This project is a fork of the pachi.git project. If you have that one already cloned locally, you can use
git clone --reference /path/to/your/pachi.git/incarnation mirror_URL
to save bandwidth during cloning.
homepage URL
last changeSun, 12 Aug 2012 14:56:28 +0000 (12 16:56 +0200)
content tags
Pachi can refer to: a simple modular framework for programs playing
the game of Go/Weiqi/Baduk, and a reasonably strong engine built
within this framework.


To build Pachi, simply type:


The resulting binary program `pachi` is a GTP client; connect to it
with your favorite Go program interface (e.g. gogui or qgo), or use
kgsGtp to connect it to KGS.  (DO NOT make the GTP interface accessible
directly to untrusted users since the parser is not secure - see the
HACKING file for details.)

The pachi program can take many parameters, as well as the particular
engine being used; the defaults should be fine for initial usage,
see below for some more tips.

In case you hit compilation issues (e.g. when building on MacOS/X)
or want to change the build configuration, check the user configurable
section at the top of the Makefile.


The default engine plays by Chinese rules and should be about
7d KGS strength on 9x9. On 19x19, it might be about KGS 1k, assuming
reasonable hardware, e.g. two-core Athlon64 machine.  On a higher-end
(e.g. six-way Intel i7) machine, it can hold a solid KGS 2d rank.
When using a large cluster (64 machines, 20 cores each), it maintains
KGS 4d and has won e.g. a 7-stone handicap game against Zhou Junxun 9p.

By default, Pachi currently uses the UCT engine that combines
Monte Carlo approach with tree search; UCB1AMAF tree policy using
the RAVE method is used for tree search, while the Moggy playout
policy using 3x3 patterns and various tactical checks is used for
the semi-random Monte Carlo playouts.  Large-scale board patterns
are used in the tree search.

At the same time, we keep trying a wide variety of other approaches
and enhancements. Pachi is an active research platform and quite a few
improvements have been already achieved. We rigorously play-test new
features and enable them by default only when they give a universal
strength boost.

How to run

By default, Pachi will run on a single CPU core, taking up to 1.4GiB
of memory and taking a little under 15 seconds per move.  You can
adjust these parameters by passing it extra command line options:

	./pachi -t _1200 threads=8,max_tree_size=3072

This will make Pachi play with time settings 20:00 S.D. with 8 threads,
taking up to 3GiB of memory (+ several tens MiB as a constant overhead)
and thinking during the opponent's turn as well.

Pachi can use an opening book in a Fuego-compatible format - you can
obtain one at and use it in Pachi
with the -f parameter:

	./pachi -f book.dat ...

You may wish to append some custom Pachi opening book lines to book.dat;
take them from the book.dat.extra file. If using the default Fuego book,
you may want to remove the lines listed in book.dat.bad.

Pachi can also use a pattern database to improve its playing performance.
You can get it at - you will also find further
instructions there.

For now, there is no comprehensive documentation of options, but you
can get a pretty good idea by looking at the uct_state_init() function
in uct/uct.c - you will find the list of UCT engine options there, each
with a description. At any rate, usually the three options above are
the only ones you really want to tweak.

Greedy Pachi

Normally, Pachi cares only for win or loss and does not take into
account the point amount. This means that it will play slack endgame
when winning and crazy moves followed with a resign when losing.

It may give you a more pleasurable playing experience if Pachi
_does_ take into account the point size, strives for a maximum
(reasonable) win margin when winning and minimal point loss when
losing. This is possible by using the maximize_score parameter, e.g.:

	./pachi -t _1200 threads=8,maximize_score

This enables an aggressive dynamic komi usage and end result margin
is included in node values aside of winrate. Pachi will also enter
scoring even when losing (normally, Pachi will never pass in that case).
Note that if you pass any 'dynkomi' parameter to Pachi, you will reset
the values set by 'maximize_score'.

Note that Pachi in this mode may be slightly weaker, and result margin
should not be taken into account when judging either player's strength.
During the game, the winning/losing margin can be approximated from
Pachi's "extra komi" or "xkomi" reporting in the progress messages.

Experiments and Testing

Except UCT, Pachi supports a simple idiotbot-like engine and an example
treeless MonteCarlo-player. The MonteCarlo simulation ("playout")
policies are also pluggable, by default we use the one that makes use of
heavy domain knowledge.

Other special engines are also provided:
* a "distributed" engine for cluster play; the description at the top of
  distributed/distributed.c should provide all the guidance
* a simple "replay" engine that will simply play moves according
  to the playout policy suggestions
* a simple "patternplay" engine that will play moves according to the
  learned patterns
* few other purely for development usage

Pachi can be used as a test opponent for development of other go-playing
programs. For example, to get the "plainest UCT" player, use:

	./pachi -t =5000 policy=ucb1,playout=light,prior=eqex=0,dynkomi=none

This will fix the number of playouts per move to 5000, switch the node
selection policy from ucb1amaf to ucb1 (i.e. disable RAVE), switch the
playouts from heuristic-heavy moggy to uniformly random light, stop
prioring the node values heuristically, and turn off dynamic komi.

You can of course selectively re-enable various features or tweak this
further. But please note that using Pachi in this mode is not tested
extensively, so check its performance in whatever version you test
before you use it as a reference.

Note that even in this "basic UCT" mode, Pachi optimizes tree search
by considering board symmetries at the beginning. Currently, there's no
easy option to turn that off. The easiest way is to tweak board.c so
that board_symmetry_update() has goto break_symmetry at the beginning
and board_clear has board->symmetry.type = SYM_NONE.


Pachi can also help you analyze your games by being able to provide
its opinion on various positions. The user interface is very rudimentary,
but the ability is certainly there.

There are currently several Pachi interfaces provided for this purpose.

Winrate Development

Pachi can evaluate all moves within a given game and show how
the winrates for both players evolved - i.e. who was winning at which
game stage. This is implemented using the `tools/` script.
See the comment on top of the script about its usage.

Move Ranking

Pachi can evaluate all available moves in a given situation
and for each give a value between 0 and 1 representing perceived
likelihood of winning the game if one would play that move. I.e. it can
suggest which moves would be good and bad in a single given situation.

To achieve the latter, note the number of move at the situation you
want to evaluate and run the `tools/` script.
See the comment on top of the script about its usage.

Pattern Move Hinting

Pachi can show instantenous pattern-based move suggestions very much
like for example Moyo Go Studio (though of course without a GUI).
You can use the Move Ranking method above (tools/,
but pass it an extra parameter '-e patternplay'.


The aim of the software framework is to make it easy to plug your
engine to the common infrastructure and implement your ideas while
minimalizing the overhead of implementing the GTP, speed-optimized
board implementation, etc.  Also, there are premade random playout
and UCT tree engines, so that you can directly tweak only particular
policies.  The infrastructure is pretty fast and it should be quite
easy for you (or us) to extend it to provide more facilities for
your engine.

See the HACKING file for a more detailed developer's view of Pachi.

Also, if you are interested about Pachi's architecture, algorithms
etc., consider taking a look at Petr Baudis' Master's Thesis:


Pachi is distributed under the GPLv2 licence (see the COPYING file for
details and full text of the licence); you are welcome to tweak it as
you wish (contributing back upstream is welcome) and distribute
it freely, but only together with the source code. You are welcome
to make private modifications to the code (e.g. try new algorithms and
approaches), use them internally or even to have your bot play on the
internet and enter competitions, but as soon as you want to release it
to the public, you need to release the source code as well.

One exception is the Autotest framework, which is licenced under the
terms of the MIT licence (close to public domain) - you are free to
use it any way you wish.
2012-08-12 Petr BaudisChange default time settings from -t =80000 to -t 15master
2012-08-12 Petr BaudisUCT pondering: Enable by default
2012-08-12 Petr BaudisUCT max_maintime_ratio: Raise back to 2.0
2012-08-12 Petr BaudisUCT virtual_loss: Revert default change
2012-08-12 Petr BaudisUCT virtual_loss: Allow setting virtual_loss to arbitra...
2012-08-12 Petr Baudisuct_leaf_node(): Fix node value debug print
2012-08-12 Petr Baudisfbook_init(): Remove bogus comment
2012-08-12 Jean-loup GaillyNew parameter defaults, obtained by tuning on large...
2012-08-09 Petr Baudischat_init(): Fix problematic ptrdiff_t format string...
2012-08-08 Neil McleanAmmend to static inline, for no gcc warning.
2012-08-08 Neil McleanGCC: -Wextra warning: ‘static’ is not at beginning...
2012-08-08 Neil McleanChange type of mercymin to 'int' to avoid gcc sign...
2012-08-08 Neil McleanWin32 gcc compiler warning fix. Added util.h prototype...
2012-08-06 Petr BaudisFix stack overflows and crashes when reading very long...
2012-08-05 Petr BaudisUCT initial_extra_komi: Way to seed adaptive dynkomi...
2012-08-01 Petr BaudisUCT spawn_thread_manager(): Make sure each thread gets...
4 years ago pachi-9.01-hakugen
4 years ago pachi-9.00-hakugen
4 years ago pachi-8.00-shuhaku
4 years ago pachi-7.00-chihaku
5 years ago pachi-6.00-dochi
5 years ago pachi-5.00-dosaku
6 years ago pachi-4.00-doetsu
6 years ago pachi-3.00-sanetsu
6 years ago pachi-2.00-sansa
3 years ago gtp
3 years ago gcc
3 years ago master
3 years ago gogui
3 years ago mob
3 years ago jmhack
3 years ago ladders
3 years ago moggypatterns
3 years ago win32
3 years ago patterns
3 years ago falseeye
3 years ago rosin
4 years ago libmap
4 years ago rldecay
4 years ago pachi-9-hakugen
4 years ago greedy2