It requires CNTK https://github.com/Microsoft/CNTK/releases
and it has a Visual Studio project file, so you can use visual studio to compile

Yes, it will be unfair allocation of CPU time (I have 4 core CPU too). And yes, statistically 24 games not so significant for this level of causality. But there isn't so important number of games, if You are randomizing the games very much. In such case You can get only statistical information about quality and number of used opening book, patterns and specific local position algorithms (I'm not guru nor in English, nor in Go, so it may be, that I give not correct terminology). And the main (for serious games with human) algorithm, that is based on simulations number and therefore is depending on time*performance settings in such sparring games isn't testing...
In one phrase: for given total number of simulations more adequate result is getting in less number of games.

And, Alex, why You are not sparring MoGo version 4.86?

I have tested leela_080_linux_x64. Go engines rate not changed (sparrings by GTP on one computer with a half of processor and memory limits for each engine):

League A
1. Leela

League B
2. Ray
3. Pachi
4. MoGo

Go engines non default compiling and start up parameters.
leela_080_linux_x64 start up parameters: -g -t 2 -b 0

Leela's developer has selected non standard way for time control. In contrast to all other Go engines in leela_080_linux_x64 it's possible get only the time per move and only by subtraction from external application time per move setting the --lagbuffer (-b) parameter value. For example, if someone adds in GoGUI a few Leela's game strength levels, then he must start gogui.jar (excepting displaying the contestants elapsed times) with "-time 1s+21474836s/1" (for example - the maximal compensable by Leela's -b parameter) and add (in Menu/ Program/ New Program), for example, the next (with path to executive file leela_080_linux_x64 in Working directory):
1) ./leela_080_linux_x64 -g -b 2147483500
2) ./leela_080_linux_x64 -g -b 2147471600
3) ./leela_080_linux_x64 -g -b 0
(1 - for 1 sec per move; 2 - for 2 min per move; 3 - for 357914 min per move)

Engines. Games.

On my computer Leela 0.8.0 is much stronger than Hiratuka 10.37B, appeared from several matches with twogtp. I have included in the GoGUI with the following command line,

C:\PPDownload\Hiratuka10_37B\gogui-twogtp.exe -black "Leela080.exe --gtp --noponder --threads 2"
-white "Hiratuka-19x19.exe -po 25000" -alternate

I don't have GPU on my AMD Athlon computer. Attached is one of the matches.

Using of this parameter isn't a good idea, because all gamers are thinking on their opponents turn...

With new version of Leela the rate is the same:
Leela 0.9.0 got stronger and, of course, remained the strongest Go engine (with GTP support). Is there any such engine, that can show at least slight resistance to Leela?

Go engines non default compiling and start up parameters.
leela_090_linux_x64 start up parameters: -g -t 2 -b 0 (with: gogui-twogtp.jar -verbose -komi 7.5 -time 1s+2/1)

Leela has new thinking process. For example, on GoGUI shell You can see: "Nodes: 248096, Win: 51.81% (MC:42.27%/VN:63.46%), PV: M8 M10 N9 N10 P9 P6",- where win percentage is now average of old Monte Carlo (MC) and some new VN algorithms. Leela's developer has selected non standard way for time control. In contrast to all other Go engines in Leela it's possible get only the time per move and only by subtraction from external application time per move setting the --lagbuffer (-b) parameter value. For example, if someone adds in GoGUI a few Leela's game strength levels, then he must start gogui.jar (excepting displaying the contestants elapsed times) with "-time 1s+21474836s/1" (for example - the maximal compensable by Leela's -b parameter) and add (in Menu/ Program/ New Program), for example, the next (with path to executive file leela_090_linux_x64 in Working directory):
1) ./leela_090_linux_x64 -g -b 2147483500
2) ./leela_090_linux_x64 -g -b 2147471600
3) ./leela_090_linux_x64 -g -b 0
(1 - for 1 sec per move; 2 - for 2 min per move; 3 - for 357914 min per move)

Engines. Games.

Hi all,

It is nice to read the old post here on these bots comparison.
I actually did similar things at end of 2016 and early 2017. I am curious how can you guys run linux version of MoGo?

There is web page for downloading mogo linux binary from the web page:
https://www.lri.fr/~teytaud/mogo.html

But actually the downloaded file "mogor" is not executable in Ubuntu 14.04 linux machine, it throw out error something like "can not find file ...."

Shoreline

This executable works here on Ubuntu 16.04. Even the windows binary works with wine.

Found the reason. This is 32 bits binary, does not work natively in 64 bits linux. By doing the compatible library installation, it trashed my virtual machine system.

Well, I will just have to give it up, not worth trying it out.

My system is also 64 Bit. Anyway, Mogo is quite similar to Fuego

Here is legacy inactive robot account at DGS using MoGo:
http://www.dragongoserver.net/userinfo.php?uid=44779

It needs 10 minutes per move set up in dual core computer and only achieved 8 kyu level at DGS by looking at the chart. Not sure your setup needs that kind of thinking time to be competitive. For fuego to achieve 5 kyu to 7 kyu at DGS, my redrobot at DGS only runs on a 10 seconds per move thinking time (default) in dual core linux computer. Here is redrobot at DGS:
http://www.dragongoserver.net/userinfo.php?uid=97563

It appears that on resource wise, and strength, MoGo is not very competitive compared to Fuego.

Try from here.
You must have x86_64 version of Linux (elsewhere You must compile MoGo 4.86).

MoGo (all versions) are stronger, than FueGo, and have a huge number of start up parameters (see link above).

See here games, including between MoGo and FueGo (all games are with the same PC resources given for each engine).

Thanks for providing MoGo X64 linux binary. Indeed, MoGo 4.86 in my machine can beat Fuego easily with similar thinking time and resource.

All MoGo versions with parameters by link above beat FueGo too...
Moreover, in games by this link You can see, that MoGo versions can't beat unambiguously each other (even 32-bit version)...

Leela in version 0.10.0 became stronger and, of course, remains the strongest Go engine (at least among engines with GTP support). Details.
But it still has two disadvantages:
1) the time control;
2) very late resigning (for example, it wasn't resign with win percentage by VN 0,00% until by MC was >20%...)

With new version of Leela the rate is the same:
Go engines non default compiling and start up parameters.
leela_0100_linux_x64 start up parameters: -g -t 2 -b 0 (with: gogui-twogtp.jar -verbose -komi 7.5 -time 1s+2/1)


Engines. Games.

Hi,

here is an Update with Leela 0.10.0, Mogo 4.86, Orego 7.08 and Matilda 1.24.

Leela 0.8.0 won 14-10 against Hiratuka, Leela 0.10.0 now wins by 16-0.

Leela vs. Hiratuka:


League A:


League B:


League C:


All engines are set up to play 1h/game with pondering off, one thread and Memory around 1GB.

Configuration:


Best,
Alex

Hi, Alex!

Can You explain, how is it possible to run Hiratuka in GTP mode (what a command is the 'taskset -c 0')?

Thanks in advance.

Hi,

I run the windows .exe under Linux. The -po parameter is the number of playouts. Hiratuka will use by default all available cores so I use the taskset command to force the core number, "taskset -c 0" means the program uses core number 0 (4 cores are numbered 0,1,2,3). You can also run two cores by "taskset -c 0,1". If you run more programs with taskset be sure to take different core numbers. There is something similar in windows afaik.

Best,
Alex