As in chess tournaments and matches...
-
OrgZ
- Posts: 944
- Joined: Thu Feb 16, 2023 2:15 pm
Post
by OrgZ » Mon Jul 29, 2024 6:34 pm
Code: Select all
TC:2m+0.20s
Suite: Ramz-Private-Suite
Score of Stockfish 240728_sse41 vs ABBY Dark Rose_sse41: 9 - 11 - 180 [0.495]
... Stockfish 240728_sse41 playing White: 8 - 2 - 90 [0.530] 100
... Stockfish 240728_sse41 playing Black: 1 - 9 - 90 [0.460] 100
... White vs Black: 17 - 3 - 180 [0.535] 200
Elo difference: -3.5 +/- 15.2, LOS: 32.7 %, DrawRatio: 90.0 %
200 of 200 games finished.
games:
https://pixeldrain.com/u/4EWQB3kP
-
OrgZ
- Posts: 944
- Joined: Thu Feb 16, 2023 2:15 pm
Post
by OrgZ » Tue Jul 30, 2024 4:38 am
Code: Select all
TC: 1m+1s
Score of caissa-1.20-x64-avx2 vs plentychess_300_dev_64_ja_bmi2: 6 - 2 - 42 [0.540]
... caissa-1.20-x64-avx2 playing White: 3 - 0 - 22 [0.560] 25
... caissa-1.20-x64-avx2 playing Black: 3 - 2 - 20 [0.520] 25
... White vs Black: 5 - 3 - 42 [0.520] 50
Elo difference: 27.9 +/- 38.1, LOS: 92.1 %, DrawRatio: 84.0 %
50 of 50 games finished.
-
Homayoun
- Posts: 1557
- Joined: Tue Mar 21, 2023 4:57 pm
- Real Name: Homayoun
Post
by Homayoun » Tue Jul 30, 2024 5:26 am
So, this Caissa is very strong.
-
OrgZ
- Posts: 944
- Joined: Thu Feb 16, 2023 2:15 pm
Post
by OrgZ » Tue Jul 30, 2024 6:00 pm
Homayoun wrote: ↑Tue Jul 30, 2024 5:26 am
So, this Caissa is very strong.
yup
-
Homayoun
- Posts: 1557
- Joined: Tue Mar 21, 2023 4:57 pm
- Real Name: Homayoun
Post
by Homayoun » Tue Jul 30, 2024 11:27 pm
I am waiting for new update of Alexandria, too. I guess the next version of Alexandria will be very good and reliable engine for analysis and we will see a revolution in it’s evaluation and move search and also NNUE structure. Let’s see what happens.
Best regards
-
OrgZ
- Posts: 944
- Joined: Thu Feb 16, 2023 2:15 pm
Post
by OrgZ » Wed Jul 31, 2024 1:26 pm
Homayoun wrote: ↑Tue Jul 30, 2024 11:27 pm
I am waiting for new update of Alexandria, too. I guess the next version of Alexandria will be very good and reliable engine for analysis and we will see a revolution in it’s evaluation and move search and also NNUE structure. Let’s see what happens.
Best regards
-
OrgZ
- Posts: 944
- Joined: Thu Feb 16, 2023 2:15 pm
Post
by OrgZ » Sat Aug 03, 2024 4:32 am
Plentychess is still stronger
Code: Select all
Score of Halogen12-windows-avx2 vs plentychess_300_dev_64_ja_bmi2: 0 - 12 - 13 [0.260]
... Halogen12-windows-avx2 playing White: 0 - 5 - 8 [0.308] 13
... Halogen12-windows-avx2 playing Black: 0 - 7 - 5 [0.208] 12
... White vs Black: 7 - 5 - 13 [0.540] 25
Elo difference: -181.7 +/- 91.8, LOS: 0.0 %, DrawRatio: 52.0 %
25 of 25 games finished.
-
OrgZ
- Posts: 944
- Joined: Thu Feb 16, 2023 2:15 pm
Post
by OrgZ » Sat Aug 03, 2024 4:56 am
Rebel 16.3
Code: Select all
Score of Halogen12-windows-avx2 vs Rebel-16.3: 0 - 7 - 18 [0.360]
... Halogen12-windows-avx2 playing White: 0 - 3 - 10 [0.385] 13
... Halogen12-windows-avx2 playing Black: 0 - 4 - 8 [0.333] 12
... White vs Black: 4 - 3 - 18 [0.520] 25
Elo difference: -100.0 +/- 67.3, LOS: 0.4 %, DrawRatio: 72.0 %
25 of 25 games finished.
-
Homayoun
- Posts: 1557
- Joined: Tue Mar 21, 2023 4:57 pm
- Real Name: Homayoun
Post
by Homayoun » Sat Aug 03, 2024 7:42 am
Thank you very much for the tests. These tests show the Halogen situation more clear. And the author claim of +180 Elo progress to Halogen 11 is not true and it is very optimistic.
Best regards