McBrain 9.0 with Smarter Tactical Setting

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
leavenfish
Posts: 282
Joined: Mon Sep 02, 2013 6:23 am

Re: McBrain 9.0 with Smarter Tactical Setting

Post by leavenfish » Sun Feb 18, 2018 8:17 pm

You do tend to play fast and loose with terms like "refuted". But interesting...always interesting, I'll say that for you.

Leo
Posts: 844
Joined: Fri Sep 16, 2016 4:55 pm
Location: USA/Minnesota
Full name: Leo

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Leo » Sun Feb 18, 2018 8:44 pm

Ovyron wrote:
Leo wrote:How are your games against Rodolfo going? Your both undefeated right?
Nah, I've been defeated many times over the years, I'm just undefeated at FICGS and the LSS. And I haven't been defeated since 2014.

We're still on the opening phase in both games, but he can tell you that I'm low on time in both, with less than 11 days remaining, but I think I'll manage (famous last words! :lol: )
Interesting. It must be some very high quality chess.
Advanced Micro Devices fan.

Rodolfo Leoni
Posts: 544
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Rodolfo Leoni » Sun Feb 18, 2018 8:51 pm

Ovyron wrote:
Rodolfo Leoni wrote:It happened once that an opponent (Roman Usmanoff, you can look at that game if you want) waited until he had 1 day, then he moved. He played the whole game in zeitnot and when his position got hopeless he lost on time. :lol:
Haha! I actually lost 15 days in 4 of my games. What happened is that I had other games with less that 15 days on the clock, and attended them first, and didn't make a move at all on those 4 until they were games with the lowest time on the clock... turned out to be a terrible idea!

But I wonder if Usmanoff did the same, having lots of games with 1 day left on the clock and letting it run until it was the one to attend, with bad consequences! :)
He was playing only 5 games... Maybe he had lot of games somewhere else, but I asked to other players and that was his usual behavior. :shock:
F.S.I. Chess Teacher

MikeB
Posts: 3559
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: McBrain 9.0 with Smarter Tactical Setting

Post by MikeB » Sun Feb 18, 2018 10:50 pm

Excuse my density here, as I'm still trying to understand what you are trying to capture here that is different that what one can get today :

If I set multiipv =2 in the commandline, and do searchmoves , say e2e4 and d2d4 at the start position, it returns the the pv for those two moves:

McBrain 9 64 POPCNT by M. Byrne, Stockfish authors and others
set multipv 2
Confirmation: multipv set to 2
go depth 15 searchmoves e2e4 d2d4

Code: Select all

info depth 15 seldepth 26 multipv 1 score cp 67 nodes 817884 nps 1439936 tbhits 0 time 568 pv e2e4 e7e5 g1f3 b8c6 f1c4 g8f6 d2d4 e5d4 e1g1 f8e7 f3d4 f6e4 d4f5 d7d5 f5g7 e8f8 d1d5 d8d5 c4d5 e4d6 d5c6 f8g7 c6f3
info depth 15 seldepth 22 multipv 2 score cp 55 nodes 817884 nps 1439936 tbhits 0 time 568 pv d2d4 d7d5 c1f4 e7e6 e2e3 g8f6 g1f3 f8d6 f4d6 d8d6 f1d3 b8c6 b1c3 c8d7 e1g1 e8g8 c3b5
bestmove e2e4 ponder e7e5
how is that different from what you are doing?

User avatar
Ovyron
Posts: 2830
Joined: Tue Jul 03, 2007 2:30 am

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Ovyron » Tue Feb 20, 2018 9:20 am

MikeB wrote:If I set multiipv =2 in the commandline, and do searchmoves , say e2e4 and d2d4 at the start position, it returns the the pv for those two moves
Already answered here:
Ovyron wrote:What's the difference between this and setting my GUI to play mode with some MultiPV setting?

There's no difference, but many GUIs for analysis don't support that (they switch to SinglePV in play mode), nor do they allow to use "go searchmoves" with MultiPV (again, it goes back to SinglePV) so hacking the Tactical setting is the solution.
Basically, an engine shouldn't depend for the user to have to type in any command to send for the engine and should be fully capable of doing what the user wants from the GUI that the user uses. So it comes down to this:

This version allows the user to do everything I explained without having to type in commands to the engine.

It's about user friendliness and there's something wrong if user is forced to use their keyboard to type stuff in. After examining hundreds of positions interactively I'd already have lost a lot of time typing in "searchmoves" in some console. My GUI allows me to use "go searchmoves" by holding the SHIFT key and dragging pieces on the board, to exclude them and hold the CTRL key and drag he moves to include them, no need to type anything in a console.

Sure, this version wouldn't be necessary for me at all if my GUI supported "MultiPV+Play Mode" or "MultiPV+Go Searchmoves", but it doesn't, and thus, this McBrain will help out people that want to do this without having to use the command line.

I also think that some MultiPV setting in engine options to set the engine to some permanent state of MultiPV to mix in with Play Mode or Go Searchmoves would work, without having to depend on the Tactical setting, but I managed to fully implement wanted behavior by just removing one line of code from Tactical, so this was clearly the most practical thing to do.

User avatar
Ovyron
Posts: 2830
Joined: Tue Jul 03, 2007 2:30 am

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Ovyron » Tue Feb 20, 2018 9:27 am

Leo wrote:Interesting. It must be some very high quality chess.
Who knows? If Rodolfo Leoni easily defeats me from both sides of the board that would prove that my methods didn't produce that much quality :lol:

But yeah, I suspect Rodolfo and me are two of the people with the highest experience using Persistent Hashes for engines, at least, I don't think you'll find anyone that has used Rybka 3 Persistent Hash more than me, and Rodolfo has an impressive record using those methods on the LSS.

I have proposed for we to continue playing games until one of us loses, to foce a kick out of being both undefeated, so we will see...
Great spirits have always encountered violent opposition from mediocre minds.

User avatar
Ovyron
Posts: 2830
Joined: Tue Jul 03, 2007 2:30 am

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Ovyron » Tue Feb 20, 2018 10:33 am

Received this query by PM:
Hello again Ovyron ..

I have another question/request for you .. :D

I'm a correspondence player as mentioned before and a Mac user but my buddy in Munich has a Windows system and ChessBase and Fritz.

I mentioned this ('With Tactical I can use "Shout out" in my GUI and have it play against itself till the end of the game, and have the GUI store all the analysis of all positions.') and he correctly I think figured out that it was the Shootout option in Fritz. He's never used it before nor any of the other options offered except for Deep Position Analysis which he executes for me.

I was wondering if you would please suggest what settings to use for the Shootout - what do you use ? I realise our requirements may differ but if I have a baseline to work from (an expert's) I could then tailor to suit my needs after examination.
Unfortunately the ChessBase GUI or the Fritz GUI don't seem suitable for using the Shootout+Tactical feature of this McBrain.

The idea is to have the engine play itself from some position and then record to your analysis file all the alternative moves shown for the main line that the engine played.

For instance, in the Shredder Classic 3 GUI (some later version doesn't allow this), with Continuous Display=ON, if I let McBrain to play with Tactical=2 from the opening position at depth 20 I get this output:

(Note I have removed output below the interesting depth 20):

Opening position

Code: Select all

 20/30	 0:08 	+0.55 	1.d4 d5 2.c4 e6 3.Nc3 c5 4.cxd5 exd5 5.Nf3 Nf6 6.Bg5 Nc6 7.Bxf6 Qxf6 8.Nxd5 Qd8 9.dxc5 Bxc5 10.e3 O-O 11.a3 Be6 (20.893.563) 2388
 20/32	 0:08 	+0.52 	1.e4 e6 2.d4 d5 3.e5 c5 4.c3 Nc6 5.Nf3 Bd7 6.Bd3 Nge7 7.O-O Ng6 8.Be3 Be7 9.Nbd2 O-O 10.a3 Rc8 11.Qb1 cxd4 (20.893.563) 2388
 20/29	 0:08 	+0.38 	1.Nf3 e6 2.c4 c5 3.e3 Nf6 4.Be2 d5 5.d4 cxd4 6.exd4 Bb4+ 7.Nc3 Nc6 8.O-O O-O 9.Be3 Bd7 10.Rc1 Bd6 11.c5 (20.893.563) 2388
 20/22	 0:08 	+0.38 	1.e3 d5 2.Nf3 c5 3.d4 Nf6 4.c4 e6 5.Be2 cxd4 6.exd4 Bb4+ 7.Nc3 Nc6 8.O-O O-O 9.Be3 Bd7 10.Rc1 Bd6 11.c5 Bc7 (20.893.563) 2388

 20/30	 0:14 	 0.00 	1...Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 Be7 5.Bg5 h6 6.Bh4 O-O 7.e3 Nbd7 8.cxd5 Nxd5 9.Bxe7 Qxe7 10.Nxd5 exd5 11.Bd3 c5 12.O-O c4 (33.606.723) 2298
 20/22	 0:14 	+0.08 	1...d5 2.c4 c6 3.Nf3 Nf6 4.cxd5 cxd5 5.Nc3 e6 6.Bf4 Bd6 7.Bxd6 Qxd6 8.e3 O-O 9.Bd3 Nc6 10.O-O Bd7 11.a3 Rac8 12.Rc1 (33.606.723) 2298
 20/33	 0:14 	+0.15 	1...e6 2.e4 d5 3.e5 c5 4.c3 Nc6 5.Nf3 Bd7 6.Bd3 cxd4 7.cxd4 Qb6 8.Bc2 Nb4 9.O-O Nxc2 10.Qxc2 h6 11.Nc3 Ne7 (33.606.723) 2298
 20/30	 0:14 	+0.30 	1...c5 2.d5 Nf6 3.Nc3 e6 4.e4 d6 5.Nf3 exd5 6.exd5 Be7 7.Bd3 O-O 8.O-O Nbd7 9.a4 b6 10.Re1 Re8 11.Bf4 Bb7 12.Bb5 Rb8 (33.606.723) 2298

 20/34	 0:12 	+0.42 	2.Nf3 e6 3.Bf4 d5 4.e3 Bd6 5.Bxd6 Qxd6 6.Be2 b6 7.Nc3 O-O 8.O-O c5 9.Nb5 Qd8 10.a3 Bb7 11.c4 dxc4 12.Bxc4 Nc6 13.dxc5 (31.502.392) 2588
 20/28	 0:12 	+0.40 	2.c4 d5 3.Nf3 c6 4.Nc3 dxc4 5.e4 b5 6.Be2 b4 7.Na4 Nxe4 8.Bxc4 e6 9.Qc2 Nf6 10.O-O Be7 11.Re1 O-O 12.Nc5 Nbd7 13.Nxe6 fxe6 14.Bxe6+ Kh8 15.Qxc6 (31.502.392) 2588
 20/27	 0:12 	+0.33 	2.e3 d5 3.c4 e6 4.Nf3 Be7 5.Nc3 O-O 6.cxd5 exd5 7.Bd3 c5 8.dxc5 Bxc5 9.O-O Nc6 10.a3 Re8 11.b4 Bb6 12.Bb2 d4 (31.502.392) 2588
 20/28	 0:12 	+0.27 	2.Bf4 d5 3.e3 Bf5 4.Bd3 Bxd3 5.cxd3 Nc6 6.Nd2 e6 7.Ne2 Be7 8.O-O O-O 9.Rc1 a5 10.Nc3 Bd6 11.Bxd6 Qxd6 12.h3 (31.502.392) 2588

 20/30	 0:20 	-0.10 	2...b6 3.c4 e6 4.e3 Bb7 5.Bd3 d5 6.cxd5 exd5 7.O-O Be7 8.b3 O-O 9.Bb2 c5 10.Nc3 Re8 11.Ne5 cxd4 12.exd4 Nc6 (52.348.340) 2528
 20/27	 0:20 	-0.08 	2...e6 3.c4 b6 4.g3 Bb7 5.Bg2 Bb4+ 6.Bd2 Be7 7.O-O O-O 8.Nc3 d5 9.cxd5 exd5 10.Nh4 Qd7 11.Qc2 c5 12.Nf5 Nc6 13.Nxe7+ Nxe7 (52.348.340) 2528
 20/28	 0:20 	-0.01 	2...d5 3.c4 e6 4.Nc3 c6 5.Bg5 Nbd7 6.e4 dxe4 7.Nxe4 h6 8.Nxf6+ Nxf6 9.Be3 Bb4+ 10.Bd2 Bxd2+ 11.Qxd2 Ne4 12.Qf4 Qa5+ 13.Nd2 Nxd2 14.Qxd2 Qxd2+ 15.Kxd2 Bd7 (52.348.340) 2528
 20/30	 0:20 	+0.14 	2...c6 3.Bf4 Qb6 4.Qc1 d5 5.c4 Bf5 6.c5 Qd8 7.h3 Nbd7 8.e3 h6 9.Be2 e6 10.O-O Be7 11.Nc3 O-O 12.b4 Rc8 (52.348.340) 2528

 20/31	 0:17 	+0.47 	3.Bg5 Bb7 4.Nbd2 d5 5.c4 e6 6.Qa4+ Qd7 7.Qc2 dxc4 8.e3 b5 9.b3 cxb3 10.Qxb3 a6 11.a4 Qd5 12.Qb2 bxa4 13.Rxa4 Bc6 14.Rc4 Qb5 (44.061.093) 2545
 20/31	 0:17 	+0.42 	3.c4 e6 4.g3 Bb7 5.Bg2 Bb4+ 6.Bd2 Bxd2+ 7.Qxd2 d5 8.O-O O-O 9.Ne5 Nbd7 10.Nc3 Ne4 11.Nxe4 dxe4 12.Qe3 f6 13.Nxd7 Qxd7 (44.061.093) 2545
 20/26	 0:17 	+0.18 	3.c3 e6 4.Bg5 d5 5.Bxf6 Qxf6 6.Nbd2 Be7 7.e4 O-O 8.Bd3 c5 9.O-O Bb7 10.exd5 Bxd5 11.Ne5 cxd4 12.Qh5 h6 13.Rae1 dxc3 (44.061.093) 2545
 20/24	 0:17 	+0.07 	3.h4 e6 4.Bf4 d5 5.e3 Bd6 6.Ne5 O-O 7.Bd3 c5 8.O-O c4 9.Be2 Bb7 10.b3 Nc6 11.bxc4 Qc7 12.Nc3 Bxe5 (44.061.093) 2545

 20/29	 0:15 	-0.03 	3...Bb7 4.Nbd2 d5 5.c4 e6 6.cxd5 exd5 7.e3 Be7 8.Bb5+ c6 9.Ba4 O-O 10.O-O Nbd7 11.Bf4 Nh5 12.Ne5 Nxf4 13.Bxc6 Nxe5 (41.863.767) 2622
 20/34	 0:15 	-0.02 	3...e6 4.e4 h6 5.Bxf6 Qxf6 6.Bd3 Qd8 7.O-O Be7 8.c4 d6 9.Nc3 Nd7 10.e5 Bb7 11.exd6 Bxd6 12.Qe2 O-O 13.Rad1 Qe7 14.Bc2 c5 15.d5 exd5 16.Qxe7 Bxe7 (41.863.767) 2622
 20/29	 0:15 	+0.10 	3...d5 4.c4 e6 5.Bxf6 Qxf6 6.Nc3 c6 7.e3 Be7 8.cxd5 exd5 9.Bd3 O-O 10.O-O Bg4 11.e4 dxe4 12.Bxe4 Nd7 13.Re1 Rae8 14.h3 (41.863.767) 2622
 20/32	 0:15 	+0.25 	3...h6 4.Bxf6 exf6 5.e4 d5 6.Bd3 dxe4 7.Bxe4 c6 8.O-O f5 9.Bd3 Be7 10.c3 O-O 11.Qc2 c5 12.dxc5 bxc5 13.Bxf5 Bxf5 14.Qxf5 Nc6 15.Re1 (41.863.767) 2622

 20/26	 0:13 	+0.25 	4.e3 e6 5.c4 h6 6.Bxf6 Qxf6 7.Be2 Bb4+ 8.Nbd2 Bxd2+ 9.Qxd2 d6 10.Qc2 Nd7 11.O-O O-O 12.Rac1 Qe7 13.Nd2 c5 14.Rfd1 Nf6 15.Bf3 Rad8 16.Bxb7 (35.446.996) 2595
 20/29	 0:13 	+0.22 	4.c4 e6 5.e3 h6 6.Bxf6 Qxf6 7.Be2 Bb4+ 8.Nbd2 Bxd2+ 9.Qxd2 d6 10.Qc2 Nd7 11.O-O O-O 12.Rac1 Qe7 13.Nd2 c5 14.Rfd1 Nf6 15.Re1 Rac8 16.Bf3 (35.446.996) 2595
 20/34	 0:13 	+0.22 	4.Nbd2 d5 5.c4 e6 6.e3 Be7 7.Qa4+ c6 8.Be2 O-O 9.O-O Nbd7 10.cxd5 exd5 11.Rfc1 c5 12.Ne5 Nxe5 13.dxe5 Nd7 14.Bxe7 Qxe7 15.f4 Rfe8 16.Qc2 (35.446.996) 2595
 20/27	 0:13 	+0.12 	4.Bxf6 exf6 5.e3 g6 6.Nc3 Bg7 7.h4 h5 8.g3 O-O 9.Bg2 d5 10.O-O f5 11.Ng5 c6 12.Ne2 Nd7 13.Nf4 Nf6 14.c3 (35.446.996) 2595

 20/30	 0:15 	-0.20 	4...e6 5.Be2 h6 6.Bxf6 Qxf6 7.O-O Be7 8.c4 c5 9.Nc3 cxd4 10.exd4 O-O 11.Ne5 d6 12.Bf3 Bxf3 13.Nxf3 Nc6 14.d5 Na5 (40.515.136) 2547
 20/32	 0:15 	 0.00 	4...h6 5.Bxf6 exf6 6.Bd3 g6 7.O-O Bg7 8.e4 O-O 9.c4 f5 10.exf5 Nc6 11.Be4 gxf5 12.Bd5 Qe7 13.Nc3 Rae8 14.a3 Bf6 15.b4 d6 (40.515.136) 2547
 20/29	 0:15 	+0.01 	4...d6 5.c4 Nbd7 6.Nc3 e6 7.Bd3 h6 8.Bxf6 Nxf6 9.d5 e5 10.O-O Be7 11.Re1 O-O 12.e4 Qd7 13.a3 Rae8 14.b4 c5 (40.515.136) 2547
 20/24	 0:15 	+0.08 	4...g6 5.c4 Bg7 6.Nc3 d6 7.Bh4 h6 8.h3 Nbd7 9.Be2 c5 10.d5 O-O 11.O-O g5 12.Bg3 Nh5 13.Bh2 Bxc3 14.bxc3 Nhf6 15.Bd3 Kg7 16.Bg3 (40.515.136) 2547
(every blank line is the engine making the top move on the board and showing analysis for the next position)

You can safely add all those moves to your analysis file. Just remember that evaluations are asymmetrical so white's mainline has a score of 0.25 and black's has a score of -0.20.

You don't need to shoot out an entire game, here I stopped the engine after 8 ply.

You can notice that 1.e4 has a higher score than 4.e3, so since this is the top score you can play this on the board and have analysis of this line up to black's 4th move:

1.e4

Code: Select all

 20/37	 0:14 	-0.05 	1...e6 2.Nc3 d5 3.exd5 exd5 4.d4 Nf6 5.Nf3 Bd6 6.Bd3 Nc6 7.O-O O-O 8.Nb5 Nb4 9.Nxd6 cxd6 10.Re1 h6 11.h3 Bd7 12.Be3 (36.501.028) 2472
 20/30	 0:14 	+0.17 	1...e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.O-O Bg4 6.h3 Bxf3 7.Qxf3 Nf6 8.d3 Bc5 9.Qg3 Qe7 10.Be3 Bxe3 11.fxe3 O-O 12.Nd2 Nh5 13.Qf3 Nf6 (36.501.028) 2472
 20/27	 0:14 	+0.24 	1...c5 2.Nf3 e6 3.Nc3 Nc6 4.Bb5 Nf6 5.Bxc6 dxc6 6.O-O Be7 7.d3 O-O 8.Bf4 b6 9.h3 Bb7 10.Re1 Nd7 11.Bg3 h6 (36.501.028) 2472
 20/32	 0:14 	+0.28--	1...Nc6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.c3 Nge7 6.Be3 f6 7.Nbd2 fxe5 8.dxe5 Ng6 9.Bb5 Be7 10.Qa4 O-O 11.Bxc6 bxc6 (36.501.028) 2472

 20/34	 0:11 	+0.46 	2.d4 d5 3.Nc3 Nf6 4.Bg5 Bb4 5.e5 h6 6.exf6 hxg5 7.fxg7 Rg8 8.a3 Bxc3+ 9.bxc3 Rxg7 10.Qh5 Ke7 11.Nf3 c5 12.Bd3 Nd7 13.Qh6 Rg8 14.Bh7 Qf8 15.Qxf8+ (28.922.053) 2491
 20/39	 0:11 	+0.40 	2.Nc3 d5 3.d4 Nf6 4.Bg5 Bb4 5.e5 h6 6.exf6 hxg5 7.fxg7 Rg8 8.a3 Bxc3+ 9.bxc3 Qf6 10.h4 gxh4 11.Qh5 Qxg7 12.Qxh4 Nd7 13.Nf3 Qg4 14.Qh2 c6 15.Qc7 (28.922.053) 2491
 20/29	 0:11 	+0.32 	2.Nf3 d5 3.exd5 exd5 4.d4 Nf6 5.Nc3 Nc6 6.Bd3 Be7 7.O-O O-O 8.a3 a6 9.Re1 h6 10.Bf4 Be6 11.h3 Bd6 12.Ne2 (28.922.053) 2491
 20/34	 0:11 	+0.31 	2.c3 d5 3.e5 c5 4.d4 Nc6 5.Nf3 Bd7 6.Be2 Nge7 7.Na3 Ng6 8.Nb5 cxd4 9.cxd4 Qb6 10.Nc3 Be7 11.O-O O-O 12.Rb1 a6 13.a3 (28.922.053) 2491

 20/31	 0:22 	-0.15 	2...d5 3.Nc3 Nf6 4.exd5 exd5 5.Nf3 Bd6 6.Bd3 O-O 7.O-O c6 8.Re1 Na6 9.a3 Nc7 10.h3 Ne6 11.Be3 Re8 12.Ne5 Qc7 13.Ng4 Nxg4 14.hxg4 Bh2+ 15.Kh1 Bd7 (53.527.562) 2428
 20/35	 0:22 	+0.13 	2...h6 3.Nc3 d5 4.Bd3 c5 5.exd5 exd5 6.Bb5+ Nc6 7.dxc5 d4 8.Qe2+ Be6 9.Ne4 Nf6 10.Nf3 Nxe4 11.Bxc6+ bxc6 12.Qxe4 Qd5 13.Qxd4 Bxc5 14.Qxd5 Bxd5 15.Be3 Bb4+ (53.527.562) 2428
 20/31	 0:22 	+0.19 	2...Be7 3.Nc3 d5 4.Nf3 Nf6 5.Bd3 c5 6.exd5 exd5 7.dxc5 O-O 8.O-O Nc6 9.Na4 Re8 10.Bf4 Ne4 11.c3 h6 12.Re1 Be6 13.Bxe4 (53.527.562) 2428
 20/32	 0:22 	+0.26 	2...c5 3.d5 Nf6 4.dxe6 dxe6 5.Qxd8+ Kxd8 6.Nc3 a6 7.Nf3 b5 8.e5 Nd5 9.a4 Nxc3 10.bxc3 bxa4 11.Rxa4 Nd7 12.Be2 Bb7 13.O-O Bc6 14.Ra5 Kc7 15.Be3 Be7 (53.527.562) 2428

 20/25	 0:08 	+0.27 	3.Nc3 Nf6 4.e5 Nfd7 5.f4 c5 6.Nf3 Nc6 7.Be3 Be7 8.Be2 O-O 9.O-O h6 10.a3 a6 11.Na4 c4 12.Nc3 b5 13.f5 Bb7 14.fxe6 fxe6 (21.763.239) 2436
 20/32	 0:08 	+0.26 	3.exd5 exd5 4.Nf3 Nf6 5.Bd3 Bd6 6.Qe2+ Be6 7.Ng5 Qe7 8.Nxe6 Qxe6 9.Nc3 Nc6 10.Nb5 O-O 11.c3 Rae8 12.Qxe6 fxe6 13.O-O e5 14.Nxd6 cxd6 15.Be3 Ng4 16.Rae1 (21.763.239) 2436
 20/29	 0:08 	+0.25 	3.Nd2 c5 4.exd5 exd5 5.Ngf3 Nf6 6.Bb5+ Bd7 7.Bxd7+ Nbxd7 8.O-O Be7 9.dxc5 Nxc5 10.Nd4 O-O 11.Nf5 Ne6 12.Nf3 Bc5 13.Be3 Bxe3 14.Nxe3 (21.763.239) 2436
 20/29	 0:08 	+0.14 	3.e5 c5 4.c3 Nc6 5.Nf3 Qb6 6.g3 cxd4 7.cxd4 Nge7 8.Bh3 Nf5 9.Bxf5 exf5 10.Nc3 Be6 11.O-O Be7 12.Rb1 O-O 13.Be3 h6 14.Kg2 Rac8 15.h3 Qc7 (21.763.239) 2436

 20/27	 0:18 	-0.03 	3...Nf6 4.exd5 exd5 5.Nf3 Bb4 6.Bd3 O-O 7.O-O c5 8.dxc5 Bxc5 9.Bf4 Nc6 10.h3 Be6 11.Re1 h6 12.a3 a6 13.Qd2 Kh8 14.Kh1 (43.738.774) 2361
 20/28	 0:18 	+0.14 	3...Bb4 4.exd5 exd5 5.Bd3 Nf6 6.Nge2 Nc6 7.O-O h6 8.Be3 a6 9.a3 Bd6 10.Qd2 Be6 11.Rae1 O-O 12.Bf4 Ne7 13.Bxd6 Qxd6 14.Ng3 (43.738.774) 2361
 20/32	 0:18 	+0.15 	3...Be7 4.e5 c5 5.Qg4 g5 6.dxc5 Nc6 7.Nb5 Nxe5 8.Qg3 f6 9.Nf3 Nxf3+ 10.gxf3 e5 11.Bxg5 Bxc5 12.Be3 Bf8 13.Qh4 Be6 14.O-O-O a6 15.Nc3 Bf7 (43.738.774) 2361
 20/34	 0:18 	+0.40 	3...h6 4.Bd3 c5 5.exd5 exd5 6.dxc5 Nc6 7.Bb5 a6 8.Bxc6+ bxc6 9.Be3 Ne7 10.Qe2 Be6 11.Nf3 Nf5 12.Nd4 Nxd4 13.Bxd4 Qg5 14.O-O Be7 15.Rfe1 O-O 16.Na4 (43.738.774) 2361

 20/26	 0:14 	+0.38 	4.Bg5 Be7 5.e5 Nfd7 6.h4 h6 7.Bxe7 Qxe7 8.f4 O-O 9.Nf3 f6 10.exf6 Nxf6 11.Qe2 c5 12.O-O-O Nc6 13.dxc5 Nh5 14.Qf2 Nxf4 15.Bb5 (37.246.709) 2507
 20/33	 0:14 	+0.30 	4.exd5 exd5 5.Nf3 Bb4 6.Bd3 O-O 7.O-O c5 8.dxc5 Bxc5 9.Bg5 Be6 10.Qd2 Nc6 11.Rae1 a6 12.a3 Bd6 13.Bf4 Re8 14.h3 Rc8 15.Bxd6 Qxd6 16.Ne2 (37.246.709) 2507
 20/34	 0:14 	+0.19 	4.e5 Nfd7 5.f4 c5 6.Nf3 Nc6 7.Be3 Qb6 8.a3 cxd4 9.Nxd4 Nxd4 10.Bxd4 Bc5 11.Na4 Qa5+ 12.b4 Qxa4 13.bxc5 b6 14.cxb6 axb6 15.c3 Qxd1+ 16.Kxd1 Ba6 17.Bxa6 (37.246.709) 2507
 20/33	 0:14 	+0.10 	4.Bd3 c5 5.Nf3 cxd4 6.Nxd4 e5 7.Nf3 d4 8.Ne2 Bg4 9.c3 Bxf3 10.gxf3 dxc3 11.Nxc3 Nc6 12.Be3 Be7 13.O-O O-O 14.Rc1 Rc8 15.Nd5 Nxd5 16.exd5 Nb4 17.Rxc8 (37.246.709) 2507

 20/34	 0:12 	-0.04 	4...Bb4 5.e5 h6 6.exf6 hxg5 7.fxg7 Rg8 8.h4 gxh4 9.a3 Bxc3+ 10.bxc3 Rxg7 11.Qh5 Qg5 12.Qxg5 Rxg5 13.Nf3 Rg8 14.Rxh4 Nd7 15.g3 Nf6 16.c4 Bd7 17.Rb1 (23.984.223) 1930
 20/30	 0:12 	+0.07 	4...Be7 5.e5 Nfd7 6.h4 h6 7.Bxe7 Qxe7 8.f4 a6 9.Nf3 b5 10.Qd2 c5 11.g4 Bb7 12.a3 O-O 13.g5 cxd4 14.Ne2 Nc6 15.Nexd4 Nxd4 16.Nxd4 (23.984.223) 1930
 20/34	 0:12 	+0.11 	4...dxe4 5.Nxe4 Be7 6.Nxf6+ Bxf6 7.Bxf6 Qxf6 8.Nf3 O-O 9.c3 Bd7 10.Qc2 Bc6 11.Ne5 Qh6 12.Rd1 Rd8 13.Qd2 Qxd2+ 14.Kxd2 Bd5 15.Bc4 Bxg2 16.Rhg1 Bd5 17.Bxd5 (23.984.223) 1930
 20/34	 0:12 	+0.41--	4...h6 5.Bxf6 Qxf6 6.exd5 Bb4 7.Bb5+ c6 8.dxc6 bxc6 9.Bc4 c5 10.Qf3 Qxf3 11.Nxf3 cxd4 12.Nxd4 Bb7 13.Bb5+ Nd7 14.O-O-O Bxc3 15.bxc3 O-O-O 16.Nc6 Bxc6 17.Bxc6 Nb6 (23.984.223) 1930
So white's mainline for 1.e4 ends with a score of 0.38 which is higher than 0.25 for 1.d4.

However, note that 1. d4 Nf6 2. Nf3 b6 3. c4 has a score of 0.42 and it's the higest score, so play a shootout from here...

1. d4 Nf6 2. Nf3 b6 3. c4

Code: Select all

 20/30	 0:10 	-0.02 	3...e6 4.e3 Bb7 5.Bd3 Bb4+ 6.Nbd2 O-O 7.a3 Bxd2+ 8.Bxd2 d6 9.Qc2 c5 10.b4 cxd4 11.exd4 Bxf3 12.gxf3 Nc6 13.Be3 e5 14.d5 Nd4 15.Bxd4 exd4 16.O-O (24.143.912) 2382
 20/26	 0:10 	+0.12 	3...Bb7 4.Bf4 e6 5.e3 Bb4+ 6.Nbd2 c5 7.Bd3 cxd4 8.exd4 O-O 9.O-O d5 10.Rc1 Nc6 11.c5 bxc5 12.dxc5 Nd7 13.Bd6 e5 14.Bxf8 Qxf8 15.Bb1 Nxc5 16.a3 (24.143.912) 2382
 20/21	 0:10 	+0.36 	3...g6 4.Nc3 Bb7 5.Qc2 d6 6.e4 Bg7 7.Be2 Nbd7 8.O-O O-O 9.Bd2 c5 10.d5 Ng4 11.Rad1 Nge5 12.Nxe5 Bxe5 13.a3 (24.143.912) 2382
 20/31	 0:10 	+0.37 	3...d6 4.Nc3 Bb7 5.Qc2 g6 6.e4 Bg7 7.Be2 Nbd7 8.O-O O-O 9.h3 Re8 10.d5 e6 11.Rd1 exd5 12.exd5 Rc8 13.b4 c5 14.dxc6 Rxc6 (24.143.912) 2382

 20/28	 0:17 	+0.33 	4.Nc3 Bb4 5.e3 Bb7 6.Bd3 O-O 7.O-O d5 8.cxd5 exd5 9.a3 Bd6 10.b4 Nbd7 11.Rb1 Ne4 12.Bb2 c5 13.Nb5 Be7 14.dxc5 bxc5 15.bxc5 (44.497.107) 2485
 20/28	 0:17 	+0.28 	4.a3 Bb7 5.Nc3 d5 6.Bf4 Bd6 7.Bxd6 cxd6 8.e3 O-O 9.Bd3 dxc4 10.Bxc4 Nbd7 11.O-O Rc8 12.Bd3 e5 13.Bc2 Re8 14.Nb5 Bxf3 15.gxf3 exd4 16.Nxd6 (44.497.107) 2485
 20/31	 0:17 	+0.20 	4.e3 d5 5.Be2 Bb7 6.Nc3 Bd6 7.b3 O-O 8.O-O Nbd7 9.Bb2 dxc4 10.bxc4 c5 11.d5 exd5 12.cxd5 Qe7 13.Nb5 Bxd5 14.e4 (44.497.107) 2485
 20/31	 0:17 	+0.17 	4.Qc2 Bb7 5.Nc3 Bb4 6.e3 c5 7.Be2 cxd4 8.exd4 O-O 9.O-O d5 10.cxd5 Nxd5 11.Bg5 Qc7 12.Bd2 h6 13.Bd3 Nf6 14.a3 Be7 (44.497.107) 2485

 20/25	 0:15 	-0.06 	4...Bb4 5.e3 O-O 6.Bd3 Bb7 7.Bd2 d6 8.O-O Nbd7 9.Re1 c5 10.Rc1 Bxc3 11.Bxc3 Ne4 12.Bxe4 Bxe4 13.b4 h6 14.dxc5 dxc5 (37.424.520) 2344
 20/24	 0:15 	-0.02 	4...Bb7 5.e3 Bb4 6.Bd3 O-O 7.Bd2 d6 8.Qc2 c5 9.a3 Ba5 10.O-O Bxf3 11.gxf3 Nc6 12.dxc5 bxc5 13.f4 Rc8 14.Rad1 h6 (37.424.520) 2344
 20/33	 0:15 	+0.22 	4...d5 5.cxd5 Nxd5 6.e4 Nxc3 7.bxc3 Be7 8.e5 Bb7 9.Bd3 Nd7 10.O-O c5 11.Bf4 O-O 12.Rc1 cxd4 13.cxd4 Rc8 14.Qa4 (37.424.520) 2344
 20/28	 0:15 	+0.51 	4...Be7 5.e4 d6 6.Bd3 Nbd7 7.O-O O-O 8.h3 Bb7 9.Qc2 e5 10.d5 Nh5 11.g3 Nhf6 12.a3 c6 13.b4 Rc8 14.Bb2 cxd5 (37.424.520) 2344
It fell down to a score of 0.33.

Now, the highest scoring variations for white are both 0.40:

1. d4 Nf6 2. c4
1. e4 e6 2. Nc3

Is up to you to decide which one you shootout. I'll go for the second one since 1.d4 has more analysis:

1. e4 e6 2. Nc3

Code: Select all

 20/28	 0:16 	-0.05 	2...d5 3.d4 Nf6 4.Bg5 Bb4 5.e5 h6 6.exf6 hxg5 7.fxg7 Rg8 8.h4 gxh4 9.Qg4 Qf6 10.O-O-O Qxg7 11.Qxg7 Rxg7 12.Rxh4 Bd7 13.Nf3 Nc6 14.Rh8+ Bf8 (39.273.828) 2401
 20/31	 0:16 	+0.07 	2...c5 3.Nf3 Nc6 4.d4 cxd4 5.Nxd4 Nf6 6.Be2 Bb4 7.Nxc6 bxc6 8.Qd3 Qc7 9.O-O d5 10.Bg5 Bd6 11.g3 O-O 12.Bxf6 gxf6 13.exd5 exd5 (39.273.828) 2401
 20/29	 0:16 	+0.10 	2...Nf6 3.e5 Nd5 4.Nf3 d6 5.d4 dxe5 6.dxe5 Bb4 7.a3 Bxc3+ 8.bxc3 Nxc3 9.Qxd8+ Kxd8 10.Bg5+ Ke8 11.Bd3 Nd7 12.Bd2 Nd5 13.c4 Nc5 14.cxd5 Nxd3+ 15.Ke2 (39.273.828) 2401
 20/28	 0:16 	+0.18 	2...h6 3.d4 d5 4.Bd3 Bb4 5.exd5 exd5 6.Nf3 Nf6 7.O-O O-O 8.a3 Bxc3 9.bxc3 Re8 10.Re1 Nbd7 11.a4 c5 12.c4 dxc4 13.Rxe8+ Qxe8 14.Bxc4 Nb6 (39.273.828) 2401

 20/38	 0:13 	+0.28 	3.exd5 exd5 4.d4 Nf6 5.Nf3 Bd6 6.Nb5 O-O 7.Nxd6 Qxd6 8.Be2 Bf5 9.Nh4 Be4 10.f3 Bg6 11.O-O Nbd7 12.Nxg6 fxg6 13.Re1 Rae8 14.Bg5 h6 15.Bh4 (28.658.709) 2158
 20/29	 0:13 	+0.25 	3.d4 Nf6 4.e5 Nfd7 5.Nce2 c5 6.c3 Nc6 7.Nf3 Be7 8.g3 O-O 9.Bg2 b6 10.O-O Ba6 11.Re1 g5 12.h3 Kg7 13.Qa4 Bb7 14.Qc2 f6 15.exf6+ Bxf6 16.Kh1 (28.658.709) 2158
 20/30	 0:13 	+0.24 	3.Nf3 Nf6 4.exd5 exd5 5.d4 Nc6 6.Bd3 Nb4 7.Bf4 Nxd3+ 8.Qxd3 c6 9.O-O Be7 10.Be5 h6 11.h3 O-O 12.Ne2 Ne4 13.Ng3 Nxg3 14.Bxg3 Be6 15.Rfe1 (28.658.709) 2158
 20/28	 0:13 	+0.09 	3.d3 d4 4.Nce2 c5 5.g3 Nc6 6.f4 Be7 7.Nf3 Nf6 8.Bg2 b6 9.c3 Ng4 10.cxd4 cxd4 11.O-O O-O 12.h3 Ne3 13.Bxe3 dxe3 14.d4 (28.658.709) 2158

 20/32	 0:11 	-0.02 	3...exd5 4.d4 Nf6 5.Nf3 Nc6 6.a3 Bd6 7.Bd3 a6 8.O-O O-O 9.Bg5 Bg4 10.Be3 h6 11.h3 Bd7 12.Re1 Re8 13.Bd2 Be6 14.Na4 Bd7 (20.930.670) 1881
 20/32	 0:11 	+0.81 	3...Be7 4.Bc4 exd5 5.Nxd5 Be6 6.Ne3 Bxc4 7.Nxc4 Nc6 8.Nf3 Bf6 9.O-O Nge7 10.d3 O-O 11.Re1 Re8 12.Bd2 b5 13.Nce5 Nxe5 14.Nxe5 Nf5 15.Nc6 (20.930.670) 1881
 20/31	 0:11 	+0.82 	3...Bd6 4.dxe6 Bxe6 5.Nf3 Nf6 6.d4 O-O 7.Be2 h6 8.O-O c6 9.Bd3 Nbd7 10.h3 a5 11.Bd2 Nb6 12.Re1 Nc4 13.Bc1 b5 14.b3 Na3 15.Kh1 (20.930.670) 1881
 20/32	 0:11 	+0.82 	3...Nf6 4.dxe6 Bxe6 5.Nf3 Bd6 6.d4 O-O 7.Bd3 c5 8.dxc5 Bxc5 9.O-O h6 10.a3 a6 11.Bf4 Nc6 12.h3 Re8 13.Bg3 Nh5 14.Bh2 (20.930.670) 1881

 20/32	 0:14 	+0.28 	4.d4 Nf6 5.Nf3 Bd6 6.Nb5 O-O 7.Nxd6 Qxd6 8.Be2 Bf5 9.O-O Nbd7 10.Nh4 Be4 11.f3 Bg6 12.c3 Rae8 13.Bg5 c5 14.Re1 Re6 15.Qd2 Rfe8 16.Nxg6 hxg6 (36.845.039) 2456
 20/35	 0:14 	+0.22 	4.Nf3 Nf6 5.d4 Bd6 6.Nb5 O-O 7.Nxd6 Qxd6 8.Be2 Bf5 9.O-O Nbd7 10.Nh4 Be4 11.f3 Bg6 12.c3 Rae8 13.Bg5 c5 14.Re1 h6 15.Nxg6 fxg6 16.Bh4 (36.845.039) 2456
 20/28	 0:14 	+0.06 	4.Be2 d4 5.Ne4 Nf6 6.d3 Be7 7.Nxf6+ Bxf6 8.Bf3 O-O 9.Ne2 Re8 10.O-O c5 11.Re1 Nc6 12.Nf4 Be5 13.Bd2 Bd7 14.a3 Kh8 (36.845.039) 2456
 20/35	 0:14 	+0.02 	4.d3 Nf6 5.Nf3 d4 6.Ne4 Be7 7.Be2 Nc6 8.O-O O-O 9.c3 Be6 10.Qc2 h6 11.Nxd4 Nxd4 12.cxd4 Qxd4 13.Be3 Qe5 14.Nxf6+ Qxf6 15.Rae1 Bxa2 (36.845.039) 2456

 20/34	 0:09 	-0.19 	4...Nf6 5.Bd3 Nc6 6.Nf3 Bg4 7.Be3 Bb4 8.O-O O-O 9.h3 Bh5 10.g4 Bg6 11.Ne5 Re8 12.Nxg6 hxg6 13.a3 Bxc3 14.bxc3 Qd6 15.Qf3 Ne4 16.c4 Nc3 (24.647.063) 2594
 20/36	 0:09 	-0.16 	4...Nc6 5.Nf3 Nf6 6.Bd3 Bg4 7.Be3 Bb4 8.O-O O-O 9.h3 Bh5 10.a3 Bxc3 11.bxc3 Re8 12.c4 dxc4 13.Bxc4 Ne4 14.Re1 Qf6 15.Bd3 Bxf3 16.Qxf3 Qxf3 17.gxf3 Nc3 (24.647.063) 2594
 20/28	 0:09 	 0.00 	4...Be7 5.Bd3 Nf6 6.Nf3 O-O 7.O-O c5 8.dxc5 Bxc5 9.h3 Nc6 10.Bf4 a6 11.Re1 Be6 12.Ng5 Re8 13.Qd2 Nb4 14.Be5 h6 (24.647.063) 2594
 20/24	 0:09 	+0.02 	4...Bb4 5.Bd3 Nf6 6.Nge2 O-O 7.O-O Na6 8.Bf4 Bd6 9.a3 c6 10.Bxa6 bxa6 11.Qd3 Bxf4 12.Nxf4 Rb8 13.b3 Qc7 14.Nce2 h6 15.h3 (24.647.063) 2594
Down to a score of 0.28.

Let's check the other one...

1. d4 Nf6 2. c4

Code: Select all

 20/25	 0:17 	+0.06 	2...e6 3.Nf3 d5 4.Nc3 Bb4 5.e3 O-O 6.Bd2 b6 7.Be2 dxc4 8.O-O Bb7 9.Ne5 Nbd7 10.Nxc4 Bxc3 11.Bxc3 Ne4 12.Nd2 Ndf6 13.Bd3 Nxc3 14.bxc3 (40.954.602) 2315
 20/31	 0:17 	+0.10 	2...d5 3.Nc3 e6 4.Nf3 Bb4 5.e3 O-O 6.Bd2 b6 7.cxd5 exd5 8.Bd3 Bd6 9.Nb5 c5 10.Nxd6 Qxd6 11.O-O Nc6 12.Bc3 c4 13.Bc2 (40.954.602) 2315
 20/28	 0:17 	+0.12 	2...c6 3.Nc3 d5 4.cxd5 cxd5 5.Nf3 Nc6 6.Bf4 Qb6 7.Na4 Qb4+ 8.Bd2 Qd6 9.e3 e6 10.Rc1 Be7 11.Ne5 Ne4 12.Bb5 Bd7 13.Nxd7 Qxd7 14.O-O (40.954.602) 2315
 20/33	 0:17 	+0.30 	2...d6 3.Nf3 Nbd7 4.Nc3 e5 5.e4 Be7 6.Be2 O-O 7.O-O b6 8.Qc2 Bb7 9.d5 a5 10.Bd2 h6 11.a3 c6 12.b4 (40.954.602) 2315

 20/29	 0:16 	+0.47 	3.Nf3 d5 4.Nc3 Bb4 5.Bg5 O-O 6.e3 c5 7.Qb3 Nbd7 8.cxd5 exd5 9.Be2 h6 10.Bxf6 Nxf6 11.dxc5 Bxc5 12.Rd1 Bb6 13.Nxd5 (42.260.271) 2486
 20/30	 0:16 	+0.41 	3.Nc3 Bb4 4.Bd2 O-O 5.e3 d5 6.a3 Be7 7.Be2 b6 8.cxd5 exd5 9.Bf3 c5 10.Nge2 Nc6 11.O-O Ba6 12.Qa4 Bb7 13.Qc2 h6 (42.260.271) 2486
 20/29	 0:16 	+0.32 	3.e3 Be7 4.Nf3 d5 5.cxd5 exd5 6.Bd3 O-O 7.O-O c6 8.a3 Be6 9.Bd2 h6 10.Bb4 Bxb4 11.axb4 Qd6 12.b5 cxb5 13.Nbd2 b4 (42.260.271) 2486
 20/26	 0:16 	+0.29 	3.Bf4 d5 4.cxd5 exd5 5.e3 Bd6 6.Bxd6 Qxd6 7.Bd3 O-O 8.Nc3 Bg4 9.Nge2 Bxe2 10.Bxe2 Nbd7 11.O-O c6 12.h3 Rae8 13.a3 c5 14.dxc5 Nxc5 (42.260.271) 2486

 20/33	 0:15 	-0.01 	3...d5 4.Nc3 a6 5.cxd5 exd5 6.Bg5 c6 7.e3 Bf5 8.Bd3 Bxd3 9.Qxd3 Nbd7 10.O-O Be7 11.e4 dxe4 12.Nxe4 O-O 13.Rae1 h6 14.Bxf6 Nxf6 15.Nxf6+ Bxf6 (36.657.400) 2385
 20/30	 0:15 	 0.00 	3...b6 4.Bf4 d5 5.cxd5 exd5 6.Nc3 Bd6 7.Be5 O-O 8.e3 Be7 9.Bd3 Bb7 10.O-O Nbd7 11.Bf4 Nh5 12.Be5 Nhf6 (36.657.400) 2385
 20/34	 0:15 	+0.06 	3...Bb4+ 4.Nbd2 O-O 5.e3 d5 6.a3 Be7 7.Bd3 c5 8.dxc5 a5 9.Rb1 Bxc5 10.Qc2 dxc4 11.Nxc4 Nbd7 12.O-O Qc7 13.b4 axb4 14.axb4 Be7 15.b5 (36.657.400) 2385
 20/29	 0:15 	+0.07 	3...Be7 4.Nc3 d5 5.Bf4 O-O 6.e3 Nbd7 7.c5 c6 8.Bd3 b6 9.b4 a5 10.a3 Ba6 11.O-O Bxd3 12.Qxd3 h6 13.Rfd1 bxc5 14.bxc5 (36.657.400) 2385

 20/30	 0:25 	+0.40 	4.Nc3 a6 5.cxd5 exd5 6.Bg5 Be7 7.e3 Be6 8.Bd3 Nbd7 9.Bxf6 Nxf6 10.O-O O-O 11.Ne2 Ne4 12.Bxe4 dxe4 13.Nd2 Bf5 14.Ng3 (56.012.778) 2211
 20/26	 0:25 	+0.20 	4.a3 Be7 5.cxd5 exd5 6.Bf4 Nh5 7.Be5 Nc6 8.e3 O-O 9.Nc3 Nxe5 10.dxe5 Be6 11.Be2 g6 12.O-O Ng7 13.Nd4 c5 14.Nxe6 fxe6 15.Bf3 (56.012.778) 2211
 20/28	 0:25 	+0.20 	4.e3 Bd6 5.Be2 b6 6.O-O Bb7 7.Nc3 Nbd7 8.b3 dxc4 9.bxc4 O-O 10.Bb2 c5 11.d5 exd5 12.a4 Re8 13.cxd5 Qe7 14.Re1 Ne4 15.Nxe4 Qxe4 (56.012.778) 2211
 20/28	 0:25 	+0.17 	4.h3 c5 5.e3 a6 6.cxd5 exd5 7.Nc3 Nc6 8.dxc5 Bxc5 9.Bd3 O-O 10.O-O Re8 11.b3 Bd6 12.Bb2 Be6 13.Rc1 h6 14.Ne2 Ne4 15.a3 (56.012.778) 2211

 20/29	 0:22 	+0.05 	4...Nbd7 5.Bf4 dxc4 6.e3 b5 7.Nxb5 Nd5 8.Nc3 Nxf4 9.exf4 Rb8 10.Bxc4 Rxb2 11.O-O Bd6 12.Ne5 O-O 13.Bd3 Bxe5 14.fxe5 Bb7 15.Nb5 Rb4 (52.836.639) 2328
 20/27	 0:22 	+0.08 	4...Bb4 5.Bg5 O-O 6.e3 c5 7.cxd5 exd5 8.dxc5 Nbd7 9.Rc1 h6 10.Bxf6 Nxf6 11.Bd3 Bxc5 12.O-O Be6 13.Na4 Bd6 14.Nc5 Bxc5 15.Rxc5 Bg4 (52.836.639) 2328
 20/29	 0:22 	+0.11 	4...c6 5.e3 Nbd7 6.Bd3 Be7 7.cxd5 exd5 8.O-O O-O 9.Qc2 Nb6 10.h3 Re8 11.Re1 h6 12.Bd2 Bd6 13.e4 Nxe4 14.Nxe4 dxe4 15.Bxe4 Be6 16.Ne5 (52.836.639) 2328
 20/32	 0:22 	+0.18--	4...a6 5.cxd5 exd5 6.Bf4 c5 7.e3 c4 8.Be2 Bb4 9.O-O O-O 10.Ne5 Nc6 11.Nxc6 bxc6 12.Be5 Bxc3 13.bxc3 Ne4 14.Rc1 Be6 (52.836.639) 2328
The score survives at 0.40! At this depth at this ply McBrain couldn't find a better line for white than this:

0.40 1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. Nc3

But you may have noticed it'd actually have been quicker for me to play to white's 4th move instead of to black's 4th. Why the time wasting? Because we also want to find out black's mainline...

So up to this point black's 4...Nbd7 has a score of 0.05. 3...b6 has a score 0.00 so you'd want to shoot out the position after 1. d4 Nf6 2. c4 e6 3. Nf3 b6, but this time you don't care about white scores, but about black's, who wants to keep the scores as close to zero as possible, and if possible, in the negative and wih a score as highly negative as possible.

Please note this is a new analysis paradigm, as before the user would keep the same score for both white and black, so a score of 0.40 would mean black really hates the position, but here we don't pay attention and backsolve a 0.05 score for black.

For instance, a GUI like Aquarium has a feature called IDEA, that would allow you to do all this automatically, for any engine that supports MultiPV (so you could do this with default McBrain), inserting the relevant lines and Minimaxing automatically. But it doesn't support the new paradigm, so you'll either have the 0.40 score for both white and black (if you didn't check black's 4th), or the 0.05 score for both white and black (if you did), which is clearly wrong.

You might note all the lines I've been posted are garbage. This is because depth 20 is too low for McBrain to produce something useful, and engines in general suck at the opening stage because rating groups use opening books that skip it and so there's no reason to improve them there. The procedure is just an example of how to use this Smarter Tactical Setting with a Shootout feature of a GUI that lets you see the analysis of previous moves (but Chessbase/Fritz clear this after each move.)
Great spirits have always encountered violent opposition from mediocre minds.

Rodolfo Leoni
Posts: 544
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Rodolfo Leoni » Tue Feb 20, 2018 11:56 am

Ovyron wrote:
Leo wrote:Interesting. It must be some very high quality chess.
Who knows? If Rodolfo Leoni easily defeats me from both sides of the board that would prove that my methods didn't produce that much quality :lol:

But yeah, I suspect Rodolfo and me are two of the people with the highest experience using Persistent Hashes for engines, at least, I don't think you'll find anyone that has used Rybka 3 Persistent Hash more than me, and Rodolfo has an impressive record using those methods on the LSS.

I have proposed for we to continue playing games until one of us loses, to foce a kick out of being both undefeated, so we will see...
I think there's nothing easy with corr. chess against an expert computer chess user. With our games I miss the advantage of PHs so I guess we could draw lots of games. But there could be a different way. If we play OTB and we drink a glass of tequila before each move... you can expect a blunder on my side at 2nd or 3rd move! :lol:

Your offer to continue playing until a loss is accepted. But you know that I'm working to build a special GUI ... be alert! :wink:
F.S.I. Chess Teacher

User avatar
Ovyron
Posts: 2830
Joined: Tue Jul 03, 2007 2:30 am

Re: McBrain 9.0 with Smarter Tactical Setting

Post by Ovyron » Tue Feb 20, 2018 12:16 pm

Rodolfo Leoni wrote:But there could be a different way. If we play OTB and we drink a glass of tequila before each move... you can expect a blunder on my side at 2nd or 3rd move! :lol:
Ha! I don't need to be drunk to play like that! Just look at this game I lost recently...

[pgn][Event "Hourly Rapid Arena"]
[Site "https://lichess.org/eOgJdpyg"]
[Date "2018.02.17"]
[Round "-"]
[White "Ovyron"]
[Black "tubaraoazul"]
[WhiteElo "1663"]
[BlackElo "1537"]
[ECO "B32"]
[Result "0-1"]

1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 e5 {B32 Sicilian
Defense: Loewenthal Variation } 5. Nf5 d6 $6 {(-0.13 ?
0.38) Inaccuracy. Best move was d5. } 6. Nc3 Bxf5 7. exf5
Nf6 8. h3 $6 {(0.58 ? -0.31) Inaccuracy. Best move was g4.
} h6 $6 {(-0.31 ? 0.32) Inaccuracy. Best move was d5. }
9. g4 $6 {(0.32 ? -0.33) Inaccuracy. Best move was Bc4. }
Qb6 $6 {(-0.33 ? 0.58) Inaccuracy. Best move was d5. }
10. Nd5 $2 {(0.58 ? -0.62) Mistake. Best move was Bg2. }
Nxd5 $2 {(-0.62 ? 0.38) Mistake. Best move was Qa5+. }
11. Qxd5 Be7 12. Bc4 O-O 13. Bb3 Nd4 $6 {(0.25 ? 0.78)
Inaccuracy. Best move was a5. } 14. Be3 $6 {(0.78 ? 0.06)
Inaccuracy. Best move was c3. } Qa5+ $4 {(0.06 ? 12.26)
Blunder. Best move was a5. } 15. Bd2 $4 {(12.26 ? -2.50)
Blunder. Best move was Qxa5. } Nxb3 $4 {(-2.50 ? 5.00)
Blunder. Best move was Nxc2+. } 16. cxb3 $4 {(5.00 ?
-14.02) Blunder. Best move was Bxa5. } Qxd5 {White resigns.
} 0-1[/pgn]

Opponent plays 14...Qa5+?? Giving away the queen, and I don't take it!

For good measure, opponent plays 15...Nxb3?? The Queen is still free for taking!

And next I just capture the knight and give away a Queen of my own! XD

Really, I was completely sober, I wonder if my opponent used voodoo or something :lol:
Your offer to continue playing until a loss is accepted. But you know that I'm working to build a special GUI ... be alert! :wink:
Yes, and don't worry, by my playing rate we'll be playing our third game by 2025! 😰
Great spirits have always encountered violent opposition from mediocre minds.

MikeB
Posts: 3559
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: McBrain 9.0 with Smarter Tactical Setting

Post by MikeB » Tue Feb 20, 2018 3:51 pm

Ovyron wrote:
MikeB wrote:If I set multiipv =2 in the commandline, and do searchmoves , say e2e4 and d2d4 at the start position, it returns the the pv for those two moves
Already answered here:
Ovyron wrote:What's the difference between this and setting my GUI to play mode with some MultiPV setting?

There's no difference, but many GUIs for analysis don't support that (they switch to SinglePV in play mode), nor do they allow to use "go searchmoves" with MultiPV (again, it goes back to SinglePV) so hacking the Tactical setting is the solution.
Basically, an engine shouldn't depend for the user to have to type in any command to send for the engine and should be fully capable of doing what the user wants from the GUI that the user uses. So it comes down to this:

This version allows the user to do everything I explained without having to type in commands to the engine.

It's about user friendliness and there's something wrong if user is forced to use their keyboard to type stuff in. After examining hundreds of positions interactively I'd already have lost a lot of time typing in "searchmoves" in some console. My GUI allows me to use "go searchmoves" by holding the SHIFT key and dragging pieces on the board, to exclude them and hold the CTRL key and drag he moves to include them, no need to type anything in a console.

Sure, this version wouldn't be necessary for me at all if my GUI supported "MultiPV+Play Mode" or "MultiPV+Go Searchmoves", but it doesn't, and thus, this McBrain will help out people that want to do this without having to use the command line.

I also think that some MultiPV setting in engine options to set the engine to some permanent state of MultiPV to mix in with Play Mode or Go Searchmoves would work, without having to depend on the Tactical setting, but I managed to fully implement wanted behavior by just removing one line of code from Tactical, so this was clearly the most practical thing to do.
Well you are free to fork McBrain , as you already know , but if you are going to release anything to the public , I ask that you name it something else. Thank you.

Post Reply