Gilchrist is a legend wrote on 05/26/12 at 01:08:54:

It would be more difficult, since the players do not know what the opponent is going to play....

However, I think that in some positions of those 960, the castling is not consistent from an aesthetic point of view, since the king has "to jump" over the rook to get to the castling position....

An idea could be to play some selected positions that are known beforehand to the players.

For example in a round robin tournament of nine games to play every round a different selected position.

Thus the players can prepare, however the preparation difficulty is strongly increased and is open to new discoveries....

I would not be surprised, especially if there is some new programme, whatever it may be called, Chess960 Opening Theory Version XVI

If grandmasters can memorise a few hundred 30-move lines in the 6. Bg5 Najdorf, Botvinnik Semi-Slav, etc., I am sure eventually they can memorise Chess960. Or perhaps if not creating theory for all 960 positions, perhaps someone will formulate a better strategy for studying opening theory of Chess960. But for example, in 2025, I am sure Chess960 will have at least a bit of opening theory. It will require more time for this to occur, due to the large amount of starting positions. As I said before, variants are supposed to be played for fun, not to be seriously studied. Yet I considered bughouse, Three Checks, giveaway, etc. too seriously and created some of more than 10-move theory lines. So likewise, sometime in the future Chess960 may accrue theory in some form and then what would be the new mode of avoiding draws via theory?

Maybe not now, but perhaps in the close future, there will be computer programmes where one can study all 960 positions concretely, or perhaps if there is a major tournament with this theme, some grandmasters might invest time to study all 960 positions and dedicate opening theory and study to each individual position. Many grandmasters can memorise many 30+ move variations, and with help of some new computer I am sure they can soon study concretely all permutations in the starting position. By this fashion, they can arrive at the board, and when their random position is generated, they will think to themselves, "I prepared for this position (1 out of 960) a few weeks ago."

Choosing one 960 position per year is just as bad, as the players can still prepare for all 960 positions by compartmentalising their study based on where a pair of pieces are placed, as I mentioned above. But 1. c3 c6/1. d3 d6 does not prevent draws from happening, if both players prepare extremely extensively, which is likely from 2700+ grandmasters.

Any type of starting position will induce opening preparation, whether in an orthodox or a more unexpected, unorthodox fashion. When I used to play variants (1 0 and 3 0) on ICC, I used to study for hours and days, sometimes even weeks, and create bughouse opening theory with friends, as well as opening theory for crazyhouse, Losers, giveaway, etc. Three Checks actually has an unexpectedly large amount of opening theory if one studies it intensively. No reason why regular chess with a different starting position would have less opening preparation.

TalJechin wrote on 05/20/12 at 07:06:19:


If players know that 1. c3 c6 is the position they will play, they will prepare for it accordingly. Or do you mean a randomiser machine chooses the position with which to start? If the theme is a randomiser chooses one of eight positions where a pair of pawns are placed on the third rank, then the players can prepare for all eight positions with a computer anyway. Or at least, if I were playing in a major tournament with that randomising theme, I would spend a few hours on 1. c3 c6, 1. d3 d6, etc. Then new theory would be created in those lines and I am sure novelties would be made.

As pointed out by Steve Giddins here: (http://stevegiddinschessblog.blogspot.se/2012/05/levellers.html) - the real problem at the top level is that both sides are analysing the same variations with the same tools, getting the same answers.

Instead of ugly jumbled 960 positions one might as well decide to start all games with 1.c3 c6, 1.d3 d6 or 1.e3 e6 and then 90% of the theory would be out the window and the game would be roughly equal already so black wouldn't need to simplify to equalise and both sides could play to win from the start. 

In 960 you would instead get an unknown position that could be simply winning for one side if played on the top level - but I assume the point of 960 is to get an equal start?

The only 960 position I'd find interesting would be the the minor change of Qd1,Ke1 vs Kd8,Qe8 - which I've already advocated here several years ago...

Tricklev wrote on 05/19/12 at 19:44:31:


No, little to no increase at all. Notice how he doesn't say "until the draws increase to critical mass" but rather "until the consensus against draws reaches critical mass"... 

Draws have got nothing to do with the lack of popularity of chess. You have to be a very good player in the first place to even understand enough of GM games to follow them. Whoever came up with this idea that draws are a reason should be fired.