Any thoughts on the pros and cons of ballots would be welcome. A forum is a good place to thrash out a topic like this.
Wouldn't this mean that "skill" in chess now consists of randomly drawing a strong ballot instead of a weak one? The best player no longer wins... it's the luckiest player who wins.
In the TCEC superfinals they impose slightly unbalanced openings to prevent all draws.
@blueemu
//Wouldn't this mean that "skill" in chess now consists of randomly drawing a strong ballot instead of a weak one? The best player no longer wins... it's the luckiest player who wins.//
Not really, because as I said in my post, the ballots would be controlled. I gave the example of no more than 0.75 pawn advantage, but that could be adjusted.
#4
In the TCEC they have to play 2 games with each slightly unbalanced opening: one with black and one with white.
0.75 pawn advantage is enough to win for a strong player. The threshold between draw and win is about 0.50 pawn.
I like the idea. It's better than Chess 960. Players would be obliged study all ECO from A00 to E99. They won't be preparing 30+ moves Opening Theory. All players will have to be playing a Universal style.
@tygxc
//In the TCEC superfinals they impose slightly unbalanced openings to prevent all draws.//
Yes, and that is probably a good thing. As chess becomes more accurate, the number of draws increases.
On the Stockfish site, comparing version 14 with 14.1, the developers say, "at this high level, there are many draws".
Actually, I don't think a draw is necessarily bad. There are many fine games which end as draws. The emphasis of my post was rather to try to reduce the influence of chess engines and tilt the game in favour of human skill at the board.
#6
Chess 960 is better: 960 positions instead of 500 ECO codes.
Besides many strong players like Carlsen have a strong grasp of all 500 ECO codes. It does not avoid theory, it only encourages broader theory.
@tygxc
//0.75 pawn advantage is enough to win for a strong player. The threshold between draw and win is about 0.50 pawn.//
Ok, thanks for that info. So my example parameter would have to be tightened up.
#7
An ideal game ends in a draw. The game quality is good then, but the competition is bad.
The Caruana - Carlsen match for the World Championship ended in 12 draws out of 12 classical games. Then they had to go to rapid tiebreakers...
In ICCF they are at like 97% draws
https://www.iccf.com/event?id=85042
The future of chess is probably 15|10 rapid time control and Chess960.
#9
All 500 ECO codes are more or less balanced. Draw a number from 1 to 500, that is your ECO code, now play a game with white and a game with black.
@tygxc
//An ideal game ends in a draw. The game quality is good then, but the competition is bad.//
Yes. However, human players are less accurate than engines, so I think increasing the variety of play with ballots would still make for good games.
To give an example, Draughts (Checkers in the US) was weakly solved by a Canadian team, and they showed that with perfect play every game ends in a draw. But humans don't play it perfectly, so there are stll many wins and losses. In fact, Draughts has used a 3-ply opening ballot for more than a century now, and its use revived the game.
An advantage of controlled opening ballots over Chess 960 is that they would be fairer. I may be wrong, but as I understand it, 960 includes quite a few opening setups which favour one side over the other, so a player is at a disadvantage right from the beginning.
#13
The 960 positions of Chess960 are not less balanced than the 500 ECO codes.
Can you provide an example of a chess960 position that you consider unbalanced?
#13
The 960 positions of Chess960 are not less balanced than the 500 ECO codes.
Can you provide an example of a chess960 position that you consider unbalanced?
There’s a few that favour white quite strongly (more than a normal opening edge) and there’s one they’re thinking of prohibiting from chess 960 events as it’s possibly a forced win.
I don’t have the position(s) to hand but will post later if I can dig it out.
also
adding atleast 460 more good positions(from played, balanced, openings) is easy, so the eco codes can be supplemented
they are fairly old
so anyway, we will have more opening ballot positions than chess960 ones easily, so why not have more variety also with more familiarity
#15
Interesting, I look forward to an unbalanced chess960 position.
Some of the 500 ECO codes are considered to lose by force.
After writing the original post about 6-ply opening ballots, quite by coincidence I came across a set of 6-ply openings on the web.
There are 1185 different opening lines, compiled by an engine enthusiast. I haven't checked them all, but most seem to give about 0.5 pawn advantage or less to one side.
This sort of work could form the basis for ballots in human play that give plenty of variety in the opening.
And even if you don't like the idea of ballots, there are a lot of opening lines if you want to play against a friend or an engine and want to mix things up to get out of an opening rut. Or you can use them for engine-engine tournaments.
On the link page, under the "Books/Start Positions" heading, select the link "My current self created pgn test book". It downloads a small zip file (16.3kb).
The file contains the opening lines as a text file and as a pgn. There is a second pgn which is a randomised version of the first. There is also an explanatory read-me and a regex macro for creating a pgn from a text file.
https://rwbc-chess.de/download.htm
I wonder why opening ballots are not used in chess, especially at the elite GM level. There has been a lot of criticism that today's GMs just play out memorised engine moves for the first 30-40 ply or more, and that the modern game has lost its human creativity.
The image of a modern GM is rather nerdy, that of a teenager or twenty-something sitting in front of a computer, studying and memorising lines the engine has found. It's not a particularly attractive image.
For some, Chess 960 is the answer to that, but not everyone is on board. I don't think five hundred years of standard chess can be brushed aside that easily.
So how about ballotted openings? I was thinking of a ballot of six ply. Obvious early losses would be weeded out of course. The included ballots could be evaluated by an engine to give no more than 0.75 pawn advantage for a side, for example.
Such parameters would produce hundreds of viable openings, one of which would be selected at random as the players sat down at the table in a tournament.
Opening ballots are already used in computer chess tournaments to give more variety of opening play and to test engines in many different midgame positions.
For humans, the effects would be interesting. On the one hand, standard opening theory would not be thrown out, as it is in Chess 960. So some opening study would still be worthwhile. On the other hand, GMs could not play their memorised pet lines, since the chance of a particular line coming up is quite small.
So there is a subtle blend of continuity with standard chess, mixed with some tweaks which refresh it without mangling the game or turning it into something else like 960.
What kind of elite player would benefit from ballotted openings? I would say it favours those who are flexible and adaptable and who have exceptional ability in positional assessment. It would favour a player like Carlsen, for example, because ballots tilt the game away from prepared lines and towards strong over-the-board skills.
It would certainly increase the variety of play, would reduce the influence of engines and most likely would reduce the number of draws, making the game more interesting for the chess public.
Any thoughts on the pros and cons of ballots would be welcome. A forum is a good place to thrash out a topic like this.