OpenTal - almost new engine

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.
chetday
Posts: 102
Joined: Sun Nov 18, 2012 8:18 pm

Re: OpenTal - almost new engine

Post by chetday » Tue Dec 19, 2017 6:57 pm

Pawel and Brendan, thank you both for this terrific new engine, which just beat the pants off me. What fun!

I'm wondering, however, if I have a problem because when OpenTal loads in Arena, I see...

...
16MB of memory allocated
17179 moves loaded from the internal book
reading book file 'rodent.bin' (failure)

Is there something I can/should do to fix the book file failure?[/img]

carldaman
Posts: 1719
Joined: Sat Jun 02, 2012 12:13 am

Re: OpenTal - almost new engine

Post by carldaman » Wed Dec 20, 2017 1:12 am

chetday wrote:Pawel and Brendan, thank you both for this terrific new engine, which just beat the pants off me. What fun!

I'm wondering, however, if I have a problem because when OpenTal loads in Arena, I see...

...
16MB of memory allocated
17179 moves loaded from the internal book
reading book file 'rodent.bin' (failure)

Is there something I can/should do to fix the book file failure?[/img]
I'm not sure why it couldn't read the rodent.bin book that's present in the same folder, either. I tried redirecting it to the ph-tal2.bin book [after right-click and 'configure engine'], and that did not give an error. Not fully sure if this is the intended optimal set-up, but it is at least using the other book.

User avatar
Ovyron
Posts: 2710
Joined: Tue Jul 03, 2007 2:30 am

Re: OpenTal - almost new engine

Post by Ovyron » Wed Dec 20, 2017 2:43 am

Can someone play a match OpenTal against the legendary Thinker 5.3b Inert? This thinker was an engine that played like Tal "accidentally". Its programmer actually made several versions, including a Passive version that was supposed to be more dynamic and an Active version specially designed to swindle agaisnt humans, but somehow this Inert version was both the strongest and most insane version of the engine.

A fluke, because Thinker dropped its style from version 5.4A onwards.

Nobody cares about actual results but, I don't recall ever seeing any chess entity as good as T5.3bI at this so if OpenTal is actually better now I wonder how such games would look like.

BrendanJNorman
Posts: 1448
Joined: Sun Feb 07, 2016 11:43 pm
Full name: Brendan J Norman

Re: OpenTal - almost new engine

Post by BrendanJNorman » Wed Dec 20, 2017 4:32 am

Ovyron wrote:Can someone play a match OpenTal against the legendary Thinker 5.3b Inert? This thinker was an engine that played like Tal "accidentally". Its programmer actually made several versions, including a Passive version that was supposed to be more dynamic and an Active version specially designed to swindle agaisnt humans, but somehow this Inert version was both the strongest and most insane version of the engine.

A fluke, because Thinker dropped its style from version 5.4A onwards.

Nobody cares about actual results but, I don't recall ever seeing any chess entity as good as T5.3bI at this so if OpenTal is actually better now I wonder how such games would look like.
OpenTal is better in style, but not in strength.

Thinker 5.3b Inert is around 2800 strength, while OpenTal is around 2500-2600, which correlates with the real Tal's FIDE (CCRL 2600 is probably equal to 2700 FIDE).

I'm a massive Thinker fan and will still run the match, but just for fun. ;)

BrendanJNorman
Posts: 1448
Joined: Sun Feb 07, 2016 11:43 pm
Full name: Brendan J Norman

Re: OpenTal - almost new engine

Post by BrendanJNorman » Wed Dec 20, 2017 4:50 am

Ovyron wrote:Can someone play a match OpenTal against the legendary Thinker 5.3b Inert? This thinker was an engine that played like Tal "accidentally". Its programmer actually made several versions, including a Passive version that was supposed to be more dynamic and an Active version specially designed to swindle agaisnt humans, but somehow this Inert version was both the strongest and most insane version of the engine.

A fluke, because Thinker dropped its style from version 5.4A onwards.

Nobody cares about actual results but, I don't recall ever seeing any chess entity as good as T5.3bI at this so if OpenTal is actually better now I wonder how such games would look like.
Game 1 is in progress! :)

Image

BrendanJNorman
Posts: 1448
Joined: Sun Feb 07, 2016 11:43 pm
Full name: Brendan J Norman

Re: OpenTal - almost new engine

Post by BrendanJNorman » Wed Dec 20, 2017 5:54 am

chetday wrote:Pawel and Brendan, thank you both for this terrific new engine, which just beat the pants off me. What fun!
Please do post the game! :lol:

User avatar
Tibono
Posts: 43
Joined: Sat Aug 01, 2015 4:16 pm
Location: France
Contact:

Re: OpenTal - almost new engine

Post by Tibono » Wed Dec 20, 2017 7:07 am

carldaman wrote:
chetday wrote:Pawel and Brendan, thank you
...

reading book file 'rodent.bin' (failure)

Is there something I can/should do to fix the book file failure?
I'm not sure why it couldn't read the rodent.bin book that's present in the same folder, either. I tried redirecting it to the ph-tal2.bin book [after right-click and 'configure engine'], and that did not give an error. Not fully sure if this is the intended optimal set-up, but it is at least using the other book.
You can select the rodent.bin book using engine configuration as well; full path to book will be set up, and it is accepted (as a complementary book to engine's internal one, if my understanding is correct).
Maybe ph-tal2.bin is the load source for the internal book? Thus selecting it in the setup may be no use (duplicate)? Intended setup is indeed a good question.
+1 regarding thanks!
BR, Eric

User avatar
Ovyron
Posts: 2710
Joined: Tue Jul 03, 2007 2:30 am

Re: OpenTal - almost new engine

Post by Ovyron » Wed Dec 20, 2017 8:59 am

Thanks Brendan, I'm looking forward to the games that ChessTal does win, this might be the battle of the top 2 engines style-wise, what if it wins the match in spectacular fashion?? Games worth watching! :)

BrendanJNorman
Posts: 1448
Joined: Sun Feb 07, 2016 11:43 pm
Full name: Brendan J Norman

Re: OpenTal - almost new engine

Post by BrendanJNorman » Wed Dec 20, 2017 9:07 am

Ovyron wrote:Thanks Brendan, I'm looking forward to the games that ChessTal does win, this might be the battle of the top 2 engines style-wise, what if it wins the match in spectacular fashion?? Games worth watching! :)
Match ended 6-0 for Thinker 5.3b Inert. :cry:

Kind of expected since it is 2840 CCRL, so I have started a gauntlet of mini-matches against weaker Thinkers (4.7a, 5.1c, 5.1e, 5.4a Passive), as well as those designed to play against humans (a few "Active" versions).

Here's the first game (just played).

[pgn][Event "OpenTal vs Classic Thinkers"]
[Site "BRENDANNORMD8A2"]
[Date "2017.12.20"]
[Round "1"]
[White "OpenTal 1.0"]
[Black "Thinker 4.7a"]
[Result "1-0"]
[ECO "E06"]
[WhiteElo "2200"]
[BlackElo "2544"]
[PlyCount "74"]
[EventDate "2017.??.??"]
[TimeControl "180+2"]

1. d4 d5 2. c4 e6 3. Nf3 Nf6 4. g3 Be7 5. Bg2 O-O 6. O-O c6 7. Qc2 b6 8. Rd1
Na6 {(8. ... Na6 9.a3 Bb7 10.Ne5 Rc8 11.Nc3 Nd7) -0.15/10 5} 9. a3 {(9.a3 Bb7 1
0.Ne5 Rc8 11.Nc3 c5 12.Bg5 Ne4 13.Bxe4 dxe4 14.dxc5 Bxg5 15.Rxd8 Rfxd8 16.c6
Bxc6 17.Nxc6 Rxc6 18.Qxe4) +1.67/16 6} Bb7 {
(9. ... Bb7 10.Ne5 Qc7 11.Bf4 Bd6 12.e3 Nh5 13.Qe2 Nxf4 14.exf4) -0.15/10 5}
10. Nc3 {(10.Nc3 c5 11.Bf4 Rc8 12.Ne5 cxd4 13.Rxd4 Nh5 14.Be3 f6 15.Ng6 Bc5 16.
Rh4 Bxe3 17.Rxh5 Qe8 18.Nxf8 Qxh5 19.Nxe6) +1.65/16 6} Nc7 {
(10. ... Nc7 11.Bf4 Nh5 12.cxd5 Nxd5 13.Nxd5) -0.12/9 5} 11. e4 {(11.e4 dxc4
12.Bf4 Rc8 13.Ne2 Qe8 14.Ne5 Nb5 15.a4 Nd6 16.a5 c5 17.axb6 axb6 18.dxc5 Rxc5)
+1.85/14 6} dxc4 {(11. ... dxc4 12.Ne5 b5 13.Be3 Nd7 14.Nxd7) +0.21/10 4} 12.
a4 {(12.a4 Na6 13.Qe2 Nb4 14.Ne5 Ba6 15.a5 Qc8 16.Be3 Rd8 17.axb6 axb6 18.d5
cxd5 19.Bxb6 Nd3 20.exd5) +1.68/16 6} b5 {
(12. ... b5 13.Ne5 Nd7 14.Be3 a5 15.Nxd7) +0.21/10 4} 13. Ne5 {(13.Ne5 b4 14.
Nb1 Na6 15.Bf4 c5 16.dxc5 b3 17.Qe2 Nxc5 18.Rxd8 Raxd8 19.Nc3 Nd3 20.Rc1 g5 21.
Bxg5 Nxe5) +1.65/18 6} Nd7 {
(13. ... Nd7 14.Be3 a6 15.Nxd7 Qxd7 16.f4) +0.21/10 5} 14. Bf4 {(14.Bf4 g5 15.
Be3 f6 16.Nxd7 Qxd7 17.axb5 cxb5 18.d5 b4 19.dxe6 Qxe6 20.Nd5 Nxd5 21.exd5 Qg4
22.d6 b3 23.Qc3 Bxg2 24.dxe7 Bf3 25.exf8Q+ Rxf8 26.Rxa7 Bxd1) +2.00/18 6} Nxe5
{(14. ... Nxe5 15.Bxe5 a6 16.axb5 axb5 17.Rxa8 Bxa8 18.Ra1 Bd6 19.Ra7) +0.30/
10 5} 15. dxe5 {(15.dxe5 Qe8 16.Rd6 Na6 17.axb5 Nb4 18.Qa4 Bxd6 19.exd6 Nd3 20.
Qxc4 Nxf4 21.gxf4 Qd7 22.e5 cxb5 23.Qc7 Qxc7 24.dxc7 Rac8 25.Nxb5 Bxg2 26.Kxg2)
+1.81/18 6} Qc8 {
(15. ... Qc8 16.Be3 Rd8 17.Rxd8+ Qxd8 18.Rd1 Qe8 19.h3 h6) +0.01/11 5} 16. Rd6
{(16.Rd6 b4 17.Ne2 Bxd6 18.exd6 e5 19.Be3 Ne6 20.Rd1 Rd8 21.Qd2 c3 22.bxc3 Ba6
23.cxb4 Rxd6 24.Qxd6 Bxe2) +2.09/17 6} Bxd6 {
(16. ... Bxd6 17.exd6 Ne8 18.axb5 e5 19.Bxe5 Qe6 20.Qa4 Qxe5 21.bxc6) +0.67/9 5
} 17. exd6 {(17.exd6 e5 18.Bxe5 Re8 19.f4 b4 20.Nd1 Rxe5 21.fxe5 Ne6 22.Ne3
Qd7 23.Qxc4 c5 24.Rf1 Re8 25.Bh3 Bc6 26.b3 a5 27.Bxe6 Rxe6) +2.21/20 5} Ne8 {
(17. ... Ne8 18.Rd1 Qd7 19.e5 f5 20.axb5 cxb5 21.Bxb7 Qxb7 22.d7) +0.47/12 5}
18. Rd1 {(18.Rd1) +3.03/18 5} Qd7 {
(18. ... Qd7 19.e5 f5 20.axb5 cxb5 21.Bxb7 Qxb7 22.d7 Nc7 23.Bg5 b4) +0.48/12 4
} 19. e5 {(19.e5) +3.03/18 5} f5 {(19. ... f5 20.axb5 cxb5 21.Bxb7 Qxb7 22.d7
Rd8 23.dxe8Q Rfxe8 24.Rd6) +0.15/12 5} 20. Be3 {(20.Be3 c5 21.Bxb7 Qxb7 22.
axb5 Rb8 23.Qe2 Qd7 24.Qxc4 Rf7 25.f3 h6 26.h4 Rb6 27.Bxc5 f4 28.g4 Rb8 29.h5)
+3.07/19 5} Rf7 {(20. ... Rf7 21.Qd2 a6 22.Bc5 Rc8) +0.72/11 4} 21. Bc5 {
(21.Bc5 Rd8 22.f4 b4 23.Ne2 Nc7 24.Bxb4 c5 25.Bxb7 cxb4 26.Qxc4 Nd5 27.Bxd5
Rc8 28.Qa6 exd5 29.Nd4) +3.15/185} Rc8 {(21. ... Rc8 22.a5 a6) +0.69/10 4} 22.
f4 {(22.f4 Rd8 23.Qd2 b4 24.Ne2 a5 25.Nd4 Nc7 26.dxc7 Qxc7 27.Qc2 Rfd7 28.Qxc4
Rd5 29.Rc1 Qd7 30.Bd6 Ba6 31.Qxa6 Rxd4 32.Bxc6) +3.59/18 5} a6 {
(22. ... a6 23.Bf3 Rd8 24.a5 Rc8 25.h3) +0.87/11 4} 23. Ne2 {(23.Ne2 a5 24.Nd4
g6 25.Qc3 Ng7 26.Rc1 Nh5 27.Re1 h6 28.Rf1 g5 29.fxg5 hxg5 30.Qxa5 f4 31.Qa7
fxg3) +3.89/18 5} g6 {(23. ... g6 24.Nd4 Ng7 25.Nf3 Rff8 26.a5) +0.75/11 4} 24.
a5 {(24.a5 h6 25.Qd2 Ng7 26.Nd4 Rd8 27.Qe3 Re8 28.Qc1 Rc8 29.Nf3 Re8 30.Qd2
Rc8 31.Nh4 Kh7 32.Qe3 Re8) +4.13/19 5} Ng7 {
(24. ... Ng7 25.Nd4 h6 26.Qe2 Kh7 27.Nf3 Nh5 28.Nh4 Rd8) +0.42/13 4} 25. Nd4 {
(25.Nd4 h6 26.Qd2 Nh5 27.Qe2 Re8 28.Qe3 Rg7 29.Qf3 Rf7 30.Rc1 Ref8 31.Qe3 Rg7
32.b3 g5 33.fxg5 hxg5 34.bxc4) +4.13/185} Nh5 {
(25. ... Nh5 26.Bf3 Kh8 27.Qg2 Ng7 28.Qh3 Kg8 29.Qh4 Rff8) +0.49/12 4} 26. Qe2
{(26.Qe2 Ng7 27.Qd2 Re8 28.Qe3 Nh5 29.Qe2 Ng7 30.Qd2 h6 31.Rc1 Rc8 32.b3 cxb3
33.Nxb3 h5 34.h4 Rcf8 35.Nd4) +4.21/17 5} h6 {
(26. ... h6 27.b3 cxb3 28.Nxb3 Qe8 29.Bb6 Ra8 30.Nc5) +0.39/13 8} 27. b3 {
(27.b3 cxb3 28.Nxb3 Rg7 29.Nd4 Re8 30.Qf3 Rf7 31.Rc1 Rg7 32.Rf1 Qd8 33.Qb3 Qd7
34.Rd1 Rf7 35.Rc1 Rg7 36.Qf3 Rf7) +4.20/205} c3 {
(27. ... c3 28.Qc2 Ng7 29.Qxc3 Qe8 30.Bb6 c5 31.Bxb7 Rxb7) -0.11/12 8} 28. Rc1
{(28.Rc1 Ng7 29.Rxc3 g5 30.h3 Re8 31.b4 Kh7 32.Rd3 Rg8 33.fxg5 hxg5 34.Rd1 f4
35.Be4+ Kh6 36.g4 f3 37.Nxf3 Rf4) +4.54/19 4} Rff8 {
(28. ... Rff8 29.Rxc3 Ng7 30.Qc2 Rfe8 31.Qc1 Nh5 32.Bb6 c5) -0.30/12 4} 29.
Rxc3 {(29.Rxc3 Ng7 30.Qd2 Rf7 31.Rc1 Rd8 32.b4 Rc8 33.Nb3 Kh7 34.Rc3 Kg8 35.
Nd4 Nh5 36.Qd1 Ng7 37.Rc1) +4.88/18 4} Ng7 {
(29. ... Ng7 30.Qc2 Kh8 31.b4 Rb8 32.Ba7) -0.75/12 4} 30. Qd2 {(30.Qd2 Nh5 31.
Rc1 Rf7 32.Qe2 Rg7 33.Qf3 Rf7 34.Qd1 Re8 35.b4 Rg7 36.Qf3 Kh7 37.Rf1 Reg8 38.
Rd1 Re8 39.Qd3) +4.85/20 4} Rf7 {
(30. ... Rf7 31.Qc2 Kh8 32.b4 Rb8 33.Ba7 Re8) -0.60/11 4} 31. b4 {(31.b4 Kh7
.Rc2 Nh5 33.Rc1 Rg7 34.Qe2 Rcg8 35.Rf1 Rf7 36.Nb3 Rc8 37.Rc1 Rg7 38.Qf3 Rf8 39.
Qd3 Rgf7 40.Bb6) +4.87/19 4} Nh5 {
(31. ... Nh5 32.Bb6 Qe8 33.Qc2 Rd7 34.Nxc6 Kh8 35.h4 Qf7 36.Bd4) -1.08/13 4}
32. Bb6 {(32.Bb6 Ng7 33.Rc1 Ne8 34.Bc5 Ng7 35.Qe3 Nh5 36.Qe2 Re8 37.Qd2 Rg7 38.
Bb6 Rc8 39.Rc5 g5 40.fxg5 hxg5 41.Qe3) +4.86/19 4} Rcf8 {
(32. ... Rcf8 33.Nxc6 Rh7 34.Bc7 Bc8 35.Ne7+ Rxe7 36.Bc6 Qxc7) -1.78/11 3} 33.
Nb3 {(33.Nb3 Qe8 34.Nc5 Bc8 35.Bf3 Rg7 36.Qd1 Rgf7 37.Rc1 Rg7 38.Qe2 Nxf4 39.
gxf4 g5 40.Bh5 gxf4+ 41.Kh1 Rff7 42.Rf1 Rg5 43.Rxf4) +5.72/20 4} Qe8 {
(33. ... Qe8 34.Nc5 Bc8 35.Rd3 Bd7 36.Nxa6 Rg7 37.Nc7 Qb8 38.Qf2 Qc8) -2.43/143
} 34. Nc5 {(34.Nc5 Bc8 35.Rd3 g5 36.Qd1 Nxf4 37.gxf4 gxf4 38.d7 Bxd7 39.Nxd7
Rxd7 40.Rxd7 Qg6 41.Rc7 f3 42.Qxf3 f4 43.Bc5 Rf7 44.Rxf7 Qb1+ 45.Kf2 Qb2+ 46.
Kf1 Qa1+ 47.Ke2 Qxe5+ 48.Kd3 Kxf7 49.Qxc6) +5.77/20 4} Bc8 {(34. ... Bc8 35.
Rd3 Bd7 36.Nxa6 Kh8 37.Qc2 g5 38.Nc7 Qc8 39.fxg5 hxg5) -2.61/14 3} 35. Rd3 {
(35.Rd3 g5 36.fxg5 f4 37.gxh6 f3 38.Bf1 f2+ 39.Kg2 Rf5 40.d7 Bxd7 41.Rxd7 Nf4+
42.Kh1 Qh5 43.Qxf2 Qxh6 44.gxf4 Rxf4) +6.05/174} Bd7 {
(35. ... Bd7 36.Nxa6 Ng7 37.Qc2 Qa8 38.Nc7 Qc8 39.a6 Ne8 40.a7) -3.04/14 3} 36.
Nxa6 {(36.Nxa6 Rg7 37.Nc7 Qf7 38.a6 g5 39.fxg5 Rxg5 40.a7 f4 41.Rf3 Qg6 42.Qa2
Nxg3 43.hxg3 Rxg3 44.Rxg3 Qxg3 45.Bf2 Qg4 46.a8Q Rxa8 47.Qxa8+ Kh7) +5.82/18 4}
Rh7 {(36. ... Rh7 37.Nc7 Qb8 38.Qc2 Rhf7 39.Bxc6 Ng7) -3.12/12 4} 37. Nc7 {
(37.Nc7 Qf7 38.a6 Rg7 39.a7 g5 40.fxg5 Rxg5 41.Ra3 f4 42.a8Q fxg3 43.hxg3 Nxg3
44.Qd3 Nf5 45.Qh3 Nh4 46.Qxf8+ Qxf8 47.Qxh4 Kh7) +6.76/184} Qb8 {(37. ... Qb8
38.Qc2 Rhf7 39.Qc5 g5 40.Bxc6 gxf4 41.a6 Bxc6 42.Qxc6) -3.64/12 3 Arena
Adjudication} 1-0[/pgn]

I think Pawel is right.

OpenTal can handle (and play beautifully) against engines (and humans even stronger) rated 2500-2600, but once they approach the 2750-2800 level, that's simply too strong.

A better opponent for Inert 5.3b would be either my Alekhine or Karpov or Kasparov Rodent personalities. ;)

PK
Posts: 824
Joined: Mon Jan 15, 2007 10:23 am
Location: Warsza
Contact:

Re: OpenTal - almost new engine

Post by PK » Wed Dec 20, 2017 10:15 am

Hi all,

Regarding problems with Fritz GUI - can somebody either confirm or deny whether Frittz GUI can accept an engine name that contains slashes?

As for playing against stronger engines, OpenTal is indeed of no use. I tested it against dev version of Rodent, and results were rather dismal. A sample game from "Luke, I'm your father" match comes below:

[pgn]
[Event "Welcome_Tal"]
[Site "DESKTOP-S1TVKAR"]
[Date "2017.12.17"]
[Round "1"]
[White "Tal-in-a-box"]
[Black "Rodent_tuned"]
[Result "0-1"]
[BlackElo "2200"]
[ECO "C03"]
[Opening "French"]
[Time "12:23:36"]
[Variation "Tarrasch, 3...Be7 4.Ngf3 Nf6 5.Bd3"]
[WhiteElo "2200"]
[TimeControl "180+2"]
[Termination "adjudication"]
[PlyCount "90"]
[WhiteType "human"]
[BlackType "human"]

1. e4 c6 2. d4 d5 3. Nd2 dxe4 4. Nxe4 Nd7 5. Nf3 Ngf6 6. Nc3 Nb6 7. a4 a5
8. h3 Bf5 9. Bd3 Bxd3 10. Qxd3 e6 11. O-O Be7 12. Re1 Nbd5 13. Bg5 {(Bc1g5
OO Ra1d1 Nd5b4 Qd3e2 Nb4d5 Bg5xf6 Nd5xf6 Nc3e4 Qd8b6 Nf3e5 Nf6xe4 Qe2xe4
Ra8d8 Rd1d3 Qb6xb2 Rd3b3 Qb2xd4 Rb3xb7) +1.15/17 6} O-O {(OO Nf3e5 Nd5b4
Qd3d1 Qd8c7 Qd1e2 h7h6 Bg5h4 Ra8d8 Ra1d1 Rf8e8 Qe2d2 Re8f8 f2f4 Nb4d5
Nc3xd5 Nf6xd5 Bh4xe7 Qc7xe7 Qd2xa5 Nd5xf4) -0.20/18 6} 14. Rad1 {(Ra1d1
h7h6 Bg5xf6 Nd5xf6 Nf3e5 Qd8c7 Nc3e4 Ra8d8 c2c3 Nf6xe4 Qd3xe4 Be7d6 Ne5f3
Bd6f4 c3c4 Rd8d7 Nf3e5 Bf4xe5 d4xe5 Rf8d8) +1.01/17 6} h6 {(h7h6 Bg5xf6
Nd5xf6 Nc3e4 Qd8b6 b2b3 Rf8d8 c2c4 Nf6xe4 Re1xe4 Qb6b4 h3h4 Qb4a3 h4h5
Qa3a2 Re4e2 Qa2a3 Qd3c2 Be7f6 Nf3e5 Bf6xe5 Re2xe5) -0.17/19 6} 15. Bxf6
{(Bg5xf6 Nd5xf6 Nc3e4 Qd8b6 c2c4 Ra8d8 Qd3c2 Qb6c7 Ne4xf6+ Be7xf6 Re1e4
h6h5 Qc2b3 Qc7e7 Qb3d3 b7b6 Re4f4 c6c5 d4d5 e6xd5 c4xd5) +1.10/18 6} Nxf6
{(Nd5xf6 Nc3e4 Qd8b6 b2b3 Rf8d8 c2c4 Qb6b4 Re1e3 Ra8c8 Qd3c2 c6c5 Ne4xf6+
Be7xf6 d4d5 e6xd5 Rd1xd5 Rd8xd5 c4xd5 Rc8d8 Qc2c4 h6h5 Kg1h2 h5h4 Nf3e5)
-0.18/20 6} 16. Ne4 {(Nc3e4 Qd8c7 c2c4 Ra8d8 Ne4xf6+ Be7xf6 Re1e4 h6h5
Qd3e3 Qc7e7 Rd1b1 Qe7b4 b2b3 Qb4a3 Rb1d1 Qa3a2 Qe3d3 b7b6) +1.01/17 6} Qb6
{(Qd8b6 c2c4 Rf8d8 Ne4c3 Be7b4 Re1e5 c6c5 d4d5 Bb4xc3 b2xc3 Qb6c7 Qd3e2
Ra8a6 Rd1d2 Nf6d7 Re5e3 Nd7b6 d5xe6 Rd8xd2 Nf3xd2 Nb6xa4 e6xf7+ Qc7xf7)
-0.16/19 6} 17. c4 {(c2c4 Ra8d8 Re1e3 Qb6c7 Qd3e2 Nf6xe4 Re3xe4 Be7f6 Qe2c2
Rd8d7 g2g3 Rf8d8 Re4f4 Bf6g5 Rf4g4 f7f5 Rg4xg5 h6xg5 Nf3xg5) +0.97/17 5}
Rfd8 {(Rf8d8 b2b3 Nf6xe4 Re1xe4 Be7f6 Qd3e3 Rd8d7 Re4f4 c6c5 d4d5 e6xd5
Rd1xd5 Rd7xd5 c4xd5 Qb6d6 Rf4f5 Qd6c7 Qe3e4 Qc7b6 Qe4d3 Ra8d8 Nf3e5 Qb6d6)
-0.16/20 6} 18. c5 {(c4c5 Qb6a6 Qd3c3 Nf6xe4 Re1xe4 Rd8d5 Nf3e5 Be7f6 Ne5g4
Bf6g5 f2f4 Bg5h4 Qc3f3 Ra8d8 Ng4e5 Rd5xe5 f4xe5) +1.03/16 5} Qb4 {(Qb6b4
Qd3c2 Nf6xe4 Re1xe4 Be7xc5 Rd1d2 Bc5b6 d4d5 Qb4e7 d5xe6 f7xe6 Rd2e2 Qe7d6
Qc2c4 Qd6d5 Qc4c3 Qd5d3 Qc3e5 Qd3d6 Qe5c3 Qd6d3) 0.00/24 6} 19. Qc2 {(Qd3c2
Nf6xe4 Re1xe4 Rd8d5 Rd1d3 Ra8d8 Rd3c3 Rd5xd4 Re4xd4 Rd8xd4 Nf3xd4 Qb4xd4
Rc3d3 Qd4b4 Rd3d7 Be7f6 Kg1h2 g7g6 Qc2c1 Bf6g5 Qc1d1 Qb4xb2) +1.00/18 6}
Nxe4 {(Nf6xe4 Re1xe4 Be7xc5 Rd1d2 Bc5b6 d4d5 Qb4e7 d5xe6 Rd8xd2 e6xf7+
Qe7xf7 Qc2xd2 Ra8d8 Qd2e2 Qf7b3 Re4e7 Bb6c5 Re7e8+ Rd8xe8 Qe2xe8+ Kg8h7
Qe8e4+ Kh7g8 Qe4e8+) 0.00/24 7} 20. Rxe4 {(Re1xe4) +1.26/20 6} Bxc5
{(Be7xc5) +0.08/23 6} 21. d5 {(d4d5) +1.49/17 5} Qb6 {(Qb4b6 Re4c4 Bc5e7
d5xe6 Rd8xd1+ Qc2xd1 f7xe6 Qd1d7 Be7f6 Nf3e5 Qb6xb2 Qd7f7+ Kg8h8 Ne5g6+
Kh8h7 Ng6f8+ Kh7h8 Nf8g6+) 0.00/22 5} 22. Rc4 {(Re4c4 Bc5xf2+ Qc2xf2
Qb6xf2+ Kg1xf2 c6xd5 Rc4c7 Ra8c8 Rd1c1 Rc8xc7 Rc1xc7 d5d4 Kf2e2 d4d3+ Ke2d2
f7f6 Rc7c3 e6e5 Rc3xd3 Rd8e8 Nf3h4 e5e4 Rd3d7 e4e3+ Kd2e2) +1.36/18 5} Be7
{(Bc5e7 d5xe6 Rd8xd1+ Qc2xd1 f7xe6 Qd1d7 Be7f6 Nf3e5 Qb6xb2 Qd7f7+ Kg8h7
Qf7g6+ Kh7g8 Qg6f7+) 0.00/20 5} 23. dxe6 {(d5xe6) +0.89/18 5} Rxd1+
{(Rd8xd1+ Qc2xd1 f7xe6 Qd1d7 Be7f6 Nf3e5 Qb6xb2 Qd7f7+ Kg8h7 Qf7g6+ Kh7g8
Qg6f7+) 0.00/22 5} 24. Qxd1 {(Qc2xd1) +0.91/20 5} fxe6 {(f7xe6 Qd1d7 Be7f6
Nf3e5 Qb6xb2 Qd7f7+ Kg8h7 Qf7g6+ Kh7g8 Qg6f7+) 0.00/24 5} 25. Qe1 {(Qd1e1
Be7f6 Qe1xe6+ Kg8h8 Rc4g4 Qb6xb2 Qe6f7 Ra8d8 Kg1h2 Bf6e5+ Nf3xe5 Qb2xe5+
f2f4 Qe5f6 Qf7xb7 h6h5 Rg4g3 Rd8f8 f4f5 c6c5 Qb7d5 Qf6d4) +0.72/18 5} c5
{(c6c5) +0.41/19 5} 26. Rg4 {(Rc4g4 Ra8d8 Qe1b1 Be7f6 b2b3 Kg8h8 h3h4 c5c4
Rg4xc4 Qb6d6 Qb1g6 Qd6d3 h4h5 Qd3xg6 h5xg6 Kh8g8 Rc4c7 Rd8d1+ Kg1h2)
+0.99/17 5} Rd8 {(Ra8d8 Qe1e3 h6h5 Rg4e4 Rd8d1+ Kg1h2 Qb6d6+ g2g3 Qd6d3
Qe3xd3 Rd1xd3 Kh2g2 Be7f6 Re4xe6 Rd3b3 Re6e8+ Kg8f7 Re8c8 Rb3xb2 Rc8xc5)
+0.56/17 5} 27. Qc1 {(Qe1c1 c5c4 h3h4 h6h5 Rg4xc4 Be7f6 Rc4c8 Rd8xc8
Qc1xc8+ Kg8f7 Qc8c2 Kf7e7 b2b3 Qb6d6 Qc2c4 Ke7f7 Qc4c1 Qd6d8 Qc1e1)
+0.78/18 5} h5 {(h6h5 Rg4e4 Qb6b3 Re4e1 Qb3xa4 Qc1e3 Qa4d7 Qe3e5 h5h4 Qe5h5
a5a4 Nf3e5 Qd7e8 Ne5g6 Be7f6 Qh5h8+ Kg8f7 Qh8h5 Bf6xb2 Ng6xh4+ Kf7g8
Qh5xc5) +0.96/20 5} 28. Rg3 {(Rg4g3 Be7f6 Qc1h6 Qb6xb2 Kg1h2 h5h4 Rg3g4
Rd8d1 Qh6e3 Qb2a1 Qe3xe6+ Kg8h8 Qe6e8+ Kh8h7 Qe8g6+ Kh7h8 Nf3g1 Bf6e5+ f2f4
Be5xf4+ Rg4xf4 Qa1e5) +0.65/17 6} Bf6 {(Be7f6 Kg1h2 h5h4 Rg3g4 Qb6d6+ Qc1f4
Qd6xf4+ Rg4xf4 Rd8d3 Rf4c4 b7b6 Nf3xh4 Rd3d2 Kh2g3 Bf6d4 Nh4f3 Bd4xf2+
Kg3h2 Rd2xb2 Rc4e4 Kg8f7 Nf3g5+ Kf7f6 Ng5xe6) +1.61/19 5} 29. Qe1 {(Qc1e1
c5c4 Qe1e4 c4c3 b2xc3 Rd8d1+ Kg1h2 Kg8f8 Nf3g5 Qb6b1 Rg3f3 Qb1xe4 Ng5xe4
h5h4 Ne4xf6 g7xf6 Rf3xf6+ Kf8e7 Rf6h6 Rd1d2 Rh6xh4 Rd2xf2 Rh4h7+ Ke7d6
Rh7xb7) +0.12/19 4} Rd5 {(Rd8d5) +1.33/16 4} 30. Qe4 {(Qe1e4) +0.08/19 4}
Kf8 {(Kg8f8 Kg1h2 Kf8e7 b2b3 Ke7d7 Nf3e1 h5h4 Rg3d3 Qb6d6+ Kh2g1 Qd6e5
Qe4xe5 Bf6xe5 Ne1f3 Be5d4 Kg1f1 Kd7c6 Nf3xh4 Bd4f6 Rd3xd5 e6xd5) +1.12/19
5} 31. Kh2 {(Kg1h2 Kf8e7 Nf3h4 Qb6b4 Nh4g6+ Ke7d6 Qe4f3 Qb4xb2 Qf3f4+ e6e5
Qf4f5 e5e4 Qf5f4+ Bf6e5 Ng6xe5 Qb2xe5 Rg3g6+ Kd6d7 Qf4xe5 Rd5xe5 Rg6xg7+
Kd7c6) +0.20/16 4} Ke7 {(Kf8e7 b2b3 Qb6b4 Qe4e3 g7g5 Rg3xg5 Bf6xg5 Nf3xg5
Qb4b6 f2f4 Qb6c6 Ng5f3 Rd5d1 Qe3c3 Qc6d6 Qc3g7+ Ke7d8 Qg7f6+ Qd6e7 Qf6h8+
Qe7e8 Qh8e5 Qe8g6 Qe5xc5) +1.33/19 4} 32. Nh4 {(Nf3h4 Qb6b4 Nh4g6+ Ke7d6
Qe4f3 Rd5d2 b2b3 Kd6c7 Ng6f4 Bf6e5 Nf4xe6+ Kc7b6 Qf3e3 Be5xg3+ Qe3xg3 h5h4
Qg3f3 Kb6a7 Qf3e3 b7b6 Kh2g1 Rd2d1+ Kg1h2) -0.13/18 5} Qxb2 {(Qb6xb2 Nh4g6+
Ke7d6 Qe4f4+ e6e5 Qf4f5 Kd6c7 Rg3f3 Rd5d2 Ng6f8 c5c4 Qf5xh5 e5e4 Qh5f7+
Kc7b8 Rf3f5 Qb2d4 Nf8e6 Bf6e5+ g2g3 Qd4a1) +1.56/17 4} 33. Ng6+ {(Nh4g6+
Ke7d7 Ng6f8+ Kd7e7 Nf8g6+) 0.00/18 4} Kd6 {(Ke7d6) +1.89/19 4} 34. Qf4+
{(Qe4f4+ Bf6e5 Qf4f8+ Kd6c7 f2f4 Be5f6 Rg3e3 Rd5d2 Qf8xc5+ Kc7b8 Qc5f8+
Kb8a7 Qf8c5+ Qb2b6 Qc5xb6+ Ka7xb6 Re3xe6+ Kb6c5 Kh2g3 h5h4+ Ng6xh4 Bf6xh4+
Kg3xh4 Rd2xg2 f4f5 Kc5b4 Re6b6+ Kb4xa4 Rb6xb7 Ka4a3 Kh4h5) -0.26/19 4} e5
{(e6e5) +2.09/19 4} 35. Qf5 {(Qf4f5 h5h4 Rg3e3 Rd5d1 Ng6f8 Kd6c7 Re3d3
Rd1xd3 Qf5xd3 Qb2xf2 Qd3d7+ Kc7b8 Qd7d6+ Kb8a7 Nf8d7 Qf2f4+ Kh2h1 Qf4f1+
Kh1h2 Qf1f4+) 0.00/19 4} Kc7 {(Kd6c7) +2.12/18 4} 36. Qe6 {(Qf5e6 Rd5d6
Qe6f7+ Rd6d7 Qf7g8 Rd7d8 Qg8e6 Rd8d6) 0.00/21 4} Rd6 {(Rd5d6) +2.21/20 4}
37. Qf7+ {(Qe6f7+ Rd6d7 Qf7e6 Rd7d6) 0.00/21 4} Rd7 {(Rd6d7 Qf7e8 h5h4
Rg3f3 Qb2a1 Rf3b3 Qa1xa4 Rb3b2 Qa4d4 Ng6f8 Qd4xb2 Qe8xd7+ Kc7b6 Qd7d6+
Kb6b5 Qd6d3+ Kb5a4 Nf8d7 b7b6 Qd3a6 Qb2xf2 Qa6xb6) +2.32/20 4} 38. Qe6
{(Qf7e6) -0.31/19 4} Qxf2 {(Qb2xf2 Rg3b3 Qf2d4 Qe6e8 e5e4 Rb3b5 Qd4d6+
Kh2h1 Qd6c6 Kh1h2 b7b6 Ng6f4 Qc6d6 g2g3 Qd6d2+ Nf4g2 h5h4 g3xh4 Qd2d6+
Kh2g1 Qd6d4+ Kg1h2 Qd4xa4) +3.41/18 4} 39. Rb3 {(Rg3b3 Rd7d6 Qe6f7+ Rd6d7
Qf7f8 Rd7d8 Qf8f7+ Rd8d7) 0.00/21 4} Qd4 {(Qf2d4 Rb3b5 Qd4d6 Qe6e8 e5e4+
Kh2h1 Qd6c6 Kh1h2 b7b6 Ng6f8 Rd7d2 Nf8e6+ Kc7d6 Qe8b8+ Kd6xe6 Rb5xb6 Bf6e5+
Qb8xe5+ Ke6xe5 Rb6xc6 Ke5d5 Rc6g6 e4e3 Rg6g5+ Kd5d4 Rg5xg7) +3.59/20 4} 40.
Rb1 {(Rb3b1 Qd4e4 Rb1b5 Qe4c6 Qe6a2 Kc7b8 Ng6f8 Rd7d1 Qa2g8 Rd1d8 Qg8e6
Qc6xe6 Nf8xe6 e5e4 Ne6xd8 Bf6xd8 Rb5xc5 e4e3 Rc5xh5) -0.75/16 4} Qe4
{(Qd4e4 Rb1b5 Qe4xg6 Rb5xb7+ Kc7xb7 Qe6xd7+ Kb7b8 Qd7b5+ Kb8c7 Qb5xc5+
Kc7d7 Qc5b5+ Kd7d6 Qb5b6+ Kd6e7 Qb6c5+ Ke7f7 Qc5d5+ Kf7f8 Qd5xa5 Qg6d3
Qa5c5+ Kf8f7 Qc5b5 e5e4 Qb5xh5+ Kf7e6 Qh5g6 Ke6d6 a4a5) +3.86/20 4} 41. Rb2
{(Rb1b2 Qe4c6 Qe6f5 Kc7b8 Rb2c2 b7b6 Qf5xh5 e5e4 Rc2e2 c5c4 Qh5f5 Qc6d6+
g2g3 e4e3 Qf5g4 Bf6d4 Kh2g1) -1.65/17 4} c4 {(c5c4 Rb2b5 Qe4xg6 Rb5xb7+
Kc7xb7 Qe6xd7+ Kb7b8 Qd7b5+ Kb8c7 Qb5xc4+ Kc7d6 Qc4a6+ Kd6e7 Qa6xa5 Qg6d3
Qa5c7+ Ke7e6 Qc7c6+ Ke6f5 Qc6b5 Qd3d4 a4a5 Bf6h4 Qb5b1+ e5e4 Qb1b5+ Kf5f4
Qb5b8+ Qd4e5 Qb8xe5+ Kf4xe5) +3.99/19 4} 42. Qe8 {(Qe6e8 c4c3 Rb2b5 Qe4c6
Rb5b3 Qc6c4 Rb3xb7+ Kc7xb7 Qe8xd7+ Kb7b8 Qd7e8+ Kb8c7 Ng6f8 c3c2 Qe8d7+
Kc7b8 Qd7e8+ Kb8a7 Qe8d7+ Ka7b6 Qd7d6+ Kb6b7 Qd6d7+ Kb7a6 Qd7d6+ Ka6a7)
-2.72/15 4} Rd8 {(Rd7d8) +7.18/18 4} 43. Qf7+ {(Qe8f7+) -3.81/17 4} Kb8
{(Kc7b8 Rb2f2 c4c3 Qf7b3 Rd8d2 Qb3g8+ Kb8a7 Qg8c8 Rd2xf2 Qc8c5+ Ka7b8
Qc5xf2 Qe4xg6 Qf2c5 Qg6d3 Qc5f8+ Kb8a7 Qf8a3 e5e4 Qa3c5+ Ka7a6 Qc5xh5 c3c2
Qh5c5 Bf6e5+ Kh2h1) +9.06/20 3} 44. Nf8 {(Ng6f8 Qe4f4+ g2g3 Rd8d2+ Rb2xd2
Qf4xd2+ Kh2h1 h5h4 g3xh4 c4c3 Nf8d7+ Kb8a7 Qf7e8 Qd2d6 Qe8c8 c3c2 Qc8xc2
Qd6xd7 Kh1h2 Qd7d4 Qc2c7 e5e4 Qc7xa5+ Ka7b8) -4.21/18 4} c3 {(c4c3 Rb2f2
c3c2 Rf2xc2 Qe4xc2 Nf8d7+ Kb8a7 Qf7d5 Qc2c7 Qd5c5+ Qc7xc5 Nd7xc5 Rd8d4
Kh2g3 Ka7b6 Nc5e6 Rd4xa4 Kg3f3 Ra4a3+ Kf3e4 Ra3g3 Ne6f8 Kb6c6 Nf8e6 Rg3xg2)
+9.88/20 6} 45. Rf2 {(Rb2f2) -5.02/18 5} c2 {(c3c2) +10.54/21 4 0-1 User
Adjudication} 0-1
[/pgn]

Playin strength could be possibly better, but improving it without sacrificing style is a delicate matter. I am pretty sure I have seen too many queen sacrifices in OpenTal games, and that can be easily fixed. There were also a coupleof half-baked attempts at exchange sacrifices, but finding a good balance here is much more difficult. Also it might need some code to avoid closed positions. This is definately an unwanted part of Rodent heritage (Rodent is actually meant to like closed games).

As for opening books, I will have to correct default path. Book setup in itself is rather odd: OpenTal first looks in its internal book (generated from games Mikhail Tal won or drew, encoded in long algebraic notation and pasted into engine's code) and only then it consults .bin book. This is deliberate: this way you may feed it a modern, strong book and still start from Tal's openings.

Post Reply