World Championship Fischer Random

[Tom O'Donnell]

If having advantage of second placement isn't enough to counterbalance the advantage of first move, then allowing the players to move in alternation, but with pairs of placements, might equal it out a bit more.
Example: White places two pieces, then Black places two, then White, etc. In any event it decreases the disadvantage to Black and makes it more likely that the better player will win with either colour. Also makes it very hard as White to just outright play for a draw.

[Pargat Perrer]

If having advantage of second placement isn't enough to counterbalance the advantage of first move, then allowing the players to move in alternation, but with pairs of placements, might equal it out a bit more.
Example: White places two pieces, then Black places two, then White, etc. In any event it decreases the disadvantage to Black and makes it more likely that the better player will win with either colour. Also makes it very hard as White to just outright play for a draw.

Ok, I see what you are getting at. But I have to wonder, is it really advantage to have the second placement (i.e. to be Black in the setup phase)?

It seems to me White always has more options than Black, if it goes White then Black then White then Black etc.

Now when placements first start, the sheer number of options isn't giving either side much advantage, but once you get down to let's say the last 3 pieces each and both of you have kept 3 distinct pieces (neither player has 2 of any piece left to place), it looks like advantage to White.

This means when White goes to do his 6th placement, he has 36 possible setups to consider( 6 arrangements for White times 6 arrangements for Black). This is a reasonable number for White to actually visuallze (if he's good at doing that) and make a very strategic or even tactical choice. Once White makes his or her 6th placement, Black has only 12 options (6 arrangements for Black, only 2 left for White). It means Black has less options to consider, and perhaps it also means that White eliminated many of the 36 options that were best for Black.

Now when both have made their 6th placement, there are only 4 remaining possibilities. White can easily visualize these and has the greater choice than Black. Once Whte makes his 7th placement, his 8th placement is also assured so Black has only 2 placement choices.

So when it really matters and visualization becomes possible, White has the greater number of choices and can limit Black's options.

If this is right, then it seems White's advantage ought to be increased rather than decreased. This would also be the case with the idea I proposed, but I still like the idea because it realy increases randomness, as the two players cannot agree before the game what arrangement to setup, for example. Neither knows what is going to come up in the random placement of 5 of their pieces.

Perhaps the solution then is to allow Black to make the first placement and then White, all way through. Then White still gets first actual move of the game.

[Brad Thomson]

It would probably make sense for both players to avoid Bishop placement, or at least try to, until they see what their opponents do with their Bishops (Sometimes this principle applies to castling in the classical game). At some point White will have only two squares left, each could be either colour, and therefore will have committed their Bishop placements whether they were already down or not, while Black will still have three squares left, which could be two of one colour and one of the other. Thus, Black will be able to either balance, or imbalance their Bishops based upon having the advantage of seeing what White has already committed to. Does this give Black at least equality, or the outright advantage? Or is this naive?

[Tom O'Donnell]

Ok, I see what you are getting at. But I have to wonder, is it really advantage to have the second placement (i.e. to be Black in the setup phase)?

It seems to me White always has more options than Black, if it goes White then Black then White then Black etc.

Now when placements first start, the sheer number of options isn't giving either side much advantage, but once you get down to let's say the last 3 pieces each and both of you have kept 3 distinct pieces (neither player has 2 of any piece left to place), it looks like advantage to White.

This means when White goes to do his 6th placement, he has 36 possible setups to consider( 6 arrangements for White times 6 arrangements for Black). This is a reasonable number for White to actually visuallze (if he's good at doing that) and make a very strategic or even tactical choice. Once White makes his or her 6th placement, Black has only 12 options (6 arrangements for Black, only 2 left for White). It means Black has less options to consider, and perhaps it also means that White eliminated many of the 36 options that were best for Black.

Now when both have made their 6th placement, there are only 4 remaining possibilities. White can easily visualize these and has the greater choice than Black. Once Whte makes his 7th placement, his 8th placement is also assured so Black has only 2 placement choices.

So when it really matters and visualization becomes possible, White has the greater number of choices and can limit Black's options.

If this is right, then it seems White's advantage ought to be increased rather than decreased. This would also be the case with the idea I proposed, but I still like the idea because it realy increases randomness, as the two players cannot agree before the game what arrangement to setup, for example. Neither knows what is going to come up in the random placement of 5 of their pieces.

Perhaps the solution then is to allow Black to make the first placement and then White, all way through. Then White still gets first actual move of the game.

Without actually testing it, I cannot say it's 100% certain that Black will actually secure an outright advantage enough to negate the advantage of first move. However, since chess is a game of information and White is the first to reveal the information, yes, I would be very certain getting to be second to make a placement is an advantage.

If White had to place all eight pieces first, I would imagine Black could secure a pretty sizable advantage when arranging his counter-setup.


As an experiment, I used the lichess engine to compare:

White: standard starting position.
Black: BKNRRQNB - not claiming this is optimal, it was just the first "good" (imo) setup I thought about.

Lichess engine gives Black a significant advantage of -2.2. For comparison, it gives the standard placement for both players at +0.3.

In this case it is like poker. Being first to act is a serious disadvantage.

[Hans Jung]

Hikaru Nakamura won the Fischer Random world championship. by beating Nepo in armageddon.

[Hans Jung]

I really enjoyed the knight play in the first game between Nakamura and Nepo.
Who was the most creative player in this world champioonship? I think Nakamura, although Magnus opening move piece sacrifice as black against Nakamura was eye popping. Although after thinking about it I guess that would be considered opening prep as Im sure Magnus came up with that before the game.

[Pargat Perrer]

Without actually testing it, I cannot say it's 100% certain that Black will actually secure an outright advantage enough to negate the advantage of first move. However, since chess is a game of information and White is the first to reveal the information, yes, I would be very certain getting to be second to make a placement is an advantage.

If White had to place all eight pieces first, I would imagine Black could secure a pretty sizable advantage when arranging his counter-setup.


As an experiment, I used the lichess engine to compare:

White: standard starting position.
Black: BKNRRQNB - not claiming this is optimal, it was just the first "good" (imo) setup I thought about.

Lichess engine gives Black a significant advantage of -2.2. For comparison, it gives the standard placement for both players at +0.3.

In this case it is like poker. Being first to act is a serious disadvantage.

It's quite an interesting question. As you say, Black has more information at each step, but White has more options at each step. I honestly don't know which is more important.

Perhaps it's a mixed bag, where maybe for the first 5 placements for each side, Black has the advantage because the extra options for White don't really come into it, then as we get down to the last 3 placements where White starts with 36 options, maybe his extra options level the advantage and maybe even give White the advantage.

I could write a short Python console program that randomly sets up 5 pieces for each side, with the following condition:
- One each of Knight, Bishop and Rook must be placed for each player, ensuring that neither player has 2 of any piece type to place

And as suggested, the King does not have to go between Rooks and there is no castling. Each player then makes selections, but rather than playing a game, we feed the FEN (with "-" for the castling options) into Stockfish and see how it evaluates the position after some depth of ply search. This could be repeated many times and maybe a pattern would develop where Black actually does bring down White's opening move advantage.

EDIT: of course the 2 players making selections should be as equally rated as possible. I am a terrible chess player so I could be the one providing the 5 piece arrangements, and the 2 players then make selections with maybe a time limit involved. This could be time-consuming since many samples would be needed, with each player getting equal times as White and Black.

[Neil Frarey]

Hikaru Nakamura won the Fischer Random world championship. by beating Nepo in armageddon.

Yesss, sensational result! ... in Reykjavik, Iceland, 50 years after another American Bobby Fischer took down another Russian Boris Spassky ... no less, ha!

[Tom O'Donnell]

It's quite an interesting question. As you say, Black has more information at each step, but White has more options at each step. I honestly don't know which is more important.

Perhaps it's a mixed bag, where maybe for the first 5 placements for each side, Black has the advantage because the extra options for White don't really come into it, then as we get down to the last 3 placements where White starts with 36 options, maybe his extra options level the advantage and maybe even give White the advantage.

I could write a short Python console program that randomly sets up 5 pieces for each side, with the following condition:
- One each of Knight, Bishop and Rook must be placed for each player, ensuring that neither player has 2 of any piece type to place

And as suggested, the King does not have to go between Rooks and there is no castling. Each player then makes selections, but rather than playing a game, we feed the FEN (with "-" for the castling options) into Stockfish and see how it evaluates the position after some depth of ply search. This could be repeated many times and maybe a pattern would develop where Black actually does bring down White's opening move advantage.

EDIT: of course the 2 players making selections should be as equally rated as possible. I am a terrible chess player so I could be the one providing the 5 piece arrangements, and the 2 players then make selections with maybe a time limit involved. This could be time-consuming since many samples would be needed, with each player getting equal times as White and Black.

In the placement phase, the players have an equal number of options.

On the first placement, White has a choice of five different units (KQRBN) and eight squares he can place them. Then Black has the exact same options when it is his turn, but he also has the information about White's first placement (e.g. he knows there is a B/g1, or N/f1 or whatever piece was placed), and can adjust accordingly. Subsequent placements would be the same, except in each case Black would have a little more knowledge than White before he makes his placement by virtue of White having already placed one more unit. The sum of all of these small amounts of knowledge is worth "something" obviously, but how much, I have no idea.

Perhaps this makes it clearer:

Let's say White and Black both independently decide to place their King and one other piece last. White has a1 and h1 open for his K, Black has g8 and b8. White decides to place his K on a1 (or place the other piece on h1 as that makes no difference). Now Black has a chance to either place his K on the opposite wing, or he can choose to place his K on the same wing. I suspect this is a significant advantage.

[Pargat Perrer]

In the placement phase, the players have an equal number of options.

On the first placement, White has a choice of five different units (KQRBN) and eight squares he can place them. Then Black has the exact same options when it is his turn, but he also has the information about White's first placement (e.g. he knows there is a B/g1, or N/f1 or whatever piece was placed), and can adjust accordingly. Subsequent placements would be the same, except in each case Black would have a little more knowledge than White before he makes his placement by virtue of White having already placed one more unit. The sum of all of these small amounts of knowledge is worth "something" obviously, but how much, I have no idea.

Perhaps this makes it clearer:

Let's say White and Black both independently decide to place their King and one other piece last. White has a1 and h1 open for his K, Black has g8 and b8. White decides to place his K on a1 (or place the other piece on h1 as that makes no difference). Now Black has a chance to either place his K on the opposite wing, or he can choose to place his K on the same wing. I suspect this is a significant advantage.

Sorry, I used the wrong word entirely. White doesn't have more options than Black, that is correct. The word I should have used was "scenarios" where a scenario is a potential complete starting position. If the placement choice goes White then Black then White then Black etc, White always has on a given turn more scenarios to consider than Black.

But I am not certain that this scenarios advantage could be advantageous overall. I don't think it really matters until the last 3 placements for each player, where White has 36 scenarios to consider (a reasonable number) and then Black will only have 12.

On the other hand, it could mean that White uses up more clock time than Black before the actual game proceeds. So Black could be starting each game with a significant time advantage.