Here is another game where the same happened I set it to reach Depth 7 and sargon always took more time, but if you noticed that the other programs always stopped at the time lime since they reached a deeper Depth than sargon
Sargon 1978 UCI Available
Moderator: Ras
-
Chessqueen
- Posts: 5685
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
-
Bill Forster
- Posts: 76
- Joined: Mon Sep 21, 2015 7:47 am
- Location: New Zealand
Re: Sargon 1978 UCI Available
Chessqueen likes to use the Sargon 1978's "FixedDepth" parameter to set a fixed depth. I implemented this for compatibility with (and harmony with /respect to) the original Z80 Sargon user interface whereby the user was asked to select the "Level" (exactly the same as depth in this context) before starting the game. The whole game would proceed at that level. If FixedDepth is set to a non-zero value, then the Sargon 1978 UCI engine will operate in a similar way, always calculating to the assigned depth, taking no notice of the time control information provided by Arena, or other GUI (it will happily just lose on time).
(In V1.01 there are some slight adjustments to this policy, Sargon quickly probes at levels 1,2 and 3 [almost instantaneous] in case there's an immediate mate available.)
Normally FixedDepth should be set to 0 (the default). Then Sargon 1978's UCI wrapper will manage time, using low depth if there's not much time left.
Incidentally, Z80 Sargon 1978 only allowed levels 1-6, and level 6 meant hours of calculation for every move. Setting level 6 with FixedDepth = 6 results in a comparatively speedy 15 secs or so per move using Sargon 1978 UCI, less on a fast machine. I constrain FixedDepth to a range of 1-20, but too much higher than 6 will get unbearably slow even on a fast machine. The time blows out exponentially as explained at length in my project documentation.
(In V1.01 there are some slight adjustments to this policy, Sargon quickly probes at levels 1,2 and 3 [almost instantaneous] in case there's an immediate mate available.)
Normally FixedDepth should be set to 0 (the default). Then Sargon 1978's UCI wrapper will manage time, using low depth if there's not much time left.
Incidentally, Z80 Sargon 1978 only allowed levels 1-6, and level 6 meant hours of calculation for every move. Setting level 6 with FixedDepth = 6 results in a comparatively speedy 15 secs or so per move using Sargon 1978 UCI, less on a fast machine. I constrain FixedDepth to a range of 1-20, but too much higher than 6 will get unbearably slow even on a fast machine. The time blows out exponentially as explained at length in my project documentation.
-
Chessqueen
- Posts: 5685
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
Re: Sargon 1978 UCI Available
Thanks for the explanation, I did noticed in most of the games that I witness that is the caseBill Forster wrote: ↑Tue Feb 09, 2021 3:29 am Chessqueen likes to use the Sargon 1978's "FixedDepth" parameter to set a fixed depth. I implemented this for compatibility with (and harmony with /respect to) the original Z80 Sargon user interface whereby the user was asked to select the "Level" (exactly the same as depth in this context) before starting the game. The whole game would proceed at that level. If FixedDepth is set to a non-zero value, then the Sargon 1978 UCI engine will operate in a similar way, always calculating to the assigned depth, taking no notice of the time control information provided by Arena, or other GUI (it will happily just lose on time).
(In V1.01 there are some slight adjustments to this policy, Sargon quickly probes at levels 1,2 and 3 [almost instantaneous] in case there's an immediate mate available.)
Normally FixedDepth should be set to 0 (the default). Then Sargon 1978's UCI wrapper will manage time, using low depth if there's not much time left.
Incidentally, Z80 Sargon 1978 only allowed levels 1-6, and level 6 meant hours of calculation for every move. Setting level 6 with FixedDepth = 6 results in a comparatively speedy 15 secs or so per move using Sargon 1978 UCI, less on a fast machine. I constrain FixedDepth to a range of 1-20, but too much higher than 6 will get unbearably slow even on a fast machine. The time blows out exponentially as explained at length in my project documentation.
sargon no longer lose to Tarrasch Toy
[Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.08"]
[Round "?"]
[White "Sargon-engine Depth6"]
[Black "TarraschToyEngineV0.906"]
[Result "1-0"]
[BlackElo "1480"]
[ECO "A03"]
[Opening "Bird Opening"]
[Time "20:36:09"]
[Variation "1...d5 2.Nf3 c5"]
[WhiteElo "????"]
[TimeControl "0+90"]
[Termination "normal"]
[PlyCount "75"]
[WhiteType "program"]
[BlackType "program"]
1. f4 d5 2. Nf3 c5 3. b3 Nc6 {(Nb8-c6 Nb1-c3 Ng8-f6) +0.17/3 0} 4. Ba3
{(Bc1-a3 Qd8-d6 Nb1-c3 Ng8-f6 g2-g3 e7-e5) -0.87/6 17} e6 {(e7-e6 Nb1-c3
Ng8-f6 e2-e3) +0.07/4 0} 5. e3 {(e2-e3 Ng8-f6 Bf1-d3 Qd8-a5 Ke1-f2 c5-c4)
0.00/6 22} Qc7 {(Qd8-c7 Nb1-c3 Ng8-f6) +0.35/3 0} 6. Bb5 {(Bf1-b5 Bc8-d7
Nb1-c3 Qc7-a5 Qd1-c1 d5-d4) +0.25/6 24} Nf6 {(Ng8-f6 Nb1-c3) +0.20/2 0} 7.
O-O {(O-O Bc8-d7 Nb1-c3 Qc7-a5 Qd1-c1 d5-d4) +0.25/6 30} Bd6 {(Bf8-d6
Nb1-c3 a7-a6 Bb5xc6+) +0.05/4 0} 8. Bb2 {(Ba3-b2 Bc8-d7 Bb2xf6 g7xf6 Nb1-c3
O-O) +0.37/6 32} O-O {(O-O Nb1-c3 e6-e5) +0.42/3 0} 9. Na3 {(Nb1-a3 Bc8-d7
Bb5xc6 Qc7xc6 Kg1-f2 c5-c4) -0.25/6 43} a6 {(a7-a6 Bb5-a4) +0.50/2 0} 10.
Bxc6 {(Bb5xc6 b7xc6 Kg1-f2 Bc8-b7 Bb2xf6 g7xf6) +0.25/6 26} bxc6 {(b7xc6
Qd1-e2 Bc8-b7) -0.02/3 0} 11. Be5 {(Bb2-e5 Bd6xe5 f4xe5 Nf6-g4 Nf3-g5
Ng4xe5) -0.25/6 24} Bxe5 {(Bd6xe5 Nf3xe5 Bc8-b7) +0.52/3 0} 12. Nxe5
{(Nf3xe5 a6-a5 Ra1-b1 Bc8-a6 d2-d3 d5-d4) -0.25/6 15} Qa5 {(Qc7-a5 Qd1-c1)
+0.07/2 0} 13. Qc1 {(Qd1-c1 Bc8-b7 Ra1-b1 Nf6-e4 Ne5-f3 Ne4-d6) 0.00/6 11}
Bb7 {(Bc8-b7 d2-d4 c5xd4) +0.17/3 0} 14. g4 {(g2-g4 Rf8-d8 Kg1-f2 Nf6-e4+
Kf2-e2 f7-f6) 0.00/6 22} Qc7 {(Qa5-c7 Qc1-b2) +0.17/2 0} 15. Kg2 {(Kg1-g2
Nf6-e4 d2-d3 Ne4-c3 Qc1-e1 Nc3-b5) +0.25/6 16} a5 {(a6-a5 Qc1-b2) +0.40/2
0} 16. Qe1 {(Qc1-e1 Bb7-a6 d2-d3 Rf8-b8 g4-g5 Nf6-d7) +0.25/6 29} Nd7
{(Nf6-d7 Ne5-d3 f7-f5) +0.62/3 0} 17. Nxd7 {(Ne5xd7 Qc7xd7 d2-d3 f7-f5
g4xf5 e6xf5) +0.37/6 16} Qxd7 {(Qc7xd7 Qe1-f2 f7-f5) +0.32/3 0} 18. d3
{(d2-d3 f7-f5 g4-g5 Kg8-f7 d3-d4 Qd7-d6) +0.25/6 21} d4 {(d5-d4 e3xd4
c5xd4) +0.47/3 0} 19. Nc4 {(Na3-c4 Qd7-d5+ Kg2-g3 d4xe3 Nc4-b6 Qd5-d4)
+1.00/6 12} f5 {(f7-f5 Qe1-e2) +0.05/2 0} 20. Nb6 {(Nc4-b6 Qd7-d6 Nb6xa8
Rf8xa8 e3xd4 c5xd4) +2.25/6 32} Qd6 {(Qd7-d6 Nb6xa8 Rf8xa8) -1.67/3 0} 21.
Nxa8 {(Nb6xa8 Rf8xa8 g4xf5 e6xf5 e3xd4 Qd6xd4) +2.75/6 15} Rxa8 {(Rf8xa8
e3xd4 c5xd4) -1.65/3 0} 22. gxf5 {(g4xf5 e6xf5 e3xd4 Qd6-d5+ Kg2-g3 Qd5xd4)
+2.75/6 15} exf5 {(e6xf5 e3xd4 Qd6xd4) -1.32/3 0} 23. exd4 {(e3xd4 c5xd4
Qe1-e5 c6-c5+ Kg2-g3 Qd6-g6+ Kg3-h4) +2.25/6 33} Qd5+ {(Qd6-d5+ Rf1-f3
c5xd4) -1.47/3 0} 24. Kg3 {(Kg2-g3 c5xd4 Qe1-e7 c6-c5 Rf1-f2 a5-a4) +2.00/6
4} cxd4 {(c5xd4 Qe1-e5 Ra8-f8) -1.67/3 0} 25. Qe2 {(Qe1-e2 c6-c5 Ra1-e1
Kg8-f7 Qe2-e7+ Kf7-g6) +2.62/6 15} c5 {(c6-c5 Qe2-h5 Ra8-c8) -1.70/3 0} 26.
Rae1 {(Ra1-e1 Qd5-c6 a2-a3 Qc6-g6+ Kg3-h3 Bb7-d5) +2.37/6 15} Rb8 {(Ra8-b8
Qe2-e6+) -1.65/2 0} 27. Rg1 {(Rf1-g1 g7-g6 Kg3-h4 Bb7-c6 Qe2-e5 Qd5xe5)
+1.50/6 24} Ra8 {(Rb8-a8 Qe2-e6+) -1.67/2 0} 28. Kh4 {(Kg3-h4 Qd5-d8+
Qe2-e7 Bb7-c6 Re1-e5 Bc6-d7) +3.12/6 15} Rb8 {(Ra8-b8 Qe2-h5) -1.67/2 0}
29. Qe5 {(Qe2-e5 Qd5xe5 Re1xe5 Rb8-c8 Re5-e7 Bb7-d5) +3.75/6 11} Qxe5
{(Qd5xe5 f4xe5) -2.55/2 0} 30. Rxe5 {(Re1xe5 Bb7-a6 Re5xf5 Rb8-c8 Rg1-g5
c5-c4) +4.00/6 3} Ra8 {(Rb8-a8 Re5xc5 Bb7-c8) -2.57/3 0} 31. Re7 {(Re5-e7
Bb7-d5 Rg1xg7+ Kg8-h8 Rg7xh7+ Kh8-g8) +4.75/6 9} Bc6 {(Bb7-c6 Rg1xg7+
Kg8-h8) -3.72/3 0} 32. Rgxg7+ {(Rg1xg7+ Kg8-h8 Rg7xh7+ Kh8-g8 Kh4-g5
Ra8-f8) +4.75/6 5} Kh8 {(Kg8-h8 Rg7xh7+ Kh8-g8 c2-c4) -4.05/4 0} 33. Rxh7+
{(Rg7xh7+ Kh8-g8 Kh4-g5 Ra8-e8 Rh7-g7+ Kg8-f8) +5.25/6 3} Kg8 {(Kh8-g8)
-3.90/2} 34. Kg5 {(Kh4-g5 Bc6-f3 Kg5xf5 Bf3-d1 Re7-g7+ Kg8-f8) +5.25/6 5}
Re8 {(Ra8-e8 Kg5xf5 Kg8-f8 Re7-c7) -5.07/4 0} 35. Reg7+ {(Re7-g7+ Kg8-f8
Kg5-f6 Bc6-d5 Rh7-h8+ Bd5-g8) +7.50/6 5} Kf8 {(Kg8-f8) -4.90/2} 36. Kf6
{(Kg5-f6 Re8-e6+ Kf6xe6 Bc6-e8 Ke6xf5 a5-a4) +10.75/6 9} Bd5 {(Bc6-d5
Rg7-f7+ Kf8-g8) -50.00/4 0} 37. Rh8+ {(Rh7-h8+ Bd5-g8 Rg7xg8+) +M2/3 0} Bg8
{(Bd5-g8) -6.87/2} 38. Rgxg8# 1-0
-
Bill Forster
- Posts: 76
- Joined: Mon Sep 21, 2015 7:47 am
- Location: New Zealand
Re: Sargon 1978 UCI Available
I used Tarrasch Toy Engine when I was bringing up Sargon. I wrote Tarrasch Toy engine myself and I was not surprised that it outclassed Sargon in all my tests. It's actually a rather similar engine to Sargon (because I only really knew the basic ideas of Shannon etc. when I wrote it - i.e. primitive engine ideas, just like those employed by Sargon). So like Sargon TTE does a broad shallow search, with SOMA to try and reduce the impact of the shallow depth. Simple minimax with alpha-beta to speed things up. Alpha-Beta is way way simpler in Sargon incidentally. If I ever write another engine I will avoid the monstrosity that was alpha-beta in TTE. Although TTE is C++ and Sargon is assembler, TTE is likely still quicker at generating moves etc because it uses lookup tables, using memory Sargon could never afford. So they are comparable in many ways, BUT Sargon has an absolutely primitive, zero knowledge leaf evaluator, whereas I spent a bit of time teaching TTE my favourite nuggets of chess wisdom. So it can play endings quite well for example, using its king, fighting for passed pawns, bullying weak opponents. Things Sargon can only dream of.
So I was unsurprised that TTE is considerably stronger than Sargon. This game of yours seems weird to me, after 19.Nc4 Sargon threatens a knight fork, and TTE just allows it ??! I set TTE as a kibitzer and just allowing the knight fork in this position doesn't appear in the analysis at all.
I have stepped through other games where you have Sargon winning, and although I am predisposed to wanting Sargon to do well of course, often the opponent engine seems to just give up and start jettisoning material. This brings me no joy. I am not sure what you are doing, but something seems off.
So I was unsurprised that TTE is considerably stronger than Sargon. This game of yours seems weird to me, after 19.Nc4 Sargon threatens a knight fork, and TTE just allows it ??! I set TTE as a kibitzer and just allowing the knight fork in this position doesn't appear in the analysis at all.
I have stepped through other games where you have Sargon winning, and although I am predisposed to wanting Sargon to do well of course, often the opponent engine seems to just give up and start jettisoning material. This brings me no joy. I am not sure what you are doing, but something seems off.
-
Chessqueen
- Posts: 5685
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
Re: Sargon 1978 UCI Available
I have set up this position 5 times under Arena and it always play 19. Nc4 if you click under game and then click Demo to let them play both sides sargon always play the same move 19.Nc4Bill Forster wrote: ↑Tue Feb 09, 2021 6:33 am I used Tarrasch Toy Engine when I was bringing up Sargon. I wrote Tarrasch Toy engine myself and I was not surprised that it outclassed Sargon in all my tests. It's actually a rather similar engine to Sargon (because I only really knew the basic ideas of Shannon etc. when I wrote it - i.e. primitive engine ideas, just like those employed by Sargon). So like Sargon TTE does a broad shallow search, with SOMA to try and reduce the impact of the shallow depth. Simple minimax with alpha-beta to speed things up. Alpha-Beta is way way simpler in Sargon incidentally. If I ever write another engine I will avoid the monstrosity that was alpha-beta in TTE. Although TTE is C++ and Sargon is assembler, TTE is likely still quicker at generating moves etc because it uses lookup tables, using memory Sargon could never afford. So they are comparable in many ways, BUT Sargon has an absolutely primitive, zero knowledge leaf evaluator, whereas I spent a bit of time teaching TTE my favourite nuggets of chess wisdom. So it can play endings quite well for example, using its king, fighting for passed pawns, bullying weak opponents. Things Sargon can only dream of.
So I was unsurprised that TTE is considerably stronger than Sargon. This game of yours seems weird to me, after 19.Nc4 Sargon threatens a knight fork, and TTE just allows it ??! I set TTE as a kibitzer and just allowing the knight fork in this position doesn't appear in the analysis at all.
I have stepped through other games where you have Sargon winning, and although I am predisposed to wanting Sargon to do well of course, often the opponent engine seems to just give up and start jettisoning material. This brings me no joy. I am not sure what you are doing, but something seems off.
[Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.08"]
[Round "?"]
[White "Sargon-engine Depth6"]
[Black "TarraschToyEngineV0.906"]
[Result ""]
[BlackElo "1480"]
[ECO "A03"]
[Opening "Bird Opening"]
[Time "20:36:09"]
[Variation "1...d5 2.Nf3 c5"]
[WhiteElo "????"]
[TimeControl "0+90"]
[Termination "normal"]
[PlyCount "75"]
1. f4 d5 2. Nf3 c5 3. b3 Nc6 {(Nb8-c6 Nb1-c3 Ng8-f6) +0.17/3 0} 4. Ba3
{(Bc1-a3 Qd8-d6 Nb1-c3 Ng8-f6 g2-g3 e7-e5) -0.87/6 17} e6 {(e7-e6 Nb1-c3
Ng8-f6 e2-e3) +0.07/4 0} 5. e3 {(e2-e3 Ng8-f6 Bf1-d3 Qd8-a5 Ke1-f2 c5-c4)
0.00/6 22} Qc7 {(Qd8-c7 Nb1-c3 Ng8-f6) +0.35/3 0} 6. Bb5 {(Bf1-b5 Bc8-d7
Nb1-c3 Qc7-a5 Qd1-c1 d5-d4) +0.25/6 24} Nf6 {(Ng8-f6 Nb1-c3) +0.20/2 0} 7.
O-O {(O-O Bc8-d7 Nb1-c3 Qc7-a5 Qd1-c1 d5-d4) +0.25/6 30} Bd6 {(Bf8-d6
Nb1-c3 a7-a6 Bb5xc6+) +0.05/4 0} 8. Bb2 {(Ba3-b2 Bc8-d7 Bb2xf6 g7xf6 Nb1-c3
O-O) +0.37/6 32} O-O {(O-O Nb1-c3 e6-e5) +0.42/3 0} 9. Na3 {(Nb1-a3 Bc8-d7
Bb5xc6 Qc7xc6 Kg1-f2 c5-c4) -0.25/6 43} a6 {(a7-a6 Bb5-a4) +0.50/2 0} 10.
Bxc6 {(Bb5xc6 b7xc6 Kg1-f2 Bc8-b7 Bb2xf6 g7xf6) +0.25/6 26} bxc6 {(b7xc6
Qd1-e2 Bc8-b7) -0.02/3 0} 11. Be5 {(Bb2-e5 Bd6xe5 f4xe5 Nf6-g4 Nf3-g5
Ng4xe5) -0.25/6 24} Bxe5 {(Bd6xe5 Nf3xe5 Bc8-b7) +0.52/3 0} 12. Nxe5
{(Nf3xe5 a6-a5 Ra1-b1 Bc8-a6 d2-d3 d5-d4) -0.25/6 15} Qa5 {(Qc7-a5 Qd1-c1)
+0.07/2 0} 13. Qc1 {(Qd1-c1 Bc8-b7 Ra1-b1 Nf6-e4 Ne5-f3 Ne4-d6) 0.00/6 11}
Bb7 {(Bc8-b7 d2-d4 c5xd4) +0.17/3 0} 14. g4 {(g2-g4 Rf8-d8 Kg1-f2 Nf6-e4+
Kf2-e2 f7-f6) 0.00/6 22} Qc7 {(Qa5-c7 Qc1-b2) +0.17/2 0} 15. Kg2 {(Kg1-g2
Nf6-e4 d2-d3 Ne4-c3 Qc1-e1 Nc3-b5) +0.25/6 16} a5 {(a6-a5 Qc1-b2) +0.40/2
0} 16. Qe1 {(Qc1-e1 Bb7-a6 d2-d3 Rf8-b8 g4-g5 Nf6-d7) +0.25/6 29} Nd7
{(Nf6-d7 Ne5-d3 f7-f5) +0.62/3 0} 17. Nxd7 {(Ne5xd7 Qc7xd7 d2-d3 f7-f5
g4xf5 e6xf5) +0.37/6 16} Qxd7 {(Qc7xd7 Qe1-f2 f7-f5) +0.32/3 0} 18. d3
{(d2-d3 f7-f5 g4-g5 Kg8-f7 d3-d4 Qd7-d6) +0.25/6 21} d4 {(d5-d4 e3xd4
c5xd4) +0.47/3 0}
-
Chessqueen
- Posts: 5685
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
Re: Sargon 1978 UCI Available
What seems to be strange is that if you start from move 19. and let them play it out after 5 times it always make the same move some but later on it change and the game is a complete different ending
Last edited by Chessqueen on Tue Feb 09, 2021 2:06 pm, edited 2 times in total.
-
Guenther
- Posts: 4718
- Joined: Wed Oct 01, 2008 6:33 am
- Location: Regensburg, Germany
- Full name: Guenther Simon
Re: Sargon 1978 UCI Available
You cannot read... Bill talked about TTE after 19. Nc4. I hope I don't need to explain what TTE means after all?Chessqueen wrote: ↑Tue Feb 09, 2021 1:13 pmI have set up this position 5 times under Arena and it always play 19. Nc4 if you click under game and then click Demo to let them play both sides sargon always play the same move 19.Nc4Bill Forster wrote: ↑Tue Feb 09, 2021 6:33 am I used Tarrasch Toy Engine when I was bringing up Sargon. I wrote Tarrasch Toy engine myself and I was not surprised that it outclassed Sargon in all my tests. It's actually a rather similar engine to Sargon (because I only really knew the basic ideas of Shannon etc. when I wrote it - i.e. primitive engine ideas, just like those employed by Sargon). So like Sargon TTE does a broad shallow search, with SOMA to try and reduce the impact of the shallow depth. Simple minimax with alpha-beta to speed things up. Alpha-Beta is way way simpler in Sargon incidentally. If I ever write another engine I will avoid the monstrosity that was alpha-beta in TTE. Although TTE is C++ and Sargon is assembler, TTE is likely still quicker at generating moves etc because it uses lookup tables, using memory Sargon could never afford. So they are comparable in many ways, BUT Sargon has an absolutely primitive, zero knowledge leaf evaluator, whereas I spent a bit of time teaching TTE my favourite nuggets of chess wisdom. So it can play endings quite well for example, using its king, fighting for passed pawns, bullying weak opponents. Things Sargon can only dream of.
So I was unsurprised that TTE is considerably stronger than Sargon. This game of yours seems weird to me, after 19.Nc4 Sargon threatens a knight fork, and TTE just allows it ??! I set TTE as a kibitzer and just allowing the knight fork in this position doesn't appear in the analysis at all.
I have stepped through other games where you have Sargon winning, and although I am predisposed to wanting Sargon to do well of course, often the opponent engine seems to just give up and start jettisoning material. This brings me no joy. I am not sure what you are doing, but something seems off.
...
Moreover, please set your Arena finally not to save full PVs into pgn, or remove it before posting here.
So often people here were told to do so (you, surely a dozen of times with various user accounts).
For a quick check that Arena crap stuff is completely distracting from reading.
@Bill: TTE was forced to play the whole game in zero seconds ;-)
(ran the crappy pgn again through a script)
[White "Sargon-engine Depth6"]
[Black "TarraschToyEngineV0.906"]
[Result "1-0"]
1. f4 d5
2. Nf3 c5
3. b3 Nc6 {+0.17/3 0}
4. Ba3 {-0.87/6 17} e6 {+0.07/4 0}
5. e3 {0.00/6 22} Qc7 {+0.35/3 0}
6. Bb5 {+0.25/6 24} Nf6 {+0.20/2 0}
7. O-O {+0.25/6 30} Bd6 {+0.05/4 0}
8. Bb2 {+0.37/6 32} O-O {+0.42/3 0}
9. Na3 {-0.25/6 43} a6 {+0.50/2 0}
10. Bxc6 {+0.25/6 26} bxc6 {-0.02/3 0}
11. Be5 {-0.25/6 24} Bxe5 {+0.52/3 0}
12. Nxe5 {-0.25/6 15} Qa5 {+0.07/2 0}
13. Qc1 {0.00/6 11} Bb7 {+0.17/3 0}
14. g4 {0.00/6 22} Qc7 {+0.17/2 0}
15. Kg2 {+0.25/6 16} a5 {+0.40/2 0}
16. Qe1 {+0.25/6 29} Nd7 {+0.62/3 0}
17. Nxd7 {+0.37/6 16} Qxd7 {+0.32/3 0}
18. d3 {+0.25/6 21} d4 {+0.47/3 0}
19. Nc4 {+1.00/6 12} f5 {+0.05/2 0}
20. Nb6 {+2.25/6 32} Qd6 {-1.67/3 0}
21. Nxa8 {+2.75/6 15} Rxa8 {-1.65/3 0}
22. gxf5 {+2.75/6 15} exf5 {-1.32/3 0}
23. exd4 {+2.25/6 33} Qd5+ {-1.47/3 0}
24. Kg3 {+2.00/6 4} cxd4 {-1.67/3 0}
25. Qe2 {+2.62/6 15} c5 {-1.70/3 0}
26. Rae1 {+2.37/6 15} Rb8 {-1.65/2 0}
27. Rg1 {+1.50/6 24} Ra8 {-1.67/2 0}
28. Kh4 {+3.12/6 15} Rb8 {-1.67/2 0}
29. Qe5 {+3.75/6 11} Qxe5 {-2.55/2 0}
30. Rxe5 {+4.00/6 3} Ra8 {-2.57/3 0}
31. Re7 {+4.75/6 9} Bc6 {-3.72/3 0}
32. Rgxg7+ {+4.75/6 5} Kh8 {-4.05/4 0}
33. Rxh7+ {+5.25/6 3} Kg8 {-3.90/2}
34. Kg5 {+5.25/6 5} Re8 {-5.07/4 0}
35. Reg7+ {+7.50/6 5} Kf8 {-4.90/2}
36. Kf6 {+10.75/6 9} Bd5 {-50.00/4 0}
37. Rh8+ {+M2/3 0} Bg8 {-6.87/2}
38. Rgxg8# 1-0
-
Chessqueen
- Posts: 5685
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
Re: Sargon 1978 UCI Available
Guenther wrote: ↑Tue Feb 09, 2021 1:59 pmYou cannot read... Bill talked about TTE after 19. Nc4. I hope I don't need to explain what TTE means after all?Chessqueen wrote: ↑Tue Feb 09, 2021 1:13 pmI have set up this position 5 times under Arena and it always play 19. Nc4 if you click under game and then click Demo to let them play both sides sargon always play the same move 19.Nc4Bill Forster wrote: ↑Tue Feb 09, 2021 6:33 am I used Tarrasch Toy Engine when I was bringing up Sargon. I wrote Tarrasch Toy engine myself and I was not surprised that it outclassed Sargon in all my tests. It's actually a rather similar engine to Sargon (because I only really knew the basic ideas of Shannon etc. when I wrote it - i.e. primitive engine ideas, just like those employed by Sargon). So like Sargon TTE does a broad shallow search, with SOMA to try and reduce the impact of the shallow depth. Simple minimax with alpha-beta to speed things up. Alpha-Beta is way way simpler in Sargon incidentally. If I ever write another engine I will avoid the monstrosity that was alpha-beta in TTE. Although TTE is C++ and Sargon is assembler, TTE is likely still quicker at generating moves etc because it uses lookup tables, using memory Sargon could never afford. So they are comparable in many ways, BUT Sargon has an absolutely primitive, zero knowledge leaf evaluator, whereas I spent a bit of time teaching TTE my favourite nuggets of chess wisdom. So it can play endings quite well for example, using its king, fighting for passed pawns, bullying weak opponents. Things Sargon can only dream of.
So I was unsurprised that TTE is considerably stronger than Sargon. This game of yours seems weird to me, after 19.Nc4 Sargon threatens a knight fork, and TTE just allows it ??! I set TTE as a kibitzer and just allowing the knight fork in this position doesn't appear in the analysis at all.
I have stepped through other games where you have Sargon winning, and although I am predisposed to wanting Sargon to do well of course, often the opponent engine seems to just give up and start jettisoning material. This brings me no joy. I am not sure what you are doing, but something seems off.
...
Moreover, please set your Arena finally not to save full PVs into pgn, or remove it before posting here.
So often people here were told to do so (you, surely a dozen of times with various user accounts).
For a quick check that Arena crap stuff is completely distracting from reading.
@Bill: TTE was forced to play the whole game in zero seconds
(ran the crappy pgn again through a script)
[White "Sargon-engine Depth6"]
[Black "TarraschToyEngineV0.906"]
[Result "1-0"]
1. f4 d5
2. Nf3 c5
3. b3 Nc6 {+0.17/3 0}
4. Ba3 {-0.87/6 17} e6 {+0.07/4 0}
5. e3 {0.00/6 22} Qc7 {+0.35/3 0}
6. Bb5 {+0.25/6 24} Nf6 {+0.20/2 0}
7. O-O {+0.25/6 30} Bd6 {+0.05/4 0}
8. Bb2 {+0.37/6 32} O-O {+0.42/3 0}
9. Na3 {-0.25/6 43} a6 {+0.50/2 0}
10. Bxc6 {+0.25/6 26} bxc6 {-0.02/3 0}
11. Be5 {-0.25/6 24} Bxe5 {+0.52/3 0}
12. Nxe5 {-0.25/6 15} Qa5 {+0.07/2 0}
13. Qc1 {0.00/6 11} Bb7 {+0.17/3 0}
14. g4 {0.00/6 22} Qc7 {+0.17/2 0}
15. Kg2 {+0.25/6 16} a5 {+0.40/2 0}
16. Qe1 {+0.25/6 29} Nd7 {+0.62/3 0}
17. Nxd7 {+0.37/6 16} Qxd7 {+0.32/3 0}
18. d3 {+0.25/6 21} d4 {+0.47/3 0}
19. Nc4 {+1.00/6 12} f5 {+0.05/2 0}
20. Nb6 {+2.25/6 32} Qd6 {-1.67/3 0}
21. Nxa8 {+2.75/6 15} Rxa8 {-1.65/3 0}
22. gxf5 {+2.75/6 15} exf5 {-1.32/3 0}
23. exd4 {+2.25/6 33} Qd5+ {-1.47/3 0}
24. Kg3 {+2.00/6 4} cxd4 {-1.67/3 0}
25. Qe2 {+2.62/6 15} c5 {-1.70/3 0}
26. Rae1 {+2.37/6 15} Rb8 {-1.65/2 0}
27. Rg1 {+1.50/6 24} Ra8 {-1.67/2 0}
28. Kh4 {+3.12/6 15} Rb8 {-1.67/2 0}
29. Qe5 {+3.75/6 11} Qxe5 {-2.55/2 0}
30. Rxe5 {+4.00/6 3} Ra8 {-2.57/3 0}
31. Re7 {+4.75/6 9} Bc6 {-3.72/3 0}
32. Rgxg7+ {+4.75/6 5} Kh8 {-4.05/4 0}
33. Rxh7+ {+5.25/6 3} Kg8 {-3.90/2}
34. Kg5 {+5.25/6 5} Re8 {-5.07/4 0}
35. Reg7+ {+7.50/6 5} Kf8 {-4.90/2}
36. Kf6 {+10.75/6 9} Bd5 {-50.00/4 0}
37. Rh8+ {+M2/3 0} Bg8 {-6.87/2}
38. Rgxg8# 1-0
So that move after 19.Nc4 Nb6 was different I do NOT know why. It seems that I just learn how to read
-
Guenther
- Posts: 4718
- Joined: Wed Oct 01, 2008 6:33 am
- Location: Regensburg, Germany
- Full name: Guenther Simon
Re: Sargon 1978 UCI Available
wrong again - it really makes no sense, it's like talking to a stone?Chessqueen wrote: ↑Tue Feb 09, 2021 2:10 pmSo that move after 19.Nc4 Nb6 was different I do NOT know why. It seems that I just learn how to read :roll: :mrgreen:
another hint: TTE != Sargon, Nb6 is not the move after Nc4
Some people here should have added a kind of warning level to their name, thus new users won't waste their time
taking them seriously.
-
Chessqueen
- Posts: 5685
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
Re: Sargon 1978 UCI Available
Guenther wrote: ↑Tue Feb 09, 2021 2:27 pmwrong again - it really makes no sense, it's like talking to a stone?Chessqueen wrote: ↑Tue Feb 09, 2021 2:10 pmSo that move after 19.Nc4 Nb6 was different I do NOT know why. It seems that I just learn how to read
another hint: TTE != Sargon, Nb6 is not the move after Nc4
Some people here should have added a kind of warning level to their name, thus new users won't waste their time
taking them seriously.
It is NOT my fault that TTE chose a different move after 19.Nc4..... what do I have to do with it, stop being so rude!