Guess what? I would like to know the point of adding the word system to certain opening designations.
From opening puzzles page, i could find 20 such systems cases.
what is systematic in openings with the system tag, is my simple question?
Not all 20 items are different systems though. I saw Colle and London more than once across different tree branches.
I have some exposure to Colle. I have been practicing finding it in opening explorer in game, not very popular compared to c4 white move. I did have some limited London exposure, and it was that subset which had common geometry with Colle (the pawn wedge covering or shielding the light bishop, at some depth).
But browsing in the puzzle page i saw positions attributed to London which were very early, and did not have much geometric signature i could recognize.
So my current set of conceptions, based on that limited experience, about what system means, needs some tuning.
It might have gotten to the point where it has lost on board meaning, and might have been used for other reasons.
I don't want a canonical truth necessarily, i want guesses, educated guesses or hypotheses.. if there is a consensus definition, then please, of course i would like that. (i will google meanwhile).
Certain recurring moves independent of the replies, no fixed order of moves.
@Sarg0n said in #2:
> Certain recurring moves independent of the replies, no fixed order of moves.
That is what i thought. So then I would expect systems with a name attached like London or Colle, to be about specifying what are those "certain pillar moves" or position configurations objectives (same thing right if about own material placement).
Could a system have sub-systems? What is the signature (minimal common set of "certain recurring moves") of London?
For a while I though Colle differed from London by the role given to the queen side B (and perhaps some ordering aboutit). but that is not all there is about London, it cover deeper sequences perhaps, and hence other pillar moves (hypothesis=question).
I am tired. and will come back tomorrow, thanks for quick reply above.
@dboing Opening "Systems" are often used by amateur players, because they use to have 1 thing in common:
You play the same opening moves (more or less) regardless of what the opponent plays - sometimes you have to vary the move orders of course to avoid some blunders. The idea is to cut down the ammount of time to learn lots different openings.
There are several systems and not every one of them is called "system".
A few examples for white:
The White Sniper
Hippopotamus Defence for White
London System
Jobava London System
Larsen's Opening
A few examples for black:
The Sniper
Owen's defence
Hippopotamus Defence
Pirc like move orders
Modern Defence
Have fun!
Actually „system“ is often misleading. A lot of concrete knowledge is essential albeit it doesn’t look like.
This is about learning what system means. not about me learning openings (not just yet or ever, i want to have some lay of the land first). I understand it is an easy assumption to make.
Thank you for the ones not called system, i will look at them as more examples, to gather what system actually means not just what it can do for me.
I am looking at opening like a zoo, maybe. Or a potential ecosystem, so I am looking for trends. openings not for me necessarily but as objets of science perhaps (just trying to explain the perspective of mine, not where i am doing science).
> So some systems do not have that tag. Are there openings with system tag, that should not have the tag? and why?
Concrete knowledge can be about knowing many names of openings, or knowing what is on the board in those openings.
If i can save memory space for my own practice, I know I would fare better with on-board chess logic, than memory enumeration of names and move sequences.
Here asking just for the logic, whether i would use it or not, does not matter. From those who know or have some communicable sense. I can work out the detail of each later. not the point.
> Question to lichess (and users that can answer too):
I also wonder if lichess puzzles by opening themes, have been made exhaustive in the naming, and could avoid the choices in designation that the opening explorer has made. Transposition in that layout, does not prevent counting twice a position in 2 different opening designation.
> So can i count on the numbers attached to each openings to actually be not accounting for transposition, so i would get all the possible opening names for the same opening position.
This is related to my system study (of the designation, remember) in the sense that for example in the opening explorer, only one name designation can happen (something like last transposition something... vague to not make error). So that actual system designation could be occult-ed by another path from non-system designation branching in a position of otherwise also system designation.
word and string of words are terrible means of communication when try to express certain simple mathematical things like tree, and transpositions..... With a graph it would be simpler to say what i just said.
<Comment deleted by user>
My writing bad. Many of my affirmations are actually questions awaiting either confirmation or correction.
Specifically about lichess puzzle and opening themes:
These were questions about whether lichess puzzles opening theme page layout with puzzle population numbers for each opening designation put there, did take advantage of not having to follow a tree anymore and allow counting transposition positions in all the path going into that position**. I mean no need to have overriding names rules for designation priority when dealing with transposition. put the puzzle with that position (before blunder i imagine) in each relevent theme.
My question to lichess, was is that so? then I would be sure that all systems would be at least counted once. Because i assume systems are going to be transposition rich, since move order is not that crucial.
> Not really on topic, but the following "technical note" might explain some past misunderstanding i have had trying to explains somethings.
** (if positions were laid in a network where the edge were the moves, so that transpositions would just be nodes with more than one branch in,, ahh i just figured out my own hidden assumption with the branching into, it was in a graph of positions, not in a tree of moves... this is to those who never understood me when i was trying to explain how the opening explorer handles transpositions). for me chess games are paths on such graph with unique positions connected by edges representing the moves. The typical chess tree is something else. it assumes a root FEN, but it creates nodes as moves are conceived.
The graph view has them all sitting there wait for the moves to connect them. they could exist without the moves displayed.
I guess I naturally and subconsciously see positions first.. like the framework behind alpha-zero.. a state space made of positions, and a set of actions on that space.. something like that. it needs the positions to be considered independently of the moves i think. (pretty sure). so the tree embedding of all openings leading to transposition headache,, maybe that's only an angle of chess openings... That graph layout, another. And i am not that flown off the handle, since it has been used in such engines.
ok wall of text interfering with my simple questions... (well the chess one at least).
are there any designated single sequence in opening tree ECO or other nomenclature, with the system tag that might be questionable as to an reader passing by here understanding of what a system might be.. taking Sargon first attempt at it (which I would find agreeable).
My plan, besides waiting for answers:
(again not canons, but subjective truths tyring to be as objective as one can, synthesis exercices can leaver out corners at first, they don't have to be complete, a wheel...)
Use the flat opening theme lichess puzzle system and population number layout, to find the least apparently transposed "system" designations to submit back here for whether they look system or not (it is about looks too).
I could even go concrete on them, with my question mode approach.. Some people might like that kind of approach.. and participate... don't be shy. I am not in a hurry, this is a long haul line of questioning for me.
separating sub thread here. some might not be curious about this, but still curious about previous.. but for me they are fundamentally related.. If my graph escapade above made your eyes stare with incredulity or boredom, skip this post.
> more technical notes in support of the pre-existing position set versus a constructed tree made of nodes representing decisions reduced to the atomic change called move (which i call SAN move here).
> In the graph case,
the moves are the edges and are actually defining position pairs (which might be another definition of move).
> In the tree case,
the tree new node are actually representing the class of moves i called SAN move. only the minimal change required to obtain the new full position if propagating the initial position of the tree (FEN). Basically that is only the material unit change board coordinates or squares. Everything else on the board could be different (within legal limit sure).
without the almost forgotten FEN assumption, the same SAN move could apply in a different line of initial SAN moves applied to the initial FEN.
And so, if the sequence of decision had started in a different graph path, in the tree it would require creating a new node for the same position. That is a transposition viewed from both "angles". in one case the same position is represented by 2 nodes, in the other the position is unique.
This is no news.. i am just spelling things out, so we can all move around a bit from framework to framework or referential to refential. or representation to representation (of the same XXXX thing).
alpha-zero and Leela chess zero mathematical framework use at its most abstract level (as per the google science paper), and or as i understood it (and still do) something called Markob Decision Processes.. Well those processes need some substrate to work on, and that is the position state space (compatible with my graph minimalist view, can be developped) and a set of actions that take a given position in that space and transition to another in that same space or set. There is more to it, but that is on top (about valuation of policies on such structure, again can be developped. trying to be readable).
The graph i am talking about, and often subconsciously conjure (and i suspect i am not the only non AB engine internal exposed human to do that), is just considering all the actions without even evaluating their potential odds in terms of WDL.
This is generating the legal position space. I don't need to have it complete in mind though, obviously, i can assume existence of things without having to construct all the instances, some of it as a foggy placeholder in mind, but some of it is clearer as my attention requires, or communicating chess things to others requires. That is the concrete things I assume Sargon was mentioning, visible data.
Even the opening explorer is compatible on top of this graph visual support... (that was my mistake assuming perhaps that the graph view was common sense).
