![]() ![]() One could argue that Blizzard held the same view when they implemented multi-building select in StarCraft 2, against the cries of Brood War fans, patriots, and a bevy of professional players. The interesting part of macro in StarCraft is the decision of what to make, not the execution of actually making it.” “I want StarCraft to be a game that is more about strategy, micro, map control and positioning rather than about performing the same repetitive actions perfectly in order to eke out the biggest army possible. ![]() Without good macro, you can’t experience a substantial part of what Starcraft offers,” Kevin ‘qxc’ Riley, a StarCraft 2 veteran, wrote in defence of Blizzard’s changes. “The problem is that macro is the gateway to the rest of the game. So understandably, people are a little upset - and angry - about the game’s future. Unfortunately, that never-ending 1a2a3a beat is considered part and parcel of StarCraft for some, it’s the very fabric of StarCraft’s soul. And the way Zerg responds to everything, in every match-up, changes massively when their ability to spread creep is crimped.īy removing the more repetitive macro elements, Blizzard hopes to open the doors to deeper strategy to those who feel left out, either through flagging flexibility in their fingers or an inability to maintain a constant physical rhythm. Protoss rushes and defences - against Zerg in particular - are massively different without the use of chrono boost. Terran build orders aren’t the same if you don’t have the massive influx of minerals MULEs can mine. It’s possibly the largest change made in the history of the SC2 franchise to date, since these three mechanics - the ability for Protoss to accelerate their speed of production, Terran’s ability to accelerate their economy and the ability for Zerg to quickly spread their creep across their map, vastly increasing the speed of their units and reconnaissance - are arguably the three silent pillars that form the foundation for the majority of all strategy in SC2.īe under no illusion: without these three mechanics, the way SC2 is played is radically different. One of those, introduced late last week, was the removal of the chrono boost and MULEs, while larva injects were reduced by half (but able to be auto-cast). The Legacy of the Void beta offers Blizzard the last chance they’ll have to revive the StarCraft 2 glory days, and they’re planning on taking advantage with some sizeable changes. But the required speed and focus on minutia turns off a lot of players from spending more time with the game, particularly in an environment where MOBAs have filled the multiplayer void that RTS games used to occupy, albeit with far less management and repetitive strain injuries. The problem is that trend runs directly counter to the macro mechanics that make StarCraft, well, StarCraft. Blizzcon attendees got a taste of that last year when they were surprised by the radical changes and the developer has continued the trend, hoping to streamline the game into something more palatable for today’s market. Blizzard has been wanting to change up the StarCraft 2 formula for a while. Various downstream tasks and analyses of the dataset are also described for the sake of research on macro-management in StarCraft II. Besides the dataset, we propose a baseline model and present initial baseline results for global state evaluation and build order prediction, which are two of the key tasks in macro-management. We also split MSC into training, validation and test set for the convenience of evaluation and comparison. MSC consists of well-designed feature vectors, pre-defined high-level actions and final result of each match. To boost the research of macro-management in StarCraft, we release a new dataset MSC based on the platform SC2LE. So most previous methods are trained with various features, evaluated on different test sets from the same or different datasets, making it difficult to be compared directly. 3) Some datasets are either too small or don't have enough labeled data for modern machine learning algorithms such as deep neural networks. ![]() 2) Some datasets are only specified for certain tasks in macro-management. But these datasets have some defects for boosting the academic and industrial research: 1) There're neither standard preprocessing, parsing and feature extraction procedures nor predefined training, validation and test set in some datasets. Various datasets together with assorted methods have been proposed in the last few years. Macro-management is an important problem in StarCraft, which has been studied for a long time.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |