Discussion about chess-playing software (engines, hosts, opening books, platforms, etc...)
Hagen
Posts: 121 Joined: Mon Jun 14, 2010 12:30 am
Post
by Hagen » Thu Jul 08, 2010 7:58 am
Fascinating results...according to this...the "modification" makes Stockfish 1.8 TinaPA 1.01 stronger than Crab, Spoon and Tactical. What would be interesting to find out is whether this modification actually makes Stockfish increase its ELO rating.
Chris Formula
Posts: 9 Joined: Sat Jun 19, 2010 7:06 am
Real Name: Chris Formula
Post
by Chris Formula » Thu Jul 08, 2010 8:25 am
hi Swami,
thanks for the quick STS testing. i wish this "sts-point" improvement translates to "elo-point" improvement.
regards,
chris
Uly
Posts: 838 Joined: Thu Jun 10, 2010 5:33 am
Post
by Uly » Thu Jul 08, 2010 9:35 am
Any plans on a 64bit release?
Eduard Nemeth
Posts: 632 Joined: Thu Jun 10, 2010 12:50 am
Post
by Eduard Nemeth » Thu Jul 08, 2010 11:06 am
The aggressiveness is too high!
I played an short engine match vs. Houdini 1.02
1-cpu vs 1-cpu and ponder ON.
After the first two variants of Nunn1 suite
Houdini won 2 games and 2 games ended draw.
Much better is this setting (only my opinion):
Code: Select all
[OPTIONS]
Mobility (Middle Game)=80
Aggressiveness=80
After first 3 games with Nunn1 suite Tinapa 1.01 won now 1 game with 2 draws.
CPU=Core 2 Duo 2.66 GHz
GUI=Fritz 10
EGTB=GUI only
Hash=128 MB, 128 MB
Timecontrol=3m + 0s
Book=Nunn1 suite
Here are the first 3 games with this settings:
And now a nice win:
In my opinion "Mobility (Middle Game)=80" (default 100) and "Aggressiveness=80" (default 100)
are much better than the default setting.
ED.
Peace be with you!
Eduard Nemeth
Posts: 632 Joined: Thu Jun 10, 2010 12:50 am
Post
by Eduard Nemeth » Thu Jul 08, 2010 12:28 pm
After any tests (games against Houdini) I prefer now fallow settings:
Code: Select all
[OPTIONS]
Mobility (Middle Game)=80
Space=80
Aggressiveness=80
Cowardice=80
CPU=Core 2 Duo 2.66 GHz
GUI=Fritz 10
Ponder=ON
EGTB=GUI only
Hash=128 MB, 128 MB
Timecontrol=3m + 0s
Book=Nunn1 suite
Not bad game with this settings,
and btw Tinapa 1.01 used also too
a much better time management in this game:
Good played against one of the strongest Engines of the world!
ED.
Peace be with you!
Eduard Nemeth
Posts: 632 Joined: Thu Jun 10, 2010 12:50 am
Post
by Eduard Nemeth » Thu Jul 08, 2010 1:12 pm
DRAW ON PLAYCHESS AGAINST A FAST RYBKA 4!
(in my first game on playchess with Tinapa 1.01)
I have played with this settings, only with my core Duo 2.66 GHz and 256 MB Hash (with my account "Solist"):
Code: Select all
[OPTIONS]
Mobility (Middle Game)=80
Space=80
Aggressiveness=80
Cowardice=80
Deep Rybka 4 played on a 8 core mashine - wow!!
I think that my settings are not bad. Tinapa 1.01 played very solid, and found
for defending with black interesting moves...like 18...Kf8! Nice game!
ED.
Peace be with you!
mariaclara
Posts: 17 Joined: Sun Jun 13, 2010 9:14 am
Real Name: clare marie
Post
by mariaclara » Thu Jul 08, 2010 6:03 pm
OT; well, in a way yes, no.
Tinapa is a Filipino fish delicacy. It's delicious smoked fish.
benstoker
Posts: 110 Joined: Thu Jun 10, 2010 7:32 pm
Real Name: Ben Stoker
Post
by benstoker » Thu Jul 08, 2010 10:35 pm
Chris Formula wrote: a minor modification of Stockfish 1.8...
thanks to Tord, Marco and Joona for the well-commented top-notch source code/engine...
thanks also to Jim for the advice on compiling...
salamat,
chris
There is only one line change:
Code: Select all
// Null move dynamic reduction based on depth
- int R = 3 + (depth >= 5 * OnePly ? depth / 8 : 0);
+ int R = 3 + (depth >= 9 ? 1 : 0) + (depth >= 15 ? 1 : 0) + (depth >= 21 ? 1 : 0) + (depth >= 27 ? 1 : 0) + (depth >= 33 ? 1 : 0) + (depth >= 39 ? 1 : 0);
Why do you think this single line change re null move reduction merits changing the name of the program? A code fork merits a name change if and when there's a
significant modification of the code. This ain't anywhere close. Nevertheless, it's interesting.
BB+
Posts: 1484 Joined: Thu Jun 10, 2010 4:26 am
Post
by BB+ » Thu Jul 08, 2010 11:19 pm
OnePly is 2, I take it?
Code: Select all
8 910 1 2 3 4 5 6 7 8 920 1 2 3 4 5 6 7 8 930 1 2 3 4 5 6 7 8 940 1 2 3 4 5 6 7 8 950 1 2
Stockfish 3 3 4 4 4 4 4 4 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 9 9 9 9 9
Tinapa 3 4 4 4 4 4 4 5 5 5 5 5 5 6 6 6 6 6 6 7 7 7 7 7 7 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 9 9 9 9
There is only one line change:
I guess we know what Tinapa 1.00 looked like then.