Comments on https://lichess.org/@/nvasquez/blog/learning-from-jacob-aagaard/mfD5nRdZ
Thanks for this. Especially the "Modern engines are crap." paragraph.
As someone who started chess old, I'm continually amazed at how much obvious stuff I missed. I wrote this chess move counting app to help "see all the moves for all the pieces":
https://bescoto.github.io/chess-count-quiz/
If you set it to "all legal moves", then it's straightforwardly forcing you to see all the moves.
The problem with easy tactics (apparently Aagaard's recommendation) is what you said about "the mind lik[ing] to automatically search for familiar moves and lines". Easy tactics may encourage this bad habit, as the tactics are easy exactly because the answer is a familiar line.
> about engine as helper tools (not as gladiator unidimensional task).
I would surmise (english? to mean I would add on top of what you said) that even when not equal, or blunder, e.g., it might be at such non-human depth that this can't be seen by any human.
really crap as human chess analysis or leanring tools.
can't differentiate between shallow blunder or deep blunders. (i.e. human chess vision breadth of evolving reach as we learn, or foresight "predicition" ability, turn by turn or not).
it seems like these guys (well-spoken, well-meaning, language masters, & chess masters) make a sincere effort to understand what we need at our level (i.e., us players below 1800 FIDE) to improve to above 1800 FIDE level, but maybe they don't. perhaps, they just cannot relate. or if they do know what we need, maybe it's not profitable....
@Inter-temporal I have been working to understand what is needed to help "regular players" like myself improve. I believe it's training visualization and tactics... but it's a slog. I have a hard time remembering all my opening lines...
@Graque that's an amazing app! Fritz17 has a training tool that is similar, but i had not seen anything like it before now. very nice!
there is one tiny recommendation i would have...
i think it's helpful to begin the list with the opponent's checks (on our K) and the opponent's captures; and then insert our checks and our captures.
> about https://bescoto.github.io/chess-count-quiz/
good on the source code availability as well. I was gonna ask about the concepts behind the tool. At least there is source code.
And having board analysis exercercies that are of an objective nature, for now, about counting things, is a good idea.
I wonder how many chess board chess theory (not "opening theory"TM) features can be counted or voted on being present or not.
Are you intending on having an open data policy as well (anonymized, as chess is a very individualistic experience minimally social in that it takes 2 persons to accept cooperating to compete, and that might be the maximal sharing intended).
About checks on own kind, I did not check myself, but I agree those are important to see very soon in a position, in ones checklist protocol, and have in the back of the mind in the foresight exploration.
as this possibility trumps many deeper calculations, offering an initiative joker to the opponent, that would make other clever shenanigans of material capture possibility not even possible, this ought to prune ones forward calculation exponential growth toward own chess vision breadth realistic limits (as we develop and learn to improve that breadth not just in size).
Otherwise we can have a wrong checklist protocol order from bad initial condition choice (that happens with lists....).