Reading that Uber are striving for the "ultimate macro game" on the one hand, and seeing the way the game is shaping up and that they "don't want to automate" beyond the queuing of commands on the other hand worries me greatly. First, all of this rests on my understanding of the term "macro" and if this differs from Uber's view my point and subsequently this thread is rendered moot. If that is the case I would love to know what macro means for Uber in regard to PA. So what is macro to me? To me it stands for the process of evaluating a current state and formulating a desired future state as well as a set of broad guidelines/constraints pertaining to how to reach that state. Micro to me is then the carrying out of actions that lead to that state in a reasonably efficient way. In my opinion a real macro game is only possible in two cases: 1) Readily available options for efficient automation of pretty much every aspect of the game 2) Limited inputs/actions per time segment Only then you don't have to exclusively decide between meaningful planning and meaningful executing. Since neither is present or even desired in PA and effective apm know no practical limit with the scale Uber are aiming for I'm asking myself if PA is really designed to be a macro game?
They did promise a lot of automation and it is slowly but surely coming. The other point is a matter of details in balance, microing bots is a matter of removing their juking of trajectory, microing multiple areas could be worked on with UI and game pace, microing too many units to control via macro can be fixed with metal and costs, microing click-intensive basebuilding speeds is a mix of automation and build rate and metal income. Generally, all being discussed or worked on. EVERY rts will have some micro, considering the definition of micro is anything click intensive for one effect and by definition is a matter of opinion. The ultimate macro rts game, is a turn-based. Even then, runescape is micro-intensive, competitive PVP must learn ALL the hotkeys because you have to immediately know to click 5 of them within 1/4th a second at times.
If we can get troops to behave smarter then supcom, but without them being fully automated like zero k (We'll mod that in, don't worry) then I feel like I might be able to stop being a helicopter parent....I mean helicopter commander.
Thanks for your replies, that is really heartening to hear. Mind you, I don't want to criticize Uber for the game's state or the way I perceive it might be heading. I am just genuinely curious about their vision of a macro game! Expanding on one of the points I made in my OP, namely hard limiting actions per minute: One game that might have profited from such a thing (and deceleration in general) is R.U.S.E.; any order you gave in this game was permanently and prominently visualized and could be intercepted and viewed in the same fashion by your enemies. If there had been more of a finality to those orders, by e.g. allowing no new orders to any given unit for a certain amount of time after it did receive one, it would have given your orders more weight and intercepting them even more meaning. Of course this is only to give an example as to why I think forced deceleration can be a good thing, not a proposal for PA!
I could never get into ruse. Something about how you need extra buildings for things I could understand why.
I think most of this is due to PA being in beta. Area commands are a huge step forward, but we're still not fully there for area commands. We're missing some and some of them still need improvements. We're also already seeing mods that help with automation.
Wow! This! All this! You hit the nail right on the head and blew it out the other side. Can't like this post enough. You are wise beyond your post count vyolin
I'm all for automation of all sorts, especially considering the ever increasing complexity found in PA matches these days. I'm absolutely fine with Uber not providing all those features themselves, iff modding support really is extensive and flexible enough to accommodate all those in client-side mods without having to jump through too many hoops. As the topic is closely related, I'll close with a quote from the Smart Combat, Unit AI thread.
You are giving me too much credit - but who am I to complain! Just to make this perfectly clear: I am not targeting the current state of the game with my OP but rather the final vision of it; That said, I think I see what Uber are aiming for and I see the tools they are providing (area commands, auto-queues?, commands as first order entities?) but I can't see the design principles behind those, i.e. how much are they willing to sacrifice for keeping it a macro game? Is there an official word on that that I missed? Can somebody shed light on this for me?
I think "micro", in RTS vernacular, is more like "directing every detail of unit action, to the extent that units are almost never doing anything you didn't explicitly tell them to do". Micromanagement. They don't take any pathing or fire on any targets that you didn't explicitly set because you are making sure to expend your 200 APM explicitly setting these things. From what I've seen of top-APM Starcraft players, the game doesn't even need "pathfinding" because no unit ever takes any step that isn't directed by the player through a series of clicks. With that in mind, I'd say there's 3 levels: Macromanagement: As you said, seeing the state of the game, thinking of the state you wish it was in and thinking of how best to head to that state. "I should probably do something about my metal income so I can afford to nuke that moon." Management: General actions in involved in heading to that state: "You, fabricator, go build on all metal deposits in this circular area." Micromanagement: Overseeing the lives of individual units to the finest detail, especially to win individual battles. "You 30 Doxes, kill THIS and move HERE and kill THIS and move HERE and kill THIS and move HERE, HERE and HERE and kill THIS..." All RTS games, I'd say, have all 3, but whereas Starcraft is very heavy on micro, PA is very heavy on "management". You tend to direct units in more general ways and rarely micro their every move because there's just too many other pressing issues to deal with. You rely more on pathfinding and automatic target selection -- something you never want to do in Starcraft.
That is such a hard question to answer before and now without going into fine details on what is getting implemented and what is getting canned for pure macro management. There has been so much progression towards eliminating busy work such as clicking on mass extractors and creating a patrol around a planet. Some of this stuff was just thought up. If I was designing a sink from scratch to be the most water efficient and water conserving, and someone asked me in the middle of working on it to give a finite number of how efficient or water conserving the system will be.. I wouldn't be able to tell you until its all said and done.
I don't think your analogy holds much water but the gist of it still applies. My curiosity in this design stuff in general and the belief that Uber were following a more rigid set of design principles than they actually seem to mainly stemmed from one of neutrinos comments where he voiced some serious dislike for automation. Bottom line: I will see it when they're done. Thanks for all your comments, though!
