Educating system 2
System 1 educates itself by looking over the shoulder of system 2. This implies that the only thing that we have to worry about is to educate system 2 well. If system 2 occupies itself with useful chess logic with a high frequency of occurrence, both systems will thrive.
How does that look like?
 |
| Black to move |
8/5N2/6P1/3p4/B7/3nPkP1/1r6/6K1 b - - 0 1
[solution]
The first thing is that you have to see what the position is about. You need to have the checkmate patterns under your belt. Only when you SEE the salient cue "Arabian mate", you can build some logic around it. 3 movers are fit to learn the patterns. But after you master those, 4 movers are very good to build your chess logic.
Without the salient clue "Arabian mate", you will have to resort to trial and error. Which probably will fail which these 4 movers.
The Arabian mate is usually in the corner of the board. But the walls of the killbox can be anything. As long as the enemy king cannot pass the walls. In the above diagram, you have to build a wall with your own king before you execute the mate. Actions:
- Bring your king in the position where it will act like a wall
- Chase the enemy king to f1
- Execute the mate
You educate system 2 by analyzing the solution. There is no point in finding the solution yourself. Of course there is nothing against finding the solution yourself, if you have enough time and are addicted to such things. But don't get distracted from your goal: educate system 2 to apply logic based on the salient cues that you SEE.
Examining the position for salient clues is a collaborative effort by both Systems. There is an enormous [HIDDEN] “back-and-forth” going on that we [UNCONSCIOUSLY] gloss over. While explaining what’s “REALLY” going on in our minds.
ReplyDeleteMy System 2 first looked at the instructions (“Black to move”; thank you for leaving out the number of moves) and inferred (solely from that datum) that there is a win (or at least a draw) for Black. Perhaps System 2 (or was it System 1?) also “suggested” that the solution requires “urgency” since White can promote the g6-Pawn in two moves, so “WATCH OUT!” System 1 was asked to find some matching pattern, ANY “matching” pattern, in long-term memory that leads to that goal.
The most recently encountered pattern(s) get a higher priority as THE potential “solution”, even if it does not match as closely as a different pattern not seen in a while. System 2 must evaluate the suggested pattern (the one that readily comes to mind), and either accept it or reject it as appropriate; that requires looking at some variations, usually short ones, to confirm if the pattern can be reached or not. If the pattern cannot be reached, the pattern is either rejected or a modification to that pattern is suggested, and System 1 is put back to work.
My first conscious “pattern” was a corridor mate: move the Black King over to g3, creating the “corridor” in conjunction with the edge of the board. System 2 took a look and said that it might be possible. Can White escape from the “corridor”? Yes, perhaps via f1 → e2. No, that escape doesn’t work: 1… Kxg3 2. Kf1 Rf2+ 3. Kg1 Re2 4. (ANY) Re1#.
Up to this point, I didn’t NOT “see” the potential Arabian Mate pattern!!
The practical mind (System 2) advised that once ONE solution has been found, that settle the issue – Black is going to win. But it’s a PROBLEM, not a game, so EVERYTHING must be examined to the end. Up to this point, I hadn’t considered any other possible “escapes” which do not involve getting the King out of the killbox zone.
With the Black King now sitting on g3, System 1 (I presume) “suggested” looking at promotion of the g6-Pawn, as a defensive try, since it only takes two moves to “queen” – WITH CHECK. System 2 nixed that idea because Black gets there first via forcing moves: 1… Kxg3 2. g7 Rb1+ 3. Bd1 Rxd1#.
What about moving the Knight? System 2 nixed that “solution” space simply by observation: the Knight cannot give check, and the first variation (corridor mate) still works.
Finally, we (me and my best-buddies, System 1 and System 2) came to the last possibility of “escape”: moving the White Bishop (with two possible locations) to block the mate.
1… Kxg3 2. Bd1 Rg2+ 3. Kf1 and System 2 STALLED! I did “see” the Arabian-like mate pattern here, but didn’t (initially) “see” that the Black King had to “seal” the killbox FIRST. 3… Kh2 4. (ANY) Rf2#.
The other Bishop move apparently merely delays the inevitable: 1… Kxg3 2. Bc2 Rxc2 3. Kf1 OOPS! How do we keep the killbox closed, boys? Since I “saw” the idea of using the Black King to “seal the deal” in the previous variation, it was not hard for one of the two Systems (I really don’t know which one it was) to “suggest” 3… Kh2 4. (ANY) Rf2#.
Aren’t we “blessed” that we don’t have to logically work through all that verbiage to find a solution?!?
Now ask yourself what is needed to alleviate the task for system 2. What should system 1 come up with? What should it absorb from this position. All tasks for system 2 are in fact pretty trivial. Create a sealed killbox, deliver mate. The only thing to look out for is whether white can prevent these tasks OR start a counter attack.
Delete