Chess2u
Would you like to react to this message? Create an account in a few clicks or log in to continue.

Chess2uLog in

StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

descriptionStockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 EmptyRe: StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

more_horiz
*** I am sorry but this Forum prevents me to modify my messages ****

roraima wrote:
RSilvian wrote:
The situation after only 7 games:

StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 Clipb147

+2 -1 =4

+49 Elo points for StockfishMZ (of course it's too early......)



How can you draw "sensationalist breaking news" with seven games??

A quick overview of the games and the engine is already inferior by -1 and also inferior openings move. Did you take the time to review the opening test setÉ

[Site "ISR 3"]
[Date "2022.05.20"]
[Round "3"]
[White "StockfishMZ 170522 bmi2"]
[Black "Fat_titz_2_win64-bmi2"]
[Result "1-0"]
[BlackElo "2200"]
[ECO "B12"]
[Opening "Caro-Kann"]
[Time "12"]
[Variation "de Bruycker Defence, 1.e4 c6 2.d4 Na6"]
[WhiteElo "2200"]
[TimeControl "240+2"]
[Termination "adjudication"]
[PlyCount "142"]
[WhiteType "program"]
[BlackType "program"]

1. e4 {(e2-e4) +0.23/60 0} c6 2. d4 {(d2-d4) +0.25/48 0} Na6 3. Nf3
{(Ng1-f3) +1.31/40 0} d6 {(d7-d6 Nb1-c3 Na6-c7 h2-h3 g7-g6 Bf1-d3 Bf8-g7
a2-a4 Ng8-f6 O-O Nf6-d7 Bd3-c4 O-O Rf1-e1 e7-e5 Bc1-g5 Bg7-f6 Bg5xf6 Qd8xf6
a4-a5 e5xd4 Nf3xd4 Nd7-e5 Bc4-f1 g6-g5 Qd1-d2 Nc7-e6) -1.53/26 7}


[Event "NN Test232"]
[Site "ISR 3"]
[Date "2022.05.20"]
[Round "7"]
[White "StockfishMZ 170522 bmi2"]
[Black "Fat_titz_2_win64-bmi2"]
[Result "1-0"]
[BlackElo "2200"]
[ECO "B07"]
[Opening "Pirc"]
[Time "13"]
[Variation "Pytel-Czech, 4.Nf3"]
[WhiteElo "2200"]
[TimeControl "240+2"]
[Termination "adjudication"]
[PlyCount "114"]
[WhiteType "program"]
[BlackType "program"]

1. e4 {(e2-e4) +0.23/60 0} d6 2. d4 {(d2-d4) 0.00/41 0} Nf6 3. Nc3
{(Nb1-c3) 0.00/41 0} c6 4. Nf3 {(Ng1-f3) +0.69/40 0} Bg4 5. h3 {(h2-h3)
+1.23/46 0} Bxf3 {(Bg4xf3 Qd1xf3 d6-d5 e4-e5 Nf6-d7 Bc1-g5 Qd8-b6 O-O-O
e7-e6 Kc1-b1 a7-a5 a2-a4 Bf8-b4 h3-h4 O-O h4-h5 h7-h6 Bg5-c1 Bb4xc3 Qf3xc3
c6-c5 Qc3-g3 Kg8-h8 d4xc5 Nd7xc5 Bf1-b5 Nc5-e4 Qg3-e3 Qb6xe3 Bc1xe3)
-1.64/25 9} 6. Qxf3 {(Qd1xf3) +1.29/22 0}

descriptionStockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 EmptyRe: StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

more_horiz
[quote="roraima"]
RSilvian wrote:
The situation after only 7 games:



1.-How can you draw "sensationalist breaking news" with seven games??

2-A quick overview of the games and the engine is already inferior by +1 and inferior openings. Did you took the same to review the opening test setÉ


bigsmile

1.-....."sensationalist breaking news" with seven games??

If you look at two posts above you can already see that there are 77 games ....... bigsmile

2.-If (if) you look at a few posts above you will see the conditions of the match: only Fat Fritz 2 is linked to SuperGM_4mvs.abk (8 plies book). bigsmile


Eman fan ? Indeed, "bad" news ........ we have a free engine (not private) that looks better! thumb up

descriptionStockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 EmptyRe: StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

more_horiz
The situation after 87 games:

StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 Y12

+15 -4 =68

+42 Elo points for StockfishMZ 170522 bmi2 Sfnnv5 net architecture.

This night will start a new match Fat Titz 2 vs. StockfishMZ 17(18)0522 bmi2 Sfnnv4 net architecture. Can't wait to see which is the better: SFMZ Sfnnv4-SFMZ Sfnnv5 or Eman 7.93 ? bigsmile

thumb up think clap

descriptionStockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 EmptyRe: StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

more_horiz
RSilvian wrote:
roraima wrote:
RSilvian wrote:
The situation after only 7 games:



1.-How can you draw "sensationalist breaking news" with seven games??

2-A quick overview of the games and the engine is already inferior by +1 and inferior openings. Did you took the same to review the opening test setÉ


bigsmile

1.-....."sensationalist breaking news" with seven games??

If you look at two posts above you can already see that there are 77 games .......  bigsmile

2.-If (if) you look at a few posts above you will see the conditions of the match: only Fat Fritz 2 is linked to SuperGM_4mvs.abk (8 plies book). bigsmile


Eman fan ? Indeed, "bad" news ........ we have a free engine (not private) that looks better! thumb up



Sorry, Silvian. I have also seen your postings for years  in Talkchess and it is the same approach.

When I agree that this is a hobby, I have checked all the opening test sets, which I use, with a strong engine to avoid bad opening resulting positions.

77 games are not conclusive, even though I am running 1000 games per match and I will never equalize the 10000 game run by the testing battery of the Stockfish Team. So I avoid drawing "sensationalist conclusions" and moderating myself to just publish the result and that people decide.

I am not an Eman fan. Sorry, other wrong affirmation. I do tests to validate to which degree a Clone Engine is "better" than the original Stockfish.

But, please, do not take me bad, just verify better the opening test sets you use, because quitting with +1 from the opening is not a valuable approach.

Thank you. (Again, sorry for the errors but this Forum prevents me to correct my messages)

descriptionStockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 EmptyRe: StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

more_horiz
[quote="roraima"][quote="RSilvian"][quote="roraima"]
RSilvian wrote:
The situation after only 7 games:

Sorry, Silvian. I have also seen your postings for years  in Talkchess and it is the same approach.

When I agree that this is a hobby, I have checked all the opening test sets, which I use, with a strong engine to avoid bad opening resulting positions.

77 games are not conclusive, even though I am running 1000 games per match and I will never equalize the 10000 game run by the testing battery of the Stockfish Team. So I avoid drawing "sensationalist conclusions" and moderating myself to just publish the result and that people decide.

I am not an Eman fan. Sorry, other wrong affirmation. I do tests to validate to which degree a Clone Engine is "better" than the original Stockfish.

But, please, do not take me bad, just verify better the opening test sets you use, because quitting with +1 from the opening is not a valuable approach.

Thank you. (Again, sorry for the errors but this Forum prevents me to correct my messages)


No problem !
I understand your position well (which is otherwise correct). It's just that the purpose of my three matches is simple: Eman7.93-SFMZ Sfnnv5 & SFMZ Sfnnv4 versus Fat Titz 2. There are only 2 goals (simple and simplistic):
1.-if Eman.exp works well with SFMZ (the answer is clear -OK);
2.-Sfnnv4 vs. Sfnnv5
Of course I do not have the time and calculation resources of CCRL-CEGT & co. However, experience has shown me that 100 games at TC = 4 '+ 2 "still means something (within the statistical margin of error). Otherwise but only the difference in Elo points calculated by the graphical interface for the 100 games "Enough for me. Chess (including computer chess) has been a passion for decades. But that's it. Passion. Nothing more. I'm not a programmer, a professional tester or a fanatic of numbers. I'm not interested in arrogance and lack of common sense ... even if it comes from a professional. They can argue with each other!

Best to you !

Well, the situation in my second match (after 91 games) is:

StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 Z13

17 -4 =70

+49 Elo points for SFMZ Sfnnv5.

thumb up

descriptionStockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2 - Page 2 EmptyRe: StockfishMZ 170522 bmi2 vs Fat_Titz_2_win64-bmi2

more_horiz
Permissions in this forum:
You cannot reply to topics in this forum