Defining micro

The word micro is being bandied about a lot here and I think it's being misused in a lot of cases. Remember it's short for micro-management: a lot of people seem to be forgetting that and assuming that it just means small adjustments to a unit's orders.

There are several types of micro-management:

PRODUCTION

Production micro is where you have to select your factory, and click the unit you want to build. Command & Conquer simplified this by having all buildable units listed on the side of the screen, and additional factories would just speed up production. Hard to justify from an internal realism point of view, but a big improvement on Dune 2 and Warcraft.

TA did away with it almost entirely with shift-clicks to add five units to a queue, and the ability to queue up as many units as you like. SupCom refined this by adding an infinite repeat button. In SupCom you can literally make an infinite number of tanks with just three clicks. It's fair to assume that PA will be the same.

StarCraft still demands one or more clicks for every unit built, an example of the developers intentionally hindering the player with an unhelpful UI. I think we can agree universally that we don't want this.


UNIT SELECTION

In Dune 2, to move a tank you had to click the tank, click the move command then click where you wanted it to go. Three clicks just to move a single unit. Command & Conquer and Warcraft both improved on this by letting you draw a box around multiple units, and having a default, context-based action such as move or attack.

One annoying flaw that all RTS games had was that you could not scroll the map whilst drawing your box, meaning that any units currently off-screen had to be selected in a separate sweep. Total Annihilation fixed this so that you could theoretically select every unit on the map in one sweep of the mouse. SupCom actually removed this ability, but it more than compensated for it with its three-dimensional camera - simply zoom out until the units you want are on screen.

Other improvements have included control groups (either C&C or Warcraft) and TA again innovating heavily with a large number of shortcut keys to select units based on type.

Again StarCraft shows us what we don't want: an artificial limit on the number of units selected in order to make wrestling the interface a large part of the game.

MOVEMENT

This seems to be the area of contention on these forums. Why should somebody who can click faster have an advantage in a strategy game? Why should the top Forged Alliance players be able to kill two tanks using a single light assault bot?

Why? Because the alternative would be much worse. This talk of automating units to the point that they auto-dodge, find the best terrain to attack from or similar is just wrong-headed. The more automation you put in, the more it comes down to a numbers game. If I have twenty tanks and my opponent has thirty, if there is no way I can win the engagement then there is no game.

Also more automation starts to render the physics simulation pointless. Again I think we all agree that simulated projectiles are one of the things that sets this sub-genre apart, and that it is awesome in every way.

Yes, your units should have a modicum of sense: they should attack anything within range certainly, and arguably they should move about a bit (I'm thinking along the lines of TA's standing orders here), but they should not play the game for you. Tactics like pincer movements and crossing the T are very important for depth, and auto-dodging, auto-arranging units would weaken that.

OTHER

The other micro-management in these games is pretty much limited to D-Gun and manually launching long-range missiles, and I don't think there's an argument to remove these.

The only other bit of micro I can think of is reclaiming. You have three options for this, either click ever tree and rock you want collected, which is relatively time-consuming but gets the resources sucked up quicker, or set a patrol route over the resource field, or use attack-move orders. I don't believe that any of these could be automated any further.

EDIT: also upgrading mexes, factories, radar etc. as I understand it these won't feature in PA, which is probably a good thing. Even with the mex manager mod, much of the player's time and effort in SupCom/FA is wasted on upgrading bloody mass extractors.

So in summary, most of the worthless micro was removed in TA, SupCom added a bit of polish and now the only micro remaining is valuable for gameplay and would harm the game if removed.

Hey, PSAs are my gig! ;p

Mike

A while back I would have agreed with you on the point of movement, but I recently gave Zero K a try, and I found that the implementation they have there (or rather, a script that seems to be available to many Spring games) works quite well. In it, the units intelligently attack based on an order you give them.
For example, I gave an attack-move command to ZK's equivalent of the Arm Hammer kbot from TA, telling it to move in toward where I suspected an enemy laser turret was. They moved forward until they gained vision of the turret, then retreated to the edge of their attack range and proceeded to demolish it.
In another instance, I gave the same order to ZK's equivalent of the Arm Peewee kbot from TA, telling it to attack a rocket unit. The rocket unit fired slow, powerful rockets, and my units were smart enough to swarm and dodge as they advanced so the rockets missed them. Then when they were in range, they took it out. One did bumble into a rocket and got destroyed, so it's not like they were pulling off perfect dodge maneuvers, just making themselves hard to hit.
In a final example, a group of units was sent to attack, and a couple of them lost health to the point that they were nearly dead. They intelligently pulled out of the fight and ran to a retreat location I had marked so they could heal up.

So does this automate the units too much? Not at all. I was able to focus on the bigger picture, of all the numerous fights happening around the map. It allowed me to multitask without having to endure the tedium of trying to keep my units alive. Sure when you have one or two units you control, you may want to control them exactly, but when you had over 100 units you were commanding in TA or Supcom, did you EVER try to control them to that detail? More likely than not, you just tossed them to the meat grinder and waged a war of attrition.
Or you did some Hawk-dancing, which is just taking micromanagement to the extreme.

Is Zero K's system perfect? Certainly not. There's plenty of room for improvement, but it's a more than sufficient proof of concept at this point. Plus you can disable the ai control completely, so your units don't try to pull of those maneuvers when you're trying to do something specific.

Overall, I think these ideas have merit, and should not be so quickly discarded, especially for a game with the scope of PA.

[EDIT] Someone posted this video http://youtu.be/8R6uq3l7gSs about retreat zones in the Zero K ideas thread. It illustrates one of the points I made. If I can find one of the intelligent attacking AI, I'll add that too. [/EDIT]

The distinction between micro and macro is very straight forward.

Micro is anything that improves the efficiency of what you have.

Macro is increasing how much you have.

There is no such thing as "production micro". Anytime you touch production you are doing macro. Even if it seems like a minor thing. Micro and macro are defined in isolation from the actual interface. The only thing that could be consider micro involving production is altering rally points to bring your production straight into battle quicker.

The need to hit a hotkey for every tank in SC2 is not micro. That is a mechanic. SC2 has macro mechanics that don't fit PA (that is individual commands for each unit built).

Lets be clear about the terminology we are using. Fiddling mechanics is not something we want in PA. However a fiddly mechanic is not necessarily micro.

In short the OP is misusing what the term micro means in a post complaining about misuse of the term micro. This is incredibly common in RTS games and we should be more specific about what we mean. Micro and macro have always referred to the type of abstract action on the field. Not how fiddly it is to execute them. Automated positioning mechanics are still about micro even though you are moving armies rather than units. Clicking on tank icons is still macro even though you are doing small scale things.

I'm inclined to agree that more intelligent units are going to make this a better game. It's going to be a next gen RTS, so we could use a next gen UI and more intelligent units. I can't see how when we have thousands of units mobile at once that we could possibly get the best out of them any other way.

The other thing that really worries me is people suggesting having local planet economies (as opposed to a persistent game-wide economy) is that it promotes individual management of these economies and distracts the player from the REAL point of the game. This IMO is something we could do without.

I agree that micro is tedious, but some people love doing it. That's why some of the above mentioned games are so popular. I personally don't believe that you should have to dictate everything about a units actions by clicking a mouse repeatedly. Issuing a behavioural order about how the unit should engage like in Dawn of War 2 for example, or the unit having a default engagement strategy that befits it's characteristics makes so much sense. That way you can concentrate on the wider strategy. The scale we are talking about here is much too large for that kind of micro management of units.

The commander in OTA is a awsome unit for micro, you can use its nano-blocking and d-gun to kill hunderds of units, hope PA would have that too.

People also seem to think that micro-management gives the game complexity and makes it more skill-based. It's true that it takes a ton of skill to use the torturous Starcraft 1 interface in a way that allows you to have total control of your army, but in the end, strategy games should not be about physical skill, but mental skill.

As for complexity, expanding the scope of the game so that it's not feasible to tinker around with units individually doesn't reduce complexity, it adds to it.

While a micromanager is busy telling strong units to advance, weak units to retreat, manually kiting, manually dodging, etc, the strategist has used this time to set up a couple of new attacks, targeting different key positions, established a new ferry route to occupy a forward position, and build up several new factories. And that's not including what plans they may have set in motion on other planets and asteroids they may occupy.
The strategist has more to think about, more to plan, whereas the micromanager is really just focusing on instincts and muscle memory. It's just such a waste of time and such a limited scope of thinking.

The problem is that people learned RTS's in an era where engines and UI's were too simple to handle anything on a grand strategic scale. So they stick to games like Starcraft, not realizing there's so much more to the genre than what they've experienced.

Defining Micromanagement alone is not enough.

You must also define Macromanagement.

I'm going to quote something that I posted a month ago, and save myself re-iterating it.

Most people here will agree that macromanagement is what we typically want to invest our energy into. The art is to reduce some of the clicking needed in certain parts of the game.

I agree with eukanuba, even though he may have called something micro when it isn't quite micro.

I don't want the game to fight my battles for me. I want to be the most effective source of leadership(or worst, the one thing that can really differentiate a good player vs. a bad player). So, I don't want to build a ton of units, send them off to fight only for them to organize themselves into formations and execute complex battle strategies.

That should be the responsibility of the player.

On the other hand, I have very mixed feelings about units dodging artillery and weapons fire.

If you have 20 tanks and the opponent has 30 tanks, he should obviously win the engagement, because he played better and brought 30 tanks where you could only bring 20.
Of course, this may be part of your strategy, and you brought only 20 tanks to slow him down while your bombers are attacking his base.

Physics simulation makes balancing units far more 'natural'. For example, an anti-heavy unit have a hard-hitting slow-firing, slow-moving no-tracking projectile instead of having 2x damage against "heavy" armour types. So more automation wouldn't make physics simulation pointless at all.

They wouldn't play the game for you. They wouldn't decide if you should go tanks or planes, if you should bring AA or not, if you should rush for the enemy base or try to extend... They wouldn't make the important stuff for you, only the small-scale stuff you can't possibly do for your thousands of units with less than 400 APM.
As others said, you should try Zero-K, which has smart units and shows how it improves the game instead of dumbing it down.

Better, have zone-specific reclaiming orders, like in Zero-K. Even better, have a default "smart" reclaim, that won't reclaim if your storage is full (and have a "force reclaim" mode, if you just want to deny wreckages).


Apart from that, I mostly agree.

People don't do better pathfinding than the flowfield pathfinding AI behavior, so I don't see much reason that they have to be better at kiting or dodging.

Dodging/Kiting = Micro
Pincer = Macro

Auto-dodge/kite is low-hanging fruit and simply a must, esp. for fast-attack units, in my book.

Including it would not negate or cheapen pincer movements or their ilk.
Quite the opposite in fact, without the need to dance you troops about to get ahead, you'd be be able to concentrate more on set up macro placements and movements for the greatest advantage.

Agreed.

IMO the OP is not a great definition of Micro because it mixes in opinion of appropriate micro and inappropriate micro, making it a biased definition.

I'm starting to be a bit more convinced that there would be some benefit in having units auto-dodge, the Zero-K example was most interesting. I did try to play Zero-K but I found the interface a bit overwhelming and then I turned the camera to a silly angle and couldn't get it back so I gave up.

I still have issues with autonomous units but I can't quite put them into words at the moment.

I would rather not use the word 'micro' because the Starcraft bunch have appropriated the word that to me does not make much sense. The word seems to work in the domain of Starcraft, maybe there are two clearly defined types of tasks.

Anyway correct me if I am wrong but here is the Starcraft definition:

• Micro: Controlling units.
• Macro: Building stuff.

To me a lot of the 'building stuff' in Starcraft2 is micromanagement. Queen vomit is something players have to keep 'micromanaging' far past the point that decisions are being made.

Instead I would rather talk about the mapping between decisions and clicks. As in how many clicks does it take to implement decisions that the player has to make within a game. If some action requires no decision then there may as well be no clicks. If a player makes a decision it should continue to be implemented without further clicks.

This position can seem to be 'anti-SC2-micro' but it is not. There are games with a lot of quick decisions which must be made in battle. Force firing can be a decision, retreating can be a decision.

(Just an aside) The default camera was fixed a while back and you could post on the other forums if you have problems.

Try ctrl+mousewheel, or press f11, disable the smoothscroll and enable the overhead free camera, then use ctrl+middle drag to rotate.

I dont see why we should not use the same definitions that are used in Starcraft, since they are known by many people.

Apart from that I agree with you about the decisions/click part.
It is not about removing micro, it is about removing the need for the player to do task X every Y seconds, even though it easily could be automated, since no decisions are involved.

Units that would move a bit more intelligent would be nice, but I doubt that ANY ai could be of a real help. I am certain I would often end up cursing the ai for doing stupid moves and in the end disable it mostly.

I think the context in which most people are talking about micro is for what amounts to small but significant things you would expect a unit to do on it's own anyway. The whole thing is about alleviating menial tasks the player shouldn't be concerned with and allow them to get on with enjoying the actual game. Not to play the game for you.

In a nut shell the game is not about micro managing units. It's about large scale strategy.

EDIT: I missed an important sentence.

I doubt you would often curse ranged units for auto kiting, they don't have better things to do in the most situations.