Berserk12 Slighty better than Berserk 11.1

[Chessqueen]

Do NOT expect me to run 1000 games, I only wanted to get an idea with a small sample of 10 games
Engine Score Mi An S-B
1: Berserk-12-x64-avx2 5.5/10 ···················· ==1=0====1 24.75
2: Berserk-11.1-x64-avx2 4.5/10 ==0=1====0 ···················· 24.7
10 games played / Tournament is finished
Name of the tournament: Berserk-12 Vs Berserk-11.1
Site/ Country: DESKTOP-4QNC0GS, United States
Level: Blitz 1/12
Hardware: Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz with 15.9 GB Memory
Operating system: Windows 10 Home Home Edition (Build 9200) 64 bit

[Ciekce]

Do NOT expect me to run 1000 games, I only wanted to get an idea with a small sample of 10 games

this is not a representative sample and does not give you any idea

[AndrewGrant]

I flipped a coin at my desk just now, and in mid air I called it as heads for Berserk 12, but it landed on tails. Hopefully Jay can address these issues.

[Chessqueen]

I flipped a coin at my desk just now, and in mid air I called it as heads for Berserk 12, but it landed on tails. Hopefully Jay can address these issues.

I got the point, and I decided to run a 40 games match at 5 seconds per move, and Berserk is leading by a great margin. But going back at a match of only 20 games if any engine win it is NOT logical for any smart and logical human to say well we need 1000 games to determine which engine version is stronger, first of all the Author would had NOT released it if he did not see any improvement with thousands of games that he tested them both, But even in 20 games you can say that a small statistical sample is enough to get an idea. For Instance, most World Chess Championship between human are determine by a small sample or amount of games such as the best of 12 games, do NOT tell me that after Carlsen beat Anand in the 1st World Championship Match you needed to wait for the 2nd encounter because 12 games were NOT enough to know that Carlsen was slightly stronger than GM Anand

NOTE: Just like the Human World Chess Championship of 12 games Carlsen did NOT lose a single game against GM Anand, neither Berserk version 12 after 30 games so far. Therefore, this small sample is way more than the 12 games that is needed to chose the Human World Champion
as well between these two versions of Berserk12 vs Berserk11.1

Rank Engine Score Be Be S-B
1 Berserk-12-x64-avx2 18/30 · ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· · 1======1=1=========11========= 189.75
2 Berserk-11.1-x64-avx2 12/30 0======0=0=========00========= · ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· · 189.75


30 of 40 games played

Tournament start: 2023.10.25, 21:16:39
Latest update: 2023.10.27, 01:21:00
Site/ Country: DESKTOP-4QNC0GS, United States
Level: 5 Seconds
Hardware: Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz with 15.9 GB Memory
Operating system: Windows 10 Home Home Edition (Build 9200) 64 bit
PGN-File: Battle of the the two Berserk Versions.pgn
Table created with: Arena 3.5.1

[Ciekce]

that's a lot of words to say "I don't have the slightest understanding of statistics"

[Guenther]

that's a lot of words to say "I don't have the slightest understanding of statistics"

You could have also said: you still don't get the point ;)
(Also this thread is again in the wrong subforum - probably for umptieth time for CQuack => Tournaments and Matches)

BTW he also has developed the new obsession (one of dozens over 25 years of computerchess quack life) with the irrelevant and unusual/weird tc of fixed time per single move in full games (and not in test positions).

[Chessqueen]

that's a lot of words to say "I don't have the slightest understanding of statistics"

Most people here can google Small statistical sample, but I decided to run a 40 games match at 5 seconds per move, and Berserk 11 did not lose a single game. But going back at a match of only 20 games if any engine win it is NOT logical for any smart and logical human to say well we need 1000 games to determine which engine version is stronger, first of all the Author would had NOT released it if he did not see any improvement with thousands of games that he tested them both, But even in 20 games you can say that a small statistical sample is enough to get an idea.

NOTE: Just like the Human World Chess Championship of 12 games Carlsen did NOT lose a single game against GM Anand. Therefore, this small sample is way more than the 12 games that is needed to chose the Human World Champion as well between these two versions of Berserk12 vs Berserk11.1

Berserk12 vs Berserk11.1

Rank Engine Score Be Be S-B
1 Berserk-12-x64-avx2 24.0/40 · ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· · 1======1=1=========11=======1======11=== 384.00
2 Berserk-11.1-x64-avx2 16.0/40 =0======0=0=========00=======0======00=== · ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· ·· · 384.00


40 games played / Tournament is finished

Tournament start: 2023.10.25, 21:16:39
Latest update: 2023.10.27, 03:31:14

Site/ Country: DESKTOP-4QNC0GS, United States
Level: 5 Seconds
Hardware: Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz with 15.9 GB Memory
Operating system: Windows 10 Home Home Edition (Build 9200) 64 bit
PGN-File: Battle of the the two Berserk Versions.pgn
Table created with: Arena 3.5.1

[Chessqueen]

that's a lot of words to say "I don't have the slightest understanding of statistics"

You could have also said: you still don't get the point
(Also this thread is again in the wrong subforum, I can give you a pawn Odds and still beat you computerchess quack.

BTW he also has developed the new obsession (one of dozens over 25 years of computerchess quack life) with the irrelevant and unusual/weird tc of fixed time per single move in full games (and not in test positions).

Go ahead lets bet $5000.00 I will even give you the plane ticket to come and challenge me with a pawn Odds. Who cares, and you are NOT such a great player either, I already beat my friend rated 2025 with a $100.00 bet after he gave me his f7 pawn, I will destroy your Ego.

I am still waiting for you to beat Stockfish or Dragon 3.3, Come on are you scare to show how Stockfish or Dragon 3.3 can destroy you with a Rook Odd ==>https://chessarena.com/profile/168714
[pgn]vent "Rook Odds"] [Date "2023.10.25"] [Round "4"] [White "Human rated 1975"] [Black "Stockfish-windows-x86-64-avx2"] [Result "*"] [BlackElo "3545"] [Time "05:10:17"] [WhiteElo "1975"] [TimeControl "60+1"] [SetUp "1"] [FEN "1nbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQk - 0 1"] [WhiteType "human"] [BlackType "program"][/pgn]

[Guenther]

that's a lot of words to say "I don't have the slightest understanding of statistics"

You could have also said: you still don't get the point ;)
(Also this thread is again in the wrong subforum, I can give you a pawn Odds and still beat you computerchess quack.

BTW he also has developed the new obsession (one of dozens over 25 years of computerchess quack life) with the irrelevant and unusual/weird tc of fixed time per single move in full games (and not in test positions).

Go ahead lets bet $5000.00 I will even give you the plane ticket to come and challenge me with a pawn Odds. Who cares, and you are NOT such a great player either, I already beat my friend rated 2025 with a $100.00 bet after he gave me his f7 pawn, I will destroy your Ego.

You have serious psychic problems, but that is known already ;)

https://www.uschess.org/msa/MbrDtlTnmtHst.php?12537202

[dkappe]

Don’t feed the troll. Every time I see one of his threads with ten posts and no replies, I smile.