Sign in with
Sign up | Sign in

Use This Algorithm, Win 'StarCraft II'

By - Source: Tom's Guide US | B 14 comments
Tags :

What's the best possible winning strategy for "StarCraft II?

Trick question: there is no single best strategy, but a new algorithm developed at North Carolina State University (NCSU) can examine the unique circumstances of your game and tell you exactly what you should focus on to maximize your chances of winning.

Computer science professor David Roberts and Ph. D. student Pu Yang analyzed thousands of professional-level matches in team-based, real-time strategy games like "StarCraft II," "Warcraft III" and "Defense of the Ancients."

From these game logs, the researchers came up with a complex set of rules and probabilities to help players develop the best possible strategies for any situation.

"Not all strategies work against all opponents," Roberts told Tom's Guide. "There's a lot going on in the environment that can contribute to the success or failure of the" match.

MORE: 13 Awesome PC Games That Have Gone Free

As professional strategy gamers know, certain strategies are more important at different times in the game. To help players determine which strategies they should employ at a given time, Roberts and Yang's program analyzes the various attributes by which a game judges progress, such as population, resources and attack power (the attributes vary from game to game). The program then tells players what level each attribute should be.

The rules generated by this program work on an "if-then" format. For example, if you're playing as X and your partner is playing as Y, then in the first fifteen minutes you should focus on A, in the second fifteen minutes on B and in the third fifteen minutes on C.

The program also tells users their likelihood of victory if the specified conditions are met.

For example, in a game of "StarCraft II," if you're playing as the Terrans and your teammate is playing as the Zerg, and you keep your population growth rate low while your teammate keeps population growth rate high, your chances of winning are over 70 percent. 

"With 'Starcraft II,' the things that we found were most predictive [of victory] had to do with population growth rates," said Roberts. But "it very much depended on team composition as well."

The researchers found that in certain games and situations within the games, some attributes are more predictive of victory and therefore more important. To help players manage their time, the program tells players which attributes will most increase the likelihood of victory in a given situation.

The algorithm also shows that sometimes you want certain attributes to be lower than your opponents'. For example, in "Defense of the Ancients," more commonly called "DotA," if the player-characters on Team A have a combined damage-point total that's 59.7 points more than Team B's, the first team has an 80 percent chance of winning.  

However, if Team A's damage-point advantage drops by just six points, to 54 more than Team B, then Team A's likelihood of victory plummets to 10 percent. In that case, Team A might want to switch strategies to something other than damage point growth.

This all might sound vague, but that's because there's no one "best strategy" — it all depends on what's happening in a given game. The program helps players juggle the many variables at play.

Roberts told Tom's Guide he's working on creating plugins for "StarCraft II," "Warcraft III" and "DotA" to help players develop the best possible strategies.

Right now, the plugin works, but it's not very easy to read and, as you can see from the above picture, takes up a large part of the screen. Roberts is now looking into usability testing to develop a user-friendly interface.

Roberts and Yang's full paper is available from NCSU's website. Professional and hardcore strategy gamers: What do you think? Do these strategies ring true for you?

Email jscharr@techmedianetwork.com or follow her @JillScharr. Follow us @tomsguide, on Facebook and on Google+.

Discuss
Ask a Category Expert

Create a new thread in the Streaming Video & TVs forum about this subject

Example: Notebook, Android, SSD hard drive

This thread is closed for comments
  • 5 Hide
    Soda-88 , August 19, 2013 4:10 PM
    What a bunch of bull... If they ever played a game of SC2 and/or watched korean pros do it, they wouldn't waste time on this 'research'..
  • 6 Hide
    Soda-88 , August 19, 2013 4:15 PM
    Also, why is there a screenshot of brood war when they talk about SC2 and Dota2
  • -6 Hide
    doomtomb , August 19, 2013 5:58 PM
    For example, in a game of "StarCraft II," if you're playing as the Terrans and your teammate is playing as the Zerg, and you keep your population growth rate low while your teammate keeps population growth rate high, your chances of winning are over 70 percent.

    These 'researchers' understand very little about the actual game. 'Population growth rates' gimme a break.
  • Display all 14 comments.
  • 2 Hide
    FloKid , August 19, 2013 8:48 PM
    I refuse to look at this map hack.
  • 1 Hide
    onichikun , August 19, 2013 9:47 PM
    Creating a statistic with so many hidden variables makes for a very useless statistic. Now if they were to give a set of simplifying assumptions that could be used as a basis for their claims, that would make it more useful. E.g., in a professional game in TvZ doing fast reaper harass leads to victory N% of the time given that the zerg player goes for fast expand etc. etc. This statistic may vary for several levels of player skill, since obviously fast reaper harass is easy to screw up, especially for new players.
  • 0 Hide
    Someone Somewhere , August 20, 2013 12:00 AM
    How much of this is cause and how much effect?
  • 2 Hide
    ddpruitt , August 20, 2013 8:15 AM
    Quote:
    student Pu Yang analyzed thousands of professional-level matches in team-based, real-time strategy games like "StarCraft II," "Warcraft III" and "Defense of the Ancients."


    I'm sure he did. That's going to be my new excuse for playing video games "But it's research!", wait I already do that.
  • 2 Hide
    jnkweaver , August 20, 2013 1:25 PM
    This would be very cool if they used it for better AIs.
  • 0 Hide
    Chaoping Guo , August 20, 2013 7:18 PM
    i lol'd
  • 0 Hide
    natcparis , August 21, 2013 6:03 AM
    I'm puzzled because if the algorithms are based on each player's unit counts, then they are of no use. As an SC2 player you do not see the opponent's units.
    At least not all of them.
  • 0 Hide
    byte_my_bits , August 21, 2013 1:26 PM
    I clicked on the screenshot/picture 3 times, and never got a larger version than the thumbnail. Please kill yourself, OP.
  • 0 Hide
    Anomandaris , October 5, 2013 12:51 PM
    @byte_my_bits I did the same, then read your post and laughed out reeeally loud. Thanks a bunch!
  • 1 Hide
    NightLight , October 18, 2013 1:18 PM
    personally, any game where i have to start using algorythms to win, is not a game anymore, it's more like work...
  • 0 Hide
    NightLight , October 18, 2013 1:18 PM
    personally, any game where i have to start using algorythms to win, is not a game anymore, it's more like work...
Tom’s guide in the world
  • Germany
  • France
  • Italy
  • Ireland
  • UK
Follow Tom’s guide
Subscribe to our newsletter
  • add to twitter
  • add to facebook
  • ajouter un flux RSS