Munich and I try to resolve which range of problems you should choose at CT. We make use of the metaphor of learning a new language.
We agree that the problems rated below 1500 can be compared to the syllables of words.
When I extend the metaphor, I compare 1500-1900 problems with whole words and >1900 with sentences.
Artificial words like desoxyribonucleic acid are to compare with composed problems with artificial clues you won't encounter in real games.
Take a look at the diagram below.
Black to move.
You can find the solution here.
The compounding patterns of the combination pop up easily enough. And I shouldn't be too worried if you need a minute or two for that in stead of 10 seconds.
Here are the patterns:
- The main weakness of white is the pinned bishop on b4. It calls for an extra attack. For instance with Be7 or Ba5
- White has a discovered attack with Bf8+. That helps you to decide between Be7 and Ba5.
- You can decoy the white king with g4+, hxg4, hxg4+, luring the monarch away from the protection of the white knight, thus creating another target for the black rook.
Both moves are forcing but one is a win and the other is a draw.
This is what I mean with "adding another element to the equation". You cannot learn anything about grammar or move order with only learning a vocabulary by heart.
When do you know you are ripe for grammar? When you don't improve from learning words anymore. That is why we seem to disagree. Munich still improves from learning more words, I want to learn the grammar. We are both right in our approach. But we have a different goal.
In the end the grammar rules are followed implicit. Just like children learn to use the grammar rules in an implicit way. Even the rules of grammar wil eventually take the form of patterns. But these kind of patterns you will not be able to find in low rated problems.
BTW I prefer the use of the abbreviation DNA ;)