Don't use Public Lichess Openings Studies To Learn Openings

Studies are one of the best and unique feature of Lichess. At least i don't remember seeing anything like this on other chess related websites at such quality, without having to pay. (Some websites even charge for downloading PGN of your own game, while another one request money to request SF analysis etc.)

This simply means Studies are one the USPs of Lichess and could only benefit from improving it even more.

My suggestion would be similar to #37, for higher quality, a human intervention is inevitable, since algorithms usually derive a value of quality from quantity. It is hard to automatise. Similar to puzzles, computer generated puzzles rarely come close to the quality of chess compositions.

On the video website "VIMEO", there is a section called "Staff Pick", where original and high-quality videos are picked by the staff. So we can have a similar section on lichess with studies suggested and selected by Lichess itself or a group of people, selected/formed and trusted by Lichess to review Studies specifically and decide. (I'd volunteer myself for such task as well, as i have high interest in studies), so this is not an extra task for devs.

The section can have 100-200 Studies collected and we can label them as "Beginner", "Intermediate", "Advanced", "Original", "Variants", "Opening", "Studies & Problems" etc. Because there are also very instructive and high-level studies in many variants, an example:

I'm also reluctant to suggest the ability to submit studies by users as in practice, i believe this will eventually be exploited in a sense that everyone will just submit their studies and lot of work.. For that, as #37 said, we have no choice to but hope for those studies to be discovered by the staff.

To sum up, a Lichess could maybe select a group of people (could be a mix of titled players, strong players, very active players who like to curate stuff etc.) for this task and they can simply discuss and review the options. For example this can easily be organised on a separate Discord Channel for example on the Discord Lichess Server etc. I believe it is not a huge task as it sounds, since there are already powerful networks established inside Lichess community and knowledge gets shared quickly. Studies could be added to the collection with time without pressure of deadline. To end up with 100-200 Studies shouldn't be that hard in such case. But the reward would be very valuable in my opinion and a something that no any other chess website has. Lichess could create a difference one more time with such initiative :)

Titled players curating lichess studies, or even making a few sounds like the right idea. I am guessing it will be a volunteer effort, without a pay for these master players. I think there are enough patrons who are strong players on lichess who will be willing to give some of their time, to curate studies.

The main post on this discussion was critical, and I agree the most starred studies are not necessarily the best ones, and a beginners can get lost when they are sorting studies in this way. If new players get this studies which are curated by titled players it will be super useful, and lichess will serve a purpose which no other chess resource has done previously.

I also think the same as @visualdennis. We should have studies labelling as 'Beginner' 'Intermediate' and so on. It is very useful for other players to know which study to choose.

i agree

#37. Good idea, but this is putting a lot of work on the inverted tip of a pyramid. Since we are concerned about content quality, a mixed protected/population based tagging system (like in the new puzzle system), could include some quality grid, meaningful enough so that superficial social media gregarious instinct reflexes à la "like", be them a source of short term happiness, would not cloud the quality signal (and avoid stock-market-like positive feedback loops runaways).

That would distributed the human thinking load, and still offer form some responsible oversight (not dictatorship, of course, more of observing than dictating, providing guidelines, for such structure to sustain itself with user-base turnover (is there?).

#38 (forgot, the one with the illusion that experts are good teachers as a logical statement). I learn from the others struggle more than from divine source dictation.

I would add though that i don't think sorting lichess studies by quantity is a bad idea. Almost all good studies I have found were over 40 chapters, and if you are making an extensive study on the theory of an opening, the limit of 64 chapters is not even near enough (personal experience). Looking through the most popular studies, most of them are around 7-15 chapters even one on the Sicillian Najdorf (

Well, yes. quantitative metrics does (EDIT: Not!) cloud qualitative ones, if both exist. and if the choices of metrics is well made. (that is general, i guess, but helpful, as it is also not general because it is general... which does not mean vague automatically, sorry, other thread bleeding through here).

Opening tree vast knowledge some call theory, is not the only topic in chess that would need proper chapterization (so that chapter titles themselves be also meaninful, not just adresses in an ordered sequence. Small brain humans like chunks.

Also, the responsive UI (not questionning that, but factuating), has made module windows for text and chapter etc... with certain sizes, that make navigation in a long sequence of chapters without threading or nesting difficult if all the chapters are part of a whole, and not just an inventory, where the ordering has no meaning.

So I know some tree aware study creators have made with that constraint by embedding those extra semantic layers using many web-linked studies.... That is a lot of work, the creator can navigate having been at the source of the categorizations, but then the semantic structure is not visible at one glance. or it can be, but not while in a study. you need a TOC consecrated study. (and one can imagine the same problem there, so only one depth of semantics? etc... ). But it is possible with extra non-lichess tools and work to start building some workable own clones subsets out of such system. But that is a LOT of work from the creator, and will not be a social media hit. it is for serious and curious, diligent, and steady. anyway not my point.

so I wonder how costly collapsible objects are at the UI level, anywhere in lichess (mostly the analysis page ecosystem, studies being part of that , or the core of it)?

augmenting the maximum size of a 1D sequence of chapter, does not seem a solution (opinion).

I agree with @visualdennis it makes sense to improve it

#46. I guess there is a problem at the chapter granularity level in common to many problems mentioned here.
The population feedback being one counter of likes assigned to the study.

This might pressure creators to actually make more studies than heavier studies. But not all topics can be spread that way, might be your point. otherwise if they can, then one can consecrate the final chapter as the teleport mechanism to the next one(s) or related ones. Note that this can sustain many inter-study network topology (chain, tree, web).

It is when topics would rather be seen together because of content, I think, that the flat 64 limit might be frustrating, right?

its not supposed to be a comprehensive resource, it is only an introduction to the opening, and i also strongly doubt you need 100s of catalan variations when you are only 1800 and no one playing into the theory