I love the "Let's Check" feature of Fritz and ChessBase. That's primarily because I run a Mac with a Windows 10 virtual machine, which limits my ability to do in-depth analysis. It is incredibly handy to have access to pre-calculated positions or to be able to pass off the calculation work to a faster, more powerful machine (plus, I save battery on my laptop and it doesn't get hot while thinking).
So I got to thinking, what if there was a database that was completely open to the public, and what if every position in that database had a standard level of reliable analysis?
Enter ChessGrid.org.
I created a simple service that runs on next-generation Intel Xeon processors with scalable hash table size.
This means every new position is calculated incredibly quickly and reliably. New positions are then added to the database for recall later.
The idea behind ChessGrid is different than that of a tablebase. A tablebase seeks to record perfect play from a given position using retrograde analysis, whereas ChessGrid is based on theoretical analysis. It is more "human" in the way it looks at a position, and considers the position in and of itself rather than on the play that led up to that position.
Please give it a try and let me know what you think. Specifically, do you think it might be useful once it reaches maturity (hundreds of millions of positions)? Could it be useful for memorizing and recalling positions, or studying famous positions (like the Lucena position)?
Note that at this point it only "knows" what you tell it. I have not "fed" the machine yet, so there are not very many positions. But the fun part is you can teach it something new!
I remember learning that the number of possible positions in the first ten moves is a 63 digit number.
I love the "Let's Check" feature of Fritz and ChessBase. That's primarily because I run a Mac with a Windows 10 virtual machine, which limits my ability to do in-depth analysis. It is incredibly handy to have access to pre-calculated positions or to be able to pass off the calculation work to a faster, more powerful machine (plus, I save battery on my laptop and it doesn't get hot while thinking).
So I got to thinking, what if there was a database that was completely open to the public, and what if every position in that database had a standard level of reliable analysis?
Enter ChessGrid.org.
I created a simple service that runs on next-generation Intel Xeon processors with scalable hash table size.
This means every new position is calculated incredibly quickly and reliably. New positions are then added to the database for recall later.
The idea behind ChessGrid is different than that of a tablebase. A tablebase seeks to record perfect play from a given position using retrograde analysis, whereas ChessGrid is based on theoretical analysis. It is more "human" in the way it looks at a position, and considers the position in and of itself rather than on the play that led up to that position.
Please give it a try and let me know what you think. Specifically, do you think it might be useful once it reaches maturity (hundreds of millions of positions)? Could it be useful for memorizing and recalling positions, or studying famous positions (like the Lucena position)?
Note that at this point it only "knows" what you tell it. I have not "fed" the machine yet, so there are not very many positions. But the fun part is you can teach it something new!