Latest Updates:
Normal Topic Komodo, Houdini 3 & GM Two Knights Defense (Read 4358 times)
sloughter
God Member
*****
Offline


I Love ChessPublishing!

Posts: 619
Location: schoharie
Joined: 12/29/08
Gender: Male
Re: Komodo, Houdini 3 & GM Two Knights Defense
Reply #4 - 08/01/13 at 17:53:14
Post Tools
Smyslov_Fan wrote on 08/01/13 at 15:37:06:
If you are going to say that a GM claims perfect analysis, then you must include that GM's name.

IF you are making an extraordinary claim, you need proof. I am not at all sure exactly what you are claiming though. You claim this replaces existing theory, show us what existing theory says and how this is an improvement.



GM Roman Dzindzichashvili, more or less, just confirms what post members have been saying for years---The main line is equal. What existing theory do you want me to reference? ECO Edition 5, MCO 15, NCO, specialty literature, on line analysis?

What Roman has done in the 8.Qf3 line is to show that Be7 is dead equal, an analytical result I have achieved independently. His handling of the opening is clearer. I thought that Qd1 was playable; it is against h6, but the critical improvement that Roman recommends is g6.

The point behind g6 is that it denies White the resource Ng5/Ne4/Ng3/Bf5 either exchanging the Bishop or forcing Black to deal with it. Once Black finds g6 it is dead equal.

I tried to salvage the opening with Qd1 but Roman instantly rejected it and provided the move order leading to a draw by repetition i.e. it is a bust for White. On a theoretical level the only "acceptable" result is to demonstrate a plus for White.
  
Back to top
 
IP Logged
 
Smyslov_Fan
YaBB Moderator
Correspondence fan
*****
Offline


Progress depends on the
unreasonable man. ~GBS

Posts: 6902
Joined: 06/15/05
Re: Komodo, Houdini 3 & GM Two Knights Defense
Reply #3 - 08/01/13 at 15:37:06
Post Tools
If you are going to say that a GM claims perfect analysis, then you must include that GM's name.

IF you are making an extraordinary claim, you need proof. I am not at all sure exactly what you are claiming though. You claim this replaces existing theory, show us what existing theory says and how this is an improvement.
  
Back to top
 
IP Logged
 
sloughter
God Member
*****
Offline


I Love ChessPublishing!

Posts: 619
Location: schoharie
Joined: 12/29/08
Gender: Male
Re: Komodo, Houdini 3 & GM Two Knights Defense
Reply #2 - 08/01/13 at 14:22:01
Post Tools
Bibs wrote on 08/01/13 at 08:07:12:
HOW DOES THIS RELATE TO CURRENT THEORY?

Please, place in context. What is known? How connect?

Then seek to discuss. Ask a question. It is not difficult to do these things. Help us to talk with you.





It replaces existing theory.
  
Back to top
 
IP Logged
 
Bibs
God Member
*****
Offline



Posts: 2120
Joined: 10/24/06
Re: Komodo, Houdini 3 & GM Two Knights Defense
Reply #1 - 08/01/13 at 08:07:12
Post Tools
HOW DOES THIS RELATE TO CURRENT THEORY?

Please, place in context. What is known? How connect?

Then seek to discuss. Ask a question. It is not difficult to do these things. Help us to talk with you.



  
Back to top
 
IP Logged
 
sloughter
God Member
*****
Offline


I Love ChessPublishing!

Posts: 619
Location: schoharie
Joined: 12/29/08
Gender: Male
Komodo, Houdini 3 & GM Two Knights Defense
08/01/13 at 07:27:43
Post Tools
According to Komodo, Houdini 3 running on 6 processors overnight supervised by a GM claiming "perfect analysis", with best play 2 variations of the main line of the Two Knights' Defense are equal: Here is his analysis verbatim:

"1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5ch c6 7.dxc6 bxc6

8.Qf3 Be7

A) 9.Bxc6+ Nxc6 10.Qxc6+ Bd7 11.Qf3 Rc8 Very accurate move, (I played this instead of O-O) since after 11...O-O 12.Nc3 white has some edge. After 11...Rc8...it's a very good  position for Black 12.Nc3 Bc6 13.Qh3 Nd5 14.d3 Nxc3 15.bxc3 Bd7 16.Qg3 O-O 17.O-O (17.c4 Ba4!!) Rxc3...



9.Bd3 O-O 10.O-O g6 11.Re1 Nd5 = (I tried to improve here with 11.Qd1, but after 11...Ng4 12.Ne4 f5 13.Nec3 Bd6 14.Be2 Nxh2 15.Kxh2 Qh4ch 16.Kg1 e4 17.g3 Bxg3 =)



8.Bd3 Nd5 9.Nf3 Bd6 10.O-O g6 11.Re1 Nxd3 12.cxd3 O-O 13.b3 Re8 Here Komodo gives 14.Nc3 c5 15.Ne4 Bf8 16.Qc2 Ba6 with equality.



14.Bb2 Houdini...the most powerful hardware and software thinks White is better by ,22 (A measureable advantage according to the GM) f6 15.Nc3 c5 (I tried Ba6 here, but according to the GM this favors White) 16.Ne4 Be6 (forced) 17.Rc1 Nb7 Komodo absolutely disagrees with Houdini's evaluation saying the position is equal after 17...Nb7 but---also suggests that 17...Nc6 is also quite good for Black [17...Nc6 18.Nxd6 Qxd6 19.Ba3 Nb4 20.Bxb4 cxb4 21.d4 exd4 22.Re4 d3 23.Rd4 Qa6 24.Rxb4 Qxa2 25.Nd4 Qa6=]



18.b4 This is why Houdini thinks white is better by ,22. but it's quite accurate. Komodo's equal evaluation is far more accurate 18...cxb4 19.d4 exd4 20.Nxd4 Bd5 not Bxa2 21.Qg4 Bxe4 22.Rxe4 Rxe4 23.Qxe4 Qd7 24.Nf5 Bf8 white has absolutely no edge.
  
Back to top
 
IP Logged
 
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo