More on calculation
Due to the work of prof. Adriaan the Groot, we nowadays mainly think of patterns when we think of calculation. The more patterns we have collected the better. Some of us (me) will add the frequency of occurrence to that.
The transfer problem is how to teleport the knowledge from one position to the other, we solved by focusing on conceptualization of the solution.
Pruning of the tree of analysis is done by logic. Logic is destructive by nature, and it shows which variations are not worth calculating.
And the latest finding is understanding. In understanding, logic and patterns are combined, while system 1 and system 2 look over the shoulder of attention.
How does this work in practice?
| Black to move |
2r5/4k1pp/p2p4/BpnPp3/4q3/5Q2/PP1K1bPP/3R1R2 b - - 5 28
The pattern at hand is the following killbox:
From this idea we can build the following logical narrative:
- To get the rook from f1 to f7 we need to clear the f-file with tempo
- Ng6+ does the job
- White is now looking for a move that both deals with the check AND covers the mating square f7
- Qxg6 does the job
- Queen g6 is now overloaded. It must cover both mating squares f7 and d6. It must prevent Rf7 and Qd6
- Rf7+ lures away the white queen
- After which Qd6 delivers mate
Comments
Post a Comment