Starcraft 2 is a program for playing against AI.

Release year: 2010
Genre:
Developer:
Publisher:
Developer site: http://us.blizzard.com/en-us/games/sc2/
Interface language: Russian
Platform: PC
System requirements:
Installed game Starcraft 2: Wings of Liberty version 1.0
Description:
StarCraft II is the continuation of the epic saga of three powerful races: Protoss, Terran and Zerg.
They will have to face each other again in the new real-time strategy sequel legendary game star craft.
In this fierce struggle for survival in space, you will have at your disposal new combat units,
as well as the previous ones, with advanced features.

Agria Valley
"Blistering Sands"
Burial Grounds
"Crossfire"
"Desert Oasis"
"Incineration Zone"
"Jungle Basin"
Junk Yard
Novice Blistering Sands
Novice Desert Oasis
Novice Steppes of War
Scrap Station
Steppes of War
"World Ship"
Xel Naga Caverns
"Elysium"
"Arid Wastes"
Debris Field
Delta Quadrant
Discord IV
"High Orbit"
"Kulas Ravine"
Lost Temple
"Metalopolis"
"Monlyth Ridge"
"New Antioch"
"Nightmare"
Novice Discord IV
Novice Kulas Ravine
Novice Metalopolis
Novice Monlyth Ridge
Novice Terminus
Novice Twilight Fortress
"Red Stone Gulch"
"Sacred Ground"
"Scorched Haven"
"Shakuras Plateau"
Tarsonis Assault
"Terminus"
"Twilight Fortress"
"War Zone"
Arakan Citadel
"Colony 426"
"Dig site"
"Dirt Side"
"Frontier"
"Moonsoon"
"Quicksand"
"Tectonic Rift"
"Bio Lab"
"Typhon"
"Ulaan Deeps"
"Abyss"
"Extinction"
"Forbidden Planet"
"Hight Ground"
"Lava flow"
"Megaton"
"outpost"
"Primeval"
Sand Canyon
"Tempest"
"Toxic Slums"

Copy maps to My Documents(Documents)StarCraft IIMaps
- If the "Maps" folder does not exist, then you need to create it.
- Go to the folder with the game and run the file "StarCraft II Editor" or "SC2Editor"
- In the program, select the map on which you want to play, then click on the "Map" tab from above,
- Select "Player Options" and starting position put Start location 1,2.
- Set AI difficulty to yourself (set in the map editor - File -
Editor Settings - Trial Run - Game Difficulty) and then click Trial Run or Ctrl + F9
- Don't forget to change control from "User" to "Computer".

renewing the most popular game StarCraft 2: Wings Of Liberty,
making changes and corrections to game process and troubleshooting the game itself.
Designed for the Russian version of the game.
With this file, you can update the game, without the need to go online,
for example, if you want to install it (the game) on a computer that does not have the ability to connect to the Internet.
List of changes:
update 1.0.3
The Friends tab on the Leagues & Rankings page now works correctly.
Fixed an issue that prevented some users from accessing single player without an internet connection.

update 1.0.2
Fixed mechanism for checking victory conditions in campaign missions.
Fixed an issue that prevented some users from accessing the single player.

update 1.0.1
Optimized the save process in campaign mode.
Fixed audio playback issue on some 7.1 systems.

The bot, created on the basis of a neural network, defeated a famous Dota 2 player on the stage of the main tournament The International (TI) 2017. The main Dota competitions take place between 5 vs 5 teams, but there is also an individual classification. Open AI created an AI that self-learned by playing with itself, 1v1. On stage, the bot beat a popular player 2-0 Daniil Dendi Ishutin from NaVi.

Before that, behind the scenes of the competition, the bot fought with SumaiL - the best 1v1 player - and with Arteezy - the best player in the world, his MMR is more than 10000. Their bot also won "dry" - 2-0.

Fun fact, after the tournament, everyone was invited to try their hand at the winning bot for rewards. By the evening of the same day, the bot was defeated by players from the Internet with an MMR of about 8000. It took several attempts to identify a weak spot. However, on the stage of the tournament, the bot won.

Why is this an important event?

The tournament is hosted by Valve, the developer of the game, with a prize pool of nearly $25 million. This biggest esports competition sets a new record for views and prize pool every year. International is watched online by half a million viewers on Twitch, several hundred thousand more through game client and millions on record.

Level 2. How does it work?

We go down a level deeper (or higher, how the imagination works). Let's figure it out!

The developers decided to approach the solution of the problem from scratch. They realized that if one tried to describe the number of rules in open game with thousands of variables, it will take several months. Instead, they set up a bot, put it into the game, and gave it free rein.

What evolution looks like:

1. The bot played against itself, the same bot. They exchanged experiences.
2. The first few games the bot stood still and did not move, after that it began to perform random actions or movements.
3. After thousands of attempts, the AI ​​came to the conclusion that you need to go to the center of the map and fight the enemy in order to win.
4. And after several weeks of continuous training, the bot began to play at the level of first-class players.

During training, AI learned:

• make the last blow - this brings bonuses to Dota;
• provoke small enemy creatures to attack;
• slow down your creatures to better group them;
• keep the enemy at a distance from their creatures, depriving the enemy of money and experience;
• cancel ability animations to avoid taking damage;
• bluff ability animation cancellation that leads to a defensive reaction of the enemy;
• adapt to new items that the enemy buys;
• pursue the enemy by predicting their possible location in the fog of war.

Visually, the process of training a neural network, look at the video:

How is a neural network different from playing against a computer?

Game against the computer

The main difference between the Open AI bot and the opponents built into the game lies in the technology on which they are based. The built-in opponents are based on a decision tree. Solutions are found by enumerating combinations of input parameters and solving optimization problems. Such decision trees include a fairly large number of parameters, so that the player will feel that he is playing against a living or adapting opponent.

Nevertheless, all possible variations in the behavior of such opponents are predetermined by the creators of the game. Embedded bots are clever at imitating intelligence, but they don't have it. Creating such artificial opponents is comparatively easier and faster for developers.

Playing against a neural network

Neural networks take a different approach. The network used by Open AI is based on a principle that is similar to evolution. By performing random actions, the network creates various solutions. Decisions are evaluated by a fitness function.

Successful solutions connect with each other, mutate and create new generations of solutions. New generations are compared with their "parents" to see if there has been progress. If there is no progress, the "parent" solutions are crossed in new proportions. The process is similar to the evolution of organisms in nature. Such an algorithm does not always lead to an ideal solution, but will be very close after passing through generations of errors.

In the video, the chief engineer of the project talks about the process of training the neural network when preparing the bot for the tournament:

A neural network-based bot requires significantly more resources than a classic decision tree. Such neural networks are used to design jet engines and find cures for Alzheimer's disease. Open AI used the computing power of the Microsoft Azure cloud server, since Microsoft is one of the founders of Open AI.

Musk's tweet after the bot won the tournament:

The whole process of training neural networks regarding the fitness function can be seen in the video about the Mario game. Using the method of neuroevolution of complementary topologies, or simply evolution by trial and error, the author of the video created a neural network that goes through the level in Mario.

At the beginning of training, such an AI cannot even walk, after 34 generations and several days of training, the AI ​​was able to finish the level without dying:

Is winning in Dota really that important? What do the experts say?

I ran through the forums of programmers, reddit, communities of lovers of neural networks. The opinions of the participants were divided: there were both enthusiasm and strong criticism.

admiration

Translation: The bot benefits by slowing down its creatures with the body of the character. It's awesome.

Translation: As I understand it, the test was to see how the bot learns to play on its own, without set rules and limits. He learned and won from the best. Understand correctly, this is a big progress.

Criticism

Translation: A bot that beats a professional in Dota is comparable to a robot that throws balls into a basketball basket from a place. This is not real basketball and not real Dota.

Translation: It sounds a lot less impressive when you realize that a bot, by definition, is equipped with the tools to have superhuman reactions and accurate knowledge of damage in numbers. This allows you to accurately calculate damage and deliver the final blow. It will be really interesting if they can teach the bot to play macro strategies that are not in the 1v1 game. If they can even win against a team of 5 people, this will be an achievement.

There were also doubters:

What is conspiracy theories?

In connection with the release of the API from Blizzard, a conspiracy version has appeared on the network. The tech giants didn't just suddenly focus on video games.

Seconds after the victory of the Open AI bot, Elon Musk tweeted that the bot organization, initiated by Elon himself, won a much more difficult game than chess or go:

The tweet is likely a boast to IBM and Google, who have mastered "easier" games. Let's not forget that Musk and his colleagues have invested more than a billion dollars in the non-profit research organization Open AI.

At a minimum, they will be able to attract the best specialists, who will subsequently be offered a job at Tesla or other sponsoring companies. On forums and on Twitter, Musk was accused of clickbait and exaggeration of the merits of Open AI.

Bot developers for Dota 2 were dissatisfied

Valve did not invite programmers who worked on third-party solutions for Dota 2 to the tournament. There are several programs based on decision trees and other technologies that are able to play even in 5v5 mode. However, an early prototype for playing 1vs 1 by Open AI. Valve did not comment on this claim.

In conclusion, I want to say: "I, for one, welcome our new robot overlords." As for me, I welcome our future robot guards. And I regularly write about them in our Robotics Channel Telegram channel. Join now!

Artificial intelligence (its weak form) is gradually becoming more and more skillful. The computer successfully solves even those problems that only a few years ago were considered accessible only to humans. One example is the game of go, where only those who have well-developed intuition and logical thinking can become a champion. Go for the car was considered an unattainable "ceiling". Now we see that there are no people on Earth who can beat the machine.

It is clear that AI is not only playing Go, the range of tasks is much wider. However, experts from various IT fields are trying to train the computer to fight on an equal footing with a person and in computer games in order to test the capabilities of their software platforms. In this case, the human and the bot are in the same conditions. Usually, opponents are pushed head-on on a strategy map with a “fog of war”, so that both opponents are not aware of what objects and where they are on this map (under normal conditions, the game “AI” knows where everything is). It was under such conditions that a man defeated a machine in StarCraft.

It's about professional StarCraft player Song Byung-gu taking on four different StarCraft bots. One of them, "CherryPi" was developed by Facebook Corporation. Other bots are created by companies from Australia, Norway and Korea.

The game was held at Sejong University, Seoul, Korea. Since 2010, StarCraft competitions involving humans and machines have been held here. The event is supported by various organizations (both commercial and research and public). This particular competition was supported, in particular, by the Institute of Electrical and Electronics Engineers (IEEE).

It is clear that Go and StarCraft matches cannot be compared. In the first case, the player sees the entire board and needs to find a winning strategy. In the case of StarCraft, you also need a winning strategy, but here you have to keep in mind many objects on the map, have a good reaction and be able to act suddenly, unpredictably. Understanding all this, the AI ​​developers decided to use the StarCraft environment to train the machine.

Notable players in StarCraft, they have previously stated that they would very much like to fight a computer opponent. AI companies such as DeepMind welcomed this desire, stating their intention to hold similar competitions in the future.

And this future has come - people were able to fight with AI. The computer did not work out - the man defeated all four bots in 27 minutes in the overall standings. The maximum duration of a match was 10 minutes and 30 seconds. The shortest match lasted 4 and a half minutes. And this despite the fact that bots performed much more actions per unit of time than a person. For example, a system created by Norwegian specialists is capable of performing 19,000 actions per minute. Cyber ​​champions from the world of people in best case capable of performing several hundred moves-actions per minute.

Son stated that bots act differently than humans, their play style is noticeably different. "We, professional players, start the battle only when we have a chance to win with our army and other elements," he said. People start a fight when they are confident in their abilities. But the bots that Son fought were somehow trying to keep their units alive without making particularly bold decisions. According to the cybersportsman, the only thing that favorably distinguished bots from people in the game was the organization. In particular, the bots managed the crowd of their units quite skillfully, trying to protect themselves from Son's attacks.

According to Kim Kyung-Jung, the organizer of the competition, the bots that competed with a person simply did not have enough experience. They played too few matches. For example, AlphaGo played tens of thousands of matches, both with human opponents and with copies of himself, before he learned to play at a high level. But the bots playing StarCraft did not have this opportunity.

However, this situation will soon change. Back in August DeepMind and Blizzard Entertainment announced AI development tools compatible with StarCraft II. On this moment it is this version of the game that is most popular among cybersportsmen. Some experts believe that AI will soon "tolerate" a person in the same way as it happened in Go. “Once bots are equipped with high-level decision-making systems like AlphaGo, people will no longer be able to win,” said Jung Han-min from the University of Science and Technology, Korea.

A Facebook spokesperson behind CherryPi said that the company has long wanted to see AI take on an esports player. After the match, Facebook received great amount useful information that will be used to improve the bot.

And so, the long-awaited Heart of the Swarm beta test. We put aside vague interviews, screenshots and battle reports, and hasten to check our inbox for the presence of the coveted beta key...

Alas.
The vast majority of casual users will be disappointed in these early days of the beta. Blizzard has officially announced that the first beta testers will include a certain group of pro-gamers, members of the press and winners of the "Trading Row" contests.

But we will not put up with such injustice, right?

And for this, especially for you, I will try to tell you as much as possible about how to play, if not on the official Battle.Net so simple at home with a computer. And there, perhaps, starfriend with his Pirate LAN & Hamachi pull up...

There is such a wonderful program: SC2 All in 1. which allows you to play StarCraft 2 Wings of Liberty without a license key and/or Internet access. True, but against artificial intelligence(AI). Which, however, at high levels of difficulty plays not much weaker than the players from the "Iron Leagues" (below the diamond, that is). Also watch replays.

Latest version at the time of writing 11.2.11 beta 8 already has support Heart of the Swarm v2.0.0.23077 beta .

Well, let's get started.

1. Download the map pack (if you are also going to play WoL offline with a computer)
2. The program may require .NET Framework 4 to run correctly

Unpack the archive with SC2Allin1 to any folder and run SC2ALLin1.exe. (By the way, if some particularly paranoid antiviruses send you to the "king of the Trojan", then first calm down the alarmists - there's nothing wrong with that.)
We have a window like this:

Despite the abundance of options, nothing will be launched, because "No map is selected!". Those. First you need to download the maps you are going to play.

Go to tab Maps .

There are four buttons to choose from:

• Add Maps from Folder- add cards from the selected folder. This is if you have previously downloaded the maps. The above link, for example.
• Select Map- select a map and go to the game settings (Game tab)
• Map Sites- launches the browser and moves us to a page with links to useful online resources where these same maps can be downloaded.
• Add Maps from Server- do not worry about searching for maps on the Internet, but simply download them from the server.

Click Add Maps from Server . In a new window you will find a list of all cards supported SC2ALLin1. We are interested in those cards on which there is a mark (hots). Mark them with checkboxes and click Download and Install Selected Maps . Downloaded maps can be found in the folder "My Documents\StarCraft II\Maps"

We click an inconspicuous cross in the upper right corner, close the window for loading maps.

Now we have the names of the maps we downloaded in the list. The number on the right indicates the number of players for whom it is designed. In the left part of the window there is a preview, information about the creators and a commentary on the map. The numbers on the diagram indicate the locations of the starting positions of the players.

Click Select Map and return to the very first tab window Game .

Now let's move on to the game settings.

The leftmost, unnamed list is responsible for choosing a race:

1. Random
2. Terran
3. Zerg
4. Protoss.

Player Type: set the player type.

1. human- player, i.e. You.
2. computer- your silicon rival
3. Referee, Spectractor- we are not interested, because we are going to play in splendid isolation.
4. closed- the slot remains empty.

You will need at least one computer opponent to play.

Difficulty: set the difficulty level. The intelligence level of your computer opponent will directly depend on this option. Options:

1. Very Easy- very easy
2. Easy- easily
3. Medium- average
4. Hard- difficult
5. very hard- very hard
6. Insaine- computer cheater.

In principle, the set of options and their difficulty depends on the version of AI that you are using. The set of this assembly includes the brain from May 10, 2012. release dates. For Human "s, the choice is prudently blocked.

In the near future I will supplement the article, explaining in more detail about setting up AI.

Color: allows you to choose a color for yourself and opponents. The number of options here is much more extensive than in most custom maps. The choice is offered:

1. White
2. Blue
3. Teal
4. Purple
5. Yellow
6. Orange
7. Green
8. light pink
9. violet
10. light gray
11. dark green
12. Brown, Light
13. Green
14. dark gray
15. Pink

Just like a children's set of 16 markers. plus more Random(random color).

location: choice of starting position 01 to 08 or Random. You start at the specified location, in accordance with the choice.

health: the health of the units of the selected player. From 100% to 5%, from the original. If you want to take out everyone as slaves or give the enemy a good head start, you can play around with these values.

There are two more options on the right side:

Game Speed: game speed. by default is Faster. If you want to see the battle of the slowpacks, you can move the slider to the leftmost position. (Slower)

Game Type: game type.

1. Team vs. Team- team-per-team
2. Free For All- aka FFA, all against all. Installed by default.
3. Use Map Settings- the mode is set by the map settings.
4. Random War- something random. :)

In case of choice Team vs. Team, additional items will appear in the settings:

1. team- allows you to choose the player's belonging to a specific team.
2. Team Shares: general review ( Vision), unit control ( control) and resources ( Spendings). Or all of this at once.

At the moment, the modes work correctly Free For All And Team vs. Team. Check mark custom AI is responsible for using the original Blizzard script (no checkbox), or a third-party script (checkmark). By default, a third-party script is used from the folder \SC2Allin1\ai

Next: go to the tab Options

In chapter Others enter your nickname (if you do not want to be just a Player). NECESSARILY check the box and specify the path to the folder with HotS Beta. (x:\Games\StarCraft II 2012 Beta- For example). Otherwise, the default program will run the program installed on your computer. Wings of Liberty. (After specifying the path and checking the box, it's better to restart SC2AllIn1 so that the path is preserved).

You can also play with the numbers in the section Cheat by giving yourself free resources.

This completes the settings.

It remains to press Launch and voila!...

Heart of the Swarm Beta at your complete offline disposal! Have a good game!

A small FAQ on popular problems and their solutions.

Q: When trying to start the game or resume an interrupted installation, an error pops up: "Faled to read required file. Please, close all other aplications, temporarily deactivate your anti-virus software, and try again"

A: Delete folder C:\ProgramData\Battle.net . If it doesn't help, start the game SC2Switcher.exe located in the folder StarCraft II 2012 Beta\Support .

Q: When trying to run it gives an error: "Unable to validate SC2 License"

A: For some reason, the game thinks that you should not licensed version. There are two treatment options:

Q: After pressing the Launch button and starting the game, it gives an error: "Unable to open map"

A: Possible solutions:

1. Check that you have started the game in Free For All or Team VS Team.
2. Check that you have downloaded and run that map - it has (hots) in its name.
3. Make sure you enter the correct path to the installation location HotS Beta and ticked Use Custom Game Installation Path.
4. There is no Cyrillic alphabet in the installation paths.

Q: There are error messages popping up in the game.

A: Just uncheck the boxes on the left and right:

And most importantly, remember that in spite of any private distribution of keys -

On October 31, 2017, a duel took place between professional StarCraft player Song Byung-gu and four different AIs. The man won every fight. One of the bots was developed by the Facebook FAIR laboratory, the rest - by programmers from Australia, Norway and South Korea.

Human versus AI

The competition took place at Sejong University in Seoul, South Korea, which has hosted an annual StarCraft AI competition since 2010. In previous years, the matches were only between bots and were organized by the Institute of Electrical and Electronics Engineers (IEEE).

Despite the fact that the duels in Seoul did not attract as much attention as between AlphaGo and go professionals, they still have a great importance in the development of AI. The research community considers StarCraft especially difficult game for bots. After AlphaGo's success in Go and some AI advances in chess, the attention of scientists turned to the idea of ​​teaching AI real-time strategies such as StarCraft.

In contrast to the same go, in which constant control of everything is possible playing field, due to which players can develop a game strategy in the first step, StarCraft requires using memory and developing tactics on the go within a limited and virtual game world. As a result, StarCraft becomes an effective tool to help AI take a step forward in development.

Past fights are just the beginning

A number of professional StarCraft players have said they are ready to host live matches against bots, similar to AlphaGo matches. The leaders of the DeepMind project have already agreed to organize similar matches in the future.

Returning to past matches, it is worth noting that the total time spent by Song in 4 matches was only 27 minutes. The longest match lasted 10.5 minutes, the shortest was only 4.5 minutes. This is despite the fact that bots could move their units much faster than a human, as well as manage multiple tasks at the same time.

The Norwegian bot, for example, at one point in time reached a speed of 19 thousand actions per minute. This is despite the fact that professional players cannot perform more than a few hundred actions per minute.

The winning man noted in a post-match interview:

Professional players start the battle only when they have enough army and unit management skills to win. The bots, on the other hand, tried to save their units without making any bold decisions. However, the way they deflected my attacks was sometimes overwhelming.

Kim Kyung-jong, a professor of computer science at Sejong University, said the bots were limited in part by the lack of data available for preparation. He also added that AlphaGo was able to defeat champions in Go only by studying a large number of matches played by human professionals.

Development perspective

Note that Google has long wanted to teach AI how to play StarCraft. Also in this quest for her Facebook and Alibaba. In August 2017, DeepMind and StarCraft developer Blizzard released a highly anticipated set of AI development tools compatible with the most popular version of StarCraft II among professionals.

Experts are already predicting a complete collapse of human dominance in StarCraft. According to Jung Han-min, a professor of computer science at South Korea's University of Science and Technology, when AI bots are equipped with high-level decision-making systems like AlphaGo, humans will no longer be able to defeat them.