Weak Opennings unearthed by Yorkman tourn?

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.
BrendanJNorman
Posts: 1801
Joined: Sun Feb 07, 2016 11:43 pm
Full name: Brendan J Norman

Re: Weak Opennings unearthed by Yorkman tourn?

Post by BrendanJNorman » Fri Jan 19, 2018 6:02 am

MikeGL wrote:
BrendanJNorman wrote: This usually comes in the form of white quickly opening things up with something like dxc5 and c4, and when black plays ...d4 white gets e4 for his knight and a king attack develops.

I have seen my engines/personalities misplay the structure like this, and indeed myself lost like this in a tournament game back in 2009.

I rediscovered this when trying to tune the Petrosian personality for Rodent...

Often engines will do this ...Ba6/...c5 stuff and beat their opponent due to being simply stronger, but against an equal opponent there might be consequences.

This kind of reminds me of 15+ years ago when engines always wanted to prematurely play ...d5 in the Boleslavsky Sicilian structures, because they thought black had no space.

So this to me, is not so much the problem with the opening, but with the fact that engines still don't play these lines well.

I just checked
[d]rnb1k1nr/p1pq1ppp/1p2p3/3pP3/3P4/P1P5/2P2PPP/R1BQKBNR w KQkq - 0 7

Engine-Engine matchs High Quality (only rated 2800 and above) from random sources CCRL/CEGT/SSDF/Talkchess Tourn Forum etc
Out of 621,000 games there were 74 French Petrosian
Ba6 line: 64
Bb7 line: 9

MillionBase by Ed Schroder (with very few engine-engine games)
out of 2.91M games only 614 French Petrosian
Ba6 line: 570
Bb7 line: 44

Bb7 (not Ba6) may be the modern way to handle this position as shown by Caruana on his latest French, but it was rapid or blitz.
I rechecked and noticed even ex world champion Tigran Petrsian himself seems to
prefer Ba6 move, so looks like engines are playing French Petrosian correctly.
The problem is that you're checking an individual position in a system.

A "system" differs from a line in that where a "line" is very move-order and position specific (hence your checking that position in the DB), a system can be played in multiple move-orders, and interpretations based on what the opponent does.

To check a single position and make a judgement shows lack of understanding of how Petrosian approached the positions.

Depending on how white played (for example the Qg4 Bf8 line) he would prefer to keep the bishop on b7, other times, he would play Ba6.

Point is, judging in which case which system/setup is better, requires Petrosian's understanding.

[pgn][Event "Belgrade"]
[Site "Belgrade"]
[Date "1954.11.02"]
[Round "11"]
[White "Joppen, Egon"]
[Black "Petrosian, Tigran V"]
[Result "0-1"]
[ECO "C16"]
[PlyCount "80"]
[EventDate "1954.10.21"]
[EventType "tourn"]
[EventRounds "19"]
[EventCountry "YUG"]
[Source "ChessBase"]
[SourceDate "1999.07.01"]

1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 b6 5. Qg4 Bf8 6. Nf3 Qd7 7. Nb5 Nc6 8. c3 a6
9. Na3 f5 10. Qg3 Bxa3 11. bxa3 Bb7 12. Ng5 O-O-O 13. h4 Nh6 14. Bd3 Kb8 15.
Qf3 Nf7 16. Nh3 g6 17. Qe2 Ka7 18. Bg5 Nxg5 19. Nxg5 h6 20. Nh3 Qe7 21. Nf4 g5
22. Nh3 Qxa3 23. Qd2 Qe7 24. O-O-O Rdg8 25. Kb1 Na5 26. hxg5 hxg5 27. f4 g4 28.
Ng5 Bc6 29. Qb2 Nc4 30. Qb4 Qd7 31. Bxc4 a5 32. Qb2 dxc4 33. Qd2 Bd5 34. Rdg1
Qc6 35. Rxh8 Rxh8 36. g3 Qe8 37. Kb2 Qh5 38. Kc2 Qh2 39. Qxh2 Rxh2+ 40. Kb1 Ka6
0-1[/pgn]

[pgn][Event "Moscow-ch34"]
[Site "Moscow (Russia)"]
[Date "1956.??.??"]
[Round "?"]
[White "Chistiakov, Alexander Nikolaevi"]
[Black "Petrosian, Tigran V"]
[Result "0-1"]
[ECO "C16"]
[PlyCount "58"]
[EventDate "1956.??.??"]
[EventType "tourn"]
[EventRounds "15"]
[EventCountry "URS"]
[Source "ChessBase"]
[SourceDate "1999.11.16"]

1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 b6 5. Qg4 Bf8 6. Nf3 Qd7 7. a3 Nc6 8. Be3 $6
Bb7 9. Bb5 $6 (9. Bd3) 9... O-O-O 10. O-O Nge7 11. b4 $6 (11. Nd2 Nf5 12. Qe2)
11... f6 $1 $132 12. Rfe1 Nf5 13. Qh3 h5 14. g3 (14. g4 $2 hxg4 15. Qxh8 Bxb4
$19 {(Moles/Wicker)}) 14... a6 15. Ba4 $2 (15. Bd3 Nxe3 (15... g5) 16. fxe3 (
16. Rxe3 g5 $1 {(Moles/Wicker)}) 16... fxe5) (15. Bxc6 Qxc6 16. Bd2 $17 {
(Moles/Wicker)}) 15... g5 $1 $17 16. g4 hxg4 $1 17. Qxh8 gxf3 $1 18. Qh5 (18.
Bxc6 Qxc6 19. Qxf6 Be7 20. Qg6 Kb8 $1 {(Moles/Wicker)}) 18... b5 19. Nxb5 axb5
20. Bxb5 fxe5 21. Kh1 Qg7 22. Bxc6 Bxc6 23. dxe5 Be7 24. b5 Bb7 $1 25. Qg4 Rh8
26. Qg1 d4 27. Bd2 g4 28. Bf4 g3 $1 29. Bxg3 Nxg3+ 0-1[/pgn]

[pgn][Event "EU Team-ch01 Final"]
[Site "Vienna"]
[Date "1957.08.28"]
[Round "6.5"]
[White "Nedeljkovic, Srecko"]
[Black "Petrosian, Tigran V"]
[Result "1/2-1/2"]
[ECO "C16"]
[PlyCount "47"]
[EventDate "1957.08.22"]
[EventType "team-tourn"]
[EventRounds "6"]
[EventCountry "AUT"]
[Source "ChessBase"]
[SourceDate "1999.07.01"]
[WhiteTeam "Yugoslavia"]
[BlackTeam "Soviet Union"]
[WhiteTeamCountry "YUG"]
[BlackTeamCountry "URS"]

1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 b6 5. a3 Bf8 6. Nf3 Ne7 7. Bd3 Nbc6 8. O-O
Bb7 9. Bd2 Qd7 10. b4 Ng6 11. Re1 O-O-O 12. Qe2 Be7 13. Na2 f5 14. a4 Nh4 15.
Nxh4 Bxh4 16. c3 Rdf8 17. Reb1 Bd8 18. a5 f4 19. f3 g5 20. Nc1 h5 21. Nb3 Kb8
22. Ba6 Ba8 23. Nc5 Qe7 24. Nd3 1/2-1/2[/pgn]

[pgn][Event "Bled 30th Anniversary"]
[Site "Bled"]
[Date "1961.09.19"]
[Round "11"]
[White "Olafsson, Fridrik"]
[Black "Petrosian, Tigran V"]
[Result "0-1"]
[ECO "C16"]
[Annotator "Bulletin"]
[PlyCount "64"]
[EventDate "1961.09.03"]
[EventType "tourn"]
[EventRounds "19"]
[EventCountry "YUG"]
[Source "ChessBase"]
[SourceDate "1999.07.01"]

1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 Qd7 5. Qg4 f5 6. Qg3 (6. exf6 Nxf6) 6... b6
7. h4 Bb7 $5 (7... Ba6 8. Bxa6 Nxa6 9. Nh3 O-O-O 10. Nf4 Kb7 11. a3 $6 (11. Bd2
) 11... Bxc3+ 12. bxc3 (12. Qxc3 c5) 12... Nb8 {Zatulovskaya-Javanovic/
Vrnjacka Banja/1965/}) 8. Bd3 $5 (8. Bd2 Nc6 9. Nf3 O-O-O 10. O-O-O {/\ 11.Ne2,
12.Nf4 +/= Clarke}) 8... Nc6 9. Nge2 (9. Nf3 O-O-O 10. Bd2) 9... O-O-O 10. Bd2
Nh6 $1 11. a3 $6 (11. O-O-O $5) (11. Bxh6 $142 $1 gxh6 12. Qe3 {/\g3,Nf4 Keres}
) 11... Be7 $1 12. Bb5 (12. Qxg7 $2 Ng4) 12... Rdg8 (12... Kb8 $6 13. Bxc6 Bxc6
14. Qxg7 Ng4 15. Bg5 $1) (12... a6 $6 13. Bxc6 Bxc6 14. Qxg7 Ng4 15. Bg5 $16 {
Clarke}) 13. Qd3 Nf7 14. O-O-O Kb8 $1 15. Nf4 Qc8 $1 16. Nce2 Ncd8 $1 $36 17.
Qb3 $2 $17 {Losing a pawn} (17. c3 c5 18. Kb1 $17 {Clarke}) 17... c6 $1 18. Bd3
c5 19. dxc5 Bxc5 $17 (19... bxc5 $6 20. c4) 20. Nh3 Nxe5 21. Bf4 Ndf7 22. Bb5 (
22. Ng5 Bd6) (22. Qc3 Bd6) 22... Ka8 23. Nd4 Ng6 24. Qa4 Bxd4 25. Bd7 Qf8 26.
Rxd4 (26. Bxe6 $5 Be5 $1 27. Bxd5 (27. Rxd5 Bxf4+ 28. Nxf4 Nxf4 29. Qxf4 Qe7 $1
$19) 27... Bxd5 28. Rxd5 Qb8 $1 29. Qc6+ (29. Rd7 Bxf4+ 30. Kb1 Rc8) 29... Qb7
30. Qxb7+ Kxb7 31. Rd7+ Kc6 32. Rxf7 Bxf4+ $19 {Clarke}) 26... e5 $1 27. Rb4 (
27. Rxd5 Nxf4 $1) (27. Bc6 Qc8) 27... exf4 28. Rxb6 Nfe5 29. Rxb7 Kxb7 30. h5
Qd6 31. hxg6 Qxd7 32. Qxf4 Nxg6 0-1[/pgn]

In fact, it seems to me that Petrosian's own score with ...Bb7 strategy was better than with ...Ba6.

In this game in response to 7.Bb5, Petrosian opts for a ...c6/...Ba6 plan, but 7...Nc6 and 8...Bb7 would revert to his normal way of playing (see above). He only drew this game.

[pgn][Event "Palma de Mallorca Gran Torneo-05"]
[Site "Palma de Mallorca"]
[Date "1969.12.01"]
[Round "8"]
[White "Damjanovic, Mato"]
[Black "Petrosian, Tigran V"]
[Result "1/2-1/2"]
[ECO "C16"]
[PlyCount "77"]
[EventDate "1969.11.23"]
[EventType "tourn"]
[EventRounds "17"]
[EventCountry "ESP"]
[Source "ChessBase"]
[SourceDate "1999.07.01"]

1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 b6 5. Qg4 Bf8 6. Nf3 Qd7 7. Bb5 c6 8. Bd3
Ba6 9. Bxa6 Nxa6 10. a3 Ne7 11. O-O Nf5 12. Ne2 Nc7 13. Qh3 O-O-O 14. g4 Ne7
15. Ng5 Ng6 16. f4 f6 17. Nf3 Ne7 18. Ng3 f5 19. gxf5 Nxf5 20. Nxf5 exf5 21.
Kh1 g6 22. a4 Ne6 23. Be3 Be7 24. b3 Kb7 25. a5 b5 26. Qg2 a6 27. c3 Rc8 28. b4
Rcg8 29. Qc2 Qe8 30. Rg1 Qf7 31. Rg2 h6 32. Rag1 Rg7 33. Bc1 Kc8 34. Rg3 Kd8
35. R1g2 Rhg8 36. h4 Rh8 37. Kg1 Rhg8 38. Kf1 Ke8 39. Ke2 1/2-1/2[/pgn]

As mentioned, theres a lot more to these systems than meets the eye, and engines are still figuring them out.

The positional play required is very subtle and move-orders of much less importance.

User avatar
Ovyron
Posts: 3344
Joined: Tue Jul 03, 2007 2:30 am

Re: Weak Opennings unearthed by Yorkman tourn?

Post by Ovyron » Fri Jan 19, 2018 6:22 am

BrendanJNorman wrote:In this game in response to 7.Bb5, Petrosian opts for a ...c6/...Ba6 plan
I think that's they keyword, "plan". Engines don't have plans, they crunch numbers, here they're put in a position they don't understand...

I have been supporting engines with learning, and manual learning for engines as opening books, so in this thread I say engines shouldn't be forced to play an opening they don't understand, they'd have learned to avoid it and never play this system, so it'd be better to replace these openings with ones the engines would handle better, otherwise it's like a thematic tournament where engines don't understand the theme.

But allow engines that handle the black side of these openings to keep playing them.

BrendanJNorman
Posts: 1801
Joined: Sun Feb 07, 2016 11:43 pm
Full name: Brendan J Norman

Re: Weak Opennings unearthed by Yorkman tourn?

Post by BrendanJNorman » Fri Jan 19, 2018 6:36 am

Ovyron wrote:
BrendanJNorman wrote:In this game in response to 7.Bb5, Petrosian opts for a ...c6/...Ba6 plan
I think that's they keyword, "plan". Engines don't have plans, they crunch numbers, here they're put in a position they don't understand...

I have been supporting engines with learning, and manual learning for engines as opening books, so in this thread I say engines shouldn't be forced to play an opening they don't understand, they'd have learned to avoid it and never play this system, so it'd be better to replace these openings with ones the engines would handle better, otherwise it's like a thematic tournament where engines don't understand the theme.

But allow engines that handle the black side of these openings to keep playing them.
I think the problem in this system, is that the engine's eval is telling it to increase mobility (the position is very cramped), and so very often to engine will do so, leaving weaknesses behind.

Against weaker engines it overcomes this via strong tactical play (outsearching the other side), but not so easy against a strong engine playing the white side.

Post Reply