The Chessgames Challenge

Team White vs Team Black
C U R R E N T P O S I T I O N

Deep Fritz 13

Can large groups of chess fans team up to create a chess masterpiece? Find out in Chessgames Challenge! You can vote for the move you think is best, and discuss the game with other members in the Kibitzer's Corner.

[Help Page]

[flip board]

GAME OVER: 1/2-1/2

[flip board]

MOVES:
1.d4 d5 2.c4 e5 3.dxe5 d4 4.Nf3 Nc6 5.a3 g6 6.Bg5 Qd7 7.Nbd2 h6 8.Bf4 g5 9.Bg3 g4 10.Nh4 Qd8 11.Qb1 Be7 12.Nf5 Bg5 13.f4 gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qd3 Qd7 17.b4 Nf5 18.b5 Na5 19.Bf2 Be3 20.g3 Bxf2+ 21.Kxf2 Ne3 22.h4 Ng4+ 23.Kg1 Rg8 24.Bh3 Qe6 25.Rc1 O-O-O 26.Nh2 h5 27.Bxg4 hxg4 28.Kg2 Qxe5 29.Rhf1 Qe6 30.Rf4 b6 31.Re4 Qd7 32.Rd1 c5 33.bxc6 Nxc6 34.Rf4 Kb8 35.Qf5 Qxf5 36.Rxf5 Na5 37.Rf4 Nxc4 38.Rdxd4 Rxd4 39.Rxd4 Nxa3 40.Nxg4 1/2-1/2
GAME OVER thank you for playing	it is now 08:43:04
[REGISTER] [HELP] [CONDITIONS] [REVIEW GAME] [ROSTERS] [DOWNLOAD PGN] [WEBMASTERS]

NOTE: You are currently not signed in. If you have a Chessgames account, you must first sign-in with your username & password to access the Chessgames Challenge area. If you do not have an account, please see our registration page.

Check out the Sticky frequently; it's used for sharing important
links and other information with your teammates. [more info]

< Earlier Kibitzing · PAGE 96 OF 240 · Later Kibitzing>

Jun-05-12

ajile:

1.d4 d5 2.c4 e5 3.dxe5 d4 4.Nf3 Nc6 5.a3 g6 6.Bg5 Qd7 7.Nbd2 h6 8.Bf4 g5 9.Bg3 g4 10.Nh4 Qd8 11.Qb1 Be7 12.Nf5 Bg5 13.f4 gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4 Qd7 17.Bf2 Qf5

click for larger view

Analysis by Rybka 2.3.2a mp 32-bit :23 ply

1. ² (0.35): 18.Qd3 Rd8 19.Rd1 0-0 20.g3 Be3 21.Bxe3 dxe3 22.Qxf5 Nxf5 23.Bg2 Na5 24.Bh3 Rxd1+

2. = (0.20): 18.Qxf5 Nxf5 19.Rd1 Be3 20.Rd3 Rd8 21.b3 Bxf2+ 22.Kxf2 Ne3 23.h3 Nc2 24.b4 f6

Jun-05-12

WinKing:

The problem with <16.Qd3 Ng6> is that white has a boatload of good responses to ...Ng6. This is where we could use one of <kutz>'s mega (12 pack) pv runs. :)

Jun-06-12

ajile:

Maybe just play to keep the extra pawn?

Jun-06-12

blue wave:

<WinKing> <The problem with <16.Qd3 Ng6> is that white has a boatload of good responses to ...Ng6. This is where we could use one of <kutz>'s mega (12 pack) pv runs. :)>

White has about 2 good replies to <16.Qd3 Ng6>.

As follows:

<hiarcs13.2 - selectivity[7] - PV[4] - playing style[active]

Depth : 28/1
Time : 08:58:41
Nodes : 76674357K
N/sec : 2372214

Score : +0.42
17. h4 Be3 18. h5 Nf4 19. Qf5 Ne6 20. Bf2 Bxf2 21. Kxf2 Qe7 22. Rd1 a5

Score : +0.20
17. Bf2 Qe7 18. Bxd4 Ngxe5 19. Nxe5 Nxd4

Score : +0.11
17. e6 Be3 18. exf7 Kxf7 19. h4 Qf6 20. Bxc7 Rac8 21. Bh2 Nge5

Score : +0.04
17. e3 Bxe3 18. Qf5 Qc8 19. Bd3 Qxf5 20. Bxf5 Nf4 21. Bxf4 Bxf4 22. O-O Be3 23. Kh1 Na5>

So <17.h4> is the most likely reply. And this leads on to a line that is pretty narrow in decent responses for both sides for a lot of moves. As far as I can see anyway.

All the other responses I would be happy to see as they have low scores for white.

I'm still looking more at the main line.

****

Currently I have this:

1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. a3 g6 6. Bg5 Qd7 7. Nbd2 h6 8. Bf4 g5 9. Bg3 g4 10. Nh4 Qd8 11. Qb1 Be7 12. Nf5 Bg5 13. f4 gxf3 14. Nxf3 Bxf5 15. Qxf5 Nge7 16. Qd3 Ng6 17. h4 Be3 18. h5 Nf4 19. Qf5 Ne6 20. Bf2 Bxf2+ 21. Kxf2 Qe7 22. b4 O-O-O 23. g3 Kb8 24. Qf6 Qxf6 25. exf6 Ng5 26. Rh4 *

click for larger view

hiarcs13.2 - selectivity[4] - PV[1] - playing style[solid]

Score : +0.54
Depth : 30/1
Time : 02:03:14
Nodes : 21674473K
N/sec : 2931055

26... Rde8 27. Rf4 Ne4 28. Kg2 Rhg8 29. g4 a5 30. Kh3 Ng5 31. Nxg5 hxg5 32. Rf5 Re3 33. Rf3 Re4 34. Rg3 Ne5 35. bxa5 Rh8 36. Rc1 c5 37. Rb1 Nxc4 38. Rgb3 Ne3 39. Rxb7 Ka8 40. Bg2 Nxg2 41. Kxg2 Rxg4 42. Kh3 Re4 43. Rxf7 Rxh5 44. Kg3 Rhh4 45. Rb2

A score of +0.54 seems reasonable, as black is still one pawn down. And from what I can tell, usually hiarcs scores are higher then Rybka4.1 in general. Hence my request to see if <RV> has any work on this line at all.

There are a few other moves in this main line that would need to be looked at, around move 22 and 23 in this line.

Jun-06-12

blue wave:

click for larger view

hiarcs13.2

Depth : 26/1
Time : 00:15:22
Nodes : 2596650K
N/sec : 2816069

Score : +0.25
30. Kh2 Nd6 31. Nxd4 Ne5 32. Kh3 Nexc4 33. Bg2 Re3

Score : +0.24
30. b5 Nd8 31. a4 c5 32. bxc6EP Nxc6 33. Kh3 Re6 34. Kh4 Rge8 35. g5 Nxg5 36. Nxg5 hxg5 37. Kxg5 Re3 38. Kh4 Rh8 39. Rf5

Score : +0.05
30. Kh3 Ng5 31. Nxg5 hxg5 32. Rf5 Re3 33. Rf3

I think black is doing OK here.

Jun-06-12

blue wave:

But I have no answer for <23.Rd1>.

click for larger view

hiarcs13.2 - selectivity[4] - PV[3] - playing style[solid]

Depth : 24/1
Time : 00:09:32
Nodes : 1520433K
N/sec : 2655159

Score : +1.16
25... a5 26. b5 Na7 27. g3 c6 28. Bh3 cxb5 29. Bxe6 fxe6 30. Rxd4 Rxd4 31. Nxd4 Rf8 32. Nxb5 Rxf6 33. Kg2 Nxb5 34. cxb5 Rf5 35. a4 Re5 36. Kf3

Score : +1.21
25... Rd6 26. Rh4 Re8 27. Re4

Score : +1.28
25... Rhe8 26. Rh4 Rd6 27. Re4 a6 28. g3 Rg8 29. Bg2 Rgd8 30. Nh4

This move appears to push the scores much higher for white, leaving black with no reply.

OK. I've seen enough of <16...Ng6> now .

Jun-06-12

RandomVisitor:

Concerning <16.Qd3 Ng6>, my deep computer run in the position after 16.Qd3 lists the 16...Ng6 refutation as 17.h4 and "isworse".

Jun-06-12

kutztown46:

After 13. f4 gxh3 14. Nxf3 Bxf5 15. Qxf5 Nge7 16. Qe4 Qd7 17. Bf2 Qf5:

click for larger view

Analysis by Stockfish 2.2.2 JA (39-ply, 23 hours):

1. ² (0.60): 18.Qxf5 Nxf5 19.g3 Ne3 20.h4 Nc2+ 21.Kd1 Be3 22.Kxc2 Bxf2 23.Bh3 Ke7 24.Rad1 Bxg3 25.Nxd4 Nxd4+ 26.Rxd4 Rad8 27.Re4 Rhg8 28.b4 h5 29.Rf1 b6 30.Bg2 Bh2 31.Bh1 Rg1 32.Rxg1 Bxg1 33.Bf3 Rd4

2. ² (0.52): 18.Qd3 Qe6 19.g3 Ng6 20.h4 Ncxe5 21.Nxe5 Nxe5 22.Qb3 Bf6 23.Bxd4 Qc6 24.Rg1 Nf3+ 25.exf3 Bxd4 26.Rg2 0-0-0 27.0-0-0 Qf6 28.f4 Rde8 29.Re2 Rxe2 30.Bxe2 c6 31.Kb1 Re8 32.Bg4+ Kb8 33.Rd3 c5

3. = (0.00): 18.Qb1 Rd8 19.Qxf5 Nxf5 20.g3 0-0 21.Bh3 Ne3 22.Bxe3 Bxe3 23.Bf5 Rfe8 24.e6 fxe6 25.Be4 Na5 26.Ne5 Nb3 27.Rd1 Rf8 28.Bc2 Bf2+ 29.Kf1 Be3+ 30.Ke1 Bf2+ 31.Kf1

4. ³ (-0.28): 18.Nxg5 hxg5 19.Qxf5 Nxf5 20.g4 Ne3 21.Bxe3 dxe3 22.0-0-0 Rd8 23.Rxd8+ Kxd8 24.Bg2 Nxe5 25.h3 c6 26.Rd1+ Kc8 27.Rd6 Re8 28.c5 Ng6 29.Bf3 Re7 30.Rf6 Nf8 31.Rf5 Ne6 32.Re5 Kd7 33.Rxe3 Nxc5

Jun-06-12

Garech:

Hi team, hope everyone has had a good day. I'm currently studying for maths exams but I haven't been able to resist looking at different variations all day as most of my maths work is located in the same place as Fritz =0)

I'm very happy that RV has pointed out the option of 13...gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qd3 Qd7 17.Bf2 Qe6!? as this is one of the main lines I have focused on today.

My last post detailed the problems with 17...Qf5; I.e. that it leads to a good endgame advantage for white. <The Long Diagonal> contests this and I leave it for the rest of the team to decide. The line goes: Qf5 18.Qd3 Rd8 19.g3 Qxd3 20.exd3 Ng6 21.h4 Be7 22.Bg2 Ngxe5 23.Nxe5 Nxe5 Kd2. This is why I looked for alternatives to 17...Qf5 and, after some considerations, Qe6 came as a viable alternative.

My mainline continues with 18.Bxd4 f5!? (0-0-0 is also possible here but the lines transpose) 19.Qd3 0-0-0 20.e3 Ng6 21.Qc3 Rhe8 22.Be2 (Nxg5 transposes) Ngxe5 23.Nxg5 hxg5 24.0-0-0:

click for larger view

an interesting position but one that is surprisingly static as it's to both sides disadvantage to exchange. We remain the pawn down but it's tough for white to make progress; we are controlling equal shares of the board. I'd be very interested to know how other engines assess the position at a decent depth. (I'm using Fritz 12).

Okay - so that's the mainline. There's a lot to consider along the way too.

For instance; 20...f4!? and it seems that white is perilously close to falling apart, but he can just hold the balance: 21.Qe4 fxe3 22.Be2 Nxd4 23.Nxd4 Qg6 24.Qg4+ Kb8 25.h4 h5 26.Qxg5 Rxd4 and white maintains an edge in all lines.

Also, there is 20...Nxe5!? 21.Nxe5 c5! (this pin will last several moves) 22.Be2 Bf6 23.Nf3 Rhg8 (Nc6 is also playable but this dares white to 0-0, or Rxg2 is coming):

click for larger view

24.0-0 Nc6 25.Kh1 cxd4 26.exd4 and we can choose how to recapture. The tactical melee has resulted in our f pawn becoming weak, which is something of a problem and means that white still maintains the edge. It's a nice-looking line but probably best avoided.

There is also 21...f4. White can answer safely with 22.0-0-0 or Be2 but there is also h4!? where we have a nice combination: 22...fxe3!

click for larger view

23.hxg5 Nxd4 24.Nxd4 Qg4 25.Qxe3 Rxd4 26.Qh3 (the only saving move) Nxe5 27.Be2 and it's about equal. (I didn't slide this line but the tactical sequence is forced in any case).

So, it seems that after 17...Qe6, the mainline (first diagram) is the best we can hope for - as I said, I would love to know what other engines (and people!) think.

-Garech

Jun-06-12

blue wave:

Spent a bit of time looking at <17.b4> in one of our main lines.

1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. a3 g6 6. Bg5 Qd7 7. Nbd2 h6 8. Bf4 g5 9. Bg3 g4 10. Nh4 Qd8 11. Qb1 Be7 12. Nf5 Bg5 13. f4 gxf3 14. Nxf3 Bxf5 15. Qxf5 Nge7 16. Qd3 Qd7 <17. b4 *>

click for larger view

hiarcs13.2 - PV[3]
Depth : 28/1
Time : 04:26:39
Nodes : 36854993K
N/sec : 2303576

Score : +0.49
17... Be3 18. b5 Na5 19. Bf2 Nf5 20. c5 Rd8 21. Rb1 Qe6 22. g3 Bxf2 23. Kxf2 Ne3 24. Rb4 Nac4 25. Bg2 Ng4 26. Kg1 Ngxe5 27. Nxe5 Nxe5 28. Qd2 Qe7 29. Bxb7 Qxc5

Score : +0.50
17... Nf5 18. Bf2 Be3 19. g3 Bxf2

Score : +0.58
17... O-O-O 18. b5 Na5 19. Bf2 Qe6 20. h4 Be3 21. Bxe3 dxe3 22. Qxe3 Kb8 23. Rc1 Nxc4 24. Qc5 Nb6 25. e4 Rhg8 26. h5 c6 27. Kf2 f6 28. Rh3 cxb5 29. exf6 Qxf6

Jun-06-12

Garech:

I agree it was strange that the lines were listed as such; it was an overnight run and so one would assume that the Fritz would have been done thinking! I'm not sure why it happened - it may have something to do with cutting and pasting the lines but I'm not sure. In any case, there is pretty much no doubt in their order.

The other line I looked at in depth was 13...gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4 Be3!?

This was actually a mistake on my part as I meant to go through the 16.Qd3 Be3 variations, but the analysis is in depth and may transpose in some lines so I thought I'd post it anyway:

17.Bh4 (maybe suboptimal due to the mistake in which line to analyse) Qd7 18.Bf6 Rg8 19.g3 (Rd1 a good alternative) Qe6 20.Qd3 0-0-0 21.b4 (Bg2!?) Rd7 - this is an interesting moment; we can very nearly play Ng6!

click for larger view

with a forcing line:

22.Bxd8 Ngxe5 23.Nxe5 Nxe4 24.Qe4 Qxc4 25.Bh3+ Kxd8 26.Rd1 (the only move to keep the advantage) Qc3+ 27.Kf1 Rg5 28.Kg2 and white survives with more material (and probably wins).

So, to continue: 22.Rb1 Kb8 23.Bg2 Ng6 24.b5 Nce7 25.Nh4 (c5!?) Nxh4 26.Bxh4 Qxe5 27.Rf1 Rd6 28.Rxf7 Ng6 29.Rf5 Qe8 30.Bf6 Nf4:

click for larger view

and white's advantage continues to grow - so we should avoid this. I believe that in the 16.Qd3 Be3 lines white will be playing this with an extra tempo and that can't be good. I will check it today to make sure there are no nuances or alternatives available to us.

I think it's important that we keep the game as complex and non clear-cut as possible in order to make it hard for the other team. I realise that we are ultimately playing for a draw but it would be truly something if we turned it into a win!

I will look at 16.Qd3 Be3 today.

-Garech

Jun-06-12

Garech:

P.s. perhaps some of you are using Fritz or other engines that allow the posting of games, if so here is my analysis that you can paste straight in to software:

[Event "?"]
[Site "?"]
[Date "2012.06.05"]
[Round "?"]
[White "Chessgames challenge"]
[Black "16...Be3"]
[Result "*"]
[ECO "D08"]
[Annotator "Godley,Sean"]
[PlyCount "50"]

1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. a3 g6 6. Bg5 Qd7 7. Nbd2 h6 8. Bf4 g5 9. Bg3 g4 10. Nh4 Qd8 11. Qb1 Be7 12. Nf5 Bg5 13. f4 gxf3 14. Nxf3 Bxf5 15. Qxf5 Nge7 Fritz 12: 1) 16. Qe4 Qd7 (16... Be3 17. Bh4 Qd7 18. Bf6 Rg8 19. g3 (19. Rd1 is a good alternative with same eval) 19... Qe6 20. Qd3 (20. Rd1 also playable) 20... O-O-O 21. b4 (21. Bg2 also playable; same eval) 21... Rd7 (21... Ng6 !? interesting; worth exploring 22. Bxd8 Ngxe5 23. Nxe5 Nxe5 24. Qe4 Qxc4 25. Bh3+ Kxd8 26. Rd1 other moves lead to a perpetual (two lines) Qc3+ 27. Kf1 Rg5 28. Kg2 c6) 22. Rb1 (22. Bg2 alternative and equal) 22... Kb8 (22... h5 also possible) 23. Bg2 Ng6 24. b5 (24. Rf1 also possible) 24... Nce7 25. Nh4 (25. c5 same eval and sharp!) 25... Nxh4 26. Bxh4 Qxe5 27. Rf1 (27. Qe4 also playabe but leads to simplification Qxe4 28. Bxe4 f5) 27... Rd6 28. Rxf7 Ng6 29. Rf5 Qe8 30. Bf6 Nf4 (30... Rf8 also playable but slightly worse) 31. Rxf4 Bxf4) 17. Bf2 Fritz 12: Qe6 sliding: 18. Bxd4 ( 18. Rd1 also possible) 18... f5 (18... O-O-O also possible 19. e3 f5) 19. Qd3 (19. Qc2 also possible but 0.0) 19... O-O-O (19... Ng6 probably worse 20. Qc3 Nxd4 21. Qxd4 Rd8 22. Qc3 Bf4 slided; better for white) 20. e3 Ng6 ( 20... Nxe5 !? 21. Nxe5 c5 22. Be2 Bf6 23. Nf3 (23. Qc3 second choice but +0. 4 instead of +0.8) 23... Rhg8 (23... Nc6 also possible and same eval but hg8 encourages 0-0 24. O-O-O cxd4 better for white but interesting) 24. O-O Nc6 25. Kh1 cxd4 26. exd4 Bxd4) (20... f4 21. Qe4 (21. Rd1 same eval) (21. Be2 same eval) 21... fxe3 22. Be2 (22. Bc3 also possible but slightly worse ) 22... Nxd4 23. Nxd4 Qg6 24. Qg4+ Kb8 25. h4 h5 (25... Rxd4 dubious but interesting 26. Qxd4 Nc6 27. Qg4 Nxe5 28. Qd4 Bf6 29. O-O-O Qxg2) 26. Qxg5 Rxd4 27. O-O (27. Qxg6 also good for an edge - white is definitely better)) 21. Qc3 Rhe8 (21... f4 !? 22. h4 (22. O-O-O also possible fxe3 23. Nxg5 hxg5 24. Qxe3 slided; better for white) (22. Be2 also possible) Fritz 12: 22... fxe3 ! 23. hxg5 Nxd4 24. Nxd4 Qg4 25. Qxe3 Rxd4 26. Qh3 Nxe5 27. Be2 Re4 28. Qxg4+ Nxg4 29. Rh4 unslided) 22. Be2 (22. Nxg5 equally good) 22... Ngxe5 23. Nxg5 hxg5 (23... Qg6 !? 24. h4) 24. O-O-O (24. O-O !?) 24... g4 ( 24... a6 !? same eval as first line 25. Rd2 (25. Rhf1 same eval as first line)) (24... Kb8) (24... f4) (24... Nd7) 25. c5 b1 he1 also possible but passive a6 *

-Garech

Jun-06-12

cro777:

<jepflast: At 17...Qf5, 18. Qd3 was the initial computer-best, however due to 18...Qe6, transposing, this was dependent on the original 16. Qd3 17. Bf2 line working out, which it didn't. Therefore, we need 18. Qxf5 to work.>

<frogbert: If we play 16. Qe4 Qd7 17. Bf2, then we either must

1) accept transposition to the <16. Qd3 & 17. Bf2-line> that we don't like, or

2) manage to create something from 18. Qxf5 Nxe5>

1. d4 d5 2. c4 e5 3. dxe5 d4 4. Nf3 Nc6 5. a3 g6 6. Bg5 Qd7 7. Nbd2 h6 8. Bf4 g5 9. Bg3 g4 10. Nh4 Qd8 11. Qb1 Be7 12. Nf5 Bg5 13.f4 gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4 Qd7 17.Bf2 Qf5 18.Qxf5 Nxf5

click for larger view

<19.g3> (+0.28 Houdini 2 at 30 ply)

Black answers <19...Rd8> or <19...Ne3>

Jun-06-12

Chessgames Challenge:

13...gxf3

FINAL VOTE TALLY:

13...gxf3

votes

(100.0%)

total # of votes: 28
draw requests: 1 (3.6%)

click for larger view

Jun-06-12

Chessgames Challenge:

13...gxf3

click for larger view

Jun-06-12

frogbert:

what a shocker!

Jun-06-12

chrisowen:

13.f4 leads to gxf3 at choose knight on da d2 adding to support in e5 fog clears up port bc8f5 at 15 Qxf5 nge7 and I'd say it is decision time on where to scatter e4 or d3 look best giving clearance to surge in queenside it means c4 becomes a nit picking issue as na5 pardons duty from c6 and looks to recapture the gain we have been trying to maintain a lead above and beyond e5 really weve got to swamp a d4 win exchange for e5 staple plan.

13.f4 gxf3 14.nxf3 bxf5 15.qxd5 nge7

click for larger view

Jun-06-12

lost in space:

Lost most the the 1,5 day run on 16. Qg4 Be3 17. Bh4 h5 18. Qe4 Qd7 19. Bf6 Rh6 20. g3 due to a electrical power outage. @#$%*&!#, have to start again

Jun-06-12

lost in space:

<<RandomVisitor:> 13...gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 and now: 16.Qe4 <Qd7><Qc8><Be3>

16.Qd3 <Qd7><Be3 is likely bad>

16.Qc2 <Be3><Qc8><Qd7>

16.Qh3 <Be3>

16.Qg4 <Be3>

Alternates? It might be too early (snip)>

I agree with

16.Qe4 <Qd7><Qc8><Be3>

16.Qd3 <Qd7>

16.Qh3 <Be3>

16.Qg4 <Be3>

16. Qc2 I have not locked at

Jun-06-12

RandomVisitor:

<Rybka4.1> Here is my understanding of the situation. Use these baselines as starting points for sliding forward analysis - <try starting a few moves into the line, never at the end>

<new<>new> indicates new updates

Variation A: 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4

Variation B: 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qd3

Variation C: 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.other

Variation D: 14.Nxf3 Nge7

*** *** *** *** *** *** *** *** *** *** *** *** *** ***

Variation A: 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4

<new<>new>[+0.42] d=30 16...Qc8 17.Bf2 Qf5 18.Qxf5 Nxf5 19.g3 Ne3 20.h4 Nc2+ 21.Kd1 Ne3+ 22.Kc1 Be7 23.Nxd4 Nxd4 24.Bxe3 Nb3+ 25.Kb1 Nxa1 26.Kxa1 Bf8 27.Bg2 c6 28.Rf1 Bg7 29.Rf5 O–O 30.Bc5 Rfd8 (19:33:43) 12328764kN <16...Qd7 transposes>

<new<>new>[+0.58] d=30 16...Be3 17.Bh4 Qd7 18.Bf6 Rg8 19.g3 Qe6 20.Rd1 Rg6 21.Qd3 Ng8 22.Bh8 h5 23.Nh4 Nge7 24.Bg2 Rg8 25.Bf6 Nxe5 26.Bxe5 Qxe5 27.Nf3 Qf5 28.Qxf5 Nxf5 29.Bh3 Nd6 30.Nxd4 Rd8 31.Nc2 Bg5 (21:47:34) 13838393kN

[+0.44] d=33 13.f4 gxf3 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4 Qd7 17.Bf2 Qf5 18.Qxf5 Nxf5 19.g3 Ne3* 20.h4 Nc2 21.Kd1 Ne3 22.Kc1 Be7 23.Nxd4 Nxd4 24.Bxe3 Nb3 25.Kb1 Nxa1 26.Kxa1 Bf8 27.Bg2 c6 28.Rf1 (210:15:25) 240609419kN

*[+0.40] d=31 20.h4 Nc2+ 21.Kd1 Ne3+ 22.Kc1 Be7 23.Bxe3 dxe3 24.Bh3 Rd8 25.Rd1 Rg8 26.Rxd8+ Nxd8 27.Kc2 Rxg3 28.Bc8 a5 29.Rb1 b6 30.h5 a4 31.Rd1 Rg2 32.Bd7+ Kf8 33.Re1 c6 34.Nd4 Rg5 35.Rf1

*** *** *** *** *** *** *** *** *** *** *** *** *** ***

Variation B: 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qd3

<new<>new>[+0.21] d=31 16...Qd7 17.Bf2 Qe6** 18.g3 Ng6 19.Bg2*** Ngxe5 20.Qb3 O–O–O 21.Nxe5 Qxe5 22.O–O Rhe8 23.Rad1 a6 24.Qa4 Qc5 25.Qc2 Qe5 26.Bf3 Kb8 27.b4 Be3 28.Bxc6 bxc6 29.Bxe3 (33:39:28) 38777686kN

**a[+0.19] d=27 18.g3 Ng6 19.Bg2 Ngxe5 20.Nxe5 Nxe5 21.Qxd4 Rd8 22.Qxa7 Bd2+ 23.Kf1 O–O 24.Qxb7 c6 25.Qb3 Nxc4 26.Qa4 Be3 27.Bf3 Nd2+ 28.Kg2 Nxf3 29.Bxe3 Nd2 30.Qf4 Rfe8 31.Bxd2 Qxe2+ 32.Qf2 Rxd2 33.Qxe2 (92:08:24) 15281601kN

b[+0.33] d=26 18.g3 Ng6 <19.h4> Ngxe5 20.Nxe5 Nxe5 21.Qb3 Bf6 22.Bxd4 Qc6 23.Rg1 Nf3+ 24.exf3 Bxd4 25.Rg2 0-0-0 26.0-0-0 [+0.33] d=33 26...Rhe8 27.Rgd2 Qf6 28.Bh3+ Kb8 29.Bf5 c5

***a [+0.26] d=28 18.g3 Ng6 <19.Bg2> Ngxe5 20.Nxe5 Nxe5 21.Qxd4 Rd8 22.Qxa7 Bd2+ 23.Kf1 0-0 24.Qxb7 c6 25.Qb3 Nxc4 26.Qa4 Be3 27.Bf3 Nd2+ 28.Kg2 Nxf3 29.Bxe3 Nd2 30.Qf4 Rfe8 31.Bxd2 Qxe2+ 32.Qf2 Rxd2 33.Qxe2

b 19.h4 [+0.00] d=26 19...Ngxe5 20.Nxe5 Nxe5 21.Qb3 d3 22.Qxb7 Bd2+ 23.Kxd2 O–O 24.exd3 Rab8 25.Qd5 Qf6 26.Rb1 Rfd8 27.Qc5 Ng4 28.Kc2 Nxf2 29.Rh2 Nxd3 30.Bxd3 Rxd3 31.Rf2 Rf3 32.Rxf3 Qxf3 33.Qxc7 Qe4+ 34.Kc3 Qe3+ (14:37:11) 8464467kN

*** *** *** *** *** *** *** *** *** *** *** *** *** ***

Variation C: 14.Nxf3 Bxf5 15.Qxf5 Nge7 16.other

<new<>new> [+0.30] d=28 16.Qc2* Be3 17.Bh4 Qd7 18.Bf6 O–O 19.g3 Ng6 20.Bg2 Rfe8 21.Rf1 Qg4 22.b4 Rab8 23.Rd1 a6 24.Qb3 (39:43:29) 27977780kN

<new<>new> [+0.12] d=28 16.Qh3 Be3 17.b4 b5 18.Bf2 bxc4 19.g3 Bxf2+ 20.Kxf2 Qd5 21.Bg2 Qe4 22.Rhc1 Rd8 23.Rxc4 Qe3+ 24.Kf1 d3 25.Re1 Nxe5 26.Rc5 d2 27.Rxe5 dxe1Q+ 28.Kxe1 Qxa3 29.Qh4 f6 30.Re4 Rd6 31.Kf2 (41:57:43) 30158529kN

[+0.22] d=21 16.Qg4 Be3 17.Bh4 h5 18.Qg7 Rf8 19.Qh7 Qd7 20.g3 O–O–O 21.Bf6 Qe6 22.b3 Qg4 23.Bg2 Kb8 24.Rd1 Nf5 (0:17:09) 268966kN

*[+0.10] d=27 16...Be3 17.Bh4 Qd7 18.Bf6 0-0 19.g3 Ng6 20.Bg2 Rfe8 21.Rd1 Ngxe5 22.Nxe5 Nxe5 23.Bd5 c6 24.Bxe5 cxd5 25.Bxd4 Rac8 26.Bxe3 Rxe3 27.0-0 Qe6 28.Rf2 dxc4 29.Qd2

*** *** *** *** *** *** *** *** *** *** *** *** *** ***

Variation D: 14.Nxf3 Nge7

15.e4 dxe3 16.Nxe7 Nxe7 <virtually forced>

[+0.88] d=24 17.Be2 Qd7 18.Qe4 Nf5 19.Rd1 Qc6 20.Nxg5 hxg5 21.Qd3 Qd7 22.Qb3 Qe7 23.0-0 Nxg3 24.hxg3 Qxe5 25.Rf3 0-0 26.Rxe3 Qf6 27.Qb4 c6 28.Rd6 Be6 29.Qxb7 Rab8 30.Qxc6 Rxb2 31.Qe4 Ra2 32.Rf3

Jun-06-12

jepflast:

Where exactly do we stand on the <16. Qd3> 17. b4 lines that karpkasp has been working diligently on?

Jun-06-12

jepflast:

<14.Nxf3 Bxf5 15.Qxf5 Nge7 16.Qe4 Qd7> I've been trying to squeeze something out of 17. Rd1 and 17. h4, but to no avail. 17. b4 is also no good. It seems that 17. Bf2 is the only move here.

Jun-06-12

AylerKupp:

<<lost in space>: Lost most the 1,5 day run due to a electrical power outage.>

Maybe you can get a good deal on an Uninterruptible Power Supply (UPS) from <RandomVisitor>. He probably buys them in quantity and gets a volume discount, and he may be willing to pass on the savings. :-)

Jun-06-12

AylerKupp:

Updated AT spreadsheet, Jun-06-12 (part 1 of 2).

The updated AT spreadsheet is ready for downloading from (part 1 of 2):

https://rapidshare.com/files/774858...

It contains all the posted analyses through Jun-06-12, 9:00 AM PDT, just before Black's 13th move (13...gxf3) was posted. And since 13...gxf3 was the only move for which we had any analyses, all the previous analyses in move tab B13 were moved to move tab W14 and move tab B13 was deleted.

White's only reasonable (dare I say "forced"?) recapture is 14.Nxf3 since both 14.exf3 and 14.gxf3 lose a piece after 14...Bxd2+ 15.Kxd2 Qg5+ and 16...Qxf5. I guess I lose my prediction of 14.exf3.

I finished making all the modifications for what I thought was a necessary fix to the logic to calculate the average evaluations for PV=1 but now I'm no longer sure that my logic was wrong to begin with. Here is my quandary and maybe you can help me out:

When you run a Multi-PV analysis with your chess engine the line with the highest evaluation (if White is making the first move) or the line with the lowest evaluation (if Black is making the first move) is the Principal Variation or PV for short. The PV represents the best play by both players and is generated using the minimax algorithm. The other non-best play lines can be valuable since they can show other interesting ideas, which lines should be avoided, and, after all, Team White may not play the best move and we need to be prepared for any reasonable eventuality. But it would be rash to assume that Team White won't play the best move so in order to make the best decision as to what move to play next we should really only consider the PV.

I record the PV number (which is really is a misnomer since there can only be one PV but we're kind of stuck with the terminology) for each line or variation posted that has an evaluation. My original version of the AT spreadsheet simply counted and averaged the evaluations for all variations which had PV=1. Then I noticed that the average evaluation for those lines with PV=1 was not that much different than the average evaluation taking all lines into account, regardless of PV number, and I thought that there ought to be a bigger difference between the two.

So I thought that I should instead include in the average PV calculation for PV=1 only those lines where the analysis started in the position reached by the previous move. Thus, if we are currently at move W13, the average evaluation calculation should only take into account the analyses that started from the position after move B12, since those are the only analyses where a direct association can be made with White's 13th move and the resulting PV=1. But that results in a much smaller number of analyses included in the calculations so that doesn't seem right either.

So now I'm thinking that my original logic was correct after all. Whether the PV was calculated starting from the immediately preceding move or not is not important since the <line> was the result of an analysis that identified it as the PV; the actual move when the analysis was started is immaterial. Of course, the concept of averaging the evaluations of different engines to reduce individual engine biases may not be valid or have any value whatsoever, so the issue may be moot. But I think that the concept of determining the "proper" evaluation of each candidate move is important and it's worth the effort to try to calculate it.

For more information on the Principal Variation refer to the AT Tutorial section at AylerKupp chessforum.

Jun-06-12

AylerKupp:

Updated AT spreadsheet, Jun-06-12 (part 2 of 2).

Any suggestions will be greatly appreciated. If my original logic was right then that was a lot of wasted effort but that's the way it goes. We have often flushed a lot of analysis down the toilet, most recently for 13.e4. And knowing when the analysis of a given line was started ( i.e. what move) is useful information since the earlier the analysis was started the less likely that this actual continuation will be played in the game.

Besides, all his work let me to do some needed cleanup:

1. There were many duplicate analyses for some reason. I've identified and deleted these. And this convinced me that I need to incorporate some duplicate analysis checking before I add a line to the AT spreadsheet.

2. There were some duplicate Analysis IDs, something I said would never happen. Hell and damnation again!, or at least virginity and abjection! Anyway, I reassigned all the Analysis IDs so that they are, once again, unique (and also sequential), and I listed the Old Analysis IDs in a separate column and highlighted them <gray italics> in case you previously used them for one reason or another. Eventually I will delete this column.

As a result of this cleanup, and still for unknown reasons, the AT spreadsheet size was cut to less than half, from more than 10 MB in the previous version two days ago to less than 5 MB in the current version, even though more analyses were added. So it was a worthwhile exercise in preventing further runaway AT spreadsheet growth. But no, I don't know how to prevent this sudden growth again (although I do know how to fix it).

I also enhanced the Focusing capability. Earlier if two highlighted moves (H-Moves) were specified and you chose to focus on the analyses containing these moves (by pressing Ctrl-Shift-F), only the analyses that contained either of these moves were displayed, and there was an implicit OR relationship between H-Move1 and H-Move2. I decided to make this relationship explicit by listing it just to the left of H-Move2, and I created a pull-down menu so that in addition to an OR relationship between the H-Moves you can specify an AND or a NOT relationship between the H-Moves as follows:

OR: The analyses will be listed if <either> H-Move1 <OR> H-Move2 are included in the analysis line.

AND: The analyses will be listed only if <both> H-Move1 <AND> H-Move2 are included in the analysis line.

NOT: The analyses will be listed only if H-Move1 <AND NOT> H-Move2 are included in the analysis line; that is, if H-Move1 is included in the analysis line and H-Move2 is not.

Just click on the cell containing the relationship and you'll see an upside down arrow. Clicking on this arrow will show the pull-down menu. Or you can just type in OR, AND, or NOT. When there are many analyses available this feature may help in reducing the number of analyses displayed even further.

< Earlier Kibitzing · PAGE 96 OF 240 · Later Kibitzing>

NOTE: You need to pick a username and password to post a reply. Getting your account takes less than a minute, totally anonymous, and 100% free--plus, it entitles you to features otherwise unavailable. Pick your username now and join the chessgames community!

If you already have an account, you should login now.

Please observe our posting guidelines:

No obscene, racist, sexist, or profane language.
No spamming, advertising, or duplicating posts.
No personal attacks against other users.
Nothing in violation of United States law.

See something which violates our rules? Blow the whistle and inform an administrator.

NOTE: Keep all discussion on the topic of this page. This forum is for this specific game and nothing else. If you want to discuss chess in general, or this site, you might try the Kibitzer's Café.

Messages posted by Chessgames members do not necessarily represent the views of Chessgames.com, its employees, or sponsors.