ChessPub Forum - ECO classification scheme is woeful

(UTC)

ChessPub Forum › General › General Chess › ECO classification scheme is woeful

(Moderators: RoleyPoley, proustiskeen, trw)

‹ Previous Topic | Next Topic ›

Pages: 1

Topic Tools

ECO classification scheme is woeful (Read 2339 times)

DrKibzwang YaBB Newbies Offline The execution soon gets worse than the threat. Posts: 26 Joined: 01/27/07	Re: ECO classification scheme is woeful Reply #3 - 10/10/11 at 17:17:16	Post Tools	Print Post
	In another discussion on Kotronias's recent book, several people have commented that the book is not systematically organized and has masses of variations. Dink Heckler adds "chess is very concrete nowadays, so perhaps organising principles are so last century..." It's a funny aside, but possibly true. Maybe we don't need much pre-prepared organization any more in our chess information. If you think about the whole academic discipline of information/library sciences, these people used to spend a lot of time organizing and discussing classification schemes for human knowledge. The assumption was that we needed something brief - an index number or key - to look up the main content. This, of course is what ECO and NIC codes do/did for chess openings. But library search has been revolutionized by free-text lookup, with Google being the best known example. We can use attributes like "keywords", "author", etc. to refine our search, but we can generally find exactly what we want by typing in a phrase and looking at the first 10-20 hits. And we can get the online text, not just a bibliographic pointer. Maybe this is what should happen to openings: instead of revising the codes and requiring people to learn a new set, which will always diverge from contemporary practice, what we need is a kind of 8x8 spatial Google that will find similar positions. Chessbase, of course, does a bit of this, but forces users to specify piece positions and wildcards so explicitly that it's a bit of a pain to do. (I know most of the 500 ECO codes, btw, which is a bizarre and sad statement coming from a terminally 1800-2000 player in his 50s. I'm not about to learn another 1000.)


	IP Logged

kylemeister God Member Offline Posts: 4989 Location: USA Joined: 10/24/05	Re: ECO classification scheme is woeful Reply #2 - 10/07/11 at 17:39:53	Post Tools	Print Post
	I'm reminded of Andrew Soltis, a few decades ago, envisaging something like ECO codes for pawn structures.


	IP Logged

Smyslov_Fan God Member Correspondence fan Offline Progress depends on the unreasonable man. ~GBS Posts: 6902 Joined: 06/15/05	Re: ECO classification scheme is woeful Reply #1 - 10/07/11 at 16:15:27	Post Tools	Print Post
	ECO is still extremely useful. Nunn's Chess Openings follows ECO almost exactly. The more modern problem with ECO classifications is that it doesn't address systemic openings such as the Hedgehog. So far, attempts to replace ECO have mostly made the same basic assumptions as ECO. Hence, the New In Chess (NIC) classifications aren't substantially better than ECO. A useful classification system would need to be able to be easily cross-referenced, not only for direct transpositions but for families of openings. I can't really visualise such a classification system. If I could, perhaps I could make a few pennies by copyrighting it!


	IP Logged

Markovich God Member Offline Posts: 6099 Location: Columbus, Ohio Joined: 09/17/04	ECO classification scheme is woeful 10/07/11 at 15:34:41	Post Tools	Print Post
	I was working on the ECO classification scheme for a website I'm building, and I was struck by its woeful inadequacy. It reflects what people in 1980 thought about the importance of given variations, not the frequency with which given positions appear today, or even over all of recorded chess history. Just to confine myself to volume A, the classification of the Modern Benoni carefully subdivides the Modern Variation as then interpreted. But this variation is very little played any more. Still worse are the Stonewall Dutch classifications, which assume that Black's KB is going to e7! The whole thing sucks majorly. What's really needed it an algorithm to produce "key positions" from a game collection encompasing past and present practice. The appearance of a given key position in a given game would be the basis of its classification. You employ the principle that the last appearing key position is the one that determines a game's classification. To determine the key positions, you would need something to balance frequency of occurrence and exhaustiveness. Say you want 500 key positions to summarize the entirety of chess practice (there are 500 ECO classifications); the question is, how to find 500 positions, from which all possible chess games descend, and each of which accounts for approximately 1/500th of all played games. That's a difficult analytical problem, but I would not think it too difficult to solve with a big database and a powerful computer -- once one cooked up an algorithm. There are some issues. For example, the position after 1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nc3 exd5 5.cxd5 g6 6.e4 Bg7 8.f4 O-O 9.Nf3 would probably have to be a "key position," though it occurs in only 6,000 positions in my 6.1 million game data base (12,000+ games per position ) would be the average for 500 key positions. Hmm...

	The Great Oz has spoken!
	IP Logged

Pages: 1

Topic Tools

‹ Previous Topic | Next Topic ›

Bookmarks:

« Board Index ‹ Board Top