2 Moves Openings Book

[fauzi]

hehe no I'm not from Morocco , I'm from Jordan

[Lyudmil Tsvetkov]

Hi again.

Here is the final version of the 2 moves book, consisting of 1000 unique positions: http://www.mediafire.com/download/jskag ... LT_1000.7z

The download includes pgn, abk, bin, epd formats.

I must say that all the credit goes to Adam Hair who did the book in its first, second and third version, but also provided the different formats available! An enormous thanks.

I hope now Larry will be happy.

Please enjoy.

PS. If anyone wants to convert the final version to other unavailable formats and post the download links, that would be great.

[fauzi]

I noticed that most of the main lines are missing , is that done intented?
For examplein French : e4 e6 there is no d4 answer, or in sicilian e4 c5 there is not Nf3 ...
I think it would be MUCH better if You introduce the main lines and remove some very weak moves like f3 or others that I found , so that the test will be much more fair

[Adam Hair]

I noticed that most of the main lines are missing , is that done intented?
For examplein French : e4 e6 there is no d4 answer, or in sicilian e4 c5 there is not Nf3 ...
I think it would be MUCH better if You introduce the main lines and remove some very weak moves like f3 or others that I found , so that the test will be much more fair

Originally, the book had these three lines:
e4 e6 d4 c5
e4 e6 d4 d6
e4 e6 d4 d5

The book also contained:
e4 c5 Nf3 Nc6
e4 c5 Nf3 d6
e4 c5 Nf3 e6

However, the book also contained lines like d4 d6 e4 e6 and Nf3 c5 e4 Nc6, which result in the same positions as e4 e6 d4 d6 and e4 c5 Nf3 Nc6. It happens that d4 and Nf3 appeared earlier in Lyudmil's book than e4 and survived the filtering process (removing transpositions so that each opening position is unique). Thus, many of the lines starting with e4 might be missing. However, they are still represented in the book because they resulted in the same position as an earlier move sequence.

I suppose that this makes the book less pleasing aesthetically, but it makes no difference for engine testing. However, if anybody would like to reorder the moves so that the more common move sequences are retained, I can breakdown the pgn and put it into a spreadsheet. Then the order can be manipulated and eventually put back into a pgn.

[Lyudmil Tsvetkov]

I noticed that most of the main lines are missing , is that done intented?
For examplein French : e4 e6 there is no d4 answer, or in sicilian e4 c5 there is not Nf3 ...
I think it would be MUCH better if You introduce the main lines and remove some very weak moves like f3 or others that I found , so that the test will be much more fair

Originally, the book had these three lines:
e4 e6 d4 c5
e4 e6 d4 d6
e4 e6 d4 d5

The book also contained:
e4 c5 Nf3 Nc6
e4 c5 Nf3 d6
e4 c5 Nf3 e6

However, the book also contained lines like d4 d6 e4 e6 and Nf3 c5 e4 Nc6, which result in the same positions as e4 e6 d4 d6 and e4 c5 Nf3 Nc6. It happens that d4 and Nf3 appeared earlier in Lyudmil's book than e4 and survived the filtering process (removing transpositions so that each opening position is unique). Thus, many of the lines starting with e4 might be missing. However, they are still represented in the book because they resulted in the same position as an earlier move sequence.

I suppose that this makes the book less pleasing aesthetically, but it makes no difference for engine testing. However, if anybody would like to reorder the moves so that the more common move sequences are retained, I can breakdown the pgn and put it into a spreadsheet. Then the order can be manipulated and eventually put back into a pgn.

Hi Adam.

I do not think this would be necessary, as you rightly point out that what matters for engine testing are positions and not sequences. You will always have to cut something, be it e4 or d4.

[Lyudmil Tsvetkov]

I noticed that most of the main lines are missing , is that done intented?
For examplein French : e4 e6 there is no d4 answer, or in sicilian e4 c5 there is not Nf3 ...
I think it would be MUCH better if You introduce the main lines and remove some very weak moves like f3 or others that I found , so that the test will be much more fair

As Adam already pointed out, the arising main positions are there, but some sequences have been removed. I think this is just fine, considering that you need positions.

I have also excluded some mainstream gambits and countergambits, like the Albin and Falkbeer, as at least personally to me it is not clear what happens there. This is not a strictly theoretical book, but the aim was to provide some variety of play by excluding positions that might be lost.

Some say Fischer won many games online, starting with 1. f3.