< Earlier Kibitzing · PAGE 140 OF 140 ·
|Nov-21-18|| ||centralfiles: <AK><<<centralfiles> I would frame the argument as follows ... Then why would we assume that white might be winning after 19...Rf6?>
We can't. But neither can we assume that White is <not> winning. There are simply too few (I think) analyses (samples of the all the possible "reasonable" lines) to reach that conclusion. And I haven't yet figured out how to determine the number of analyses that would need to be done in order that the conclusion we reach is statistically significant. It could be a lot!>
I don't really know what to make of this.
If all lines the engines are evaluating as possibly winning <+.7 or higher for arguments sake> can be shown to actually peter out to a draw<less than +.5> with best play, then we are left with lines the engines are evaluating as drawish why would we assume that some +.41 line is actually winning? it's seems terribly unlikely.
Statistical analysis of lines evaluated at +.7 or higher doesn't seem necessary, they can individually all be shown to peter out to lower evals or provable endgame draws for human analysts.
|Nov-21-18|| ||centralfiles: <AK> <Probably because no one seems to have anywhere near the fervor in favor of 19...h6 leading to a draw that you and <kwid> (and possibly others) feel that 19...Rf6 leads to a draw. So the effort that has been spent to "show" that 19...h6 (or any other 19th move by Black for that matter) leads to a draw is nowhere near as much as has been spent to "show" that 19...Rf6 leads to a draw.>|
Lest we forget that ...h6 was actually played. Neither team found a line that might draw, nor did subsequent engine analysis find any real improvements for black.
At the same time no one has yet been able to show a line after ...Rf6 where black doesn't seem to be holding on to the half point though some posters might have dropped some obfuscating posts that imply otherwise without providing any analysis. (only real attempt to show win for white was by Ceri but those lines petered out to draw as well <Ceri was unfortunately given inferior black moves to analyze>) I don't understand how this little fact can be ignored.
|Nov-21-18|| ||centralfiles: <About the only anywhere near overwhelming evidence that tournament provided, I believe, is that after 24 games were played, Black was not able to win even once. While this may also not technically be "conclusive" either, it does seem that if Black has any expectation of winning, improvements for Black need to be found between 5.Bxf7+ Ke7 and 20.Nb5. Because, why would you play the Traxler if the best you can hope for is a draw?>|
You play the traxler for the same reasons you play other gambits - great results against players that are not great with their king under pressure and not booked up to the hilt. The real Question of this whole exercise was more whether Black is busted theoretically or not.
Though in a practical sense I would agree black is somewhat busted even though 19...Rf6 is a draw. With 8.b4 White has successfully turned the tables with black being the one under pressure to defend accurately. It's hard to imagine anyone wanting to play these positions OTB as black.
|Nov-21-18|| ||centralfiles: <AK><I agree, that's what I've been saying all along. The position after 20.Nb5 is extremely complex as is the Traxler in general. So to "conclusively" prove anything is at least very hard, and possibly impossible. The best we could hope for, I think, is to show that a particular move at a particular position leads to the desired result in a statistically significant way to a level of confidence we can live with, and accepting the non-zero margin of error in the estimate.>|
I can't imagine <KWID> means "conclusively" as 100% definitive mathematical proof.
|Nov-21-18|| ||AylerKupp: <<centralfiles> As was said earlier the engine tournaments serve almost no purpose here. If the black side deviated at all form our analysis the result is just about worthless.>|
Well, I certainly disagree. It may not be conclusive but it gives some indication, statistically speaking if you play enough games, of Black's drawing chances. If you keep an open mind and try different moves at different stages of the game, it will give you some idea of what moves work better than others as far as results are concerned.
<If all lines the engines are evaluating as possibly winning <+.7 or higher for arguments sake> can be shown to actually peter out to a draw<less than +.5> with best play, then we are left with lines the engines are evaluating as drawish why would we assume that some +.41 line is actually winning? it's seems terribly unlikely. >
You don't need to restrict the analyses to lines evaluated as possibly winning, best to look at all lines. And lines that originally did not seem to be winning could, depending on how deep you search, could indeed turn out to be winning. Again, not likely, but you just can't rule them out.
|Nov-21-18|| ||AylerKupp: <<central files?> Lest we forget that ...h6 was actually played.>|
Sure, but afterwards it was not subjected to anywhere the depth of investigation that 19...Rf6 was. There were no champions for 19...h6 like there were for 19...Rf6. For all we know if 19...h6 had been subjected to the same level of analysis as 19...Rf6 it might have also shown that all lines petered out to a draw. And just because Ceri wasn't able to show a win for White, given the limited scope of his analysis, that again doesn't mean that the win wasn't there.
So I don't think that this little fact was ignored, it was simply the fact that no one really cared a lot about showing a certain win for White, so the matter was dropped due to lack of interest.
<Though in a practical sense I would agree black is somewhat busted even though 19...Rf6 is a draw.>
Ha! You and <kwid> can keep on insisting that's the case until the cows come home but that doesn't make it a fact. Given the limited number of lines that have been analyzed compared to the number of reasonably alternative lines there is no conclusive evidence as far as I'm concerned that 19...Rf6 leads to a draw or, for that matter, that it's superior to 19...h6 as far as providing Black with the best practical chances for a draw. 19...h6 has simply not been investigated anywhere as thoroughly as 19...Rf6 by anyone who has the same motivation as you and <kwid> have to insist that 19...Rf6 leads to a draw.
|Nov-21-18|| ||AylerKupp: <<I can't imagine <KWID> means "conclusively" as 100% definitive mathematical proof.>|
Me neither, but that's a question to ask <kwid> as to what he meant. I personally would not expect a 100% definite mathematical proof but only that the conclusion (whatever that is) is statistically significant. The most typical answer to what that means is a 95% confidence level that the result was not due to chance; i.e. a 5% or ± 2.5% that it could be.
Exactly what that means in this case, if anything at all, I'm not sure. Confidence levels are typically used when the data (population) that you are looking at is Normally distributed and you sample that population at random. And I have not been able to come up with a was to cast situations like this into a situation like that. Clearly the analyses will not be selected at random although, if you have enough analyses, they could theoretically be. And I don't know if the analysis results can be considered Normally distributed since there are only 3 of them; White wins, White loses, or the game is a draw.
So, to be frank, I don't know how to use the term "Confidence Level" in this situation other than intuitively, although I would not require that 100% of a "significant" number of analyses indicate that the result was a draw to accept that hypothesis.
As a reminder, I'll be out of town for a few days so my ability to respond will be limited. So don't feel that I have lost interest if I don't respond to anyone's posts in a timely manner. I'll be taking a probability book with me (no joke!) so maybe by the time I come back I'll a better ideas of how to deal with the problem. I know, I have to get a life.
|Nov-21-18|| ||centralfiles: <AK><I know, I have to get a life.>
No No No...
Nobody wants to see you doing that... ;)
|Nov-21-18|| ||centralfiles: <AylerKupp: <<central files?> Lest we forget that ...h6 was actually played.>
Sure, but afterwards it was not subjected to anywhere the depth of investigation that 19...Rf6 was. There were no champions for 19...h6 like there were for 19...Rf6. For all we know if 19...h6 had been subjected to the same level of analysis as 19...Rf6 it might have also shown that all lines petered out to a draw. And just because Ceri wasn't able to show a win for White, given the limited scope of his analysis, that again doesn't mean that the win wasn't there.|
So I don't think that this little fact was ignored, it was simply the fact that no one really cared a lot about showing a certain win for White, so the matter was dropped due to lack of interest.>
Sounds to me like you may have been absent during the game it was analyzed quite extensively. Do you any possible improvements?
|Nov-21-18|| ||AylerKupp: <centralfiles> Yes, I was absent for the greater part of the game. I glanced through the comments after the game and the number of analyses of 19...h6 did not strike me to be nearly as many as for 19...Rf6, particularly after the game ended. So I will review the posts once I get back home and compare their relative number and, if I can, their quality.|
|Nov-21-18|| ||kwid: <AylerKupp:> The arguments before us as I see it should be framed as : h6 was the losing move while Rf6 instead could have led to a forced draw.
You keep insisting that this claim is based on unsubstantiated proof or high degree of uncertainty.|
Well than challenge me with my cards open on the table:
19.Nb5 < Rf6 > 20.Rb1 Kd8 21.Qd5 Bd7 22.Red1 g6 23.Ne4 Rf5 24.Qd3 Rxb5 25.Rxb5 Kc7 26.Qxd6+ Qxd6 27.Nxd6 Rf8 28.Be2 a6 29.Rb2 Rxf4 30.Nxb7 Rf2 31.Nd6 Ne5 32.Ne4 Rf4 33.Ng5 h6 34.Nf3 Nxf3 35.Bxf3 Bb5 36.Rb4 Rc4 37.a4
[37.Rdb1 Bc5 38.Rxc4 Bxc4 39.Re1 Kd6 40.a4 a5 41.g3 h5 42.Kg2 Bb3 43.Rb1 Bc2 44.Rb2 Bf5 45.Rb8 Kc7 46.Rb7+ Kd6 47.Rb2 Bd7 48.h4 Bxa4 49.Ra2 Bb5 50.Rxa5 Bd3 51.Ra4 Ke5 52.Ra5 Kd6 53.Ra8 Ke6 54.g4 hxg4 55.Bxg4+ Ke5 56.Ra5 Kd6 57.Bf3 Bc4 58.Kg3 e2 59.Ra1 Be3 60.Bxe2 Bxe2 61.Re1 g5 62.Rxe2 gxh4+ 63.Kxh4]
37...Rxb4 38.cxb4 Bxa4 39.Rc1+ Kd6 40.Be2 Bb5 41.Bxb5 axb5 42.g3 Bd4 43.Kg2 Bb2 44.Rc8 Ba3 45.Rd8+ Ke6 46.Re8+ Kf6 47.Rxe3 Bxb4 48.Rb3 Bc5 49.Rxb5 Bd4 50.Kf3 h5 51.Kf4 Bc3 52.Rb6+ Kg7 53.Rd6 Bb2 54.Rb6 Bc3
|Nov-21-18|| ||kwid: Well I just saw a computer analysis from this game a minute ago and thus looked at my data with Rf6 followed by Qd3 as shown at this board. Here is part of my assessment of Qd3:|
[Event "Traxler Bxf7+/Bc4/b4"]
[Black "Analysis 19...Rf6 20.Qd3"]
1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. Ng5 Bc5 $2 5. Bxf7+ Ke7 6. Bc4 Rf8 7. O-O Qe8
8. b4 Bxb4 9. c3 Ba5 10. d4 d6 11. f4 Bb6 12. Kh1 exd4 13. e5 $1 Ng4 14. exd6+
cxd6 15. h3 Ne3 16. Re1 Kd7 17. Bxe3 dxe3 18. Na3 Qe7 19. Nb5 Rf6 20. Qd3 Kd8
21. Rad1 (21. Ne4 Bf5 22. Nbxd6 Kc7 23. Nxf5 Rxf5 24. Ng5 Rf6 25. Qxh7 Rd8 26.
Rad1 Kb8 27. g3 Rfd6 28. Rxd6 Qxd6 29. Ne4 Qd7 30. Kg2 Na5 31. Be2 Qd5 32. h4
Nc4 33. Kh3 Qd7+ 34. Kh2 Nd2 35. Ng5 Rc8 36. Rc1 a6 37. Qg6 Ka7 38. Qe6 Qd8 39.
c4 Rc6 40. Qf5 Bc5 41. Nf7 Qd4 42. Rc2 Rc7 43. Ne5 Bd6 44. Nf3 Nxf3+ 45. Bxf3
Rxc4 46. Rxc4 Qxc4 47. Qd7 Qc7 48. Qxc7 Bxc7 49. Kg2 b5 50. Be4 a5 51. h5 Bb6
52. Kf3 Kb8 53. g4 Kc8 54. f5 Bd8 55. Kxe3 Bg5+ 56. Kd4 b4 57. Kc5 Kd7 58. Kb5
Bd8 59. f6 Bxf6 60. Kxa5 Be7 61. Ka4 Ke6 62. g5 Ke5 63. g6 Kf6 64. Bf3 Kf5 65.
Kb3 Bd6 66. Be2 Kg5 67. Kc4 Kf6 68. Kb5 Kg5 69. Bd1 Kf6 70. Bc2 Kg5 71. Kc6 Bf8
72. Bd1 Kf6 73. Kd5 Be7 74. Kc4 Bd6 75. Be2 Bf8 76. Kb5 Bd6 77. Kc6 Bf8 78. Bd1
Be7 79. Kd5 ½-½) 21... a6 22. Nxd6 Kc7 23. Nge4 (23. Nxc8 Rxc8 24. Qxh7 Kb8
25. Qb1 Na5 26. Bf1 Nc4 27. Qb3 Qc7 28. Bxc4 Qxc4 29. Qxc4 Rxc4 30. g3 Rfc6 31.
Rd3 Rxc3 32. Rxc3 Rxc3 33. Re2 Kc7 34. Kg2 Kc6 35. Kf3 Kd5 36. f5 Rc1 37. g4
Rf1+ 38. Kg2 Rc1 39. Kf3 ½-½) 23... Rh6 24. Rb1 Ba7 25. Red1 (25. Nxc8 Rxc8
26. Bd5 Rd8 27. Re2 Qa3 28. Reb2 b5 29. Rb3 Qf8 30. c4 Qf5 31. Qf1 Rxd5 32. Ng3
e2 33. Qxe2 Rxh3+ 34. gxh3 Qxh3+ 35. Qh2 Qxh2+ 36. Kxh2 Rd2+ 37. Kh3 b4 38. a3
a5 39. axb4 axb4 40. Kg4 Bd4 41. Rh1 h6 42. Ne4 Rc2 43. Rd3 Rxc4 44. Rh5 Be5
45. Kf3 Nd4+ 46. Kg4 Nc6) 25... Rb8 26. Nxc8 Rxc8 27. Bd5 Rd8 28. Rb2 (28. Kh2
b5 29. a4 Qh4 30. Qf1 Qh5 31. Ng3 e2 32. Qxe2 Qxe2 33. Nxe2 Rhd6 34. Bb3 bxa4
35. Bxa4 Rd2 36. Ng3 Rxd1 37. Rxd1 Rxd1 38. Bxd1 a5 39. Ne4 Ne7 40. Ba4 Nd5 41.
Kg3 Nb6 42. Be8 a4 43. Bf7 Nc8 44. Kf3 Nd6 45. Ba2 Nxe4 46. Kxe4 Bf2 47. Bb1 h5
48. g4 Be1 49. Kd3 hxg4 50. hxg4 Kd6 51. Kc4 Bd2 52. f5 Ke5 53. Bc2 a3 54. Kb3
a2 55. Kxa2 Bxc3 56. Kb3 Bd2 ½-½) 28... Rd7 29. Kh2 b5 30. a4 (30. Ng3 Qf8
31. Qe4 Qe7 32. Qf3 Rf6 33. Qg4 g6 34. Bxc6 Rxc6 35. a4 h5 36. Qxd7+ Qxd7 37.
Rxd7+ Kxd7 38. axb5 axb5 39. Rxb5 Rxc3 40. Rb7+ Rc7 41. Rxc7+ Kxc7 42. Ne2 h4
43. g4 hxg3+ 44. Kxg3 Kc6 45. Kf3 Kd5 46. Ng3 Kc4 47. f5 gxf5 48. Nxf5 Kd3 49.
Nxe3 Bxe3 50. Kg2 ½-½) 30... Ne5 31. fxe5 Qxe5+ 32. Ng3 Rhd6 33. axb5 Qxd5
34. Qxd5 Rxd5 35. Rxd5 Rxd5 36. bxa6 Bb6 37. Re2 Kc6 38. Re1 Ra5 39. Nf1 Re5
40. Kg3 e2 41. Nd2 Re8 42. Nc4 Ba7 43. Nd2 h5 44. h4 Re3+ 45. Kf2 Re6+ 46. Kg3
Kd5 47. Kf3 Re3+ 48. Kf2 Re7+ 49. Kg3 Re8 50. Kf3 Re3+ 1/2-1/2
|Nov-21-18|| ||centralfiles: <Kwid> I'm afraid we might just have to agree to disagree with <AK> here. We don't seem to be making much progress convincing him of the merits of 19...Rf6.|
Did you ever get a chance to do a deep analysis of 9...h6!?
I haven't spent much time on it but I'm not convinced of its merits at this point.
Also I never quite did get over 8...Bb6 9.a4 a6 Which i still think is probably adequate for black and much more "fun" <some aggressive double edged lines> than 8...Bxb4 though you never agreed with me on that one. You supplied positions where white was winning but i found black avoids those positions with best play.
All this is a bit silly though I was replying to the variations you posted as reasons to avoid 8...Bb6 and simply did not realize the strength of 8.b5! which might be just as hard for black as game continuation
<6...Rf8 7.0-0 Qe8 8.b4>
<52/63 +1.09 8. ... Bb6 9.b5 Na5 10.Be2 d6 11.Nc3 Bd4 12.Nf3 Bxc3 13.dxc3 b6 14.Nd2 Kf7 15.Nb3 Nb7 16.Bd3 Kg8 17.a4 Bg4 18.f3 Bd7>
|Nov-21-18|| ||kwid: Yes h6 was a must for us.
<Since we have no h6 voters but me I have to emphasis that moving our bishop twice just gives the game away.
I am not implying that we could win against a sound move order which we should expect with RV's lines as a guidance.
But with h6! and not Ba5? or Bc5? we will have a fighting chance to survive and deny white for being credited to come up with the Traxler refutation line.>
|Nov-23-18|| ||diceman: <<AK>
<Probably because no one seems to have anywhere near the fervor in favor of 19...h6 leading to a draw that you and <kwid> (and possibly others) feel that 19...Rf6 leads to a draw.>
You sure got that right.
When I first proposed 19...h6
I said I thought it lost.
(not exactly a winning endorsement)
kwid and centralfiles seem to want to rewrite history.
They also don't take into account,
the real world and the psychology involved.
(I probably would have been a fan of
19...Rf6 had kwid not been on my team,
with him, I wanted to run from it.)
kwid's the reason 19...h6 happened,
kwid's the reason it lost so effortlessly.
(kwid was a big fan of 24...Rxe6,
our guaranteed draw)
(the real moment we were done)
You see part of the problem with move
...19 is you have some responsibility
what you play after it.
19...h6 was never in the mix of a better move or great move.
The only real questions are:
1 Does 19... Rf6 draw?
2 Is there a better move?
...and even if found to be true, it
has no relation to what we actually knew at the time, or who would have been convinced to vote for the best moves.
What we saw wasn't a failure of chess moves, but a failure of psychology.
Too bad it never happened.
|Nov-23-18|| ||diceman: By the way, there is a poster named:
If you look at the 1972 Fischer/Spassky match. He's done excellent work on many games with new computer analysis.
He seems to be a less verbose/better computer/ version of Ayler.
(Wow, you win on both sides of the equation :) )
Maybe he can be recruited for some testing?
I am always willing to give other people projects!
|Nov-23-18|| ||AylerKupp: Sorry guys, I check this page from time to time but without my computer or my data handy (and limited time), I don't have much to add at this point.|
But, for <diceman>, as far as <Charles Sullivan> being a less verbose/better computer/version of me, none of them are too hard to accomplish. :-(
|Nov-24-18|| ||centralfiles: <Diceman>
<kwid and centralfiles seem to want to rewrite history.>
Am i too brazen If i can ask you to point to a single instance of something i said that could be seen as rewriting history?
I never took any part in any kind of blame game I did not vote for 19...Rf6 myself, nor do i profess to see its merits without the benefit of millions of silicon transistors.
I am simply maintaining that post game it appears pretty clear that 19...Rf6 draws from a theoretical standpoint. I am happy to analyze other plausible alternatives as well.
I would rather not have to point out some of the more blatantly erroneous posts disparaging <Kwid>s lines that were written with shallow engine analysis and were never recanted.
|Nov-25-18|| ||centralfiles: <Kwid><all>
How many of you would be willing to take on the white/black sides of the Frankenstein/Dracula variation in the Vienna?
Blacks results in practice in high level correspondence play have been pretty good I think, with several fine games by Timmerman.
Yet theoretically many still insist white is better. So why not have a world game to see if white can get a healthy advantage or is he destined to suffer through an anaconda's grasp just to salvage the half point.
It would be nice to see a fine example of how to play the Frankenstein as white.|
|Nov-30-18|| ||Damenlaeuferbauer: <centralfiles> As a 1.e4,e5 player with the black pieces, I am always ready to take the black side in the Frankenstein/Dracula variation of the Vienna game in correspondence chess. I think theoretically the chances are almost equal. I am eagier to play a Team White vs. Team Black game on Chessgames or a correspondence game against you! Your / Team White's turn!|
|Dec-02-18|| ||centralfiles: <Damenlaeuferbauer> I wont be strong enough to have meaningful correspondence game on my own :(|
|Dec-04-18|| ||kwid: <centralfiles:> could I be of any help to you? from what position would we start? it could be of interest to the chess community if we could discover new grounds in this Frankenstein Dracula opening variation.
if this site is inappropriate and chessgame not interested to open an other forum for it why not use your own site for it?|
|Dec-04-18|| ||centralfiles: <Kwid>
If we could organize another team game i think it would be much more interesting than the Traxler variation game.
Unfortunately Daniel is no longer with us so who knows when chessgames will be able to organize another team game.
If we had two private pages each which can be made to be accessible only to members of one team, then we can set up unofficial game ourselves.
However in absence of such pages it won't be possible to prevent teams from seeing each other's analysis.
Starting position should probably be here:
click for larger view
|Dec-05-18|| ||Tiggler: <kwid> <To close out the discussion if Rf6 holds I searched in support of the white side and could not come up with any winning lines.>|
If a member of team white might be permitted to comment on a thread that has produced scores of pages of commentary, mostly about the mythic (unplayed) 19.. Rf6, I would observe that the players of team white have evidently gone away content, but that frustrated team black players continue to fight the good fight. No aspersion on the respective team members: if the result had been a draw no doubt team white members would be the ones here battling over what went wrong!
Under the circumstances, including psychological ones, I don't think <kwid> searching in support of the white side is a substitute for actual team white members doing so, notwithstanding the respect in which <kwid> was known to be universally held by every team white member who is acquainted with him.
|Dec-05-18|| ||Tiggler: <diceman: By the way, there is a poster named:|
If you look at the 1972 Fischer/Spassky match. He's done excellent work on many games with new computer analysis.>
I'd like to follow <diceman> in urging people here to take a look at this poster's exemplary work.
Here is a link to the point in one particular game of the match mentioned where his analysis begins: Fischer vs Spassky, 1972 (kibitz #292),
and this is a later post in which he responds to my request for information about his methods and hardware: Fischer vs Spassky, 1972 (kibitz #303)
A few lines further down is my comment on what he told us. It is short enough to repost here:
<Tiggler: <CS> I appreciate your detailed answer. You should not be even <almost> embarrassed to have spent so long. I would be amazed if your, seemingly comprehensive, analysis in some cases had required less. I know you could not have done it so fast with any current engine without your own skill in forward and backward sliding, as you briefly indicated.
Do you play correspondence chess? I would not want to take you on in that arena!>
< Earlier Kibitzing · PAGE 140 OF 140 ·