Proposing to expand puzzles’ possible types of scenario

@Hott said in #28:
>
There are many learning purposes to the puzzles system beside showing off how good one is on puzzles..

There is the theme system.. and the relative difficulty control given a player level basic pairing band puzzle challenge offering system I find that given that many themes can be sequential spread in a longer puzzle, and that longer puzzles also following an AND or conjunction type of probablity composition (to succees one needs Theme1=T * Theme2=T * ....).
to fail, one can lose on any of the themes1.

when putting the relative difficulty to the max, one would get longer puzzles (as can one theme on short puzzle be intrinsically ranging all the difficulties?). But there is the deliberate work about the puzzle, that is becomes problem solving more than recognition or memory performance, on can start forging pattern formation in mind. And one loses less rating that way.. as the occasional skill set non-gap ability might make a puzzle win worth all the many losses.

for thematic improvement area diagnostic that is worthless though.. in theory to make the thematic dimension diagnostic optimal one would have to go completely the other way, hoping for my assumption of shorter puzzles on average, and less hypothetic idea that shorter puzzle has less theme items to fail in a bunch..

anyone understand my point about the lack of symetries success vs failure on the thematic spectrum performance rating pedagogical feedback power. (it would be nice in passing if lichess would also put the sample sizes of their thematic performance ratings, or questions marks, since prefering binary variables).

so @Hott , you can do what you need for your priorities about the puzzles. success and failure are just tools... not objectives.

To @OctoPinky, #30.

Yes, both your examples are instances of what I meant. They are also instances of what I want as puzzles at the same time, though those particular two can hardly be entertaining—some filtering would be necessary.

@Hott said in #32:
> those particular two can hardly be entertaining—some filtering would be necessary.

Yes, they were just two examples fitting your requirements. I usually review GM games to find interesting moves or ideas. Many of them look just like a boring, error-free series of moves but there are some mistakes or (like in these examples) positions in which only one choice keeps the balance.

I would like to have more knowledge to build a coherent showcase with explanations.

Here’s a primitive example of an instance of what I talked about this whole thread: