Chess analysis project

[jdart]

I found this on Github:

https://github.com/ChessAnalysis/chess-analysis.

They are using a giant cluster to analyze all the FENs from 5 million or so historical games.

--Jon

[bob]

I found this on Github:

https://github.com/ChessAnalysis/chess-analysis.

They are using a giant cluster to analyze all the FENs from 5 million or so historical games.

--Jon

I assume there is no mathematician among them? That is such a tiny faction of the total chess positions one might actually reach...

[mar]

I assume there is no mathematician among them? That is such a tiny faction of the total chess positions one might actually reach...

Where did they claim they're trying to solve chess?

[AlvaroBegue]

I assume there is no mathematician among them? That is such a tiny faction of the total chess positions one might actually reach...

Where did they claim they're trying to solve chess?

"We are writing a technical report on various statistics of the chessgame database (e.g., number of unique positions);"

I have no idea what a database of games can contribute to computing the number of positions.

[Dann Corbit]

I assume there is no mathematician among them? That is such a tiny faction of the total chess positions one might actually reach...

Where did they claim they're trying to solve chess?

"We are writing a technical report on various statistics of the chessgame database (e.g., number of unique positions);"

I have no idea what a database of games can contribute to computing the number of positions.

Their goal is to analyze every single unique position from every high level game ever played.

It is a big effort.

It is also worthwhile.

Games from sources like TCEC, CCRL and CEGT used a small fixed set of openings. Games from all the games ever played will exercise all the openings.

They are not idiots who imagine that they are analyzing all possibilities.

[jdart]

Still, the objectives are a bit unclear. There are a lot of things you could do with such a database. But which ones they are actually doing .. I don't know.

--Jon

[bob]

I assume there is no mathematician among them? That is such a tiny faction of the total chess positions one might actually reach...

Where did they claim they're trying to solve chess?

I assume they believe this will be significant. But how, I don't follow.

[Ferdy]

Depth 20 may not be enough.

Code: Select all

Objectives

We hope to gather various interesting insights on the skills, ratings, or styles of (famous) chess players. In fact numerous applications can be and have been considered such as cheat detection, computation of an intrinsic, "universal" rating, or the determination of key moments chess players blunder. For instance we would like to answer a question like "Who are the best chess players in history?"

[...]

In average, Stockfish calculates 7 millions combinaisons at depth 20. So, we can think Igrida has, at total, calculates more than 2e15 nodes (~2 000 000 000 000 000).

[Dann Corbit]

Depth 20 may not be enough.

Code: Select all

Objectives

We hope to gather various interesting insights on the skills, ratings, or styles of (famous) chess players. In fact numerous applications can be and have been considered such as cheat detection, computation of an intrinsic, "universal" rating, or the determination of key moments chess players blunder. For instance we would like to answer a question like "Who are the best chess players in history?"

[...]

In average, Stockfish calculates 7 millions combinaisons at depth 20. So, we can think Igrida has, at total, calculates more than 2e15 nodes (~2 000 000 000 000 000).

Actually, depth 20 is pretty useless unless you mini-max the whole mess when you are done.

On high end hardware, SF gets to 20 in a sneeze. So what did we gain?

[bob]

Depth 20 may not be enough.

Code: Select all

Objectives

We hope to gather various interesting insights on the skills, ratings, or styles of (famous) chess players. In fact numerous applications can be and have been considered such as cheat detection, computation of an intrinsic, "universal" rating, or the determination of key moments chess players blunder. For instance we would like to answer a question like "Who are the best chess players in history?"

[...]

In average, Stockfish calculates 7 millions combinaisons at depth 20. So, we can think Igrida has, at total, calculates more than 2e15 nodes (~2 000 000 000 000 000).

I have posted more than one 1-2 trillion node searches here. at 100M, that only needs 2 hours are so. so 2,000 trillion will need about 2000 * 2 hours = 4000 hours or 23 days, using just one 20 core box...