Winboard Forum

Hi all ,

a database with 203 short tactical games is updated. All games with 40/40 (adapted to 2 Ghz P4 CPU).

Two examples, in one Spike with a devastating sacrifice against Junior, the other Glaurung at its best:

Code: Select all


[Event "CEGT 3"]
[Site "CEGT"]
[Date "2005.04.28"]
[Round "2.7"]
[White "Spike 0.9a"]
[Black "Junior 9"]
[Result "1-0"]
[ECO "C49"]
[PlyCount "35"]
[EventDate "2005.04.23"]
[EventType "tourn"]

1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. Bb5 Bb4 5. O-O O-O 6. d3 d6 7. Bg5 Bxc3 8.
bxc3 h6 9. Bh4 a6 10. Bxc6 bxc6 11. d4 g5 12. Nxg5 hxg5 13. Bxg5 Qe7 14. f4 Qe6
15. f5 Qe7 16. Qe1 exd4 17. Rf4 Qe5 18. Qg3 1-0



Code: Select all


[Event "CEGT 4"]
[Site "CEGT"]
[Date "2005.05.13"]
[Round "2.4"]
[White "Shredder 9"]
[Black "Glaurung 0.2.3"]
[Result "0-1"]
[ECO "B06"]
[PlyCount "60"]
[EventDate "2005.05.12"]

1. e4 g6 2. d4 Bg7 3. Nc3 d6 4. f4 c6 5. Nf3 Bg4 6. Be3 Qb6 7. Qd2 Nf6 8. O-O-O
Qa5 9. Qe1 O-O 10. e5 Nd5 11. Nxd5 Qxd5 12. c4 Qe6 13. Qh4 f6 14. Rg1 Bxf3 15.
gxf3 fxe5 16. fxe5 Rxf3 17. Bg5 dxe5 18. d5 cxd5 19. Bg2 e4 20. Bxf3 exf3 21.
Rge1 Qb6 22. b3 Nc6 23. Rxd5 Qb4 24. Re3 Rf8 25. Qf2 e6 26. Rd7 Qc5 27. Rxg7+
Kxg7 28. Qb2+ Kg8 29. Bh6 Ne5 30. Rxe5 Qg1+ 0-1



http://www.husvankempen.de/nunn/downloads.htm

Best Regards
Heinz

Hi all ,

just to add that a lot of programs play the "suicidal" move g5? like in the game Spike against Junior. And Spike won more of this games with this position that way.

An average human player does not have to calculate, he even feels that pushing the pawn shield of the King is much too dangerous here. I think this is a general weakness in many chess programs.

Best Regards
Heinz

I can verify that Movei also does this mistake even after a long search to depth 16

Here is analysis by version 295 and version 310
295 seem to avoid g5 but it is only an illusion because at depth 16 it finds g5 again.

310 always like g5 but need less nodes to get depth 16 with the same pv


version 295:

setboard r1bq1rk1/2p2pp1/p1pp1n1p/4p3/3PP2B/2P2N2/P1P2PPP/R2Q1RK1 b - - 0 11
analyze
1 -19 0 85 g7g5
2 4 0 197 g7g5 d4e5 f6e4
2 4 0 304 g7g5 d4e5 f6e4
3 4 0 534 g7g5 d4e5 f6e4
3 4 0 1172 g7g5 d4e5 f6e4
4 -1 0 1752 g7g5 d4e5 f6e4 d1d3
4 -1 0 2853 g7g5 d4e5 f6e4 d1d3
5 0 0 3920 g7g5 d4e5 f6e4 d1d3 c8f5
5 0 0 5740 g7g5 d4e5 f6e4 d1d3 c8f5
6 27 1 10177 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 e4g3 f2g3
6 27 1 11554 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 e4g3 f2g3
7 37 4 23488 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 d6d5
7 37 4 26937 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 d6d5
8 7 10 56615 g7g5
8 16 15 82235 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 a1b1
8 16 20 108764 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 a1b1
9 -14 23 126099 g7g5
9 -20 29 164882 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 a1b1 e5d4 c3d4
9 -20 101 539261 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 a1b1 e5d4 c3d4
10 10 126 679808 g7g5
10 30 162 875799 g7g5 f3g5 h6g5 h4g5 g8g7 a1b1 f8h8 f2f4 e5f4 f1f4
10 30 176 944165 g7g5 f3g5 h6g5 h4g5 g8g7 a1b1 f8h8 f2f4 e5f4 f1f4
11 16 234 1256075 g7g5 f3g5 h6g5 h4g5 g8g7 f2f4 e5d4 c3d4 f8h8 e4e5 d6e5 f4e5
11 16 378 2029137 g7g5 f3g5 h6g5 h4g5 g8g7 f2f4 e5d4 c3d4 f8h8 e4e5 d6e5 f4e5
12 -14 451 2438978 g7g5
12 -2 598 3243471 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 g8g7 a1d1 e5d4 c3d4 e8h8 f3f4
12 -2 1323 7159853 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 g8g7 a1d1 e5d4 c3d4 e8h8 f3f4
13 21 1926 10435320 g7g5 f3g5 h6g5 h4g5 f8e8 a1b1 g8g7 d1f3 e8h8 f1d1 e5d4 c3d4
c8e6
13 21 2203 11959203 g7g5 f3g5 h6g5 h4g5 f8e8 a1b1 g8g7 d1f3 e8h8 f1d1 e5d4 c3d4
c8e6
14 -9 2471 13454953 g7g5
14 -41 3353 18406546 g7g5 f3g5 h6g5 h4g5 f8e8 a1b1 e5d4 d1f3 g8g7 c3d4 e8h8 e4e
5 d6e5 d4e5
14 -40 8240 45309757 d8e7
14 -17 19609 106274994 d8e7 d1d3 e5d4 h4f6 e7f6 c3d4 f8e8 e4e5 f6f5 d3c3 c8b7 a
1b1 a8b8 g1h1 d6d5
14 -17 21142 114537803 d8e7 d1d3 e5d4 h4f6 e7f6 c3d4 f8e8 e4e5 f6f5 d3c3 c8b7 a
1b1 a8b8 g1h1 d6d5
15 -23 33835 182283341 d8e7 d1d3 e5d4 a1e1 d4c3 d3c3 g7g5 e4e5 f6h5 h4g3 h5g3 f
2g3 a8b8 c3c6 c8b7
15 -23 44023 238760196 d8e7 d1d3 e5d4 a1e1 d4c3 d3c3 g7g5 e4e5 f6h5 h4g3 h5g3 f
2g3 a8b8 c3c6 c8b7
16 -31 126876 670964183 d8e7 d1d3 f8e8 d3c4 c6c5 d4c5 d6c5 h4f6 e7f6 f1b1 f6f4
c4c5 c8g4 f3e1 a8c8 f2f3
16 -30 152862 811643640 g7g5
16 -29 160257 852523251 g7g5 f3g5 h6g5 h4g5 d8e7 d1f3 g8g7 f3g3 e7e6 d4e5 d6e5
g5f6 g7f6 f2f4 f6e7 f4e5 e7e8
16 -29 202137 1074439451 g7g5 f3g5 h6g5 h4g5 d8e7 d1f3 g8g7 f3g3 e7e6 d4e5 d6e5
g5f6 g7f6 f2f4 f6e7 f4e5 e7e8

version 310:
new
setboard r1bq1rk1/2p2pp1/p1pp1n1p/4p3/3PP2B/2P2N2/P1P2PPP/R2Q1RK1 b - - 0 11
analyze
1 -19 0 84 g7g5
2 4 0 196 g7g5 d4e5 f6e4
2 4 0 303 g7g5 d4e5 f6e4
3 4 0 533 g7g5 d4e5 f6e4
3 4 0 1171 g7g5 d4e5 f6e4
4 -1 0 1751 g7g5 d4e5 f6e4 d1d3
4 -1 0 2852 g7g5 d4e5 f6e4 d1d3
5 0 0 3919 g7g5 d4e5 f6e4 d1d3 c8f5
5 0 1 5667 g7g5 d4e5 f6e4 d1d3 c8f5
6 27 1 10086 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 e4g3 f2g3
6 27 1 11451 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 e4g3 f2g3
7 37 4 22550 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 d6d5
7 37 4 25960 g7g5 d4e5 f6e4 d1d4 c8f5 h4g3 d6d5
8 7 10 52977 g7g5
8 8 15 76677 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 g5f6 e6f6
8 8 20 102590 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 g5f6 e6f6
9 -20 26 140981 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 a1b1 e5d4 c3d4
9 -20 76 408765 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 e8e6 a1b1 e5d4 c3d4
10 10 96 511475 g7g5
10 30 118 630062 g7g5 f3g5 h6g5 h4g5 g8g7 a1b1 f8h8 f2f4 e5f4 f1f4
10 30 129 686680 g7g5 f3g5 h6g5 h4g5 g8g7 a1b1 f8h8 f2f4 e5f4 f1f4
11 16 178 952317 g7g5 f3g5 h6g5 h4g5 g8g7 f2f4 e5d4 c3d4 f8h8 e4e5 d6e5 f4e5
11 16 287 1534947 g7g5 f3g5 h6g5 h4g5 g8g7 f2f4 e5d4 c3d4 f8h8 e4e5 d6e5 f4e5
12 -14 317 1696100 g7g5
12 0 403 2168000 g7g5 f3g5 h6g5 h4g5 f8e8 a1b1 g8g7 d1d3 e8h8 f2f4 e5d4 c3d4
12 0 876 4741961 g7g5 f3g5 h6g5 h4g5 f8e8 a1b1 g8g7 d1d3 e8h8 f2f4 e5d4 c3d4
13 13 1310 7097765 g7g5 d4e5 f6e4 f1e1 d6d5 h4g3 e4g3 h2g3 c6c5 f3d2 g5g4 a1b1
c8f5
13 13 1525 8265452 g7g5 d4e5 f6e4 f1e1 d6d5 h4g3 e4g3 h2g3 c6c5 f3d2 g5g4 a1b1
c8f5
14 -17 1857 10132927 g7g5
14 -2 2420 13339534 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 g8g7 a1b1 e8h8 d4e5 d6e5 f1d1
d8e7 h2h3
14 -2 6637 36348280 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 g8g7 a1b1 e8h8 d4e5 d6e5 f1d1
d8e7 h2h3
15 0 8510 46901608 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 g8g7 a1b1 e8h8 d4e5 d6e5 f3g3
d8d6 g5f6 g7f6 f1d1
15 0 13443 73848905 g7g5 f3g5 h6g5 h4g5 f8e8 d1f3 g8g7 a1b1 e8h8 d4e5 d6e5 f3g3
d8d6 g5f6 g7f6 f1d1
16 -30 15521 85439134 g7g5
16 -29 20701 114482876 g7g5 f3g5 h6g5 h4g5 d8e7 d1f3 g8g7 f3g3 e7e6 d4e5 d6e5 g
5f6 g7f6 f2f4 f6e7 f4e5 e7e8
16 -29 75103 413548382 g7g5 f3g5 h6g5 h4g5 d8e7 d1f3 g8g7 f3g3 e7e6 d4e5 d6e5 g
5f6 g7f6 f2f4 f6e7 f4e5 e7e8
exit

Hi Heinz,

thanks for posting the two nice games!

In fact Spike gives a high bonus for a broken opponents pawn shield, especially for an advanced (or even missing) pawn on G. Doing this is very risky, because sometimes the sacrifice is very unsound like in the games that were posted by Jan Klima (http://wbforum.volker-pittlik.name/viewtopic.php?t=2366), just looking like a bug, on the other hand one can watch some nice games even against the commercials like in this game.

Our goal is (like all engine programmers I guess) to build an engine which plays interesting chess, so we consider this as a feature not a bug at the moment

I've analyzed the game some minutes ago with my newest version of Spike, it would also play the sacrifiece on g5, but the score would be far less optimistic, Spike 0.9a thinks white is nearly a pawn up while the score of the development version is nearly a draw. Anyway, this game would not have happened with the newer version, because it would play now 11. Rb1 instead of 11. d4

Best wishes
Ralf

Hi all,

it seems that I missed the point, the sacrifice was of course forced in this position. It was 11. d4 which invited the opponent to play g5.

The reason that the new version does not play 11. d4 anymore is the introducing of a "bad trade" code, which tells Spike that having a piece less than the opponent is not that good. Well, this is not very good tested at the moment and I'm pretty sure that I will change and test this part of the evaluation many times before the next release.

Keeping a desired playing style and additionally improving the engine is one of the toughest challenges in development of a chess engine

Best wishes
Ralf

Hi Ralf ,

yes, it was a sacrifice anyway, because Spike somehow was "aware" of the consequences after 11.d4, maybe already earlier.

When you give the game to a GM, he would swear that the white pieces were lead by a human and maybe the black ones by a computer or very weak player.

The astonishing thing also in the similar games is that it can?t be calculated with all consequences (maybe only after hours), because it is too deep (calculation horizon). But of course you can just configure the evaluation that way, that pieces, even rook and queen, are sacrificed for attack values like disrupted king position.

What many players do not bear in mind: it does not count the material balance over the whole board...in case of an attack it only counts how many pieces can attack and go to decisive squares and how many pieces can defend at the right time. A rook more on a8, that cannot defend, counts nothing, as the game will be over, before it comes into play. So when the attack is successful you can give away all material you have and give a mate with the last remaining pawn .

Best Regards
Heinz

Hi Heinz,

When I analyze the game I find that g5 was only one mistake of black.
Black could defend better later by Nxe4 instaed of hxg5 as Shredder9 suggests.

I am not sure if white wins easily after Nxe4 so blaming only g5 for the loss of black is not fair

Uri

Uri Blass wrote:Hi Heinz,

When I analyze the game I find that g5 was only one mistake of black.
Black could defend better later by Nxe4 instaed of hxg5 as Shredder9 suggests.

I am not sure if white wins easily after Nxe4 so blaming only g5 for the loss of black is not fair

Uri

Hi Uri ,

when you have a look at the database you will find that also other engines like Ktulu, SlowChess and Naum lost here in similar position against Spike.

If you, the Spike authors or others are interested, the CEGT testers could give deeper analyses in a few days to this interesting position. Something I already wanted to do for a longer time, but never had the impression that it interests a lot of people.

Having played Correspondence Chess at higher level for more than 25 years I should hopefully be able to analyze, if Black can survive after g5. At first sight I would say no, but maybe you are correct and I know that you are also a very strong correspondence chess player.

Best Regards
Heinz

Heinz van Kempen wrote:

Uri Blass wrote:Hi Heinz,

When I analyze the game I find that g5 was only one mistake of black.
Black could defend better later by Nxe4 instaed of hxg5 as Shredder9 suggests.

I am not sure if white wins easily after Nxe4 so blaming only g5 for the loss of black is not fair

Uri

Hi Uri ,

when you have a look at the database you will find that also other engines like Ktulu, SlowChess and Naum lost here in similar position against Spike.

If you, the Spike authors or others are interested, the CEGT testers could give deeper analyses in a few days to this interesting position. Something I already wanted to do for a longer time, but never had the impression that it interests a lot of people.

Having played Correspondence Chess at higher level for more than 25 years I should hopefully be able to analyze, if Black can survive after g5. At first sight I would say no, but maybe you are correct and I know that you are also a very strong correspondence chess player.

Best Regards
Heinz

If black can survive with exact moves is one question and if black has practical chances to survive is another question.

I did not analyze enough to know.

I now took a vacation from my correspondence games from the europe championship 15.5-13.6 inclusive.

I am board 3 of Israel.
I have 12 games at the same time and no game is decided at this moment.

I hope to get rid of part of them until the end of the year.

Uri

Hi all ,

while Uschi is dreamimg the stories for the next matches (according to her motherly obligations she will surely need one week or two for each match), a very serious match takes place in the Generalife of the palace:

http://www.husvankempen.de/nunn/spicer.htm

Best Regards
Heinz

Hi Heinz,

attracted by the bright sunshine, I hobble (it?s really hard to walk wearing a stiletto on the right and a haggish half-boot on the left ) to the Generalife seeing the splendid fight. Placing genteelly under a blossoming tree (...honestly it was more some form of tumbling...) I?m watching the game with more and more intentness.

Magnitised I?m mumbling: "Come on, Uri,play up, lest you all miss the thrilling women?s fight in the Hall!" being interrupted by my displosive sneeze caused by the odour of the gaudy flowers.

Many thanks for your wonderful story, Heinz !
It?s very animating to get such a fine oasis for dreaming about the ladies? fight.

Best Regards,

Uschi

Heinz van Kempen wrote:Hi all ,

while Uschi is dreamimg the stories for the next matches (according to her motherly obligations she will surely need one week or two for each match), a very serious match takes place in the Generalife of the palace:

http://www.husvankempen.de/nunn/spicer.htm

Best Regards
Heinz

Thanks
Thanks also to Uschi for her support.

I have one correction:

I found that Junior can also find Nxe4 on my machine at longer time control.

Junior does not seem to me absent-minded playing with a stick and some bugs that are crawling through the sand.


Here is some analysis of Junior(with slightly more time than the time that it had in the game it can both avoid g5 and hxg5)

,a - Junior 9, CEGT 3 2005
r1bq1rk1/2p2pp1/p1pp1n1p/4p3/3PP2B/2P2N2/P1P2PPP/R2Q1RK1 b - - 0 1

Analysis by Junior 9:

11...exd4 12.Bxf6 Qxf6 13.cxd4 Bg4
? (0.54) Depth: 3 00:00:00
11...g5 12.Bxg5 hxg5 13.Nxg5
= (-0.09) Depth: 3 00:00:00 1kN
11...g5 12.Bg3 Nxe4 13.Qd3 Bf5 14.dxe5
= (-0.10) Depth: 6 00:00:00 7kN
11...g5 12.Nxg5 hxg5 13.Bxg5 Kg7 14.dxe5 dxe5 15.Rb1 Qxd1
? (0.58) Depth: 9 00:00:00 288kN
11...g5 12.Nxg5 hxg5 13.Bxg5 exd4 14.Qxd4 Kg7 15.e5 dxe5 16.Qxe5 Qd6 17.Bxf6+ Qxf6 18.Qxf6+ Kxf6 19.Rad1
? (0.45) Depth: 12 00:00:01 2295kN
11...exd4 12.Qxd4 c5 13.Qd3 Bb7 14.Rae1 g5 15.Bg3 g4 16.Nd2 Rb8
? (0.32) Depth: 12 00:00:02 4700kN
11...exd4 12.cxd4
? (0.62) Depth: 15 00:00:17 30938kN
11...g5 12.Nxg5 hxg5 13.Bxg5 Qe7 14.Qf3 Kg7 15.Qg3 Qe6 16.Bxf6+ Kxf6 17.Rfb1
? (0.47) Depth: 15 00:00:21 37531kN
11...g5 12.Nxg5 hxg5 13.Bxg5 Qe7 14.Qf3 Kg7 15.Qg3 Rh8 16.Bxf6+ Kxf6 17.h4 Rh5 18.Rab1
? (0.53) Depth: 17 00:00:57 99595kN
11...Qe7 12.Re1 Bg4 13.h3 Bxf3 14.Qxf3 Rab8 15.Rab1 Qe6 16.Bxf6 Qxf6 17.Qxf6
? (0.49) Depth: 17 00:01:22 141656kN
11...Qe7 12.Re1 Bg4 13.Qd3 Bxf3 14.Qxf3 Rab8
? (0.47) Depth: 18 00:02:08 229594kN
11...Qe7 12.Re1 Bg4 13.Qd3 Bxf3 14.Qxf3 Rab8
? (0.47) Depth: 18 00:02:27 257109kN

(, 26.05.2005)

,a - Junior 9, CEGT 3 2005
r1bq1rk1/2p2p2/p1pp1n1p/4p1N1/3PP2B/2P5/P1P2PPP/R2Q1RK1 b - - 0 1

Analysis by Junior 9:

12...hxg5 13.Bxg5 Bg4 14.Qd3
? (-0.71) Depth: 3 00:00:00
12...hxg5 13.Bxg5 Kg7 14.dxe5 dxe5 15.Rb1 Qxd1
? (0.43) Depth: 6 00:00:00 14kN
12...Nxe4 13.Nf3 Qd7 14.Qd3 Qf5 15.Rab1
? (0.41) Depth: 6 00:00:00 20kN
12...Nxe4 13.Nf3 Qd7 14.Qd3 Qg4 15.h3 Qf4 16.Be7 Be6 17.Bxf8 Rxf8
+- (1.58) Depth: 9 00:00:00 165kN
12...hxg5 13.Bxg5 Re8 14.Qf3 Kg7 15.Bxf6+ Qxf6 16.Qxf6+ Kxf6 17.Rab1
= (0.14) Depth: 9 00:00:00 231kN
12...hxg5 13.Bxg5 Qe7 14.Qf3 Kg7 15.Qg3 Qe6 16.Bxf6+ Kxf6 17.Rfb1
? (0.32) Depth: 12 00:00:01 1806kN
12...hxg5 13.Bxg5 Qe7 14.Qf3 Kg7 15.Qg3 Rh8 16.Bxf6+ Kxf6 17.h4 Rh5 18.Rab1
? (0.38) Depth: 15 00:00:08 13468kN
12...hxg5 13.Bxg5 Qe7 14.Qf3 Kg7 15.Qg3 Qe6 16.Bxf6+ Kxf6 17.f4 Qg4 18.fxe5+ Ke7 19.Qf2 Be6 20.Qf6+ Kd7 21.exd6 cxd6
? (0.92) Depth: 17 00:00:42 66735kN
12...hxg5 13.Bxg5 Qe7 14.f4 Qe6 15.f5 Qe7 16.Qe1 Kg7 17.Qg3 Rh8 18.dxe5 dxe5 19.Bxf6+ Kxf6 20.Rad1
? (1.09) Depth: 19 00:02:39 263323kN
12...Nxe4 13.Nf3 Ng5 14.Bxg5 hxg5 15.dxe5 g4 16.Nd4 dxe5 17.Nxc6 Qf6 18.Nb4 Bf5 19.Qe2 Rad8 20.Rae1 e4 21.Qc4 Qe5
? (0.95) Depth: 19 00:07:29 803046kN

(, 26.05.2005)

Hi Uri ,

yes, that?s correct. Even the top programs need to submerge deep into the position to detect that 12...Nxe4 is practically the only move here, only Shredder found quickly that 12...hg5: is bad, while Junior also changes quickly to 12...Nxe4 but then again for a long time prefers 12...hg5:.

12...hg5: leaves Black with no chances after 13.Bg5: all moves like 13...Qe7, 13...Kg7, 13...exd4 or 13...Re8 seem to lose rapidly. If anyone has doubts here I could give a few examples.

So 12....Nxe4 and Spike continues after one hour with
13.Nf3 (20-07 0.67) 13..f6 14.dxe5 dxe5 15.Qd8: Rd8:

I will give this position to Movei 00_8_310, Shredder 9 and Junior 9 with one hour each, as 13...Qd7 seems to be an alternative to 13...f6 and maybe there are also other alternatives.

There will not be a lot of support for Spike as Ralf is on honeymoon, but he is looking forward to seeing what happened when he will be back.

By the way a big thanks to Lex Domingo from the Philippines who added some amusing information regarding Spike (comics/TV), where I am not really familiar with.

Best Regards
Heinz

Hi all ,

while Brujatrace is casting love spells on SMK (who seems to be attracted by the combination of warts and bloodred lipstick) and Spike distracts Junior by carrying the stick to and fro, the master of Cerebro 3 (triple brain) will have to take a tough decision now. So let?s give Uri some more time....

http://www.husvankempen.de/nunn/spicer.htm

Analyses from Shredder will be added this evening after some hours of calculation.

Best Regards
Heinz

I do not know which move is better(f6 or Ng5)
I did not do much analyze but at least the line of Junior does not seem convincing to me and sacrificing the rook at f1 seems bad and I do not see more than a draw for white.


After 13.Nf3 f6 14.Qd3 Bf5 15.Nd2 Nxd2 16.Qxf5 Nxf1 17.Qg6+ Kh8
I believe that it is a draw(Shredder gives 0.11 for white at depth 19).

Of Course Shredder does not agree with almost nothing of Junior's moves.

It gives 14.Nd2 in your line that should be analyzed(I did not look at this line at the time of writing this post) and also
white can choose 16.Qxd2 that shredder evaluates as 0.29/20 but I also do not believe that white is better

Earlier white may try to improve by 15.Qe3 when Shredder gives 0.57/17,0.61/18,0.60/19


I can also add that I expected SOS5.1 to give an opinion after
Hearing the SOS calls from Movei.

Uri

meanwhile Junior9 analyzed the position after f6

After 2 hours of search it gives the following analysis

Spike 0.9a - Junior 9, CEGT 3 2005
r1bq1rk1/2p5/p1pp1p1p/4p3/3Pn2B/2P2N2/P1P2PPP/R2Q1RK1 w - - 0 1

Analysis by Junior 9:

1. ? (0.82): 14.Qd3 Bf5 15.Nd2 Nxd2 16.Qxf5 Nxf1 17.Qg6+ Kh8 18.Qxh6+ Kg8 19.Qg6+ Kh8 20.dxe5 dxe5 21.Rxf1 Qd6 22.c4 Rad8 23.c5
2. ? (0.80): 14.Qe2 Ng5

(, 27.05.2005)

It seems that Junior simply did not search deep enough when it claimed 1.21 pawns against f6

After 13...Ng5 14.Bxg5 hxg5 15.dxe5 g4 (the line that is accepted as main line both by Movei Junior and Shredder)

Junior says:


Spike 0.9a - Junior 9, CEGT 3 2005
r1bq1rk1/2p2p2/p1pp4/4P3/6p1/2P2N2/P1P2PPP/R2Q1RK1 w - - 0 1

Analysis by Junior 9:

1. ? (0.90): 16.Nd4 c5 17.Nb3 Re8 18.exd6 cxd6 19.f3 c4 20.Nd2 g3 21.Ne4 gxh2+ 22.Kxh2 Bf5 23.Qe1 Bxe4 24.fxe4 Qg5 25.Rd1
2. ? (0.70): 16.Nd2 dxe5 17.Qe2 Qg5 18.Ne4 Qe7 19.Qc4 Qe6 20.Qc5 f5 21.Ng5 Qd6 22.Qc4+ Qd5 23.Qa4 f4 24.Rad1 Qb5 25.Qa3

(, 27.05.2005)





It seems even based on the scores of Junior that f6 is better.

I am not sure about nothing but I guess that Shredder is right and f6 is better.

Uri

Hi all ,

"It is still not easy", the confused reporter from hologram television dares to interrupt. "To me it seems that after 13...f6 14.Qe2!? white has still some edge, for example":

14...Bf5 15.de5: de5: 16.Ne5: Qd2 17.Qd2: Nd2: 18.Rfd1 fe5: 19.Rd2: (in this line the bishops of opposite colours do not guarantee a draw, as rooks are also present and could penetrate, what changes a lot)

or 14...Ng5: 15.Bg5: hg5: 16.Qc4+ Kg7 17.Qc6:

"On the other hand after 14..Ng5 15.Bg5: hg5: 16.de5: g4 17.Nd2 or 17.Nd4 Black?s position is also unpleasant."

I added a first part replay to the site, but so far did not analyze a lot, for example not the moves 14.Nd2 or 14.dxe5 after 13...f6.

http://www.husvankempen.de/nunn/tabellen/spicer2.htm

http://www.husvankempen.de/nunn/spicer.htm

Just to add for those who do not play correspondence chess, that a correspondence player here would need many days to analyze all consequences for the four main alternatives after 13.Nf3, using as well different engines and analyzing step by step bringing in his own ideas and positional feeling.

Best Regards
Heinz

Some analysis by Shredder at depth 21 after the line
13...f6
14.Qe2 Ng5: 15.Bg5: hg5: 16.Qc4+ Kg7 17.Qc6 Rb8:

Spike 0.9a - Junior 9, CEGT 3 2005
1rbq1r2/2p3k1/p1Qp1p2/4p1p1/3P4/2P2N2/P1P2PPP/R4RK1 w - - 0 1

Analysis by Shredder 9 UCI:

1. ? (0.82): 18.Qc4 Qd7 19.Nd2 Qf7 20.Qd3 Rb2 21.Nc4 Bf5 22.Qxf5 Qxc4 23.dxe5 dxe5 24.Rfd1 Rf7 25.a4 Qf4 26.Qd3 a5 27.f3 c5 28.g3
2. ? (0.82): 18.Nd2 Qd7 19.Qc4 Qf7 20.Qd3 Rb2 21.Nc4 Bf5 22.Qxf5 Qxc4 23.dxe5 dxe5 24.Rfd1 Rf7 25.a4 Qf4 26.Qd3 a5 27.f3 c5 28.g3

(, 27.05.2005)


The score does not improve for white when the depth becomes bigger so it seems that black has good chances for a draw in that line

Uri

Looking for some refreshment of the sweltering heat I?m searching a shadily place in the Generalife, watching curiously the running game.

Attentively hearing all the analyses, I?m babbling: "Hm, yes...I like the idea to preserve the knight after f6 more than changing it - if I was white, I would try to keep up advantage by looking for tactical threats, so I wouldn?t like to exchange pieces."

"Qd3 seems to be an active alternative, is there any possibility to avoid the exchange of knights here, too, like 15. .... Rf-e1?", I?m asking for help of all the experts in the garden. As Qe2 allows Ng5, I?m keeping on to fancy Qd3, realising,that it is too stiflingly hot today for some careful consideration.

"Tempo, tempo, Spike, don?t give him time for consolidation!" I?m mumbling while refreshing with divine nectar, settling back, waiting relaxed for some more interesting explanations.

It?s great fun, to follow your analyses - I?m looking forward to see more of them

Uschi

Hi all ,

so here is what really happens.

http://www.husvankempen.de/nunn/spicer.htm

Update of replay and some forced moves tomorrow. Then we will see....changing Queens now is a bit boring, shame on Spike .


Best Regards
Heinz