On Opening books in 2015

[Lyudmil Tsvetkov]

The Two Knights' is probably equal but the path to equality is a narrow one and there are a lot of ways to go wrong.

This is an interesting observation, since it suggests that we can quantify something. For example, one could attempt to measure the fraction of legal moves that alter the theoretical value of the position. Some positions would then be more "volatile" than others ---but still each position has a theoretical evaluation of +∞, -∞ or 0.

Precisely the case: for a long time people have wondered if d4 or e4 is the best first move for white, with sharply differing opinions.

Well, the truth is that d4 is more successful in a wide range of lines with no perfect black play, while e4 is successful in a relatively fewer number of lines, but with perfect play from black.

That is why, with perfect play, e4 should be better than d4, though, in terms of the width of winning continuations with imperfect black answers, d4 certainly performs much better than e4.

[Laskos]

One more iteration, but what does it mean anything more than the last one?

Code: Select all

info depth 43 seldepth 67 multipv 1 score cp -26 nodes 70963710485 nps 18178544 hashfull 999 tbhits 0 time 3903707 pv d7d5 e4d5 c6a5 c4b5 c8d7 d1e2 f8e7 g5f3 f6d5 f3e5 c7c6 b5d3 d5f4 e2e4 f4d3 e4d3 d7e6 d3d8 a8d8 d2d3 e7f6 f2f4 c6c5 c1e3 f6e5 f4e5 b7b6 e1g1 a5c6 b1c3 e8g8 e3f4 c5c4 d3c4 e6c4 f1d1 c6b4 d1d6 c4e6 h2h3 d8d6 e5d6 f8d8 a1c1 e6f5 c1d1 h7h6 d1d2 b4c2 b2b3 c2b4

Now it is getting closer to the truth.

I think Lyudmil is right. From this position, I built an opening book depth=6, and played 400 fast self-play games with Stockfish. The outcome was 132 white wins, 205 draws, 63 black wins. Performance ~59%, much higher than an average opening (53%). The Stockfish eval of this position, according to my data of performance versus eval should be ~45 cp.

[zullil]

Let's be specific.

Two Knight's Defence:

[d]r1bqkb1r/pppp1ppp/2n2n2/4p1N1/2B1P3/8/PPPP1PPP/RNBQK2R b KQkq - 0 4

Which side favours that one?

What do you mean when you say a position "favors" one side or another? (Serious question---please consider deeply.) The value of any position is simply White wins, Black wins or draw.

More specifically, what does "White is better by 60 cp" mean? Is this different from "White is ahead by 20 cp"? Is it different from "Black is ahead by 20 cp"?

Yeah, I think white wins that, as 60cps, if the assessment is even very slightly correct, should be just above or near the winning margin in the opening.

Fischer also claimed white wins that.
And Fischer was able to whitewash his direct opponents at the height of his strength, so I am inclined to believe even only his words here.

Don't think you've really addressed my question. What do humans mean when they say a position favors White or favors Black? What does the statement mean?

If I understand correctly, your assertion is that with optimal play by both sides, the Two Knights position you posted is a win for White. My guess is that it is a draw, as are almost all standard opening positions.

In any case, Stockfish's evaluation based on its search is currently

Code: Select all

info depth 42 seldepth 67 multipv 1 score cp -16 nodes 46315755063 nps 17942433 hashfull 999 tbhits 0 time 2581353 pv d7d5 e4d5 c6a5 d2d3 h7h6 g5f3 e5e4 d1e2 a5c4 d3c4 f8c5 f3d2 d8e7 e1g1 e8g8 b1c3 c8g4 e2e1 e4e3 f2e3 e7e3 e1e3 c5e3 g1h1 e3d4 h2h3 d4c3 h3g4 c3e5 d2f3 f6g4 f3e5 g4e5 b2b3 f8e8 c1f4 a8d8 a1e1 f7f6 h1h2 c7c6 f4e5 f6e5 d5c6 b7c6 a2a4 e5e4 h2g3 e4e3 b3b4

So we might say that "the position favors White by a bit", but I'm not sure that statement has any well-defined meaning.

A human thinks the same as an engine: maybe I have an advantage, it is small, so probably not winning, but I am not certain even about that...

Still, I stick by my opinion.

"maybe" implies some sort of probabilistic interpretation. But what exactly?
The true value of the position is either ∞, -∞ or 0. With perfect players, "maybe" has no place. There are no degrees of advantage---the position is a win, a loss or a draw.

So maybe a win against whom/what? Fischer's ghost? Stockfish? Skipper? Me?

[Michel]

"maybe" implies some sort of probabilistic interpretation. But what exactly?
The true value of the position is either ∞, -∞ or 0. With perfect players, "maybe" has no place. There are no degrees of advantage---the position is a win, a loss or a draw.

You can assign elo to a position. The idea is that the elo of a position should be added to the elo of the player to move to compute his expected score from that position. The elo of the opening position is commonly called the "white advantage".

The elo of a position can be determined using the standard maximum likelihood methods. Of course this depends on the validity of the elo model and the latter holds only approximately.

Elo can be converted to an advantage in "pawns". Indeed elo can be converted to an expected score and similarly there exist models that convert an advantage in pawns to an expected score (I forgot how that goes).

[Lyudmil Tsvetkov]

Let's be specific.

Two Knight's Defence:

[d]r1bqkb1r/pppp1ppp/2n2n2/4p1N1/2B1P3/8/PPPP1PPP/RNBQK2R b KQkq - 0 4

Which side favours that one?

What do you mean when you say a position "favors" one side or another? (Serious question---please consider deeply.) The value of any position is simply White wins, Black wins or draw.

More specifically, what does "White is better by 60 cp" mean? Is this different from "White is ahead by 20 cp"? Is it different from "Black is ahead by 20 cp"?

Yeah, I think white wins that, as 60cps, if the assessment is even very slightly correct, should be just above or near the winning margin in the opening.

Fischer also claimed white wins that.
And Fischer was able to whitewash his direct opponents at the height of his strength, so I am inclined to believe even only his words here.

Don't think you've really addressed my question. What do humans mean when they say a position favors White or favors Black? What does the statement mean?

If I understand correctly, your assertion is that with optimal play by both sides, the Two Knights position you posted is a win for White. My guess is that it is a draw, as are almost all standard opening positions.

In any case, Stockfish's evaluation based on its search is currently

Code: Select all

info depth 42 seldepth 67 multipv 1 score cp -16 nodes 46315755063 nps 17942433 hashfull 999 tbhits 0 time 2581353 pv d7d5 e4d5 c6a5 d2d3 h7h6 g5f3 e5e4 d1e2 a5c4 d3c4 f8c5 f3d2 d8e7 e1g1 e8g8 b1c3 c8g4 e2e1 e4e3 f2e3 e7e3 e1e3 c5e3 g1h1 e3d4 h2h3 d4c3 h3g4 c3e5 d2f3 f6g4 f3e5 g4e5 b2b3 f8e8 c1f4 a8d8 a1e1 f7f6 h1h2 c7c6 f4e5 f6e5 d5c6 b7c6 a2a4 e5e4 h2g3 e4e3 b3b4

So we might say that "the position favors White by a bit", but I'm not sure that statement has any well-defined meaning.

A human thinks the same as an engine: maybe I have an advantage, it is small, so probably not winning, but I am not certain even about that...

Still, I stick by my opinion.

"maybe" implies some sort of probabilistic interpretation. But what exactly?
The true value of the position is either ∞, -∞ or 0. With perfect players, "maybe" has no place. There are no degrees of advantage---the position is a win, a loss or a draw.

So maybe a win against whom/what? Fischer's ghost? Stockfish? Skipper? Me?

You know if the position is a W,D or L in simple positions.
For those positions you also know what the perfect play/optimal moves is.

For more complicated positions, like openings, you might reasonably suspect that a certain position is a W,D or L, but can not be absolutely sure, as tremendous depths are required to confirm this, whether by human players or engines.

Similarly, in such positions there is a perfect sequence of moves, but it is very difficult to reproduce this sequence, as it is very long and often non-trivial. A perfect player might be able to do so, but, as far as I know, the perfect player, whether engine or human, still has not been born.

When this player is born, you might address this question to him, for the time being - back to Gligoric-Smyslov.

[jefk]

[quote="zullil"]
In any case, Stockfish's evaluation based on its search is currently
[code]info depth 42 seldepth 67 multipv 1 score cp -16 nodes 46315755063 nps 17942433 hashfull 999 tbhits 0 time 2581353 pv d7d5 e4d5 c6a5 d2d3 [/quote]

depth 42 looks like an impressive search, but it illustrates again
you cannot look deep into the pv.
At ply 4 it already goes wrong, in my book Bb5+ is better
than d3, and also if you check on the chessbase online database

http://database.chessbase.com/js/apps/database/

checked it again, and my new multimove analysis ends
in a slight advantage of 0.20 for White. Provided both sides
play accurately. So this preliminary end result (at least for me)
is not such a big advantage, but still 'proving' that ...Bc5 i.e.
Italian fundamentally is more solid for Black.
But the Two knights defense is more double edged, and
thus might be interesting against a weaker opponent.

jef

[Lyudmil Tsvetkov]

In any case, Stockfish's evaluation based on its search is currently

Code: Select all

info depth 42 seldepth 67 multipv 1 score cp -16 nodes 46315755063 nps 17942433 hashfull 999 tbhits 0 time 2581353 pv d7d5 e4d5 c6a5 d2d3 [/quote]

depth 42 looks like an impressive search, but it illustrates again
you cannot look deep into the pv.
At ply 4 it already goes wrong, in my book Bb5+ is  better 
than d3, and also if you check on the chessbase  online database

http://database.chessbase.com/js/apps/database/

checked it again, and my new multimove analysis ends  
in a  slight advantage of 0.20 for White. Provided both sides
play accurately. So this preliminary end result (at least for me)
is not  such a big advantage, but still 'proving' that ...Bc5 i.e.
Italian  fundamentally is more solid for Black.
But the Two knights defense is more double edged,  and 
thus might be interesting against a weaker opponent.

jef[/quote]

0.20 for white?
That is almost a draw.

[d]r1bqkb1r/p4pp1/2p2n1p/n3p3/8/7N/PPPPBPPP/RNBQK2R b KQkq - 1 9

This is a critical position arising after white's Nh3 - the Fischer suggestion.
Previous moves are more or less best and forced for both sides.

Any clue how black equalises here, as I really can not see?
All I see is very big white advantage, just as Fischer thought. 

For reference, you might also want to look at this exemplary Fischer game:

[pgn][Event "Tournament"]
[Site "New York (USA)"]
[Date "1963.??.??"]
[EventDate "?"]
[Round "5"]
[Result "1-0"]
[White "Robert James Fischer"]
[Black "Arthur Bisguier"]
[ECO "C59"]
[WhiteElo "?"]
[BlackElo "?"]
[PlyCount "57"]

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5+ c6
7.dxc6 bxc6 8.Be2 h6 9.Nh3 Bc5 10.O-O O-O 11.d3 Bxh3 12.gxh3
Qd7 13.Bf3 Qxh3 14.Nd2 Rad8 15.Bg2 Qf5 16.Qe1 Rfe8 17.Ne4 Bb6
18.Nxf6+ Qxf6 19.Kh1 c5 20.Qc3 Nc6 21.f4 Nd4 22.Qc4 Qg6 23.c3
Nf5 24.fxe5 Rxe5 25.Bf4 Re2 26.Be4 Rxb2 27.Be5 Re8 28.Rxf5
Rxe5 29.Rxe5 1-0


[/pgn]

[jefk]

[quote="Lyudmil Tsvetkov"]
Any clue how black equalises here, as I really can not see?
All I see is very big white advantage, just as Fischer thought.
For reference, you might also want to look at this exemplary Fischer game:
1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 Na5 6.Bb5+ c6
7.dxc6 bxc6 8.Be2 h6 9.Nh3 Bc5
[/quote]

ok well i have 9...Nb7! in my book, checked it with the Chessbase
livebook, it there is followed by 10.d3 with only 50 pct score.
Also 9...Bd6 might be ok, it appears with preliminary SF
analysis, but then 10.d4 is scoring 60 pct with the live-book..
So without going to analyse it all i prefer 9...Nb7!

And because 9...Nb7! = is good enough for me (for Black)
after 9.Be2, I have 9.Bd3 instead of 9.Be2 in my book, and
then you can get quite some complicated lines. Not going
to show them all to you, but like above if you have a
question about a specific move i can answer.

NB from this 'live book' it's not directly apparent how
to obtain such a slight advantage for White, it again requires
careful analysis, and backsolving, to obtain such a result..
But it ( aslight advantage) makes sense though, here is the
complicated and rather double-edged end-position (move 23):

r1b5/p4kpp/3n1b2/q1p3p1/8/BP1P4/P2P1PPP/1RQ1R1K1 b - - 1 1

best regards from a 'veterate' theoretician; although i
don't get a heart attack from such lines, though


jef

[Graham Banks]

Another one.

[d]r1bqkbnr/1ppp1ppp/p1n5/1B2p3/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 0 4

Ruy Lopez, Morphy Defence/a6

Which is the best move here?

As, according to my SF, Bc6, the Exchange Variation, is the best continuation, with Ba4 ranking many centipawns behind.

So, if I have to trust SF in the opening, I have to play Bc6 every single time.

Those positions have been analysed extensively by humans, and it is proven beyond any doubt that Ba4 is better than Bc6, much better indeed.

How do you want me to trust SF here?

I know where to trust SF: in open middlegames and endgames with abundant tactics; there the engine excels.

However, I do not trust SF in closed positions, specific theoretical endgames, as well as in a large bunch of openings.
The search depth and understanding of engines for such positions is simply too low.

[pgn][Event "Komodo 1392 vs Stockfish 280"]
[Site "Chess"]
[Date "2015.03.05"]
[Round "9.2"]
[White "Stockfish 280215 x64 8CPU"]
[Black "Komodo 1392 x64 8CPU"]
[Result "1/2-1/2"]
[Time "04"]
[TimeControl "7200+30"]
[PlyCount "111"]
[Number "18"]
[Termination "GUI TB adjudication"]
[WhiteType "program"]
[BlackType "program"]
[Variant "normal"]

{ AMD FX-8350 }
1.e4 {[%eval 0,1] [%emt 00:00:00]} e5 {[%eval 0,1] [%emt 00:00:00]}
2.Nf3 {[%eval 22,31] [%emt 00:02:57]} Nc6 {(Nc6) [%eval -14,24] [%emt 00:01:50]}
3.Bb5 {(Bc4) [%eval 25,33] [%emt 00:02:05]} a6 {(a6) [%eval -17,28] [%emt 00:04:41]}
4.Bxc6 {(Ba4) [%eval 23,34] [%emt 00:02:16]} dxc6 {(dxc6) [%eval -14,28] [%emt 00:02:15]}
5.O-O {(O-O) [%eval 20,34] [%emt 00:02:34]} Be6 {(Be6) [%eval -13,30] [%emt 00:03:53]}
6.c3 {(d4) [%eval 23,34] [%emt 00:06:20]} Qd3 {(Nf6) [%eval -1,30] [%emt 00:04:03]}
7.Nxe5 {(Nxe5) [%eval 4,33] [%emt 00:02:29]} Qxe4 {(Qxe4) [%eval 2,28] [%emt 00:01:35]}
8.d4 {(d4) [%eval 2,33] [%emt 00:03:37]} O-O-O {(Qf5) [%eval 1,29] [%emt 00:03:29]}
9.Nd2 {(Nd2) [%eval 19,31] [%emt 00:02:31]} Qd5 {(Qd5) [%eval 0,27] [%emt 00:01:59]}
10.Nb3 {(Re1) [%eval 15,33] [%emt 00:05:04]} f6 {(Bd6) [%eval 14,28] [%emt 00:03:16]}
11.Nd3 {(Nd3) [%eval 14,30] [%emt 00:01:56]} Bf5 {(Bf7) [%eval 14,28] [%emt 00:02:41]}
12.Re1 {(Re1) [%eval 19,31] [%emt 00:01:56]} g5 {(g5) [%eval 9,28] [%emt 00:03:15]}
13.Be3 {(Qe2) [%eval 13,33] [%emt 00:05:05]} b6 {(b6) [%eval 11,26] [%emt 00:03:37]}
14.f3 {(f3) [%eval 31,31] [%emt 00:03:19]} h5 {(h5) [%eval 6,27] [%emt 00:08:07]}
15.Nf2 {(Nf2) [%eval 0,31] [%emt 00:02:19]} g4 {(Ne7) [%eval 9,25] [%emt 00:05:24]}
16.Rc1 {(Rc1) [%eval 17,29] [%emt 00:03:18]} Ne7 {(Qd7) [%eval 15,25] [%emt 00:02:33]}
17.c4 {(c4) [%eval 46,31] [%emt 00:05:40]} Qg8 {(Qg8) [%eval 7,24] [%emt 00:04:38]}
18.d5 {(Qe2) [%eval 44,32] [%emt 00:14:47]} g3 {(g3) [%eval 24,23] [%emt 00:01:44]}
19.Nd3 {(Ne4) [%eval 8,30] [%emt 00:02:15]} gxh2 {(gxh2) [%eval -13,20] [%emt 00:00:47]}
20.Kh1 {(Kh1) [%eval 0,34] [%emt 00:01:39]} h4 {(Bxd3) [%eval -13,24] [%emt 00:02:08]}
21.Bf4 {(Bf4) [%eval 60,30] [%emt 00:01:50]} Qg6 {(Qg6) [%eval -13,27] [%emt 00:03:48]}
22.Nb4 {(Nb4) [%eval 37,32] [%emt 00:02:26]} Rg8 {(Rg8) [%eval -13,26] [%emt 00:00:53]}
23.Re2 {(Qd2) [%eval 0,33] [%emt 00:01:54]} a5 {(a5) [%eval 13,24] [%emt 00:01:57]}
24.Nd4 {(Nd4) [%eval 0,35] [%emt 00:02:18]} h3 {(h3) [%eval 0,27] [%emt 00:02:49]}
25.g3 {(g3) [%eval 0,36] [%emt 00:01:45]} Bh6 {(Bh6) [%eval 0,28] [%emt 00:01:18]}
26.Nbxc6 {(Qd2) [%eval 0,37] [%emt 00:01:57]} Nxc6 {(Nxc6) [%eval -13,29] [%emt 00:01:46]}
27.Nxc6 {(Nxc6) [%eval 0,41] [%emt 00:02:10]} Qh5 {(Rde8) [%eval -5,30] [%emt 00:02:52]}
28.Ne7 {(Ne7) [%eval 0,39] [%emt 00:01:57]} Kb7 {(Kb7) [%eval 0,31] [%emt 00:02:25]}
29.Nxf5 {(Nxf5) [%eval 0,40] [%emt 00:01:09]} Qxf3 {(Qxf5) [%eval 0,32] [%emt 00:02:56]}
30.Kxh2 {(Kxh2) [%eval 0,40] [%emt 00:01:58]} Bxf4 {(Bxf4) [%eval 0,32] [%emt 00:01:31]}
31.Rc3 {(Rc3) [%eval 0,42] [%emt 00:02:07]} Qg4 {(Qg4) [%eval 0,31] [%emt 00:03:42]}
32.Qd3 {(Qf1) [%eval 0,43] [%emt 00:02:22]} Be5 {(Rg5) [%eval 0,30] [%emt 00:03:34]}
33.Rb3 {(Ra3) [%eval 8,37] [%emt 00:02:00]} Rg5 {(Rg5) [%eval -12,29] [%emt 00:07:13]}
34.Rf2 {(Rf2) [%eval 6,39] [%emt 00:02:28]} Rxf5 {(Rxf5) [%eval -12,31] [%emt 00:05:07]}
35.Qxf5 {(Qxf5) [%eval 0,41] [%emt 00:01:38]} Qxc4 {(Qxc4) [%eval -10,29] [%emt 00:00:59]}
36.Rd2 {(Rd3) [%eval 6,42] [%emt 00:01:18]} Kb8 {(Rxd5) [%eval 0,27] [%emt 00:03:57]}
37.Qf3 {(d6) [%eval 20,36] [%emt 00:05:55]} Ka7 {(Qa4) [%eval 0,25] [%emt 00:01:12]}
38.Qe2 {(Re3) [%eval 36,35] [%emt 00:01:32]} Qa4 {(Qxe2) [%eval -3,27] [%emt 00:01:06]}
39.Rbd3 {(Rf3) [%eval 2,34] [%emt 00:02:32]} Re8 {(Re8) [%eval 0,29] [%emt 00:02:42]}
40.Qd1 {(Qf2) [%eval 0,36] [%emt 00:01:15]} Qd7 {(Qd7) [%eval 0,29] [%emt 00:02:58]}
41.a3 {(b4) [%eval 0,39] [%emt 00:02:02]} f5 {(f5) [%eval 0,28] [%emt 00:01:30]}
42.Rf3 {(Rf3) [%eval 0,39] [%emt 00:00:50]} Rf8 {(Rg8) [%eval 0,30] [%emt 00:03:03]}
43.Qf1 {(Qf1) [%eval 0,40] [%emt 00:01:16]} Qg7 {(f4) [%eval 0,30] [%emt 00:02:11]}
44.b4 {(b4) [%eval 0,41] [%emt 00:01:22]} Qg4 {(Qg5) [%eval 0,32] [%emt 00:01:57]}
45.Qe2 {(Re2) [%eval 0,41] [%emt 00:01:18]} Bd6 {(Re8) [%eval 0,32] [%emt 00:01:53]}
46.bxa5 {(bxa5) [%eval 0,43] [%emt 00:01:09]} f4 {(f4) [%eval 0,34] [%emt 00:01:16]}
47.axb6 {(axb6) [%eval 0,49] [%emt 00:01:05]} Kb7 {(cxb6) [%eval 0,34] [%emt 00:01:27]}
48.Rxf4 {(Rxf4) [%eval 0,49] [%emt 00:01:15]} Qxe2 {(Qxe2) [%eval 0,42] [%emt 00:01:25]}
49.Rxe2 {(Rxe2) [%eval 0,63] [%emt 00:01:17]} Bxf4 {(Bxf4) [%eval 0,43] [%emt 00:00:28]}
50.gxf4 {(gxf4) [%eval 0,64] [%emt 00:00:59]} Rxf4 {(Rxf4) [%eval 0,49] [%emt 00:02:13]}
51.bxc7 {(Re6) [%eval 0,78] [%emt 00:01:05]} Kxc7 {(Kxc7) [%eval 0,59] [%emt 00:00:30]}
52.a4 {(Kxh3) [%eval 0,121] [%emt 00:00:42]} Rxa4 {(Rxa4) [%eval 0,83] [%emt 00:03:22]}
53.Re7 {[%eval 0,64] [%emt 00:00:43]} Kd6 {(Kd6) [%eval 0,37] [%emt 00:00:21]}
54.Rh7 {(Rh7) [%eval 0,70] [%emt 00:00:41]} Ra1 {(Kxd5) [%eval 0,41] [%emt 00:00:48]}
55.Rh8 {(Rh6) [%eval 0,64] [%emt 00:00:41]} Ra2 {(Kxd5) [%eval 0,33] [%emt 00:00:18]}
56.Kxh3 {(Kxh3) [%eval 0,71] [%emt 00:00:41]}
1/2-1/2[/pgn]

[Lyudmil Tsvetkov]

Another one.

[d]r1bqkbnr/1ppp1ppp/p1n5/1B2p3/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 0 4

Ruy Lopez, Morphy Defence/a6

Which is the best move here?

As, according to my SF, Bc6, the Exchange Variation, is the best continuation, with Ba4 ranking many centipawns behind.

So, if I have to trust SF in the opening, I have to play Bc6 every single time.

Those positions have been analysed extensively by humans, and it is proven beyond any doubt that Ba4 is better than Bc6, much better indeed.

How do you want me to trust SF here?

I know where to trust SF: in open middlegames and endgames with abundant tactics; there the engine excels.

However, I do not trust SF in closed positions, specific theoretical endgames, as well as in a large bunch of openings.
The search depth and understanding of engines for such positions is simply too low.

[pgn][Event "Komodo 1392 vs Stockfish 280"]
[Site "Chess"]
[Date "2015.03.05"]
[Round "9.2"]
[White "Stockfish 280215 x64 8CPU"]
[Black "Komodo 1392 x64 8CPU"]
[Result "1/2-1/2"]
[Time "04"]
[TimeControl "7200+30"]
[PlyCount "111"]
[Number "18"]
[Termination "GUI TB adjudication"]
[WhiteType "program"]
[BlackType "program"]
[Variant "normal"]

{ AMD FX-8350 }
1.e4 {[%eval 0,1] [%emt 00:00:00]} e5 {[%eval 0,1] [%emt 00:00:00]}
2.Nf3 {[%eval 22,31] [%emt 00:02:57]} Nc6 {(Nc6) [%eval -14,24] [%emt 00:01:50]}
3.Bb5 {(Bc4) [%eval 25,33] [%emt 00:02:05]} a6 {(a6) [%eval -17,28] [%emt 00:04:41]}
4.Bxc6 {(Ba4) [%eval 23,34] [%emt 00:02:16]} dxc6 {(dxc6) [%eval -14,28] [%emt 00:02:15]}
5.O-O {(O-O) [%eval 20,34] [%emt 00:02:34]} Be6 {(Be6) [%eval -13,30] [%emt 00:03:53]}
6.c3 {(d4) [%eval 23,34] [%emt 00:06:20]} Qd3 {(Nf6) [%eval -1,30] [%emt 00:04:03]}
7.Nxe5 {(Nxe5) [%eval 4,33] [%emt 00:02:29]} Qxe4 {(Qxe4) [%eval 2,28] [%emt 00:01:35]}
8.d4 {(d4) [%eval 2,33] [%emt 00:03:37]} O-O-O {(Qf5) [%eval 1,29] [%emt 00:03:29]}
9.Nd2 {(Nd2) [%eval 19,31] [%emt 00:02:31]} Qd5 {(Qd5) [%eval 0,27] [%emt 00:01:59]}
10.Nb3 {(Re1) [%eval 15,33] [%emt 00:05:04]} f6 {(Bd6) [%eval 14,28] [%emt 00:03:16]}
11.Nd3 {(Nd3) [%eval 14,30] [%emt 00:01:56]} Bf5 {(Bf7) [%eval 14,28] [%emt 00:02:41]}
12.Re1 {(Re1) [%eval 19,31] [%emt 00:01:56]} g5 {(g5) [%eval 9,28] [%emt 00:03:15]}
13.Be3 {(Qe2) [%eval 13,33] [%emt 00:05:05]} b6 {(b6) [%eval 11,26] [%emt 00:03:37]}
14.f3 {(f3) [%eval 31,31] [%emt 00:03:19]} h5 {(h5) [%eval 6,27] [%emt 00:08:07]}
15.Nf2 {(Nf2) [%eval 0,31] [%emt 00:02:19]} g4 {(Ne7) [%eval 9,25] [%emt 00:05:24]}
16.Rc1 {(Rc1) [%eval 17,29] [%emt 00:03:18]} Ne7 {(Qd7) [%eval 15,25] [%emt 00:02:33]}
17.c4 {(c4) [%eval 46,31] [%emt 00:05:40]} Qg8 {(Qg8) [%eval 7,24] [%emt 00:04:38]}
18.d5 {(Qe2) [%eval 44,32] [%emt 00:14:47]} g3 {(g3) [%eval 24,23] [%emt 00:01:44]}
19.Nd3 {(Ne4) [%eval 8,30] [%emt 00:02:15]} gxh2 {(gxh2) [%eval -13,20] [%emt 00:00:47]}
20.Kh1 {(Kh1) [%eval 0,34] [%emt 00:01:39]} h4 {(Bxd3) [%eval -13,24] [%emt 00:02:08]}
21.Bf4 {(Bf4) [%eval 60,30] [%emt 00:01:50]} Qg6 {(Qg6) [%eval -13,27] [%emt 00:03:48]}
22.Nb4 {(Nb4) [%eval 37,32] [%emt 00:02:26]} Rg8 {(Rg8) [%eval -13,26] [%emt 00:00:53]}
23.Re2 {(Qd2) [%eval 0,33] [%emt 00:01:54]} a5 {(a5) [%eval 13,24] [%emt 00:01:57]}
24.Nd4 {(Nd4) [%eval 0,35] [%emt 00:02:18]} h3 {(h3) [%eval 0,27] [%emt 00:02:49]}
25.g3 {(g3) [%eval 0,36] [%emt 00:01:45]} Bh6 {(Bh6) [%eval 0,28] [%emt 00:01:18]}
26.Nbxc6 {(Qd2) [%eval 0,37] [%emt 00:01:57]} Nxc6 {(Nxc6) [%eval -13,29] [%emt 00:01:46]}
27.Nxc6 {(Nxc6) [%eval 0,41] [%emt 00:02:10]} Qh5 {(Rde8) [%eval -5,30] [%emt 00:02:52]}
28.Ne7 {(Ne7) [%eval 0,39] [%emt 00:01:57]} Kb7 {(Kb7) [%eval 0,31] [%emt 00:02:25]}
29.Nxf5 {(Nxf5) [%eval 0,40] [%emt 00:01:09]} Qxf3 {(Qxf5) [%eval 0,32] [%emt 00:02:56]}
30.Kxh2 {(Kxh2) [%eval 0,40] [%emt 00:01:58]} Bxf4 {(Bxf4) [%eval 0,32] [%emt 00:01:31]}
31.Rc3 {(Rc3) [%eval 0,42] [%emt 00:02:07]} Qg4 {(Qg4) [%eval 0,31] [%emt 00:03:42]}
32.Qd3 {(Qf1) [%eval 0,43] [%emt 00:02:22]} Be5 {(Rg5) [%eval 0,30] [%emt 00:03:34]}
33.Rb3 {(Ra3) [%eval 8,37] [%emt 00:02:00]} Rg5 {(Rg5) [%eval -12,29] [%emt 00:07:13]}
34.Rf2 {(Rf2) [%eval 6,39] [%emt 00:02:28]} Rxf5 {(Rxf5) [%eval -12,31] [%emt 00:05:07]}
35.Qxf5 {(Qxf5) [%eval 0,41] [%emt 00:01:38]} Qxc4 {(Qxc4) [%eval -10,29] [%emt 00:00:59]}
36.Rd2 {(Rd3) [%eval 6,42] [%emt 00:01:18]} Kb8 {(Rxd5) [%eval 0,27] [%emt 00:03:57]}
37.Qf3 {(d6) [%eval 20,36] [%emt 00:05:55]} Ka7 {(Qa4) [%eval 0,25] [%emt 00:01:12]}
38.Qe2 {(Re3) [%eval 36,35] [%emt 00:01:32]} Qa4 {(Qxe2) [%eval -3,27] [%emt 00:01:06]}
39.Rbd3 {(Rf3) [%eval 2,34] [%emt 00:02:32]} Re8 {(Re8) [%eval 0,29] [%emt 00:02:42]}
40.Qd1 {(Qf2) [%eval 0,36] [%emt 00:01:15]} Qd7 {(Qd7) [%eval 0,29] [%emt 00:02:58]}
41.a3 {(b4) [%eval 0,39] [%emt 00:02:02]} f5 {(f5) [%eval 0,28] [%emt 00:01:30]}
42.Rf3 {(Rf3) [%eval 0,39] [%emt 00:00:50]} Rf8 {(Rg8) [%eval 0,30] [%emt 00:03:03]}
43.Qf1 {(Qf1) [%eval 0,40] [%emt 00:01:16]} Qg7 {(f4) [%eval 0,30] [%emt 00:02:11]}
44.b4 {(b4) [%eval 0,41] [%emt 00:01:22]} Qg4 {(Qg5) [%eval 0,32] [%emt 00:01:57]}
45.Qe2 {(Re2) [%eval 0,41] [%emt 00:01:18]} Bd6 {(Re8) [%eval 0,32] [%emt 00:01:53]}
46.bxa5 {(bxa5) [%eval 0,43] [%emt 00:01:09]} f4 {(f4) [%eval 0,34] [%emt 00:01:16]}
47.axb6 {(axb6) [%eval 0,49] [%emt 00:01:05]} Kb7 {(cxb6) [%eval 0,34] [%emt 00:01:27]}
48.Rxf4 {(Rxf4) [%eval 0,49] [%emt 00:01:15]} Qxe2 {(Qxe2) [%eval 0,42] [%emt 00:01:25]}
49.Rxe2 {(Rxe2) [%eval 0,63] [%emt 00:01:17]} Bxf4 {(Bxf4) [%eval 0,43] [%emt 00:00:28]}
50.gxf4 {(gxf4) [%eval 0,64] [%emt 00:00:59]} Rxf4 {(Rxf4) [%eval 0,49] [%emt 00:02:13]}
51.bxc7 {(Re6) [%eval 0,78] [%emt 00:01:05]} Kxc7 {(Kxc7) [%eval 0,59] [%emt 00:00:30]}
52.a4 {(Kxh3) [%eval 0,121] [%emt 00:00:42]} Rxa4 {(Rxa4) [%eval 0,83] [%emt 00:03:22]}
53.Re7 {[%eval 0,64] [%emt 00:00:43]} Kd6 {(Kd6) [%eval 0,37] [%emt 00:00:21]}
54.Rh7 {(Rh7) [%eval 0,70] [%emt 00:00:41]} Ra1 {(Kxd5) [%eval 0,41] [%emt 00:00:48]}
55.Rh8 {(Rh6) [%eval 0,64] [%emt 00:00:41]} Ra2 {(Kxd5) [%eval 0,33] [%emt 00:00:18]}
56.Kxh3 {(Kxh3) [%eval 0,71] [%emt 00:00:41]}
1/2-1/2[/pgn]

Interesting, I observe your games get more and more interesting, but for the time being that tells us nothing how good Bc6 actually is, apart from confirming that top engines like Bc6 even at long time control and with many threads, that is still a very useful info.

I wonder, without book, how many engines will choose Ba4 instead?