find bigchess more and more fascinating.. I Think it is a wonderful creation of Thibault's (I presume it is Thib. who created it: any way he offers it seriously on this site...) - the starting position is very well-concieved..

I think Bigchess needs more publicity. This is about the only place one can play it - and here there are 2-3 top-class players; less than 20 middle-standard players (including me); others try it once or twice and for some reason get scared or overwhelmed and give up - I see no reason why..

Bigchess gives no advantage on account of huge memorisation of theory, or of better engines: there are *no* theory books; and there are no known engines in existence (probably there isnt one - too little demand, and writing a *good* engine is somewhat laborious, coming up with a *good* static-eval function is tricky, fast board-implementation issues...), so it is all wits...

In fact last week I spend part of two days writing down whatever theory I could discover [with help from top games], it comes to half a page..

Tips for people who want to try bigchess:

1. Bishops are much more powerful than Knights. (because of much longer range compared to 8x8 chess).

The consensus on the values of bigchess pieces is David Grosdemange's valuation:

knight=2.5 (written 2,5 in the continent, of course)

2. In the opening position, the c,f,L and o-pawns are unprotected.

So, if white's opening move is with the j2-Knight ( freeing the queen), then on move 2 white can move the Queen and threaten to pick up a pawn by forks.. Similarly for black.

*However*, such pawn gambits are quite playable because the Queen can be forced to make many moves to capture a pawn, while the gambitting side develops their pieces.

3. Most Important For Many People: board for offline analysis.

Best of course, is to take time to draw a 16x16 board on paper and stick it on cardboard. And get hold of four sets of chess pieces.

Another way: print a position, and after a move is made - just update the position using correction fluid (typewriter/printer-ink erasing fluid) or something. That way you don't have to keep printng a lot of positions.

Daniel Parmet    (2009-05-11 03:00:59)

I only played 6 games of bigchess but I did go 6/6! I will say the values I used without consulting to anyone was
P 1
N 2.5
B 4
R 4.5
Q 10

Daniel Parmet    (2009-05-11 03:02:30)

oh also the reason I don't play more big chess is cause the average game goes over 200 moves and that assumes they resign once its clear they are lost... I can see games going to 600 moves if you play to mate.

Normajean Yates    (2009-05-11 03:52:01)
Daniel, I think you underestimate Rs...

you give B=4, R=4.5 but in the endgame specially, the Rooks pull their weight..15th ('7th' file), escorting Ps to promotion, ...

About people playing on to mate after being a Q down or so without counterchances, well a bigger player-pool is needed ... plus I understand some newbie players not being so sure that mate is easy in a given pos (though much lengthier than in chess) - so, we need *more* non-newbie players!

And a non-newbie begins as a newbie: however, I agree that's not the whole story -- some people *will* play on to sees that in chess (8x8) too -

lets see: my bigchess record: 17/17 - wow! (with 1 pending: it is still in the gambit-opening stage)

Thibault de Vassal    (2009-05-11 21:27:26)
Bigchess pieces values

Actually the value of the knight seems to change a lot during a game, according to the left pawns positions.

200 or 300 moves is not so much compared to a poker game (that can be over 1000), but some games may be really time consuming when the position gets really complex... Well, we play for this kind of fun after all ;)

Normajean Yates    (2009-05-12 00:33:50)
response to Thib's post:

But that is true of 8x8-chess also (about value of knights) - there, 'P=1, B=N=3,R=5,Q=9' is a very rough guideline. Ditto for bigchess piece values.

What is it that Reti (or was it Breyer said) - about move depending on particular position, *not* general principles? ;)

Normajean Yates    (2009-05-12 00:47:43)
more bigchess thoughts..

compared to 8x8 chess, bigchess has twice the number of men but four-times the number of squares. Hence, initial and average board-population density is half that of 8x8chess. So, games are more commonly 'open' - games as closed as an 8x8-chess closed game will be comparatively rare.

Hence, also, (because of low population-density) long-range pieces Rook, Bishop and Queen are *in general* -- Thib., agreed: only 'in general' :) --- much more powerful than their 8x8-chess counterparts.

However, the queen can still get trapped early - see my game number 31191 - opp resigned on my [black's] 7th move because of trapped Queen..

Normajean Yates    (2009-05-12 01:02:03)
long bigchess .. and chess games..

from the english wikipedia: "in May 2006, Bourzutschky and Konoval discovered a KQNKRBN position with an astonishing DTC of 517 moves." That is distance-to-conversion! (and that was 2006!) So, you can figure out distance-to-mate and the overall length of a game that reaches this endgame ;)

Normajean Yates    (2009-05-12 01:15:30)
bigchess thoughts+tips, #3: R v B

[re Daniel Parmet's valuation in his post] - both are longrange pieces but:

1. a Bishop can reach only 128 squares, a Rook can reach all 256. (and all the other standard reasons why the R is [in general] much more powerful than the B, they carry over to bigchess eg a R can confine the opp's K to an edge; K+R v K is standard easy mate, etc.)

2. In 8x8 chess, once you have a semi-open file, one tries to 'boost' a rook up that file (even if one cannot reach the 7th file). This is much more common, and much more commonly advantageous, in bigchess, because one easily creates a 'quasi-semi-open' file by pushing a pawn 4-5 squares ahead, and boosts a Rook up this 'quasi-semi-open' file.

Hence, I feel that just as in 8x8 chess, the Rook is nearly twice as powerful as the Bishop in bigchess.

Daniel Parmet    (2009-05-12 02:04:56)
Queen trap

yes I must say these are quite common in bigchess. I trapped my opponents queen in 2 games and I think if I recall I used the threat of trapping in one game to win massive material.

Normajean Yates    (2009-05-12 07:21:02)
to Daniel: and to Thib and programmers

The queen-traps - of the patzer kind as in my game at any rate - will become less common once we have more middle-level players, I think.

Also, [to thib and non-retired programmers] in bigchess game records, it would help to indicate the piece moved [tiny array in the code so that the piece display is in the language one wants, or in figurine notation], to indicate captures by 'x' and the captured piece, and to indicate promotion.

This is trivial [to write a converter from present notation to this more human-freindly one, given a game from the starting position -- 10 years ago I would have written and uploaded it (C/Haskell source code, command-line window) in 3 hours - but now I feel sooo lazy to write a single line of code - my programming brain-cells are dead or in a coma :)

Sophie Leclerc    (2009-05-12 18:49:59)
Too much pieces and too big board

I understand BigChess is a good invention, But for me, the board and the pieces are now too much to think about.

Thibault de Vassal    (2009-05-16 19:41:58)
When the knight overplays the bishop

Yes, it is possible at Big Chess too !

An interesting game to watch : Pichelin vs. Legrand 0-1

What do you think ?

Normajean Yates    (2009-05-17 02:38:48)
wow - great game!

bigchess game # 31148 Pichelin-Legrand 0-1; as Thib. posted above... I voted for it as 'besr game' - people please follow it and consider voting for it!

Pavel Hase    (2009-11-30 23:56:42)

Value is higher, my guess.
N - all fields, but horde moves for displacement, very slow piece.
B - only 128 fields, but only 2 moves for displacement (if clear board)
R - all fields, only 2 moves for displacement
Q - only 2 moves for displacement, but over one move, than Rook.

N - 3 (2-4) (From two pawns other sides any chance, but if pawns nearly, anyway 8x8 chess. Board 8x8 Notin, needed max. 4 moves, here?) Other tip? Mutually support afore own pawns.
B - 6(!) Very higher movement, than Notin. Other tip? Between own pawns, menace opponent piece.
R - 11 (10-12) Anyway 8x8 chess, pieces for middle game and endings. Interplay here is heavy work.
Q - 23 (20-30!) If interplay Rooks is heavy work, then Queen probably better, than two Rooks. Anyway 8x8, attention, traps and time for raven.

Sorry, my english language is weak.

Pavel Hase    (2009-12-01 01:25:54)
Notin movement

Scheme is here:

Thibault de Vassal    (2009-12-01 09:08:38)
Notin movement

What what ? What is this Notin movement ? Can anyone explain ?

About big chess pieces values, I agree that the pawn value is most probably less than considered before... Now I would say something like Q = 15.