- Blind mode tutorial
lichess.org
Donate

Computer eval of Fischer-Random (960) starting positions?

Observation: top players start thinking from move one, as opposed to standard. Conclusion: in standard, there are possibly the most alligators in the water with the kings smack dab in the centre and a long time from castling... 1.e4 was particularly treacherous, as the equalising try 1..e5 could open up the king and be met in a variety of fairly forcing ways including the Scotch and King's Gambits. While the 1.d4 /1.c4/1.Nf3 set was promising, 1.e4 was truly dangerous. Eventually it was studied and now computers advise the first moves and black diverges from the computer lines only unwillingly as they might be losing then. If white does not press critically black will be equal. Facing a 960 position with white there are those kind of options again. Should I go for a gambit (in case of a symmetrical reply) or a slow build-up? And importantly, the opponent's probable reactions are not intensively known already. Black in 960 has a similar choice, aiming for symmetrical equality often makes sense, but dynamic openings if feasible are preferable for winning chances. In standard the Pirc and others are precarious, but in 960 the venom can strike more often, explaining the black wins.

The stronger the engine, the closer to 0.0 the best play will be evaluated as.

Observation: top players start thinking from move one, as opposed to standard. Conclusion: in standard, there are possibly the most alligators in the water with the kings smack dab in the centre and a long time from castling... 1.e4 was particularly treacherous, as the equalising try 1..e5 could open up the king and be met in a variety of fairly forcing ways including the Scotch and King's Gambits. While the 1.d4 /1.c4/1.Nf3 set was promising, 1.e4 was truly dangerous. Eventually it was studied and now computers advise the first moves and black diverges from the computer lines only unwillingly as they might be losing then. If white does not press critically black will be equal. Facing a 960 position with white there are those kind of options again. Should I go for a gambit (in case of a symmetrical reply) or a slow build-up? And importantly, the opponent's probable reactions are not intensively known already. Black in 960 has a similar choice, aiming for symmetrical equality often makes sense, but dynamic openings if feasible are preferable for winning chances. In standard the Pirc and others are precarious, but in 960 the venom can strike more often, explaining the black wins. The stronger the engine, the closer to 0.0 the best play will be evaluated as.

Yeah. Many people seem to think in 960 there would be just "that one optimal opening line" for each start position, while in reality there should be a humongous amount of possibilities / different systems for each.

And if you consider the vastness of choices in the standard chess starting position, and how many valid options there are, and how long it took to discover them, it is clear that 960 will offer the opportunity for fresh choices for decades to come. A bit like in the beginnings of standard chess, when opening theory was not available.

Yeah. Many people seem to think in 960 there would be just "that one optimal opening line" for each start position, while in reality there should be a humongous amount of possibilities / different systems for each. And if you consider the vastness of choices in the standard chess starting position, and how many valid options there are, and how long it took to discover them, it is clear that 960 will offer the opportunity for fresh choices for decades to come. A bit like in the beginnings of standard chess, when opening theory was not available.

Lichess Multiverse (921K chess): This 921K Variant is the "Asymmetrical Revolution". With an asymmetrical starting position memorising an opening would be a waste of time, especially when you don't know what you will have as a starting position. Press the game button and voila, it becomes a lichess multiverse marathon.

Chess might be better off using asymmetrical starting positions. In Chess960, there are only 960 symmetrical positions, where white and black mirror each other. But what if white and black had completely different starting setups? That would be the end of the copy cats.

This would add a new layer of complexity and strategic depth, pushing players to rely on pure chess understanding rather than memorizing openings. Every game would be a fresh puzzle to solve, requiring adaptability and core principles instead of rote theory.

If we use the 960 symmetrical positions as a reference, there would be 921,600 distinct asymmetrical starting positions (960 for white × 960 for black). That’s a massive increase in variety and a whole new dynamic for Lichess !

This would shift the focus entirely from memorisation to creative problem-solving, making every game unique.

https://lichess.org/analysis/chess960

If many love the idea maybe someone will submit it in for a feature request.

Lichess Multiverse (921K chess): This 921K Variant is the "Asymmetrical Revolution". With an asymmetrical starting position memorising an opening would be a waste of time, especially when you don't know what you will have as a starting position. Press the game button and voila, it becomes a lichess multiverse marathon. Chess might be better off using asymmetrical starting positions. In Chess960, there are only 960 symmetrical positions, where white and black mirror each other. But what if white and black had completely different starting setups? That would be the end of the copy cats. This would add a new layer of complexity and strategic depth, pushing players to rely on pure chess understanding rather than memorizing openings. Every game would be a fresh puzzle to solve, requiring adaptability and core principles instead of rote theory. If we use the 960 symmetrical positions as a reference, there would be 921,600 distinct asymmetrical starting positions (960 for white × 960 for black). That’s a massive increase in variety and a whole new dynamic for Lichess ! This would shift the focus entirely from memorisation to creative problem-solving, making every game unique. https://lichess.org/analysis/chess960 If many love the idea maybe someone will submit it in for a feature request.

could lichess put 960 stuff in editor tool.. and as it has for standard a drop down list.. perhaps a hoverable list of say all minor piece laterally symetric set ups.. if not 960... decent subset that can be reasoned with to make start studying rationally with both brains involved...

I find this current hush hush about setups to be an irrationnal extreme take against the opening theory spolied standard chess current emerges preparation strategy problem... (that it is the entry limiting factor for a generalist or frontal chess learning plan or principle... like for me all i can hope to improve, and actually have curuisity for).

but 960 behind the random dogma veil from traumatized social level compeition rules seems like some kind of attavistic mental blockage.. got my floury words out.. I have some deep conceptual frustration about this.. Feeling crazy to find it so obvious and that I might appear not making sense... really worrying......

could lichess put 960 stuff in editor tool.. and as it has for standard a drop down list.. perhaps a hoverable list of say all minor piece laterally symetric set ups.. if not 960... decent subset that can be reasoned with to make start studying rationally with both brains involved... I find this current hush hush about setups to be an irrationnal extreme take against the opening theory spolied standard chess current emerges preparation strategy problem... (that it is the entry limiting factor for a generalist or frontal chess learning plan or principle... like for me all i can hope to improve, and actually have curuisity for). but 960 behind the random dogma veil from traumatized social level compeition rules seems like some kind of attavistic mental blockage.. got my floury words out.. I have some deep conceptual frustration about this.. Feeling crazy to find it so obvious and that I might appear not making sense... really worrying......

Standard Chess is like a job with well-established routines, while Chess960 offers a similar structure but with a fresh, symmetrical starting point. The Chess Multiverse (921K Chess) takes it a step further. It’s like showing up to work and facing an entirely new project each day, forcing you to rely entirely on your core problem-solving skills and it has an asymmetrical starting point.

Chess960 and the 921K Chess Multiverse aren’t about playing memorized sequences of moves. These variants force players into pure problem-solving from move one. They challenge players to think creatively, rather than relying on historical rote knowledge. Chess needs to move away from memorized openings and rushing through games. It needs a reset, starting with a clean slate and figuring things out from scratch.

The clock needs a reset too. The time pressures between the different versions need to be adjusted. In standard chess, players rely on known paths, allowing for faster play. But in 960 and the Multiverse, the starting position is unfamiliar. Players need more time to think through each type of variant. To make these variants truly viable, they need increment time or additional time compared to standard chess time controls, allowing for deeper thinking and reducing the pressure to rush. Without this adjustment, ratings will always be skewed lower because the clock interferes with the endgame. It’s not about time management when you’re not given the time needed to think things through. It’s like an unreasonable boss making the tempo so fast that people would rather quit than keep up.

Standard Chess is like a job with well-established routines, while Chess960 offers a similar structure but with a fresh, symmetrical starting point. The Chess Multiverse (921K Chess) takes it a step further. It’s like showing up to work and facing an entirely new project each day, forcing you to rely entirely on your core problem-solving skills and it has an asymmetrical starting point. Chess960 and the 921K Chess Multiverse aren’t about playing memorized sequences of moves. These variants force players into pure problem-solving from move one. They challenge players to think creatively, rather than relying on historical rote knowledge. Chess needs to move away from memorized openings and rushing through games. It needs a reset, starting with a clean slate and figuring things out from scratch. The clock needs a reset too. The time pressures between the different versions need to be adjusted. In standard chess, players rely on known paths, allowing for faster play. But in 960 and the Multiverse, the starting position is unfamiliar. Players need more time to think through each type of variant. To make these variants truly viable, they need increment time or additional time compared to standard chess time controls, allowing for deeper thinking and reducing the pressure to rush. Without this adjustment, ratings will always be skewed lower because the clock interferes with the endgame. It’s not about time management when you’re not given the time needed to think things through. It’s like an unreasonable boss making the tempo so fast that people would rather quit than keep up.

massive learning program. .we already have too small a brain for a single set an now some want all possible permutations and not symetrical. C'est une fuite en avant. I have not found any english equivalent expression, and I did listen for it (reading too) for many years..

I am for chess ideas based chess theory, and that is not what Fisher was trying to circomvent, it was the strict turn by turn knowledge specialization.. We don'T need that many other single game random user choice world to make that preparation winning learning strategy at entry in standard chess, become a learning strategy that tumps all others.

Just having a differnt social competition unit of 2 games as the fair challenge, where each side chooses among a subset of 960 or even smaller set, would at some cardinality of that subset make the learning depths first turn by turn knowledge specialization strategy an increasingly feeble learning program..

it might not take that many to make such knowledge only usable for its ideas that might be useful or ratoinally adaptible ideas to generalize to other set ups....

but this complete randomization over such a huge space just to keep fairness, while preventing like an absolutist impossibility that the socical competitive chess culture to go deep lines competition, seems like a hammer for a fly. no offense to the analogy fly...

So going even further in not having any logic possible is not just about preventing that but also any both brain learning of patterns...

unless we really want pure non learnable caluculation power competitions.. That is also legitimate. but it should be clear that it would make for a huge gap in types of chess.. That one is just not curious to find out where the limit was in terms of setup size extending from the standard one.

I bet this is because, nobody can think of chess as sometimes that can be studied many heads at a time. That it always has to be an individualistic preformance improving task. That the only way to improve is by individual cheating with insider knowledge about the some thin streak of single game amist the divergence.

I am glad that there is some pressure from on high, altthough derivative of same mentality I might be pointing at, in that it makes an artificial curiostiy push in the direction of curiosity.. but overshoot much might make it a dud. only valid for already learned standard chess players. A game for the higher strates... is it?

no need for a plebian bottom of the pyramid?

massive learning program. .we already have too small a brain for a single set an now some want all possible permutations and not symetrical. C'est une fuite en avant. I have not found any english equivalent expression, and I did listen for it (reading too) for many years.. I am for chess ideas based chess theory, and that is not what Fisher was trying to circomvent, it was the strict turn by turn knowledge specialization.. We don'T need that many other single game random user choice world to make that preparation winning learning strategy at entry in standard chess, become a learning strategy that tumps all others. Just having a differnt social competition unit of 2 games as the fair challenge, where each side chooses among a subset of 960 or even smaller set, would at some cardinality of that subset make the learning depths first turn by turn knowledge specialization strategy an increasingly feeble learning program.. it might not take that many to make such knowledge only usable for its ideas that might be useful or ratoinally adaptible ideas to generalize to other set ups.... but this complete randomization over such a huge space just to keep fairness, while preventing like an absolutist impossibility that the socical competitive chess culture to go deep lines competition, seems like a hammer for a fly. no offense to the analogy fly... So going even further in not having any logic possible is not just about preventing that but also any both brain learning of patterns... unless we really want pure non learnable caluculation power competitions.. That is also legitimate. but it should be clear that it would make for a huge gap in types of chess.. That one is just not curious to find out where the limit was in terms of setup size extending from the standard one. I bet this is because, nobody can think of chess as sometimes that can be studied many heads at a time. That it always has to be an individualistic preformance improving task. That the only way to improve is by individual cheating with insider knowledge about the some thin streak of single game amist the divergence. I am glad that there is some pressure from on high, altthough derivative of same mentality I might be pointing at, in that it makes an artificial curiostiy push in the direction of curiosity.. but overshoot much might make it a dud. only valid for already learned standard chess players. A game for the higher strates... is it? no need for a plebian bottom of the pyramid?

however, about vertically asymetric IC. Those are also good questions as study perturbations. But as social competition package, they would augment even further the big chess world. Maybe we need to clarify what it is that we want to compete about, and what we allow practice or experienec and reasoning toward learning to get better, to allow improving.

That there may be different factors not made clear. Often stating the goals and questions that we assume shared obviously can shed light on discussion, and let people discuss or argue better on what then becomes arguable.

however, about vertically asymetric IC. Those are also good questions as study perturbations. But as social competition package, they would augment even further the big chess world. Maybe we need to clarify what it is that we want to compete about, and what we allow practice or experienec and reasoning toward learning to get better, to allow improving. That there may be different factors not made clear. Often stating the goals and questions that we assume shared obviously can shed light on discussion, and let people discuss or argue better on what then becomes arguable.

Themed chess games are like starting positions. It's really no different. It forces a person to think immediately, because the prior moves might have been memorised.

Some people like stability and so they live in their chess world with only one opening. While others like the adventures and not the routines and so they are like vagabonds of the chess world, where the starting position is always different.

Themed chess games are like starting positions. It's really no different. It forces a person to think immediately, because the prior moves might have been memorised. Some people like stability and so they live in their chess world with only one opening. While others like the adventures and not the routines and so they are like vagabonds of the chess world, where the starting position is always different.

I have not managed to find https://lichess.org/@/justiniano565 's article online. Anyway, one solution could be to make it chess 958 with position 80 and standard excluded. Basically people are looking for equal positions to play from without having to study theory. Unfortunately the Casablanca Chess event was unsatisfactory for me, playing exclusively from positions that occurred in world championships is actually more restrictive than standard chess. Especially the Evans position was very badly chosen, after some mistakes a typical Evans position was found equal according to engines. However, for white the feasible moves were very easy to find, while black was burdened with a humanly impossible defensive task. Another solution can be to generate equal positions with some moves having been made from any position.

I have not managed to find https://lichess.org/@/justiniano565 's article online. Anyway, one solution could be to make it chess 958 with position 80 and standard excluded. Basically people are looking for equal positions to play from without having to study theory. Unfortunately the Casablanca Chess event was unsatisfactory for me, playing exclusively from positions that occurred in world championships is actually more restrictive than standard chess. Especially the Evans position was very badly chosen, after some mistakes a typical Evans position was found equal according to engines. However, for white the feasible moves were very easy to find, while black was burdened with a humanly impossible defensive task. Another solution can be to generate equal positions with some moves having been made from any position.

Engine evaluations of starting poistions may be considered, especially if we include non-symmetrical but equal random starting positions we we will have around 960 times 800 starting positions.

Engine evaluations of starting poistions may be considered, especially if we include non-symmetrical but equal random starting positions we we will have around 960 times 800 starting positions.

This topic has been archived and can no longer be replied to.