Unit skirmish button!

There are not many circumstances in pa where advanced maneuvers like flanking, wedging and the rest you mentioned will work. The splitting of groups into easily manageable flanking groups can lead you to be steam rolled by the blob. Flanking can work but most effective with faster units like air or bots to kill shellers at the back of their blob or their bombers detached from their fighters.

Which speaks volumes of PA's current popcorn balance. If such 'advanced' manoeuvres - I would call them pretty basic, really - are ineffectual in PA this should finally be addressed. And advanced unit behaviour might be just the tool for showing as well as fixing those flaws. That, plus an HP-multiplier. Good point, though, unfortunately.

@elodea : Thanks for clarifying that for me, much appreciated. It seems we are from entirely different schools of thought, though, so I will not try to convince you of the merits of getting rid of micro as much as possible.

On a more general note, Rise of Nations provided tons of automation. AI War has lots of automation, too, but my experience there is limited. In both cases this resulted in and was necessary for the vast scope they deliver. So I am flabbergasted when confronted with the notion of PA aspiring to become a game of huge scope yet avoiding automation like the plague.

And lastly, a sincere question to all the proponents of obligatory micro: Most of you seem to genuinely enjoy babysitting all your units - or at least combat units. What is the appeal in that for you? Because I for one would rather spend that time on orchestrating grand manoeuvres and outwitting my opponent. So, shed some light on this for me, if you would be so kind. Thanks!

Some people like starcraftian anachronisms, where the UI is actually a test of your APM and muscle memory.

I don't like fighting such an outdated UI... but I know that such people exist.

If I remember correctly, Uber used exactly that as an example for what they want to avoid.
They were talking about not restricting the player by the UI but instead giving them as many tools as possible..I hope they let actions follow these words before the 1.0 release.

Such tools might make individual battles less about skill but any sufficiently skilled player should still be able to improve his units' effectiveness by microing.
Things like certain units keeping their distance to certain other units is not about attacking or defending, it is about stopping certain types of units from enforcing micro.

@elodea is right in many ways when it comes to the downsides of providing such features, but he does not seem to take into account that PA takes place on a much larger scale and that this scale requires the gameplay to work a bit differently, that not having such tools has more downsides to the whole game than having them does.

Also, I don't understand why you believe that some simple behaviours like these will always be superior to human input ? We are not talking about full skirmish ai that takes control over all of your units here.

And what's with all of this talk about attack move vs kiting ? Attack move could benefit from such a feature, as well, as tanks wouldn't walk into Infernos and so on.
Things like AA not charging into that turret and exploding while the actual force it is supposed to protect stops and shoots, would work a bit differently but is also some "automation" that would greatly benefit the game.
Or will you tell me that constantly having to manually move the aa is something that benefits the gameplay ?

Firstly could everyone please stop talking about the argument "Automation is fine because it is just an option. If you don't like automation then don't use it"? It is a stupid argument. Most people who play at all competitively are going to use all the (legitimate) options they can in order to get ahead. Even if you do without automation your opponent will get some sort of advantage from it.

This assumes singificantly oversimplified army and combat mechanics. You write as if battles are executed elsewhere. As if a self contained battle is played out whenever two armies meet. As if the strategic level and tactical level are as separate as they are in Total War games. I know your view is likely not to be this extreme but it is still what you sound like.

I think battles can be a lot more complicated than you have implied. This complication is probably swamped under the efficiency gains bought about by micromanagement. This effect can be so great that few people bother with the complications. They spend all their effort on micromangement to 'brute force' a win outright instead of delving into the complications for a trickier win. In my experience your fear that battles will become trivial is unjustified because once micro is reduced there are more complicated tactics to apply in order to get an edge. Micromanagement may look like it contains all the depth of the current battles because it is by far the most useful.

I do not claim that the current implementation of PA would not lose a lot of depth with the addition of automation. The balance may be such that a particular unit is too powerful when it kites competently on a large scale. To me the existence of such a unit is just a timebomb waiting for someone with ridiculous APM to exploit it. Automation would reveal these issues and I know from experience that it is possible to balance diverse units with automation involved.

Here is one way that battles can be complicated and it provides a good way to prevent someone solving battles and producing some optimum unit composition. Battles do not all have to have the same aim. Here are some examples:

• If you are fighting a slow war of attrition you may want units to individually retreat when they have low health. A retreating army will lose ground but will have fewer loses. Whether you want to fight this type of battle depends on your valuation of your units vs. the ground they defend.
  
• If your opponent is about to receive reinforcements from elsewhere to defend an important target you probably want a fast battle. To do this it could be best to run your units in rapidly to close the distance and reduce the effectiveness of dodging projectiles. This tactic makes your units more vulnerable to fire but finishes the battle more quickly.
• Target prioritization within the opposing army can change depending on external circumstances. If you have a powerful airforce and your opponent has some glass cannon anti-air you can use a ground force to destroy those units and retreat. The defender wants to protect high priority targets but has to read their opponent to figure out what the high priority targets are. This depends on their scouting.
  
• A raiding force wants to run through territory and kill mexes. When it is engaged in battle it wants to move in a desirable direction away from the opposing army. The opposing army has to decide which directions it wants to block.

There are a lot more aims for battles. They can be very subtle and depend on a players overall strategy as well as what they think their opponent is doing. These battles could be managed with orders such as "stay at max range", "retreat when damaged", "advance carefully" and "prioritize killing this unit over your own safety". These orders are simple to implement in terms of unit movement and do not take the main choice away from the player; deciding the aim of the battle in the first place. It is still important to think and implement things quickly as the aim of a battle can change rapidly with new information.

The aim of a battle is also related to your unit composition and it is important to choose units which work well in the types of battles you plan to fight. A raiding force should be made of fast units. A slow attrition retreat force would have weak long ranged units. A dive and snipe type battle depends on what you intend to dive.

If it takes a lot of micromanagement to implement orders such as "stay at max range" then battle plans which require this have a higher cost. High micromanagement requirements for basic concepts forces people to practice implementing only a few of the most useful basic concepts. If someone wants to change their battle plan they must learn a new bit of micromanagement. This puts a high cost on changing your battle plans as the improvement gained from using a plan which fits the situation will be swamped by the complexity of implementing a less familiar combination of basic orders. In games where battle micromanagement is very important the aim of many large battles seems to be "destroy the other army" because it is a decent plan and can be achieved through superior micromanagement.

I am not saying battles with a large basic micromanagement component are an inherently bad thing. There are many games with that type of battle. It just does not seem to be what PA claimed to be.

This thread is worrying because it makes me think that Uber will not add this type of automation. They have often said that they want to make a powerful UI but have previously taken steps in the opposite direction. They also seem to hate automation with justification that it violates WYSIWYG. So unless there is large community support (which there apparently isn't) they will not add this. But it would be really nice if they would chip in.

Also I dislike Maneuver and Roam. I set all my units to Hold Position because I do not want them baited and killed.

First off, awesome writeup.
Secondly, I only agree with your quoted argument in the case of automation strictly dominating manual control. If the automated behaviour is very basic and thus bound to be less efficient than manual control I consider the the tradeoff that optional automation presents - namely between attention spent and efficiency achieved - to be a valid argument in favour of it. It opens up a new choice which is a good thing in my book.

If I might illustrate it with a Rise of Nations example? If not, feel free to skip the rest of this post.
There are designated scout units that you can order to scout the map autonomously - among other nifty things. They will try to explore as much territory yet unexplored as possible and will generally roam the map in a circular fashion. If you have a bunch of them that might be sufficient for your scouting needs, given that they are fast and stealthy. Controlling them manually on the other hand allows you to explore the most likely enemy positions first, as well as collect the various one-time resource pickups scattered all over the map - a huge advantage over the auto-roaming scouts which do not hunt them down nearly as effectively.
So why is the presence of this option a good thing?

• Everyone has easy access to scouting. (Build Scout. Click 'auto-scout'. Get free intel.)
• Thus everyone gets to experience the usefulness of scouting. (Tutorial function for newbies: Scouting ---> Good)
• Manual control is worthwhile. (Auto-scouting very basic. Resource pickups good early game. Useful abilities on scouts)

I am sure that not all of this is applicable for PA. I do think, though, that automation can be implemented in such a way that it is not only helpful but also meaningful. Your examples regarding battle-stances show ways of achieving this.

How I at least visualize it is something along the lines of how units had "stances" (or whatever they called them) in AoE II.
In PA I could see it being something like:
All units have 3 possible "stances" they can be in, which dictates their idle behavior (i.e. what they're doing when not given other orders (things like attack and patrol would be separate from these))
A "default" stance basically being how the units operate now when they see enemies
A "aggressive stance" that would have units move towards to close to engage enemies that they see
and a "defensive/skirmish" stance that has them try to move away from/out of range of enemies that come too close.

Seems relatively simple, straightforward, and just gives more strategic versatility without over-complicating the game.

So long as the balance is changed to compensate for this.

Yes, you're correct. automated skirmish ensures the optimal outcome will always occur.

No not really. People are doing this because even though there are so many other directions to attack from, they as players are simply not practiced enough at thinking about sending units around the flanks. It has nothing to do with attention and time management but all about the trap of reactionary tunnel visioning that many players fall into. Just move order queue 3 tanks to the mex fields behind your opponent from the left, right, and behind zones so they plow through every single one of them. Set and forget, no micro or babysitting required. It's really that simple in PA, but players just don't have the confidence to spare a measly 3 tanks for some reason.

I guarantee you, if you played games focusing entirely on your macro with heavy reliance on radar (expanding metal safely, building lots of factories and energy), and only used queued move orders to send your tanks out in groups around the flanks of your opponent, you will win alot more of your games than you are now. I'm not asking you to babysit them or perform any ammount of combat micro at all. Just spend two seconds to set a move queue so your opponent has to spend time dealing with them at no time cost to you.

This tunnel vision tug of war is not an accessibility issue that requires automation, it is a player skill issue. In PA, much more so than any other RTS game out there, time spent on micro is of less value than time spent on macro. If you are losing because you are "trying to keep your units alive", then maybe you are spending your time very very wrongly (neglecting macro) or playing against opponents who are just better than you. Both of which are not valid reasons to automate skirmish. I guarantee you, if you collated all the data on pa stat games, you will see that the winner almost 99% always has a higher metal spent number.

I want to very strongly re-iterate that RTS is all about a game of resources. That includes not only ingame resources, but 'out of game' human resources. Otherwise what is the point of a human playing the game? Why would someone play a game if he cannot bring his own unique resources to bear? Why does the game need a human to play it in the first place? The developer decides how much weight to give each type of resource depending on the target audience, but at the end of the day you need to have a healthy balance of both.

This doesn't justify automated skirmish. Yes it is an interesting maneuver indeed, but I don't see how this is now suddenly possible because skirmish is automated - there's nothing stopping you from doing this exact thing now if your human opponent retreats his units like this. All you're doing is fighting the ai and not the player. This is like two people playing a single player game but in multiplayer.

@GoogleFrog
It isn't an oversimplification. Skirmish automation gives the player both unlimited time resources and optimal movement. It is absolutely without a doubt cold and calculated what units you build and where you use them. You are basically accelerating one aspect of the game to it's end point by taking out human error and limitation. There will be only one equilibrium.

If i were to ask 1000 people to solve 48474*28273 in their head, some will get it wrong. When i ask a calculator to solve this, it will get it right 100% of the time.

I am not dodging macro complexity. What do you mean by this anyway? You're assuming players are not playing equally skilled players to begin with, and also assuming that PA is so much more deep than the tank/bomber spam it is.

Even if i wanted to, PA is hardly the game you can bruteforce a win with micro. If i wanted to play that sort of game i might be inclined to switch completely over to sc2, and even then, it would still be very very difficult to 'brute force' micro there. In ANY rts, to become a good consistent player you must have solid macro.

I think you are gravely mistaken to think the rewards of micro management actions are currently so great as to allow macro management actions to become swamped/irrelevant.

I'm sorry, but this is blatantly false. No one is asking for smart unit AI here. Most of your arguments seem to hinge on the 'this ensures optimal outcome' argument, and it's starting to bug me how you keep bringing it up as though it were undisputed fact.

Some of us are just asking for the option to set units to back away from the range of enemies in a straight line, whether that be coupled with the pre-existing auto-attack mechanics (as skirmish), or if that seems like too much, possibly just have it available as an alternative option to the pre-existing auto-attack modes (as evading / retreating), which would force you to selectively choose your units' engagement behavior.

Anything that isn't "slowly back up" (or "move into range of the enemy", which is already in-game) still wouldn't be automated in either case. If the second option is used, each behaviour type would have major pitfalls, and choosing which one to use for a set of units is a strategic choice.

But only having the option to automate one half of the move in/move out dichotomy has caused and will continue to cause problems, especially with units like vanguards and infernos.

Ok respectfully, I'm done. Involving myself in threads like this where some posts can't even stay internally consistent has never been good for my sanity.

I wish you luck and hope someone makes a mod for automatic skirmish.

Honestly, I'm glad. You've had a constant insulting tone and demonstrated blatant disregard of people's arguments without addressing them, calling them things like 'pseudo-intellectual', 'fools-gold logic', 'internally inconsistent' without addressing a single point. Saying people are 'in an idealist bubble', equating arguments to 'saying the earth is flat', 'sticking their heads in the sand and repeating themselves', usually without any counter-arguments beyond insisting in your own correctness a priori.

The pot calling the kettle black

Well, he turned to a more constructive tone and his arguments are well thought out, actually.

But he still seems to revolve around some assumptions which i don't believe to be true for PA.

The only comments I make that could carry any kind of demeaning tone will have a follow up explaining them. You can look through the thread if you want; I don't call people pseudo-intellectual, I don't tell people they have their heads stuck in sand, I also don't accuse people of hypocrisy with a one line statement.

Would you call that last post well thought out or constructive?
It's little more than a thinly veiled insult.

The pot calling the kettle black.

Do you think calling out a post as insulting is hypocritical? Read through those quotes. What kind of response do you think that would warrant, none? For fear of hypocrisy? You could explain your reasoning instead of using unsupported one liners.

I'm a bit fuzzy on this how does it eliminate the meta-game? I ask because I want to know the reasoning behind it. The definition of meta-game is the game that is around the game proper as long as Uber continues to update PA by adding new units, etc. The meta-game will continue to evolve and new strategies, tactics, etc will arise. Just like it does in games like Magic: The Gathering.

Secondly, I want to address this to everybody this need to reduce micro was evident while PA was still in alpha and before that time. You can see this by threads popping up back then about this topic. specifically a thread called Mechanics to reduce micro. Where I am quoting this from:

I believe one through four are very necessary for PA but not five in most circumstances. Due to the fact that people cannot be everywhere at once. Your brain just simply can't do it; PA is simply planned to be too massive for it to be reasonable and practicable.

Nope, not worth the time, pot.

The attempt, not the success.

Then you've given me no reason to take you seriously.


Now, I would like to flesh out my argument for a separate 'evade' engagement mode, to complement the current 'maneuver' one.
This first bit will be a bit repetitive to anyone who's read most of my posts here, but I think it's more concise than my earlier posts, and the rest follows from it:

So skirmish mode may lead to too many hard counters and too few tactical options when it comes to unit positioning / micro. The current system is very one-sided as to what unit types require micro, so countering those units requires babysitting and wasting time on trivial retreating commands.

So now I'm going to go into why I think a new engagement mode could be a solution.

Let's say you have 3 options for engagement modes (I'm removing 'roam' because it isn't too different from maneuver):


Maneuver: Still the default engagement mode. Units automatically move to put enemies in their guard radius in range and auto-attack.

• Pros: Higher DPS for group encounters due to swarming, good for guarding an area against scouts, etc. Good for attacking with low and medium range units.Good for attacking with large forces.
• Cons: It's possible to bait units that are set to maneuver. Enemies are not kept at optimal range, and may close in on longer ranged units if not microed.

Hold Position: Units do not react to enemies besides auto-firing.

• Pros: Cannot be baited.
• Cons: Can be avoided by enemies without worrying about their guard radius. Usually much lower DPS in groups, as they will not swarm enemies.

Evade: Units automatically move away from enemies when they are too close to their attack range (so long as you have vision of them).

• Pros: Moderate DPS for group encounters as units will get 'pushed' into swarms by approaching enemies. Good for survival, especially against slower and shorter ranged units. Good for attacking with long ranged units and support units such as combat fabbers.
• Cons: Can be exploited by displacing units with longer ranged ones or artillery. Not effective with only radar coverage (cannot tell enemy range). Bad for attacking in most cases.

This gives you an option which can be used to avoid close-range threats, and to preemptively set some of the behaviour within groups of units (for instance, clicking 'attack' on a turret with a group of units wouldn't send your combat fabbers or AA veheciles into it when using this option; right now they have to be de-selected manually and moved around separately to do this).
It also addresses the concern about hard counters and stale combat, since each option is situational depending on unit type and switching between them is manual. If a group of dox is set to evade, they may rip through an army of infernos, but supporting the infernos with laser tanks may allow you to push the dox aside and move into enemy territory relatively unhindered if your enemy isn't paying attention.

Of course, lots of theory-crafting here, I don't have the tools to test something like this (at least I don't think so). And while the usage of 'evade' would be quite limited when compared to 'maneuver', I think this could address some of the concerns about the current engagement modes.

Edit: Now with fancy attention-grabbing icons.