< Earlier Kibitzing · PAGE 3 OF 3 ·
|Mar-23-08|| ||sombreronegro: I have to think 21 Ng5 with the mate threat to force open the h file is the first move. Declining the knight sac is a must with 21 ... N-f6 this gives up the exchange in at least in two places. The rook on c8 or the knight fork on f7.|
|Mar-23-08|| ||Anatoly21: This was just too easy for a Sunday puzzle. I've already missed two this week, and it didn't even take me five minutes to calculate all the major variations for this one.|
|Mar-23-08|| ||MostlyAverageJoe: <Eyal> Good comments about passed pawn and active/well-placed pieces that black has in the exchange lines. Thanks. I also see that in these lines, black has some space advantage.|
<MostlyAverageJoe: Analysis of the exchange lines:...> ADDENDUM: I run an overnight forward scan (21-plies per move) on the 22.Bxc8 line. The analysis progress was very slow, indicating great complexity of the variants, but It did appear to be a draw. The last evaluation after 5 full moves was +0.07. Alas, I wiped out the results of 65 CPU hours of analysis by mistakenly hitting ctrl-V instead of ctrl-C, and pasting the GMT game on top of my analysis. Yuck.
<johnlspouge: <<MostlyAverageJoe> wrote: [snip] 22.Qg6 is the best move, no question there.>
Being just a Mom and Pop operation, <MostlyAverageJoe>, I do not have access to any turbocharged, terabyte computers (at least not legally). Poor old Toga II 1.3.1 does beg to differ after 21.Ng5 Nf6, however, so you might want to run the computation a little longer just to be sure>
But I most certainly did run it longer. At low plies, Bxc8 is seen a lot in the lead, with one or two occurrences of Nf7+, then Qg6 moves up front, then f3 becomes a leader for a while, but eventually Qg6 regains the lead and stays there for good, while ratings of f3 drop down.
15 plies you're quoting for Toga is just not deep enough. I usually run at least 18-20 plies of analysis (note that Hiarcs has expensive evaluation function, so it is more accurate at the same depth compared to many other engines). It is also possible that one engine is blind to certain lines, although note that RV's Rybka agrees with Hiarcs here, so it is probably Toga (and related Fruit) that may be missing something. Fruit at 19 plies deep still preferred f3 over Qg6. Qg6 moved to front only after 20 plies (it needed 70 CPU minutes to get there at 3GHz).
The pattern you described, <from <22.Bxc8> to <22.Qg6> to <22.f3>> shows up in other engines too (Glarung, Shredder). In all cases, however, there is one more change, returning <22.Qg6> to the top.
<johnlspouge: After a little reflection (not an intensive analysis, however), the point of the mysterious move 23.f3 seems to be to play Nf7+, but with two preconditions: (1) Black cannot play ...Be4+>
I had the same impression that the goal of f3 (which shows up in many lines) is to prevent Be4 incursion. One technique to see what a move is supposed to threaten is to have the engine give two consecutive moves to one side (this reveals the main threat arising from the first move) and I did it quite a lot. However, Be4 did not really show up as a move that Black would want to play, so I did not mention it.
|Mar-23-08|| ||al wazir: <Samagonka>: You're posting in the wrong forum:|
|Mar-23-08|| ||Samagonka: <Al Wazir< , You're posting in the wrong forum:
Though we all share the same passion for this game, why shouldn't I share an even more valuable passion with you? Take it, or leave it but the feedback itself shows me that you heard and that's a mission accomplished.|
<ounos> time will tell...
|Mar-23-08|| ||012: Saturday puzzle <34. ...?> Mar-22-08 Keres vs L Popov, 1973|
|Mar-23-08|| ||gaatab: i think 21:ne5 is strong enough...
no need to be insane:)
|Mar-23-08|| ||wals: Noting think:- In the Lord I trust, all others, cash cown.|
Forward to TPOTD
Castled on opposite sides means go for broke in attack according to the pundits, throw everything, including the kitchen sink
Thinking 21.Bxd7 with idea of Ne5 to g6+ but this is shot down in flames by Bxf3.
Beyond me, I surrender.
Ah so that's how it's done Yes, I;m sure to remember that combination for a few hours.
|Mar-23-08|| ||Jason Frost: <anatoly21> Wow, you must calculate faster then a computer then. After like 10 minutes of thought my 20th century fritz gives |
(1.38) 22. Qg6 Rce8 23. Bf5 Bc6 24.Nf7+ Rxf7 25.Qxf7 Ne4 26. g4 Re5
(1.34) 22.f3 b5 23. cxb5 Rc7 24.Qg6 Qb6 25.Nf7 Rfxf7 26. Bxf7 Qxb5
(1.22) 22.Bxc8 Bxc8 23. Nf3 Nh5 24.Nd2 Bf5 25.Ne4 Qc6 26. f3 Nf6
Seems like a wierd puzzle, with so many reletivly = lines none of which seem to lead to an easy win.
|Mar-23-08|| ||wals: Erling Mortensen - John Rodgaard, Denmark Ch 1
Analysis by Fritz 11:
1. (-3.72): 21...Nd7xe5 22.Be6xc8 Bb7xc8 23.Kb1-a1 Ne5-c6 24.Rd1-d2 Bc8-f5 25.Qc2-a4 d4-d3 26.Ka1-b1 Qd6-f4 27.Re1-d1 Nc6-a5 28.b2-b3 Qf4xh4 29.Rd1-e1 Bf5-g6 30.Qa4-d7 Rf8xf2 31.Rd2xf2
<gaatab> take note
|Mar-23-08|| ||DarthStapler: I kept thinking Bxd7 followed by Ne5|
|Mar-23-08|| ||JSYantiss: 27. Rxf6 was a nice shot. I might have missed that OTB myself.|
|Mar-23-08|| ||ounos: <JSYantiss> Umm, define "yourself"?|
|Mar-23-08|| ||benveniste: <luddite>, My theme song is "Never on Sunday."|
I got ♘g5 easily enough, and even found ♕g6. But I never was never in the same zip code as f3, and without that move it all comes to nothing.
So put me down in the "not solved" category.
|Mar-23-08|| ||Jason Frost: <benveniste> 23. Bf5(not f3) is by far the best move in the position, although I still think that if you found Ng5 followed by either Bxc8, Qg6, or f3 you have solved it.|
|Mar-23-08|| ||Jason Frost: 23. Bf5 Rg8 24. g4 Bc8 25. Ne6 Bd7 26. Nxc7(g5? Be8)Bxf5+ 27. Qxf5 Qxc7 28. g5 and black can resign|
The reason I don't like this puzzle is that 22...Rc7 is a mistake black should have played 22...Rce8 when it unclear whether white can force a win
|Mar-24-08|| ||MostlyAverageJoe: Final update on the exchange line 21. Ng5 Nf6 22. Bxc8: it appears that it is NOT winning. I re-run forward analysis, 21 plies per move, and got the line below. Black appears to have a slight advantage at the end (-0.52).|
21. Ng5 Nf6 22. Bxc8 Bxc8 23. Nf3 Nh5 24. Ka1 Nf4 25. Rg1 d3 26. Qa4 Bg4 27. Qxa7
Qf6 28. Rd2 Nxg2 29. Rxg2 Bxf3 30. Rg3 Rf7 31. Qa4 Be2 32. Qe8+ Rf8 33. Qg6 Qe7
34. Rg1 Rxf2 35. Qxb6 Qe3 36. Qg6 Rf7 37. a3 Re7 38. Rdd1 Bxd1 39. Rxd1 Rd7 40.
h5 Qe2 41. Ka2 d2 42. Qc2 Rd4 43. b3 Kg8 44. a4 Rd6 45. a5 Kh8 46. Kb2 Rd3 47.
Ka3 Rd8 48. Ka2 Kg8 49. Ka3 Rd3 50. Kb2 Rd7
click for larger view
Anyone who wants to find improvements for white - be my guest.
|Mar-24-08|| ||MostlyAverageJoe: <Jason Frost: ... 22...Rc7 is a mistake black should have played 22...Rce8>|
This is a pretty strong claim. What is the evidence for it?
|Mar-24-08|| ||znprdx: Well it was difficult enough to get to 5ply at which point I'd worked out 23.Bf5 which I figured forced Bc8 24. Bx[B]c8 leaving Black a difficult choice of how to recapture - in face of the unstoppable Re6 threat, the point of Qg6. |
I broke my timing out rule - being well past my second coffee - and was already moving the pieces when I finally gave in only to have to deal with the incomprehensible 23.f3 and Ba6 and I thought is the same game?...HELP
|Mar-24-08|| ||znprdx: OMG what is happening to me?"the lunatic is in my head" Now I'm finding a Rybka move (according to <parzer2>:)|
<johnlspouge:> f3 makes no sense whatsoever,(Be4+ is hardly a threat) but its psychological value (and gain in time on the clock) must have been phenomenal. The text response Ba6 showed no understanding of the impending doom. Bc8 might have held on a little longer -
|Mar-24-08|| ||znprdx: <ounos:The actual games were never meant to be puzzles ETC> You have raised an interesting point. However you have to consider that the greater number of users are introduced to the game in question specifically because of the features recognized by <CG editors> presented as a puzzle or problem position. Arguably there are potentially half a dozen such potential stop-points in a game depending upon your playing strength |
Actually nothing stops you from just playing through the game - >making your own notes as you go< - as if you were in the tournament hall doing live-action commentary. Then you could check and compare with kibitzer inputs....but many of may not have the time or like me are to lazy to do so. Those that do often find spoilers to the daily challenge - which opens another doorway of perception, sometimes more rewarding than the actual game as played.
|Mar-24-08|| ||znprdx: BTW the threat after 23.Bf5 is Nh7!! with a view to the crushing Nx[N]f6. It looks like the best try seems to be ...23.Re7 24. Rx[R]e7 Qx[R]e7 Now f3 would make sense but is unnecessary since Nx[N]f6 forces Be4+ 25.Bx[B]e4 and Black must surrender the Queen to stop mate.|
Yup ...insane to believe one could see all this OTB. However...f3 shows a certain degree of self control...it sets up absolute domination of the board - the sign of a potential grandmaster...which unfortunately it seems was not to be for Mortensen despite becoming Danish Champion the following year.
|Mar-24-08|| ||Jason Frost: <mostleyaveragejoe: <Jason Frost: ... 22...Rc7 is a mistake black should have played 22...Rce8>
This is a pretty strong claim. What is the evidence for it?>|
As I stated earlier after 22...Rc7 23.Bf5!(not the weaker f3) Rg8 24. g4 Bc8 25. Ne6 Bd7 26. Nxc7(g5? Be8)Bxf5+ 27. Qxf5 Qxc7 28. g5 black can resign.
So, considring 22...Rc7 is lost 22...Rce8 seems like the only other logical move
|Mar-24-08|| ||znprdx: <CG> what has happened to the 'flip board' bar?-|
|Mar-26-08|| ||malvar: Insane, 21, Whites move.
This is very strong and precise play. Excellent play. For me Ng5 was a possibility. My main worry was blacks bishop controlling the long diagonal. Couldn't decide on what plan to follow.
< Earlier Kibitzing · PAGE 3 OF 3 ·