The strongest engines ... for opening analyzes ... statistic!

[Frank Quisinsky]

Hi Ferdy,

Today (FCP Tourney's) I am using for the ECO names ...
ChessPad 2.09
http://www.wmlsoftware.com/index.html

Three steps I do with Shredder GUI *.pgn files:
1. Original Shredder PGN sort with name of players and number of rounds with Chess Explorer by Mark Uniacke
2. Set ECO codes with ChessPad 2.09
3. CleanUp, part of the *.pgn utilities by Norm Pollock

I believe what the programmer from ChessPad do is to use the eco.pgn from David J. Barnes.
Also not perfect but much better as chessbase eco classification.

Most of time I am working with Chessbase, but ECO classification from Chessbase is a big chaos.
Not sure, over 200 grandmasters are working all the time for chessbase with the results that the eco classification is bad?

What Shredder GUI do with PGN is also not OK.
The "space" on the end of each line has what sense?

Better is to sort and fix, set the correct ECO, with one tool only.

SCID eco.pgn never I try out.

Best
Frank

Not sure what I do 2016 with FEOBOS (ECO names).
Must looking in my own documentation.

[Ferdy]

ChessPad 2.09 is not accurate on A04/11 ECO code, so I will not use it.

I make my own based on pgn-extract eco.pgn. It is in github, changes are visible at https://github.com/fsmosca/fcp/commits/main. I will be updating it.

[Frank Quisinsky]

Ferdinand,

it is possible to set your ECO names later with a tool in PGN?
The biggest problem ... the transposition of move.

Best
Frank

[Ferdy]

Ferdinand,

it is possible to set your ECO names later with a tool in PGN?
The biggest problem ... the transposition of move.

Best
Frank

We can use pgn-exract and eco.pgn to set eco, opening name and variation name in the pgn. That is what I use.

I can create a tool to add transposition eco code.

In the pgn standard:

Code: Select all

9.4.1: Tag: ECO

This uses a string of either the form "XDD" or the form "XDD/DD" where the "X"
is a letter from "A" to "E" and the "D" positions are digits; this is used for
an opening designation from the five volume _Encyclopedia of Chess Openings_.
This tag pair is associated with the use of the EPD opcode "eco" described in a
later section of this document.

There is XDD/DD, that is fine but what if it transposed to a new opening system. That will become XDD/YDD. But I will not use that formatting. I will use the following instead.

Code: Select all

[Event "?"]
[Site "?"]
[Date "?"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]

1.c4 Nf6 2.d4 g6 3.Nc3 d5 *

It is English and become a Gruenfeld.
1.c4 Nf6 is A15
1.c4 Nf6 2.d4 g6 3.Nc3 d5 is D80

That will become:

Code: Select all

[Event "?"]
[Site "?"]
[Date "?"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[ECO "A15"]
[ECOT "D80"]

1.c4 Nf6 2.d4 g6 3.Nc3 d5 *

The T in ECOT is transposition or final transposition because an opening may start at opening1, transposed to opening2, and finally to opening3, the ECO in the opening3 will be the value in ECOT tag.

The output of pgn-extract is:

Code: Select all

[Event "?"]
[Site "?"]
[Date "?"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[ECO "D80"]
[Opening "Gruenfeld defence"]

1. c4 Nf6 2. d4 g6 3. Nc3 d5 *

pgn-extract is using the transposition method of identifying the ECO, so I will just create a tool to generate from sequence + transposition.

[Frank Quisinsky]

Hi Ferdinand,

brilliant!
New possibilites ...

a. *.pgn sort by ECO at first
b. *.pgn sort by ECOT at first

Will help a lot during a work with the database for some reasons!

Best
Frank

[Ferdy]

Hi Ferdinand,

brilliant!
New possibilites ...

a. *.pgn sort by ECO at first
b. *.pgn sort by ECOT at first

Will help a lot during a work with the database for some reasons!

Best
Frank

Sorting will be provided.

I change the repo of eco changes. It is now in https://github.com/fsmosca/eco.
I also just got a permission to use the eco.pgn from David.

[Ferdy]

This is the algorithm.

1. Parse the game in the pgn file for up to 5 plies (settable), save the position and look it up in eco db. If found save the eco, opening and variation.
2. Continue parsing the move, save the position and check it on eco db. If there is save it as the eco_t, opening_t, variation_t. Do this for up to 24 plies or 12 moves, this is settable too. The last saved eco_t, opening_t, variation_t will be our values for transposition.
3. Then save those info in a pgn.

Example output.

Code: Select all

[Event "FEOBOS, Rank 00068"]
[Site "Trier"]
[Date "2017.01.29"]
[Round "20"]
[White "FEOBOS"]
[Black "FEOBOS"]
[Result "1/2-1/2"]
[BlackElo "3000"]
[ECO "E00"]
[ECOT "A64"]
[Opening "Catalan opening"]
[OpeningT "Benoni"]
[PlyCount "24"]
[Source "Frank Quisinsky"]
[VariationT "fianchetto, 11...Re8"]
[WhiteElo "3000"]

{ Game 28832 } 1. d4 Nf6 2. c4 e6 3. g3 c5 4. d5 exd5 5. cxd5 d6 6. Nc3 g6 7. Bg2 Bg7 8. Nf3 O-O 9. O-O Re8 10. Nd2 a6 11. a4 Nbd7 12. Ra2 Nh5 1/2-1/2

So what we have are these:

Code: Select all

[ECO "E00"]
[ECOT "A64"]
[Opening "Catalan opening"]
[OpeningT "Benoni"]
[VariationT "fianchetto, 11...Re8"]

There is no Variation value but we have VariationT. This is because in the first 5 plies (1. d4 Nf6 2. c4 e6 3. g3) the catalan in eco db has no variation.

Checking the entry we can see that the value is null.

Code: Select all

    "rnbqkb1r/pppp1ppp/4pn2/8/2PP4/6P1/PP2PP1P/RNBQKBNR b KQkq -": {
        "eco": "E00",
        "opening": "Catalan opening",
        "variation": null
    },

[fen]rnbqkb1r/pppp1ppp/4pn2/8/2PP4/6P1/PP2PP1P/RNBQKBNR b KQkq -[/fen]

[Frank Quisinsky]

Hi Ferdinand,

for me looks perfect.
You make ECO to an event and very important for people working on opening books.

BTW:
Another wish ...
If ready in around three weeks ...

Could you switched the tourney from FCP Tourney-2022 to FCP Tourney-Ki, run-6?
Hope this is not many work for you but after all I understand you have to switched only the PGN?

This would be great!

At the moment I am in round 6 of 30 ... around 21 hours my systems need for new games for one round.

Code: Select all

FCP Tourney-KI
run-6 after round 06 out of 30

         Name                       Games     Won     Draw     Lose       Pts         S-B         %

01. Stockfish 020622 NN dev       :   240  :  138+  :  102=  :    0-  :   189.0  :  21598.75  :  78.75%
02. Dragon 3 NN (Komodo)          :   240  :  131+  :  109=  :    0-  :   185.5  :  21296.00  :  77.29%
03. Fire 8 MC.3 NN                :   240  :  103+  :  134=  :    3-  :   170.0  :  19426.00  :  70.83%
04. SlowChess Blitz 2.9 NN        :   240  :  107+  :  126=  :    7-  :   170.0  :  19071.50  :  70.83%
05. Koivisto 8.9 NN               :   240  :  101+  :  136=  :    3-  :   169.0  :  19040.00  :  70.42%
06. Berserk 9 NN dev3             :   240  :   94+  :  139=  :    7-  :   163.5  :  18393.25  :  68.13%
07. Revenge 2.0 NN                :   240  :   93+  :  137=  :   10-  :   161.5  :  18086.00  :  67.29%
08. rofChade 3.0 NN               :   240  :   91+  :  132=  :   17-  :   157.0  :  17690.75  :  65.42%
09. RubiChess 20220223 NN         :   240  :   83+  :  146=  :   11-  :   156.0  :  17496.75  :  65.00%
10. Seer 2.5.0 NN                 :   240  :   79+  :  143=  :   18-  :   150.5  :  16775.00  :  62.71%
11. Arasan 23.3 NN                :   240  :   67+  :  160=  :   13-  :   147.0  :  16601.75  :  61.25%
12. Minic 3.22 NN                 :   240  :   64+  :  150=  :   26-  :   139.0  :  15418.25  :  57.92%
13. Igel 3.1.0 NN                 :   240  :   52+  :  163=  :   25-  :   133.5  :  14925.00  :  55.63%
14. Lc0 0.28.2 752187 CPU         :   240  :   56+  :  143=  :   41-  :   127.5  :  14182.75  :  53.13%
15. Halogen 10.23 NN dev          :   240  :   60+  :  135=  :   45-  :   127.5  :  13840.25  :  53.13%
16. Ethereal 13.07                :   240  :   47+  :  157=  :   36-  :   125.5  :  13954.00  :  52.29%
17. Clover 3.1 NN                 :   240  :   51+  :  148=  :   41-  :   125.0  :  13714.50  :  52.08%
18. Wasp 5.54 NN dev              :   240  :   49+  :  151=  :   40-  :   124.5  :  13812.50  :  51.88%
19. Nemorino 6.09 NN dev          :   240  :   47+  :  152=  :   41-  :   123.0  :  13578.50  :  51.25%
20. Rebel 15x2 NN                 :   240  :   51+  :  143=  :   46-  :   122.5  :  13500.25  :  51.04%
21. Booot 7.0 NN dev              :   240  :   41+  :  161=  :   38-  :   121.5  :  13463.25  :  50.63%
22. Uralochka 3.36c NN            :   240  :   41+  :  157=  :   42-  :   119.5  :  13412.25  :  49.79%
23. Tucano 10.00 NN               :   240  :   32+  :  160=  :   48-  :   112.0  :  12230.75  :  46.67%
24. Fritz 18 NN (Ginkgo)          :   240  :   37+  :  147=  :   56-  :   110.5  :  12029.00  :  46.04%
25. Zahak 10.0 NN                 :   240  :   35+  :  135=  :   70-  :   102.5  :  11183.25  :  42.71%
26. Velvet 3.3.0 NN               :   240  :   28+  :  142=  :   70-  :    99.0  :  10846.50  :  41.25%
27. Xiphos 0.6                    :   240  :   32+  :  128=  :   80-  :    96.0  :  10450.25  :  40.00%
28. Combusken 2.0.0               :   240  :   28+  :  133=  :   79-  :    94.5  :  10230.25  :  39.38%
29. Weiss 2.0                     :   240  :   25+  :  131=  :   84-  :    90.5  :   9970.75  :  37.71%
30. DanaSah 9.0 NN                :   240  :   21+  :  137=  :   82-  :    89.5  :   9833.00  :  37.29%
31. Schooner 2.2 XB               :   240  :   19+  :  141=  :   80-  :    89.5  :   9697.50  :  37.29%
32. Bit-Genie 9.19 dev            :   240  :   27+  :  122=  :   91-  :    88.0  :   9619.25  :  36.67%
33. Hiarcs 15.1                   :   240  :   20+  :  132=  :   88-  :    86.0  :   9598.75  :  35.83%
34. Laser 1.7                     :   240  :   20+  :  132=  :   88-  :    86.0  :   9358.00  :  35.83%
35. Defenchess 2.3 dev            :   240  :   21+  :  130=  :   89-  :    86.0  :   9333.25  :  35.83%
36. Stash 33.0                    :   240  :   21+  :  127=  :   92-  :    84.5  :   9272.75  :  35.21%
37. Fizbo 2.0                     :   240  :   27+  :  115=  :   98-  :    84.5  :   9092.75  :  35.21%
38. Shredder 13                   :   240  :   19+  :  127=  :   94-  :    82.5  :   9031.25  :  34.38%
39. Marvin 5.2.0 NN               :   240  :   22+  :  121=  :   97-  :    82.5  :   8941.25  :  34.38%
40. Chiron 5                      :   240  :   20+  :  123=  :   97-  :    81.5  :   8728.00  :  33.96%
41. Expositor 2WN29 NN            :   240  :   13+  :  107=  :  120-  :    66.5  :   7508.25  :  27.71%

Best
Frank

[Frank Quisinsky]

Hi Ferdinand,

if you are ready ...
Short question ...

The best place for your new "eco.pgn" is the site from David, think so!

I believe that the tool programmers Norm Pollock or David Barnes have to changed different parts of the own tools for the new ECO information?

Best
Frank

[Frank Quisinsky]

Hi Ferdinand,

sorry for my messages ...

Note:
In the time I am working on FEOBOS main database I made an important mistake.
Better, no available tools can help me to make it better.
Need to many work to do it from hand.

I am sorting the main database without move transposition and got 41.634 position (from over 85.000 positions in database).

The problem here is the direct way to the ECO-Codes ...
Speaking from the ECO main lines ...

In final FEOBOS database (41.634 games) are very often not the direct way to the ECO codes. OK, not important for eng-eng matches but a very big "blemish".

So, the direct way to the ECO codes is very ... really very important.

After your work ...

Code: Select all

[ECO "E00"]
[ECOT "A64"]
[Opening "Catalan opening"]
[OpeningT "Benoni"]

The direct way should be "ECO" ... the official way to build the ECO-code
The move transposition is "ECOT".

Now ...

Later if you are ready ...
Sorting phase ...

It should be possible that your tool can do this one ...

PGN with the direct way to ECO code select in a database ...
Without "ECOT"

PGN with move-transposition in an second database.

So, the tool must be able ...

1.
Sort the *.pgn database to ECO-Code ... A00 at first, A01, A02 ...

2.
Sort the *.pgn database to ECOT ... A00 at first, A01, A02 ...

And ... this is important!!

Build two PGN databases ...

The database with ECO and no information for ECOT ... the direct way to ECO-Codes
And the ECOT database with all the move-transposition in a second database!

Think so ...
With such a tool I can make the next FEOBOS clearly better with the result ...

In the openings books for all the available GUIs later the direct way to ECO-Codes at first ...
And I won many hours, days for set priorities between the lines ... priorities are important for give the popular systems more attention.
This is very complicated if I am working with a database the direct way is not clear, to many move transposition inside (without doubles) ... the problem from the 41.634 FEOBOS database without doubles.

Best
Frank