GGT + Firebird wins Amar Gambit + related chess notes +

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Rebel, chrisw

rainhaus
Posts: 185
Joined: Sun Feb 01, 2009 7:26 pm
Location: Germany
Full name: Rainer Neuhäusler

GGT + Firebird wins Amar Gambit + related chess notes +

Post by rainhaus »

GGT + Firebird wins Amar Gambit + related chess notes +

Hi,
comrade, if you are not interested in chess but only in scores, jump to the last cross table, please :)

Eco A00 (Irregular Openings)
The starting position file for the GGT (Great Gambit Tournament) contains 50 gambits which are sorted by the Eco Code and have to be played with switched colours. If you like, the A00-section is a cabinet of horror-moves which occurs rarely among serious chess players. Off and on, such moves will be good for an effect of surprise. Play 1. a3 or h4 or Nh3 and it will be hard for your opponent keeping his countenance, because he feels not taken seriously. Is anybody there who wants to say: just as Tony Miles has won 1980 against Karpov with 1.a3..? I have to say, no, but it's nearly right. It was 1.- a6, after Karpov opened with 1. e4, therefore it was not an A00-game, but the effect was the same :)
However, there are in the A00-section a few wacky gambits too which were really played in former tournaments!
But does this concern computer and programs at all? They cannot be offended. Each legal move has to be evaluated, it's as simple as that. Good and well, but a gambit always means a sharp challenge and a precarious test position from the very beginning. Particulary, when there is no help by an opening book. With irregular moves the gambit get an additional positional pepperiness. In this regard, let's go running hot the algorithms :)

Amar Gambit
1. Nh3 d5 2. g3 e5 3. f4. Often cited as Paris Gambit too, but in the Wikipedia "List Of Chess Gambits" the Paris Gambit is quoted with 1.Nh3 d5 2.f3 e5 3.e4 f5. I don't know..
The Parisian Amateur Charles Amar came up with his gambit in the 1930s. Probably nobody would have communicated it to the posterity, but Savielly Tartakower, a French and Polish Grandmaster and also the King of Chess Journalism of that era, enjoyed this gambit and applied it successfully. Please take pleasure of some of his famous and often quoted aphorisms:

- It's always better to sacrifice your opponent's men.
- The blunders are all there on the board, waiting to be made."
- The winner of the game is the player who makes the next-to-last mistake.
- It is not enough to be a good player; you must also play well.
- The move is there, but you must see it.
- No game was ever won by resigning.
- I never defeated a healthy opponent
- Tactics is knowing what to do when there is something to do
- Strategy is knowing what to do when there is nothing to do."
- Moral victories do not count.

Andor Lilienthal(* 1911) is the oldest living grandmaster. He played in tournaments with many world champions and won several games against legends like Emanuel Lasker, José Raúl Capablanca, Alexander Alekhine, Max Euwe, Mikhail Botvinnik and Vasily Smyslov.

Tartakower,Saviely (HBE 2719) - Lilienthal,Andor (HBE 2710)
Paris, 1933 (HBE is an abbreviation of mine and means: Historic best ELO)
1.Sh3 d5 2.g3 e5 3.f4 Lxh3 4.Lxh3 exf4 5.0-0 fxg3 6.hxg3 Sf6 7.d3 Sc6 8.Sc3 Ld6 9.Lg5 Lxg3 10.Lxf6 gxf6 11.e4 Tg8 12.Sxd5 Le5+ 13.Kh1 Dd6 14.c3 Tg3 15.Dh5 Txd3 16.Tad1 Txd1 17.Txd1 Se7 18.Se3 Dc5 19.Dxh7 1-0

Back to the electronic competitors. Looking at the Amar starting position you may suppose that Black's king safety, the restricted mobility of the knight and the imminent loss of an accepted gambit pawn, might be followed by a negative evaluation over one pawn. But look yourself, the engines are evaluating that ripped king wing relatively harmless.

Starting evaluation after 3. f4
t=30 sec. Black to move, but the Fritz Gui gives the values always in the view of White! The nodes are noted only for interest.

Code: Select all

                 cp    kN/s 3CPU
Deep Shredder  - 0,34  2500
Firebird       - 0,39  5750
Rybka          - 0,44  220
Naum           - 0,50  4200
Zappa          - 0,61  2900
Deep Fritz     - 0,62  8600
Stockfish      - 0,64  4950
With the free database program SCID I've got the following table of the first ten moves played by all engines

Theory Table SCID

Code: Select all

-------------------------------------------------------------------------------
  1.Sh3 d5 2.g3 e5 3.f4: +8 =10 -24 (13/42: 31%)
-------------------------------------------------------------------------------
       3        4        5        6        7        8        9       10     
-------------------------------------------------------------------------------
 1  f4       fe5      Lg2      c3[2]    Sf4      ed3      Sd3      Df3       5:
    Sc6[1]   d4       Se5      d3[3]    Sf6      Sd3      Dd3      Da6[4]   20%

 2  ...      ...      c3       cd4[6]   Sc3      Sf4      e3       Lb5       5:
    ...      ...      Se5[5]   Dd4      Sf6[7]   Lg4      Dd7      Ld1[8]   70%

 3  ...      c4[9]    Sc3[11]  Sf2      e3[13]   cd5      Lg2      g4        8:
    e4       c6[10]   Sf6      Le7[12]  OO       cd5      Db6      Td8[14]  44%

 4  ...      ...      Lg2      OO[15]   d3[16]   Da4[17]  Sa3      Sg5       5:
    ...      d4       Sc6      Lc5      e3       Ld7      Sge7     h5[18]   20%

 5  ...      Lh3      c4       Da4      Dc4      hg3[20]  Sc3[21]  Kd1       6:
    Lh3      ef4      dc4[19]  c6       fg3      Ld6      Lg3      Se7[22]  17%

 6  ...      ...      d4       hg3      Dd3      Tf1[24]  a3       Sc3       4:
    ...      ...      fg3      Sc6[23]  Ld6      h6       Sf6      OO[25]   50%

 7  ...      ...      ...      Dd3[26]  c3[27]   hg3      OO       Sd2       4:
    ...      ...      Ld6      Sc6      fg3      h6       Sf6      De7[28]  13%

 8  ...      ...      ...      ...      Sc3      hg3      Dd2      dc5       5:
    ...      ...      ...      ...      fg3[29]  Sb4      c5[30]   Lg3[31]  10%

---------------------------
And now, finally
The scores

Code: Select all

GGTJan10_4221_10+12_A00Amar 2010.01.24 - 2010.01.26
                                    Score     Fi DS DR Za Na St DF
-------------------------------------------------------------------
 1: FireBird 1.0 beta x64 mp      10.0 / 12   XX 11 =1 1= 10 11 11
 2: Deep Shredder 12 x64 FrGui     7.5 / 12   00 XX 10 1= 11 11 10
 3: Deep Rybka 3 x64               7.0 / 12   =0 01 XX == 1= 10 11
 4: Zappa Mexico II x64            6.0 / 12   0= 0= == XX 1= 1= 10
 5: Naum 4 MP x64                  4.5 / 12   01 00 0= 0= XX 01 1=
 6: Stockfish 1.6 JA x 64          4.0 / 12   00 00 01 0= 10 XX 1=
 7: Deep Fritz 11                  3.0 / 12   00 01 00 01 0= 0= XX
-------------------------------------------------------------------
42 games: +8 =10 -24
[/b]

Winning %

One variation of Shredder I liked very much, pushing his connected center pawns consequently into White's half like a deadly sting (game 3+,4+,24+,29-). This was Shredder's right way to catch the second place and it's recommendation for countering the Amar Gambit.
But in total, the probability of winning is daunting:

Code: Select all

1-0       8 games 19% 
1/2-1/2  10 games 24%
0-1      24 games 57%
Unfortunately there is no more time to have a closer look at the games, because the tournament must go on!

Here is the pgn-link for the interested folk:
http://www.file-upload.net/download-221 ... r.pgn.html

Next: Ware Gambit

Test conditions
Starting positions: pool of 100 gambits sorted by Elo number.
Time Control: short tournament level with 40/20', 20/10', 10'+12''.
System: Intel Core i7 920, oc 3700 MHz, 6 GB RAM. Hyperthreading off, Turbo Mode off. Vista 64Bit
Parameter: 3 Threads, ponder off, 1.2 MB Hash, EGTB 3-4-5 (not used by Firebird and Stockfish, but sometimes used by the GUI, even for both engines)
Fritz11-GUI: remis never, resign late.

Rainer Neuhäusler
beram
Posts: 1187
Joined: Wed Jan 06, 2010 3:11 pm

Re: GGT + Firebird wins Amar Gambit + related chess notes +

Post by beram »

Great stuff mate. Good Work and very interesting.
As it is also nice to see Firebird fly so well.

BTW. In my Nunn2 match FB 64 - Rb 64 FireBird wins 50 games match, 4m 2sec, with +24-7 =19 on my HP laptop with T4300processor Fritzmark 5,7

grts Bram
rainhaus
Posts: 185
Joined: Sun Feb 01, 2009 7:26 pm
Location: Germany
Full name: Rainer Neuhäusler

Re: GGT + Firebird wins Amar Gambit + related chess notes +

Post by rainhaus »

beram wrote:Great stuff mate. Good Work and very interesting.
As it is also nice to see Firebird fly so well.

BTW. In my Nunn2 match FB 64 - Rb 64 FireBird wins 50 games match, 4m 2sec, with +24-7 =19 on my HP laptop with T4300processor Fritzmark 5,7

grts Bram
Thx for your feedback, Bram :) Previously I've often used the Nunn positions, a very sophisticated suite by the English Grandmaster. He released it 1998 (Nunn 1) and in 2000 (Nunn 2) in the German Chess Journal CSS. Upcoming accusations, his selection would prefer certain chess programs has proved to be irrelevant. The rumours came from this well known mistrustful species of the human race who can see through a millstone and hear the grass grow:)
To be a little creative, I prefer today test suites from my own. Nunn's philosophy about preferable neutral starting positions seems rather opposed to what is represented by my gambit suite. But if one see both methods not one sided but supplemental, I guess he is back on track.
I don't like Blitz, because I've a higher level of quality with a longer time control and then I have also more time to watch the games and to understand what' s going on.
In the current stand I guess, Firebird will win most of the gambits. We will see...

Cheers Rainer
ernest
Posts: 2041
Joined: Wed Mar 08, 2006 8:30 pm

Re: GGT + Firebird wins Amar Gambit + related chess notes +

Post by ernest »

beram wrote:In my Nunn2 match...
As I said elsewhere, those 25 positions are from ChessBase, not from Nunn... :)
rainhaus
Posts: 185
Joined: Sun Feb 01, 2009 7:26 pm
Location: Germany
Full name: Rainer Neuhäusler

Re: GGT + Firebird wins Amar Gambit + related chess notes +

Post by rainhaus »

ernest wrote:
beram wrote:In my Nunn2 match...
As I said elsewhere, those 25 positions are from ChessBase, not from Nunn... :)
select case "what does he mean"
case puzzle
print: Nunn 2 = 20 positions not 25
case joke
print: lukewarm :?
case else
wait for input
end select
User avatar
notyetagm
Posts: 253
Joined: Mon Jan 25, 2010 3:11 am

Re: GGT + Firebird wins Amar Gambit + related chess notes +

Post by notyetagm »

Wow, when will the Ware Gambit results become available?

Thanks
rainhaus
Posts: 185
Joined: Sun Feb 01, 2009 7:26 pm
Location: Germany
Full name: Rainer Neuhäusler

Re: GGT + Firebird wins Amar Gambit + related chess notes +

Post by rainhaus »

notyetagm wrote:Wow, when will the Ware Gambit results become available?

Thanks
Today or tomorrow
rn