Starcraft Brood War Competition

broken image


Starcraft Brood War Competition
  1. Computer

Rules

2016 Rules(※Base Rules come from AIIDE StarCraft AI Competition Rules)

Tournament Format

(Updated in July 13th) Chosen maps will be announced after submission deadline.
(Updated in July 13th) We will announce the list (not complete, will be updated) of AI bots resubmitted from the last year competitions without change at July 13th 0:00 A.M (UTC +9:00)

(Updated in July 13th) Last year bots resubmitted without change

(Updated in July 31st) 'Overkill' added in last year bot entries


(Updated
in August 2nd) Final Entry in this year.
(Updated in August 1st) 17 Entries submitted. However, 'bin heo ' and 'Marcus Lanvers' don't submit bot yet. So, we will use 8 seed number except Heo's and Lanvers's number(10, 1337) for selecting maps.

Random Seed

(Added in June 29th) We will get 'Seed Number' from each participant at Submission Questionnaire. Accumulated Seed Number will be used as 'Random Seed' to select maps.

Map (※IMPORTANT)

(Updated in July 13th) 5 maps will be selected randomly from the Map-pack and used in the competition.
(Added in June 27th) '(2)Neo Heartbreaker Ridge2.0.scx' - File name is too long to be recognized by BW, we'll use this map after renaming to '(2)Neo Heartbreaker Ridge.scx'.
(Updated in July 13th) We will randomly select one 2-players map, two 3-players maps and two 4-players maps.
(Added in June 30th) We will use Java Standard Random Generator to choose maps.
(Added in August 1st) Through total seed number and standard java random generator,
'(2)RideofValkyries1.0,(3)NeoAztec2.1,(3)Plasma1.0, (4)ArcadiaII2.02, (4)Andromeda1.0', choosed in this year.
(
Updated in August 2nd)
'(2)RideofValkyries1.0, (3)Alchemist1.0, (3)TauCross1.1, (4)LunaTheFinal2.3, (4)Python1.3'through XOR combinations. https://jjpf.over-blog.com/2020/12/cloudmounter-cloud-encryption-3-2.html.


Game Type

The only game type for the competition will be 1 vs. 1 full game StarCraft: BroodWar 1.16.1 with fog of war enabled.

Time Limit

Games will have a 'frame limit' of 86400 frames, to simulate one hour of gameplay. If a game goes this long, it will be stopped and the in-game score will be used to determine the winner.

Tournament Structure (Updated in July 13th)

  • This year competition is in two stages.
  • The 1st stage will select top half from the all entries. The results will not be used in the 2nd stage.
  • In the 2nd stage, only the top half entries will compete to produce the final rank.
  • The final ranking will be determined just based on the 2nd stage results.
  • In stage 1 of the competition, all entries will participate in a round-robin tournament. Top half will advance to the second stage (for example, 3 best entries from 5 submissions). In stage 2 of the competition, the top half will compete in a round-robin tournament. The final rank will be given based on the stage 2 results.
  • The file I/O will be deleted between the 1st and 2nd stages. The results of the 1st stage will not used in the 2nd stage competition ranking.
  • The two stages will use a round-robin tournament style with as many as possible.
  • The two stages will be set as the same number of rounds.


Tournament Rules

Persistent File I/O

File I/O will work the same way as AIIDE StarCraft AI Competition Rules. You will have read access to folder 'bwapi-data/read/' and write access to folder 'bwapi-data/write/', both of which will be in the standard location under the StarCraft root directory.

IMPORTANT: File I/O works as follows:

  1. Before each game, the contents of the server-side read directory for your bot are copied to the client machine under 'bwapi-data/read/'. For the first round of the tournament this directory will be blank since a full round has not yet been played.
  2. During a game, you have write access to the 'bwapi-data/write/' folder and read access to the 'bwapi-data/read/' folder on the client you are currently playing on
  3. After the game ends, the contents of the 'bwapi-data/read/' folder are deleted, and the contents of the 'bwapi-data/write/' folder are sent to the server and stored in your bot's server-side write folder.
  4. WARNING: Step 3 may overwrite previous results if you are not careful in specifying unique filenames for each opponent that you play against.
  5. After each bot has played each other bot on the current map (one round robin on that map) the contents of each bot's server-side write folder are copied into the respective server-side read folder and the write folder is cleared.
  6. WARNING: Step 5 will over-write previous round data inside the server-side read folder. Your file output must either be cumulative, or have a filename scheme such that no overwrites will happen in order for your data to be preserved.
  7. The bot's server-side read folder is NEVER deleted throughout the tournament
  8. The next round starts, and from step 1 you will have access to your server-side read data from the previous rounds inside the client's 'bwapi-data/read/' folder

Again, due to the nature of overwriting the read folder with the contents of the write folder, your file names should be unique at least to the current opponent. Two popular choices for file naming schemes are:

  1. Results_EnemyName.txt - Inside this file you store all data about your matches against bot EnemyName. This file will be overwritten every round on the server, so make sure to read its contents each game, and write out the cumulative results into the write folder so you do not lose data after a round
  2. For example, you can store one line per map played inside this file so you have something like 'EnemyName MapName Wins Losses ExtraData', then if you win a game on that map against that enemy, you read in the contents and then output 'EnemyName MapName Wins+1 Losses ExtraData+NewData'
  3. Results_EnemyName_MapName_TimeStamp.txt - If you don't want to have a single file per enemy storing cumulative results, you can instead have a single file per game to store results with a unique time stamp or random number suffix. Please not that this is much slower, as your bot will play thousands of games and then have to process each file in the directory each time. Any time-out due to file processing is still a timeout and you will get a game loss. I strongly recommend the first option.
  4. Any reading or writing of other directories is forbidden. You have a 1GB limit for all files in the read directory. If you go over this, it will be deleted.

Any reading or writing of other directories is forbidden. You have a 1gb limit for all files in the read directory. If you go over this, it will be deleted.

BWAPI Version

The following BWAPI release versions will be supported this year:

Versions 3.7.4, 4.0.1 Beta, 4.1.0 Beta, 4.1.2.

Software Requirements

Open-Source

All source code and data files needed to compile and run the competition entries will be made available to the public as free software with a license that is compatible with the GNU public license.
We received a variety of opinions on this matter and decided, in the end, that to foster research it is best to have the next generation of programmers stand on the shoulders of giants, rather than re-invent the wheel. Please note, that program authors retain the copyright of their code. Also, releasing code as free software will allow new authors to take code from the previous competition and improve it. This has worked well in the past in the International Planning and ORTS competitions.

Game Speed

All games will be played at setLocalSpeed(0) (fastest setting). Please make sure that your bot functions correctly on this setting. To avoid problems, base your 'timings' on unit counts or currentFrame() and not machine time.

Software Time-Out

Make sure that each onframe call does not run longer than 42ms. Entries that slow down games by repeatedly exceeding this time limit will lose games on time. In particular:
In a match a bot LOSES (immediately) on time iff
>= 2 frames exceed 10 seconds, or
>= 10 frames exceed 1 second, or
>= 200 frames exceed 55ms

Code Obfuscation

We reserve the right to reject entries if we feel the authors attempt to obfuscate their code (like using silly or deceiving variable names or encoding their algorithm in a 10MB FSM transition matrix stored on disk). Authors who are concerned about the source code leaking information that has not been published yet, are encouraged to submit a technical report to their institutions that describes their entries prior to submitting the code.
To simplify this process we ask authors to only use standard development tools, that can be invoked via shell commands (such as Visual C++). In case remote access is not feasible, we will need detailed instructions for building each DLL.

Conflict of Interest

Organizing AND participating in events like this can be tricky, in particular if some choices, like which maps we use, are made 'randomly', and we get access to competitors source code prior to running the tournament.


Oblivion elder scrolls logo.

We address these problems as follows:

  1. For choosing maps we will use the standard Java pseudo random number generator that will be seeded by the XOR combination of 'random numbers' all program authors provide when they submit their entry. Our random number is (TBA).
  2. (Updated in July 19th) Before accepting zip files, we will distribute the MD5 SHA-1 check sums of our zip files to all participants. This way, participants can be confident that we don't change our entry after seeing their source code.
  3. (Added in July 20th) Until the deadline (31st July), we'll not open source code or use execution files of all entries submitted. We'll start to check the files after the deadline.

Detailed rules

  • No entry fee
  • Maybe some prizes
  • All replay files and source code will be made publicly available after the contest
  • StarCraft Brood War version 1.16.1 will be used for all games
  • StarCraft is a trademark of Blizzard Entertainment
  • Blizzard will not held liable for any damage caused
  • We will not be held liable for damage caused. This includes getting your CD key banned on Battle.net due to running a 3rd party StarCraft launcher
  • Entries must use the version of the Brood War API provided on this site
  • Bots that perform malicious behavior will be disqualified and banned from all future contests. This includes but is not limited to:
    1. Intentionally crashing StarCraft
    2. Installing worms/viruses/malware on the host machine
    3. Malicious utilization of resources such as sockets, files, zombie processes (Using 100% RAM and 100% CPU is permitted)
    4. Spamming the in-game console
  • Programs that attempt to cheat will be disqualified. Bots must disable the perfect information flag in the full-game tournament
  • Games in which an agent crashes StarCraft will be counted as a loss
  • Bots will be run in a native Windows XP SP3 environment.
    1. Entries must be tested on native Windows XP SP3 before submission
    2. We will not be held responsible for entries that crash StarCraft due to our configuration of the system
  • Games will be run in a semi-automated fashion.
  • Internet access is forbidden, bots will be run on a LAN
  • Only 1 entry is permitted per affiliation. Different bots can be submitted for each of the tournaments.
  • Bots must not slow down the game speed
    1. Games will be run on the FASTEST setting, which is 24 fps
    2. See the details on writing a remote process for your AI here
    3. Games in which a bot slows down the game speed significantly will be counted as a loss
  • Participants are required to disclose all source code and to provide a short description of implementation details.
  • Entries are only allowed to make use of free software components - including BWTA and BWSAL
  • Entries are allowed to utilize a single socket to communicate with a remote process
    1. The remote process must be a 1-click to run application (e.g. we are not going to install a LISP compiler on the host machine to run a bot)
    2. Additional details on writing a remote process are available here
  • Bots are not allowed to pause the game, games in which a bot pauses the game will be counted as a loss
  • The following StarCraft bugs/tricks are permitted:
    1. Plague on interceptor
    2. Units pressed through
    3. Drops to defuse mines
    4. Mineral walk
    5. Manner Pylon
    6. Lurker hold position
    7. Observer over turret
    8. Stacking air units
  • All other bugs/exploits are forbidden. Bots caught attempted these exploits will be disqualified. This includes but is not limit to:
    1. Flying drones and templars
    2. Terran sliding buildings
    3. Stacking ground units
    4. Allied mines
    5. Gas walk, to get through blocked entrances or ramps
  • Bad mannered in game behavior is discouraged, but not forbidden

Welcome to the home of the annual IEEE CoG StarCraft AI competition which is organized by the Cognition & Intelligence Lab (CILab) at GIST and Sejong University, Seoul, Korea. It is sponsored by the IEEE Conference on Games (CoG).

Starcraft brood war map editor
  1. Computer

Rules

2016 Rules(※Base Rules come from AIIDE StarCraft AI Competition Rules)

Tournament Format

(Updated in July 13th) Chosen maps will be announced after submission deadline.
(Updated in July 13th) We will announce the list (not complete, will be updated) of AI bots resubmitted from the last year competitions without change at July 13th 0:00 A.M (UTC +9:00)

(Updated in July 13th) Last year bots resubmitted without change

(Updated in July 31st) 'Overkill' added in last year bot entries


(Updated
in August 2nd) Final Entry in this year.
(Updated in August 1st) 17 Entries submitted. However, 'bin heo ' and 'Marcus Lanvers' don't submit bot yet. So, we will use 8 seed number except Heo's and Lanvers's number(10, 1337) for selecting maps.

Random Seed

(Added in June 29th) We will get 'Seed Number' from each participant at Submission Questionnaire. Accumulated Seed Number will be used as 'Random Seed' to select maps.

Map (※IMPORTANT)

(Updated in July 13th) 5 maps will be selected randomly from the Map-pack and used in the competition.
(Added in June 27th) '(2)Neo Heartbreaker Ridge2.0.scx' - File name is too long to be recognized by BW, we'll use this map after renaming to '(2)Neo Heartbreaker Ridge.scx'.
(Updated in July 13th) We will randomly select one 2-players map, two 3-players maps and two 4-players maps.
(Added in June 30th) We will use Java Standard Random Generator to choose maps.
(Added in August 1st) Through total seed number and standard java random generator,
'(2)RideofValkyries1.0,(3)NeoAztec2.1,(3)Plasma1.0, (4)ArcadiaII2.02, (4)Andromeda1.0', choosed in this year.
(
Updated in August 2nd)
'(2)RideofValkyries1.0, (3)Alchemist1.0, (3)TauCross1.1, (4)LunaTheFinal2.3, (4)Python1.3'through XOR combinations. https://jjpf.over-blog.com/2020/12/cloudmounter-cloud-encryption-3-2.html.


Game Type

The only game type for the competition will be 1 vs. 1 full game StarCraft: BroodWar 1.16.1 with fog of war enabled.

Time Limit

Games will have a 'frame limit' of 86400 frames, to simulate one hour of gameplay. If a game goes this long, it will be stopped and the in-game score will be used to determine the winner.

Tournament Structure (Updated in July 13th)

  • This year competition is in two stages.
  • The 1st stage will select top half from the all entries. The results will not be used in the 2nd stage.
  • In the 2nd stage, only the top half entries will compete to produce the final rank.
  • The final ranking will be determined just based on the 2nd stage results.
  • In stage 1 of the competition, all entries will participate in a round-robin tournament. Top half will advance to the second stage (for example, 3 best entries from 5 submissions). In stage 2 of the competition, the top half will compete in a round-robin tournament. The final rank will be given based on the stage 2 results.
  • The file I/O will be deleted between the 1st and 2nd stages. The results of the 1st stage will not used in the 2nd stage competition ranking.
  • The two stages will use a round-robin tournament style with as many as possible.
  • The two stages will be set as the same number of rounds.


Tournament Rules

Persistent File I/O

File I/O will work the same way as AIIDE StarCraft AI Competition Rules. You will have read access to folder 'bwapi-data/read/' and write access to folder 'bwapi-data/write/', both of which will be in the standard location under the StarCraft root directory.

IMPORTANT: File I/O works as follows:

  1. Before each game, the contents of the server-side read directory for your bot are copied to the client machine under 'bwapi-data/read/'. For the first round of the tournament this directory will be blank since a full round has not yet been played.
  2. During a game, you have write access to the 'bwapi-data/write/' folder and read access to the 'bwapi-data/read/' folder on the client you are currently playing on
  3. After the game ends, the contents of the 'bwapi-data/read/' folder are deleted, and the contents of the 'bwapi-data/write/' folder are sent to the server and stored in your bot's server-side write folder.
  4. WARNING: Step 3 may overwrite previous results if you are not careful in specifying unique filenames for each opponent that you play against.
  5. After each bot has played each other bot on the current map (one round robin on that map) the contents of each bot's server-side write folder are copied into the respective server-side read folder and the write folder is cleared.
  6. WARNING: Step 5 will over-write previous round data inside the server-side read folder. Your file output must either be cumulative, or have a filename scheme such that no overwrites will happen in order for your data to be preserved.
  7. The bot's server-side read folder is NEVER deleted throughout the tournament
  8. The next round starts, and from step 1 you will have access to your server-side read data from the previous rounds inside the client's 'bwapi-data/read/' folder

Again, due to the nature of overwriting the read folder with the contents of the write folder, your file names should be unique at least to the current opponent. Two popular choices for file naming schemes are:

  1. Results_EnemyName.txt - Inside this file you store all data about your matches against bot EnemyName. This file will be overwritten every round on the server, so make sure to read its contents each game, and write out the cumulative results into the write folder so you do not lose data after a round
  2. For example, you can store one line per map played inside this file so you have something like 'EnemyName MapName Wins Losses ExtraData', then if you win a game on that map against that enemy, you read in the contents and then output 'EnemyName MapName Wins+1 Losses ExtraData+NewData'
  3. Results_EnemyName_MapName_TimeStamp.txt - If you don't want to have a single file per enemy storing cumulative results, you can instead have a single file per game to store results with a unique time stamp or random number suffix. Please not that this is much slower, as your bot will play thousands of games and then have to process each file in the directory each time. Any time-out due to file processing is still a timeout and you will get a game loss. I strongly recommend the first option.
  4. Any reading or writing of other directories is forbidden. You have a 1GB limit for all files in the read directory. If you go over this, it will be deleted.

Any reading or writing of other directories is forbidden. You have a 1gb limit for all files in the read directory. If you go over this, it will be deleted.

BWAPI Version

The following BWAPI release versions will be supported this year:

Versions 3.7.4, 4.0.1 Beta, 4.1.0 Beta, 4.1.2.

Software Requirements

Open-Source

All source code and data files needed to compile and run the competition entries will be made available to the public as free software with a license that is compatible with the GNU public license.
We received a variety of opinions on this matter and decided, in the end, that to foster research it is best to have the next generation of programmers stand on the shoulders of giants, rather than re-invent the wheel. Please note, that program authors retain the copyright of their code. Also, releasing code as free software will allow new authors to take code from the previous competition and improve it. This has worked well in the past in the International Planning and ORTS competitions.

Game Speed

All games will be played at setLocalSpeed(0) (fastest setting). Please make sure that your bot functions correctly on this setting. To avoid problems, base your 'timings' on unit counts or currentFrame() and not machine time.

Software Time-Out

Make sure that each onframe call does not run longer than 42ms. Entries that slow down games by repeatedly exceeding this time limit will lose games on time. In particular:
In a match a bot LOSES (immediately) on time iff
>= 2 frames exceed 10 seconds, or
>= 10 frames exceed 1 second, or
>= 200 frames exceed 55ms

Code Obfuscation

We reserve the right to reject entries if we feel the authors attempt to obfuscate their code (like using silly or deceiving variable names or encoding their algorithm in a 10MB FSM transition matrix stored on disk). Authors who are concerned about the source code leaking information that has not been published yet, are encouraged to submit a technical report to their institutions that describes their entries prior to submitting the code.
To simplify this process we ask authors to only use standard development tools, that can be invoked via shell commands (such as Visual C++). In case remote access is not feasible, we will need detailed instructions for building each DLL.

Conflict of Interest

Organizing AND participating in events like this can be tricky, in particular if some choices, like which maps we use, are made 'randomly', and we get access to competitors source code prior to running the tournament.


Oblivion elder scrolls logo.

We address these problems as follows:

  1. For choosing maps we will use the standard Java pseudo random number generator that will be seeded by the XOR combination of 'random numbers' all program authors provide when they submit their entry. Our random number is (TBA).
  2. (Updated in July 19th) Before accepting zip files, we will distribute the MD5 SHA-1 check sums of our zip files to all participants. This way, participants can be confident that we don't change our entry after seeing their source code.
  3. (Added in July 20th) Until the deadline (31st July), we'll not open source code or use execution files of all entries submitted. We'll start to check the files after the deadline.

Detailed rules

  • No entry fee
  • Maybe some prizes
  • All replay files and source code will be made publicly available after the contest
  • StarCraft Brood War version 1.16.1 will be used for all games
  • StarCraft is a trademark of Blizzard Entertainment
  • Blizzard will not held liable for any damage caused
  • We will not be held liable for damage caused. This includes getting your CD key banned on Battle.net due to running a 3rd party StarCraft launcher
  • Entries must use the version of the Brood War API provided on this site
  • Bots that perform malicious behavior will be disqualified and banned from all future contests. This includes but is not limited to:
    1. Intentionally crashing StarCraft
    2. Installing worms/viruses/malware on the host machine
    3. Malicious utilization of resources such as sockets, files, zombie processes (Using 100% RAM and 100% CPU is permitted)
    4. Spamming the in-game console
  • Programs that attempt to cheat will be disqualified. Bots must disable the perfect information flag in the full-game tournament
  • Games in which an agent crashes StarCraft will be counted as a loss
  • Bots will be run in a native Windows XP SP3 environment.
    1. Entries must be tested on native Windows XP SP3 before submission
    2. We will not be held responsible for entries that crash StarCraft due to our configuration of the system
  • Games will be run in a semi-automated fashion.
  • Internet access is forbidden, bots will be run on a LAN
  • Only 1 entry is permitted per affiliation. Different bots can be submitted for each of the tournaments.
  • Bots must not slow down the game speed
    1. Games will be run on the FASTEST setting, which is 24 fps
    2. See the details on writing a remote process for your AI here
    3. Games in which a bot slows down the game speed significantly will be counted as a loss
  • Participants are required to disclose all source code and to provide a short description of implementation details.
  • Entries are only allowed to make use of free software components - including BWTA and BWSAL
  • Entries are allowed to utilize a single socket to communicate with a remote process
    1. The remote process must be a 1-click to run application (e.g. we are not going to install a LISP compiler on the host machine to run a bot)
    2. Additional details on writing a remote process are available here
  • Bots are not allowed to pause the game, games in which a bot pauses the game will be counted as a loss
  • The following StarCraft bugs/tricks are permitted:
    1. Plague on interceptor
    2. Units pressed through
    3. Drops to defuse mines
    4. Mineral walk
    5. Manner Pylon
    6. Lurker hold position
    7. Observer over turret
    8. Stacking air units
  • All other bugs/exploits are forbidden. Bots caught attempted these exploits will be disqualified. This includes but is not limit to:
    1. Flying drones and templars
    2. Terran sliding buildings
    3. Stacking ground units
    4. Allied mines
    5. Gas walk, to get through blocked entrances or ramps
  • Bad mannered in game behavior is discouraged, but not forbidden

Welcome to the home of the annual IEEE CoG StarCraft AI competition which is organized by the Cognition & Intelligence Lab (CILab) at GIST and Sejong University, Seoul, Korea. It is sponsored by the IEEE Conference on Games (CoG).

During this competition, programs ('bots') will play 1v1 Starcraft Broodwar games against each other using BWAPI, a software library that makes it possible to connect programs to the Starcraft: BroodWar game engine. Cocktail 10 3 1 download free.

The purpose of this competition is to foster the development & evaluation of progress in AI development applied to real-time strategy (RTS) games and solve challenging issues in RTS game AI such as uncertainty, real-time processing, and managing & coordinating agents. Where feasible, the competition strives for openness, transparency, reusability and reproducibility, both in the way the competition rules are defined and evaluated, and in the bots themselves.

RTS games pose a much greater challenge for AI research than chess because of hidden information, vast state and action spaces, and the requirement to act quickly. The best human players still have the upper hand in RTS games, but in the years to come this will likely change, thanks to competitions like this one. IEEE CoG StarCraft competitions have shown significant progress in the development and evolution of new StarCraft bots. For the evolution of the bots, participants have used various approaches to write AI bots and it has enriched game AI and methods such as HMM, Bayesian model, CBR, Potential fields, and reinforcement learning. However, it is still quite challenging to develop AI for the game because it should handle many units and buildings while considering resource management and high-level tactics.

StarCraft and Brood War are now completely free, get first patch in 8 years By Wes Fenlon 18 April 2017 Patch 1.18 brings StarCraft up to speed with modern systems. The 2005 Sky ProLeague Grand Final between SKT T1 and KTF StarCraft has an active professional competition circuit, particularly in South Korea. The two major game channels in South Korea, Ongamenet and MBCGame, each run a Starleague, viewed by millions of fans. StarCraft: Brood War. EffOrt 226 sas.Sziky 86 Oya187 77 Iris 68 KwarK 43 Hmarnc 12. Counter-Strike: Global Offensive. Fl0m 2080 byalli 1217 rGuardiaN 955.

Getting Started

The competition will use StarCraft Brood War 1.16.1. You must possess a legal copy of the game.

Macbook windows right click. Bots for StarCraft shall use the Broodwar API, which provides hooks into StarCraft and enables the development of custom AI for StarCraft. Interfaces in many languages are available to query the current state of the game and issue orders to units.

The following material is a bit outdated, as it was prepared by Ben Weber for the AIIDE StarCraft competition in 2010 (and not everything has been updated since then). However, it still provides a valuable starting point if you want to start designing and implementing a bot from scratch. Please note, that the referenced BWAPI versions are not used any more for our competition. (we go with three versions of BWAPI that is 3.7.4, 4.1.2, 4.2.0, and 4.4.0)

  • Instructions for setting up the environment are available Video, Download 'Hellow World!' Project
  • An introduction to the Broodwar API is available here
  • Instructions for building a bot that communicates with a remote process are available here.
  • A JNI (Java) bridge to the BWAPI by Ben Weber is available here

Additional Software

Computer

  • Sparcraft : an open source StarCraft combat simulater. It can be used to create standalone combat simulations or be imported into an existing BWAPI-based StarCraft bot to provide additional AI functionality.
  • Tournament Manager Software : This program helps to play a lot of games of StarCraft automatically.
  • BWSAL (BWAPI Standard Add-on Library) : This library aims to develop several add-ons for BWAPI, which are scour manager, worker mamanger, supply manager, etc.
  • BWTA2 (Brood War Terrain Analyzer): a fork of BWTA, an add-on for BWAPI that analyzes the map and computes the regions, chokepoints, and base locations.
  • BWEM (Brood War Easy Map): C++ library that analyzes Brood War's maps and provides relevant information such as areas, choke points an base locations.




broken image