|
< Earlier Kibitzing · PAGE 507 OF 838 ·
Later Kibitzing> |
| Apr-22-07 | | TefthePersian: 10 hours?! My system is much slower and reaches 20 ply in maybe an hour in complex positions. What do you have your selectivity set to, and what nodes or kilonodes/second are you getting? |
|
| Apr-22-07 | | Knightlord: < afioonov: and when im playing i hear the symphony No.7 by beethoven. >
Now you're talking! |
|
| Apr-22-07 | | RandomVisitor: After 29...a6:
(21-ply)Rybka2.3.1
1. (-0.03): 30.b5 axb5 31.axb5 Nc5 32.Ba3 Re8 33.Bxc5 bxc5 34.b6 Rxe5 35.Qa4 Qe8 36.Qxe8+ 2. (-0.30): 30.h3 b5 31.axb5 axb5 32.Bc3 Qe6 33.Bb2 Rf7 34.Re1 h6 35.Qd1 d4 36.Re4 d3 3. (-0.30): 30.Bf1 b5 31.axb5 axb5 32.Bg2 Qe6 33.h3 Rf7 34.Re1 h6 35.Qd1 d4 36.Re4 d3 4. (-0.30): 30.Bh3 b5 31.axb5 axb5 32.Bg2 Qe6 33.h3 Rf7 34.Re1 h6 35.Qd1 d4 36.Re4 d3 5. (-0.30): 30.Bh1 b5 31.axb5 axb5 32.Bg2 Qe6 33.h3 Rf7 34.Re1 h6 35.Qd1 d4 36.Re4 d3 6. (-0.36): 30.Bc3 Qe6 31.Bb2 Kh8 32.b5 axb5 33.axb5 Rf4 34.Qc3 d4 35.Qb4 h6 36.h3 Rf5 7. (-0.36): 30.Bd4 Qe6 31.Bb2 Kh8 32.b5 axb5 33.axb5 Rf4 34.Qc3 d4 35.Qb4 h6 36.h3 Rf5 |
|
| Apr-22-07 | | whiteshark: <We need to add more protection to our c4 Bishop>, but not with (tempo wasting moves like) 29... a6.
Instead 29...Kh8 is an <indirect protection> move which avoids the Qxc4 threats. Thereafter Bc4 is well protected and safe. Fear or concerns are bad advisors. |
|
| Apr-22-07 | | Thorsson: An interesting choice.
29...a6. At first sight this gives White another bite at the Bh3 cherry, but Black profits from the pawn no longer being on a7. Of course if a4 was to prevent b5 then a6 allows us to play it again, but 30.b5 axb5 31.axb5 Nc5 looks playable for White. 29...Kh8. The point of this is to avoid the tactics where B takes on d5 after our B is distracted, and also it makes Bh3 less potent. It seems slow, but it's claimed that White has no good move. Is 30.Bf1 playable? 29...Qe6 prevents pawn advances and Bh3, but removes the Queen from the f-file. 30.h3 is claimed as equal, but possibly 30...Kh8 is a good reply. |
|
| Apr-22-07 | | maxundmoritz: <Hugin> What are your thoughts regarding a continuation after 29...Kh8 30. f3 Qe6 31. Bf1 Bxf1 32. Rxf1 d4 33. f4? |
|
| Apr-22-07 | | Hugin: afioonov: is it a draw?:
29...Kh8 30. Bf1 a5 31. Bxc4 dxc4 32. Ba3 Re8 33. bxa5 bxa5 34. Rd6 Qf435. Qc3 h6 36. Bb2 Nc5 37. Qg3 Qf8 38. Rc6 Ne4 39. Qg2 c3 40. Bxc3 Nxc3 41. Rxc3 Rxe5 * <<<
30 Bf1 in that variation is meet with Qe6 Qh5 Qf3 or Qf4 all a lot better then 30 - a5? |
|
| Apr-22-07 | | Boomie: (Tef: <Boomie>: Check out 32...Qh5.> Yep. Removed my message. Thanks. |
|
| Apr-22-07 | | TefthePersian: <maxundmortiz>: Don't discount 30...Kg8. Black doesn't need to fear Qxc4/Bxd5 anymore. |
|
| Apr-22-07 | | Hugin: maxundmoritz: <Hugin> What are your thoughts regarding a continuation after 29...Kh8 30. f3 Qe6 31. Bf1 Bxf1 32. Rxf1 d4 33. f4?<<<<<<< Interesting...hm white supports the e pawn..I need more time on that one.....jesus i hope it's not a refutation....posting some tomorrow..) |
|
| Apr-22-07 | | ganstaman: <RandomVisitor: After 29...a6: (21-ply)Rybka2.3.1
1. (-0.03): 30.b5 axb5 31.axb5 Nc5 32.Ba3 Re8 33.Bxc5 bxc5 34.b6 Rxe5 35.Qa4 Qe8 36.Qxe8+> Bah, I still don't like it! (31...Nc5, that is). |
|
| Apr-22-07 | | TefthePersian: I don't either. Right now I'm leaning toward 29...Kh8. |
|
| Apr-22-07 | | ganstaman: Oh, goody. Checking RV's forum I see that after 29...a6 30. b5 axb5 31. axb5, 31...Nc5 doesn't seem as good with sliding analysis. 31...Re8 or oddly 31...h6 seem better (or I'm not reading things right). |
|
Apr-22-07
 | | GoldenKnight: Interesting talk about engine speeds. I have a 2.8 MHz hyperthreading Pentium 4 with only 1G of RAM with my hash tables set at the default (750, I think). Yet I get to 16 ply in about ten minutes and 20 ply in about an hour. Getting to 24 ply, however, requires an overnight run. I like <Thorsson>'s points. I'm going to investigate the last two: <29...Kh8. The point of this is to avoid the tactics where B takes on d5 after our B is distracted, and also it makes Bh3 less potent. It seems slow, but it's claimed that White has no good move. Is 30.Bf1 playable?29...Qe6 prevents pawn advances and Bh3, but removes the Queen from the f-file. 30.h3 is claimed as equal, but possibly 30...Kh8 is a good reply.> |
|
| Apr-22-07 | | Thorsson: <Hugin: 30 Bf1 in that variation is meet with Qe6 Qh5 Qf3 or Qf4 all a lot better then 30 - a5?> 30.Bf1 Qh5 31.Bxc4 seems to lead to a draw. If 30...Qe6 is good, then 29...Qe6 30.h3 Kh8 must be better! |
|
Apr-22-07
 | | chancho: I guess 29...Qe6 30.Qxc4 dxc4 31.Bd5 Qxd5 32.Rxd5 is not favorable for white. Specially after the black rook goes to c8, no? |
|
| Apr-22-07 | | Boomie: Here is RV's only Kh8 line:
(21-ply)
1. = (-0.15): 29...Kh8 30.f3 Qe6 31.Qf2 d4 32.Bxd4 Rd8 33.Kh1 Qf5 34.e6 Rxd4 35.Rxd4 Nxd4 36.Qxd4 Carrying it out a bit further:
36...Qb1+ 37. Qg1 Qxb4 38. Bh3 Qxa4 39. Qd4 Bb5 -0.29/15 Not a spectacular rating for Fritz but if black can survive the white initiative, he wins the endgame. However so far white seems to have enough for the draw.
click for larger view |
|
| Apr-22-07 | | maxundmoritz: <TefthePersian> I'm not discounting. I'm just trying to prune lines. By the way, I'm still looking for the "Teft endgame" reference. In my own analysis, 29. ... Qe6 just leads to a draw, but I'm curious what I'm missing. |
|
Apr-22-07
 | | Tabanus: Before 28.exd5 was played, we had studied our position in detail for at least a week, and as far as I know, when we were hoping for 28.a4 we had only two options left: a6
Qe6
I have not followed Qe6 (Teft endgame) very closely, in fact almost not, and just rely on others than it is a draw. Regarding a6, we had established that 29...a6 30.b5 axb5 31.axb5 <Nc5> was a draw because of 32.Ba3 h6 33.Bxc5, and we had discarded other 31st moves as well. We were then starting to look at RV's comp. move <31...Re8>, with this in mind: 32.h3 Nc5 33.Qc3 Ne6, and found that white could hardly play 34.Rxd5 here. RV gives 34.Qe3 Nf4 etc. and <izimbra>'s dream of freeing the b3 knight has come true! Thus the main a6 line so far is
29...a6 30.b5 axb5 31.axb5 Re8 32.h3 Nc5 33.Qc3 Ne6
Kh8 I don't know... |
|
Apr-22-07
| | Tabanus: <ganstaman: Oh, goody. Checking RV's forum I see that after 29...a6 30. b5 axb5 31. axb5, 31...Nc5 doesn't seem as good with sliding analysis. 31...Re8 or oddly 31...h6 seem better (or I'm not reading things right).> I think you are reading well. I don't remember what was wrong with 31...h6 though. We analyzed 31...h6, but where? |
|
| Apr-22-07 | | TefthePersian: <Thorsson>: 29...Kh8 30. Bf1 Qh5 31. Bxc4 dxc4 32. Rd6 a5  33. bxa5 Qg4 
33. Bc3 Qg4+
33. e6 axb4 |
|
| Apr-22-07 | | Boomie: An improvement for black in the RV Kh8 line.
29...Kh8 30.f3 Qe6 31.Qf2 d4 32.Bxd4 Rd8 33.Kh1 Qf5 34.e6 Rxd4 35.Rxd4 Nxd4 36.Qxd4 Qb1+ 37. Qg1 Qxb4 38. Bh3 <Bd5> 39. Qe3 Qxa4 40. e7 Bc6 41. Qe3 Qxa4 42. e7 -0.46/15 |
|
| Apr-22-07 | | Thorsson: <TefthePersian: 29...Kh8 30. Bf1 Qh5 31. Bxc4 dxc4 32. Rd6 a5 > I assume you're right, but what about 32.h3? |
|
| Apr-22-07 | | Artar1: <Hugin: <<<Kh8 is simply a weak move does not solve backrank problems for black. Surely h6 is much better in that variation.. put that into Dr fritz and it will do a lot better.> Hugin:
Your suggestion did not help matters for Black:
<27...Qf7 28.exd5 exd5 29.Bh3 b5 30.e6 Qh5 31.Bg2 Re8 32.Re1 d4 33.Qe4 Bxe6 34.Bxd4 Bf7 35.Be5 Nd2 36.Qf4 Nc4 37.Bc3 Rxe1+ 38.Bxe1> h6 39.Qb8+ Kh7 40.Qxa7 
click for larger view20-ply measured at move 40 by Deep Fritz 10:
1.  (0.36): 40...Qg6 41.Qa8 Qg5 42.h3 Qc1 43.Qe4+ Bg6 44.Qe6 Nxa3 45.Bd5 h5 46.Kg2 Qg5+ 47.Kh1 Qc1 48.Kh2 Qf4+ 2. (0.44): 40...Qe2 41.Qxf7 Qxe1+ 42.Bf1 Qe4 43.h3 Ne5 44.Qb3 Nf3+ 45.Kh1 Nd2+ 46.f3 Qe1 47.Qd3+ g6 48.Kg2 Nxf1 49.Qxf1 Qd2+ 50.Qf2 Qg5+ 51.Kf1 Qc1+ 52.Qe1 Qxa3 3. (0.44): 40...Qd1 41.Qxf7 Qxe1+ 42.Bf1 Qe4 43.h3 Ne5 44.Qb3 Nf3+ 45.Kh1 Nd2+ 46.f3 Qe1 47.Qd3+ g6 48.Kg2 Nxf1 49.Qxf1 Qd2+ 50.Qf2 Qg5+ 51.Kf1 Qc1+ 52.Qe1 Qxa3 4. (0.44): 40...Qe5 41.Bc3 Qxc3 42.Qxf7 Qe1+ 43.Bf1 Qe4 44.h3 Ne5 45.Qb3 Nf3+ 46.Kh1 Nd2+ 47.f3 Qe1 48.Qd3+ g6 49.Kg2 Nxf1 50.Qxf1 Qd2+ 51.Qf2 Qg5+ 52.Kf1 Qc1+ 5. (0.45): 40...Be6 41.Qd4 Qg6 42.Qe4 Nxa3 43.Qxg6+ Kxg6 44.Bc3 Kf7 45.f4 g5 46.Be4 gxf4 6. (0.53): 40...Be8 41.Qd4 Qg6 42.Qd5 Nxa3 43.Kh1 Bf7 44.Qa8 Qe6 45.Bc3 Qe2 46.Qe4+ Qxe4 47.Bxe4+ Bg6 48.Bc6 Bd3 49.Kg2 Nc2 When using Deep Fritz 10, I do not let the machine run unattended. In a given position, for example, I will let Fritz run until it has reached 20-ply, and then I will copy the best results out of a list of six candidate lines into the notation field. (In this position, it only takes my machine about seven minutes to hit 20-ply.) From there, I check each ply in the notation field to a depth of 20-ply to check the validity of each move. If need be, I repeat the process. This is how I conduct sliding forward analysis. Thus, each ply is examined on an individual basis to a depth of 20-ply. Unfortunately, your suggestion is not favorable for Black in my opinion. But now that Yury has played a4, all of this analysis is moot. |
|
| Apr-22-07 | | Artar1: <Hugin: 27 - Qf7 28 exd5 exd5 29 a4 Kh8* 30 f3 Qe6 31 Qf2 d4 32 Qc2 d3 33 Rxd3 Bxd3 34 Qxd3> White does not have to play 30.f3. Did you also test: 30.Kh1
30.Bf1
30.h3
These moves may be better for White.
|
|
 |
|
|
< Earlier Kibitzing · PAGE 507 OF 838 ·
Later Kibitzing> |
Kibitzing
