Upgrade to Chess.com Premium!

Descriptive Chess Notation

  • 12 months ago · Quote · #21


       Done & dusted - for now.   "Chess Translator and PGN Manager 1.05" is downloadable.
       "Mate (Checkmate, Superior Legal Mobility, or Stalemate) Search"
       at some future date.   I need a short sabbatical.
       What do we have?   Any number of PGN files can be compiled (memory permitting).   The compilation can be filtered by any number of Tagname - Tag pairs (inclusive OR).   It can be written out after changing Tagname filters to create different collections.   It can be written (but not compiled) using Opening and/or Endgame filters as well, to create different collections.   The output PGN's can be produced in various formats, Algebraic, Descriptive, Sorted Algebraic, Ordered Descriptive (sorting based on openings), and a concise sorted algebraic format for indexing to find any opening variation.   They can be written in Export Format or include an indexing Tagname - Tag pair such as [Game "11057"].
       A compilation can be composed, inspected, and/or edited from the keyboard, or using menus, with ease.   The inspection option
       "14 REMOVE duplicate games from CURRENT COMPILATION"
       is facilitated by a colour coding which identifies differences in game moves and rosters easily.   The facility also does much of the work.   Moves will quickly be seen to vary somewhat between sources of the same game.   Also, there are many helpful settings available.  For instance, the compiling of a game, and its translation in progress, can be viewed on a display board, at a reduced rate, or toggled into high speed (without the display) while in progress.   FEN positions are no problem.  And much more.
       As an interesting aside, the Endgames comprised in any particular game, are stored as a set.   By using N=13, B=15, R=20 and Q=39, all piece reductions (not pawns) create a unique value.   Considering P=4, the hierarchy given in Jose Capablanca's "Value of the Pieces" from "Chess Fundamentals" is automatically created, an assessment supported by Reuben Fine's "Basic Chess Endings" tome.   It is only upset by promotions yielding 3 of the same piece type, which is saved as such without further distinction.   Of course the peculiarities of the Pawn can render it worth more than a queen.   Also, the hierarchical values of positional anomalies (eg: Q & N, B vs N, etc) must be ignored here.
       There are some good PGN viewers downloadable, even for free, such as "Haundrix Chess".   It asks for a donation.   The single game PGN packs generated by the option
       accommodates its unique requirements.   I hope my keyboard driven program doesn't cause too much frustration before you master it.   You may eMail any complaints via the
       "18 HELP"
       option, offer ideas, or report bugs.   Thank you all for your interest.

  • 12 months ago · Quote · #22


    Besides any reported bugs in due time, be aware that 1.05 does presently experience memory limitations around 2000 game compilations for normal PC's, best checked by recompiling the produced PGN.   A future refinement will relegate much storage to hard drive, practically eliminating this limitation.

  • 6 months ago · Quote · #23


    Hi Folks.   Back from Scotland & Greece.   Trying to get my head around "Chess Translator & PGN Manager" once again.    It makes endgame study a breeze, so I've decided to include about 2010 openings.   These may be listed in several formats, turned into PGN files, or have their names included in PGN games as initial comments in the movetext.   The Openings will be identified by their final positions, so that transpositions will be no problem.   The ECO tag will be adjusted A00 to E99, but the names will include subvariants of these.   Whoever compiled the codes has had a mighty job.   They've made this easy for me.   CT&PM Version 1.14 is out soon.   I'm even including little diagrams in the PGN of the final game positions, and they will still recompile.   I'll give you all a cooey soon as it's done.   Seems this is the way chess study is going, and the games are already out there.

  • 6 months ago · Quote · #24


    This is a sample from many possible formats (5x2x2x2x2) by CT&PM Version 1.14 (somewhat stripped of spaces to preserve formatting.   Only Courier New font will preserve spacing for the board position.):

    [Game "23"]                             #   #   #   #
    [Event "Birmingham sim"]      p   # n R r k p
    [Site "Birmingham"]                 # p #   # p #
    [Date "????.??.??"]                   #   # p #   # 
    [Round "?"]                              #   #   #   #
    [White "Morphy, Paul"]          # P #   # N # 
    [Black "Avery"]                      P # P #   # K P
    [Result "1/2-1/2"]                 #   #   #   # 
    [ECO "B21"]          
    [FIN "8/p2nRrkp/2p3p1/3p4/8/1P3N2/P1P3KP/8 w - - 11 34"]                     

    {Sicilian Defence, Smith-Morra Gambit}

  • 5 months ago · Quote · #25


    Hi folks, there seems to be some interest in what is becoming a full-fledged PGN manager. Thank you for your interest. Version 1.15 is ready to download, and incorporates the capacity to identify ECO openings and sub-variants by name, from resultant board positions. This will be a taste of what is to come. V1.20 will be next, with the capacity to isolate 48 categories of ECO openings into separate packs, as with the endgames (plus associated enhancements). This will virtually be the computer equivalent of the OPENINGS and ENGAME index's (indices?) of the usual chess book. Refinements and testing will need to be completed. Forgive any current bugs for the moment. Feedback is welcome (see Help option).

  • 3 months ago · Quote · #26


    Hi folks. SWISSIMMACULATE is temporarily under reconstruction - it will be back, and should return in time for PGN Manager Version 1.20.   Some delightful improvements are being worked on, but also some difficult coding problems.   It is my intention that no limitation remains to the size of a generated PGN file, with all formats immediately available when compilation completes.   A couple of weeks should do it.

    Something interestion to look forward to.   A compilation of 11408 games from downloaded Bishop's Openings was created, and these were separated according to the program's 48 opening categories.   Since transpositions are recognised, these were divided into foldered single game PGN for Haundrix Chess as follows:

    Alekhine's Defence 1
    Bishop's Opening 8443
    Centre Game 17
    Evans Gambit 41
    Four Knights Game 68
    Giuoco Pianissimo 322
    Giuoco Piano 573
    King's Gambit 6
    King's Knight Opening 198
    Latvian Counter-Gambit 4
    Petrov's Defence 81
    Philidor's Defence 12
    Scotch Gambit 175
    Two Knights Defence 502
    Vienna Game 965

    Progressive positions are turned to FEN notation and these compared with the opening repetoire.   The last match in a game defines the opening.   This required a departure from the current standard FEN which I think is not so good.   The program will supply the following explanation when it becomes relevant:

    NB: This program's FEN notation e.p. target square implies that such a capturing pawn exists, whether execution is positionally legal or not.   The equivalence or otherwise of positions (same moves available) can only be guaranteed this way, either for the purposes of recognising opening move transpositions, or as the first known occurrence in a draw claim.   Input FEN notation will be corrected accordingly if necessary.

  • 6 weeks ago · Quote · #27


    http://www.swissimmaculate.com/?justsaying=1 is back up, looking better than ever.   Work commitments have delayed a (debugged &) revamped PGN compilation editor.   The "PURGE..." option (to become "LIQUIDATE or EDIT games from CURRENT COMPILATION") requires sophisticated integration with the database files to work properly.   'Till then, what's there is quite a useful tool for chess players.   Not yet a fully reliable one for PGN management, as the current release does not select the best choice of identical games as intended.   Hopefully soon...have fun with this little keyboard driven application.   Regards, JC

  • 3 weeks ago · Quote · #28


    Done at last! Compilers of PGN files (thank you for your efforts) should be able to use "Chess Translator & PGN Manager" with a degree of confidence. In time, use & reported bugs (see HELP) will eliminate problems and enhance utility. The option CREATE CURRENT COMPILATION will reveal errors in a diagnostic text file when we try to compile PGN. Compiling eliminates duplicated games automatically, optimally merging elements of both. Those that require human intervention are amenable to LIQUIDATE (PURGE) & EDIT. (This option also allows a perusal of all games in a compilation by setting percentage identity to zero.)

    A log file of the compilation or editing process allows inspection to ensure that the results are satisfactory. Many settings may be applied. PGN files can be optionally produced with opening names in comments, opening positions and game termination positions aligned with the rosters, in many formats. The roster, with optional settings, will produce all tags correctly if this can be determined by the program. ECO tags, for instance, are defined by the most advanced opening transposition found, rather than by move sequences. It has been a challenge, but we can now produce a single PGN Compilation as a database folder, available whenever the program is started in the same directory, and with many different filters and output modes. This database can be appended to at any time, and ultimately contain all games that can possibly interest us. Its size is only limited by disk space.

    It is my hope that chess intelligence and fair competition become available to as many as could love the game as I have - and in time, an up to the minute world chess champion be maintained on an official game site - to be verified, of course, by FIDE. Once Capablanca failed to lose for six years. That may be much more difficult with the access allowed to a champion by Odds Ratings on the net. But I am not convinced that post-Einstein physics (and much more) hasn't gone dark for advantage. It seems relatively uninspired and inconsistent. Then, who knew that the Horse that destroyed Troy was set on fire, to divert the guards at the gate? Military secrets are nothing new. Human nature, I guess.
       I have left myself the MATE SEARCH to tinker with at leisure. It is beyond the design scope after all, and but for simple mates, best left to Deep Blue descendants for anything more. Updates will follow testing through my own continued use, as with all of the
    applications. Just check the version numbers from time to time. Now for a well-earned rest...

  • 3 weeks ago · Quote · #29


    Amazing that they didn't stumble upon algebraic notation much sooner

  • 3 weeks ago · Quote · #30


    Maybe it's a bit like giving directions to your home to a cabbie by supplying longitude and latitude, at least before GPS.

  • 3 weeks ago · Quote · #31


    Algebraic is easier for computer programs to parse. exd4 in a causual descriptive notation might just be PxP (being the only pawn that could take another pawn). Similarly, you might notate 4. BxN QxB

  • 3 weeks ago · Quote · #32


    mike_tal wrote:

    Amazing that they didn't stumble upon algebraic notation much sooner

    Sooner than the 18th century? (Stamma's version wasn't exactly the same as the current notation, but close enough.)

  • 3 weeks ago · Quote · #33


    I believe algebraic was designed to be easily parsed (decoded) by the early computer programs, written in assembler or worse (machine code). Coordinates and a simple 'language' syntax were required. An unfortunate side effect of mandatory algebraic game scores, however, was loss of familiarity with descriptive, and the resulting obselescence of the chess classics, like Capablanca's "Chess Fundumentals". I have not seen a better description of the relative value of the pieces than that given here. The proof of Capablanca's assessment is Reuben Fine's "Basic Chess Endings. If I were to choose to teach chess, they would be the text books.

  • 3 weeks ago · Quote · #34


    Ivan Butrimov (born in St. Petersburg March 7, 1782) produced Russia's first chess book, "Chess Play," in 1821.  Below is a sample from that book. Notice he used very easily recognizable algebraic notation that employed the latin alphabet (as opposed to the cyrillic alphabet he used in the text).

    This was about 75 years after Phillip Stamma, but also a great improvement:

  • 3 weeks ago · Quote · #35


    Thanks for that Batgirl. It looks like he gets the honours then. A bit like Batiste Galoise inventing abstract algebra in 1830, and it didn't receive acknowledgement by the academic community until someone bothered to check it out - in 1968! On the same theme, how many people are aware that the Russian sputnik discovered that the far side of the moon looks like a different planet altogether, or that the ocean basins have been mapped for age by the French "Geological Map of the World", indicating global expansion at an exponential rate to double since the Jurrasic. But our theories are sadly lacking in the explanation department!?

  • 3 weeks ago · Quote · #36


    I think you mean Evariste Galois, and he didn't invent abstract algebra. He used groups, but they were always treated as sets of permutations (i.e. it wasn't abstracted yet).

  • 3 weeks ago · Quote · #37


    Well, you learn something new every day.  Thanks

  • 3 weeks ago · Quote · #38


    That's him. It's been a while. Groups, rings etc. I used it in deriving solutions (very efficiently) to:

    p & q = r

    where A has probability p of winning against B, and B has probability q of winning against C, while A has probability r of winning against C, given by the operator & defined:

    p & q = pq / (pq + (1 - p)(1 - q)).

    The solution involves solving for q given p and r. Directly, using algebra only, it is a pain, and the kind of solution that prevailed before abstract algebra. Using Abelian group theory, it is a few lines and simple. Incidentally, this statement of the relationships between probabilities in games of skill, is the axiomatic basis of Probabilistic Rating Theory which allowed me to derive its rating transformation formula:

    R = R(r, s) = 1/k x log((exp(k x r x (r >= 0)) + s) / (exp(-k x r x (r <= 0)) + 1 - s))

    where R is the new rating difference

               r is the old rating difference

               s is the game score 0 <= s <= 1, for chess s is 0, 1/2 or 1

               k is a scaling constant to create the familiar Elo range

               and the inequalities are True = 1 and False = 0.

    You will notice that it is NOT a linear function by any stretch, and for that reason, needs no provisional ratings, limitations or k factors to stop it falling over in practice. Ignoring transitivity, and maintaining a system using simple integer ratings without any other data, I doubt a better axiom can be found to describe the nature of the reality being idealised in ratings. Notably, being derived from probabilities, it also allows the calculation of probabilities from its rapidly converged ratings:

    p = 1 / (1 + exp(-kr))

    with pronumerals as defined before.

    Solving for k with p = 2/3 and r = 100 is recommended.

  • 13 days ago · Quote · #39


    I have compiled the following into a single CURRENT COMPILATION in the database folder. From this it may be filed as a PGN in many formats and with many filters to make a selection, just by cranking up the PGN Manager. After cleaning it up with the editor (LIQUIDATE ...) this complex routine should be bug free, and have revealed a few convenient enhancements for an update. It will be possible to sequence the games chronologically (input sequence) and alphabetically (concise algebraic ordering). And the editor will have revealed identical games and names with minor but insignificant (eg. move order) differences. Many of the pioneering champions were sourced from multiple PGN files. When I'm through and the update is ready to download, I will notify the four and a half thousand interested viewers here. The current version messes up if two identical names are set to be replaced. Just alter one. This is made mandatory in V1.31. Now check out the list, with currently 16876 games. About a thousand will prove to be duplicated but requiring human decisions to select various combinations of items.

    1550 - 1610 Polerio, Giulio Cesare 7 (algebraic).pgn
    1600 - 1634 Greco, Gioachino 79 (algebraic).pgn
    1698 - 1769 Lolli, Giambattista 2 (algebraic).pgn
    1719 - 1796 Ponziani, Domenico Lorenzo 4 (algebraic).pgn
    1726 - 1795 Philidor, François André 52 (algebraic).pgn
    1795 - 1840 de La Bourdonnais, Louis-Charles Mahé 123 (algebraic).pgn
    1818 - 1879 Anderssen, Karl Ernst Adolf 880 (algebraic).pgn
    1830 - 1908 Bird, Henry Edward 566 (algebraic).pgn
    1836 - 1900 Steinitz, Wilhelm 940 (algebraic).pgn
    1837 - 1884 Morphy, Paul Charles 653 (algebraic).pgn
    1841 - 1924 Blackburne, Joseph Henry 1093 (algebraic).pgn
    1850 - 1908 Chigorin, Mikhail Ivanovich 923 (algebraic).pgn
    1868 - 1925 Teichmann, Richard 602 (algebraic).pgn
    1868 - 1941 Lasker, Emanuel 1563 (algebraic).pgn
    1872 - 1906 Pillsbury, Harry Nelson 586 (algebraic).pgn
    1874 - 1918 Schlechter, Carl 842 (algebraic).pgn
    1877 - 1944 Marshall, Franklin James 1369 (algebraic).pgn
    1880 - 1961 Rubinstein, Akiba Kiwelowicz 1134 (algebraic).pgn
    1888 - 1942 Capablanca y Graupera, José Raúl 991 (algebraic).pgn
    1889 - 1929 Réti, Richard 643 (algebraic).pgn
    1892 - 1946 Alekhine, Alexander 2687 (algebraic).pgn
    1896 - 1975 Sämisch, Friedrich (Fritz) 615 (algebraic).pgn
    1905 - 1966 Khan, Sultan Malik Mir 141 (algebraic).pgn
    1914 - 1993 Fine, Reuben 546 (algebraic).pgn
    1943 - 2008 Fischer, Robert James (Bobby) 1150 (algebraic).pgn

Back to Top

Post your reply: