NEWS: FinalGen chess endgame tablebase generator.

[Terry McCracken]

Is the software free of malware? After installing, Panda Antivirus says virus: Trj/CI.A

Get rid of Panda...I have used it in the past...false positives and even deletion of crucial files can occur.

[IGarcia]

BTW: Any know other end game tools, like endgamne browser nalimov (or gaviota) bases

Long time ago there existed a program called "Wilhelm' which had many functions for endgame tables, but I doubt you'll still find a link for it (it was on geocities which has been closed a few years ago).

For analysing 3-6 men positions you can also use my engine ChestUCI - although it's usually a matesolver I've also included a special mode 'EgtbAnalysis' that shows you the complete solution for such endgame positions (of course only if you have installed the needed Nalimov-EGTBs).

Franz

Great!! I read the manual looks fantastic. Tomorrow will instal in the only windows notebook I have. Thanks.

BTW, SCIDvsPC has a nice tool too. More simple tham this one but very instructive to learn endgames positions.


Thanks all for links and screenshots etc!


Whillhem + FinalGen + Scidvs PC combined: endagame fan dream!!
Ignacio

[Arpad Rusz]

I was curious if FinalGen solves the famous "Szén position" (which is a win for White) and some other similar positions:

1.Carrera (1617) and Szén (1840)
[D]4k3/5ppp/8/8/8/8/PPP5/3K4 w - - 0 1

2.Greco (1621)
[D]4k3/5ppp/8/8/8/8/PPP5/4K3 w - - 0 1

The first endgame is unsolvable with FinalGen, but it proves the win for White in the second position. The problem with the first one is that the win involves the following position which is unsolvable by the software:

1a.
[D]8/8/8/PkP5/1P4p1/5pKp/8/8 b - - 0 15

15...h2!? 16.Kh2 Kc4! 17.c6 Kd3 18.c7 f2 19.c8Q. Now White is winning (19...f1Q 20.Qa6+ +-) but this position cannot be evaluated by the program.

The next position is very easy to solve but I think it shows an original positional draw:
3.Rusz (2012)
[D]8/6p1/8/k1P5/5p1p/8/PP4K1/8 w - - 0 1

[Robert Flesher]

Hello:

I was browsing in CCRL Endgame Tablebases Forum, when I read this topic:

7 8 or more pieces tablebase generator

According to the author (Pedro Pérez Moreno), this tool (which reminds me to Freezer) was developped in the last four years. It can be downloaded for free from this URL:

http://www.mtu-media.com/finalgen

The web is well documented with a tutorial and some examples. The downloaded file is an installer, and once installed, FinalGen is translated into English, French and Spanish.

IMPORTANT:

However, you cannot give the castling status.

Remark: Tablebases assume that castling is not possible for two reasons:
First, in practical endgames, this assumption is almost always correct. However, castling is allowed by convention in composed problems and studies.

Second, if the king and rook are on their original squares, castling may or may not be allowed. Because of this ambiguity, it would be necessary to make separate evaluations for states in which castling is or is not possible.

The same ambiguity exists for the en passant capture, since the possibility of en passant depends on the opponent's previous move. However, practical applications of en passant occur frequently in pawn endgames, so tablebases account for the possibility of en passant for positions where both sides have at least one pawn.

This new tool seems interesting to me, although good hardware and nice amounts of time are required. I will not use it, but I post this release here because I know that some people will find it useful.

@Moderation team: I do not know if this is the best subforum for this thread. Please feel free of move it if necessary.

As a side note: the name of the tool may be overlooked for a non-Spanish speaker: Final means End; applied to chess, Final means Endgame. Gen are the first letters of Generador ('Generator' in English). So, FinalGen can be translated as Endgame Generator. Enjoy this new endgame tablebase generator!

Regards from Spain.

Ajedrecista.

Thank-you very much. This software is amazing!

[Jouni]

Yes very interesting, but after 1 correct ran I got always "The position is not valid" message whatever I try

[fern]

You are an incredible nice guy, Ruxy.
If you only let me know the girl in your avatar.


Yum yum regards
Fern

[Jouni]

Or I get "unexpected error - FinalGen will be closed". In first try FG analysed 5 piece ending in about 1 second - it was too good to be true

[Robert Flesher]

I was curious if FinalGen solves the famous "Szén position" (which is a win for White) and some other similar positions:

1.Carrera (1617) and Szén (1840)
[D]4k3/5ppp/8/8/8/8/PPP5/3K4 w - - 0 1

2.Greco (1621)
[D]4k3/5ppp/8/8/8/8/PPP5/4K3 w - - 0 1

The first endgame is unsolvable with FinalGen, but it proves the win for White in the second position. The problem with the first one is that the win involves the following position which is unsolvable by the software:

1a.
[D]8/8/8/PkP5/1P4p1/5pKp/8/8 b - - 0 15

15...h2!? 16.Kh2 Kc4! 17.c6 Kd3 18.c7 f2 19.c8Q. Now White is winning (19...f1Q 20.Qa6+ +-) but this position cannot be evaluated by the program.

The next position is very easy to solve but I think it shows an original positional draw:
3.Rusz (2012)
[D]8/6p1/8/k1P5/5p1p/8/PP4K1/8 w - - 0 1

I find it very intersting that you state Final Gen cannot solve the first position. If it is a tablebase generator, then the tablebases base should look at every possible move, and thus be correct. If that is true, then either the tablebase it not complete/incorrect, or the winning position you provided may not be forced. I am running Final gen now, this has me very curious. Thanks

[Arpad Rusz]

I find it very intersting that you state Final Gen cannot solve the first position. If it is a tablebase generator, then the tablebases base should look at every possible move, and thus be correct. If that is true, then either the tablebase it not complete/incorrect, or the winning position you provided may not be forced. I am running Final gen now, this has me very curious. Thanks

That's the truth, you will see!
FinalGen's tablebases are not complete like Nalimov's etc. That is the prize paid for being able to generate tablebases with 7, 8 or even more pieces. There will be positions (the author estimates ~5%) with some kind of promotions that will not be solved. I don't fully understand it, but try this:

[D]6Q1/8/8/k7/8/8/p1K5/8 b - - 0 1

It is unsolvable only because the Knight promotion with check! (A bug or this is the natural thing with the author's algorithm? I don't know...)

[Robert Flesher]

I find it very intersting that you state Final Gen cannot solve the first position. If it is a tablebase generator, then the tablebases base should look at every possible move, and thus be correct. If that is true, then either the tablebase it not complete/incorrect, or the winning position you provided may not be forced. I am running Final gen now, this has me very curious. Thanks

That's the truth, you will see!
FinalGen's tablebases are not complete like Nalimov's etc. That is the prise for being able to generate tablebases with 7, 8 or even more pieces. There will be positions (the author estimates ~5%) with some kind of promotions that will not be solved. I don't fully understand it, but try this:

[D]6Q1/8/8/k7/8/8/p1K5/8 b - - 0 1

It is unsolvable only because the Knight promotion with check! (A bug or this is the natural thing with the author's algorithm? I don't know...)

I have always had the understanding that a tablebase includes all possible moves for a given position. If this understanding is correct, then the software is not generating the tablebase for every possible position; and thus the software is broken. I hope someone with a greater understanding can explain this further.