Battles without micromanagement mode ?

This is a brief overview of what i would expect from AI to be able to micromanage and also the idea of a command center included in UI (1 icon for each pool of units) with reporting about the battlefield.


I see this as being an option, not a removal of micro. But for large planets I hardly see how one could manage hundred of units accross many planets without having such a system taking advantage of A/I for micromanagement.

i think somebody actually made something like what you describe for forged alliance

you could assign an AI to take over control of a base or units. i think you also assign AI to tasks like building bases and harassing your opponent.

i can't really remember how to find it though, it was on the GPG forums

edit: found http://forums.gaspowered.com/viewtopic. ... cb67390ebd

Apparantly the original idea behind support commanders was to have them manage a base for you?

SACUs could resurrect structures if they were directly assisting something when it dies. It was a tad silly, because Supcom resurrection costs 50% of a unit's mass, when reclaim+rebuild only costs 20% mass.

If I'm a bad enough player that I can't pay attention then yes I will accept that as a loss.

I'm okay with there being built in AI routines for certain things, AS LONG AS no control is ever taken away from me. Manual play will always be more effective than automated so a player with good micro will always come out on top in roughly equal confrontations.

Why would you want that? In any engagement you want to take down the incoming DPS as quickly as possible. So you would either want to target the Loyalists first or the MMLs if they were in range.

If you're focus firing you always want the quickest kills possible, except if there are some truly negligible units in there (given a Percival's 1600 damage shot, killing T1 with it is a bad idea, for example).

------

(for the benefit of anyone who doesn't know FA, the Loyalist has 1/3 the health of the Brick, and about half of its DPS. The MML is basically an artillery support unit with double the range of the Brick, but 4½ times less health than a Loyalist so cannot take much punishment. That all sounds as confusing as hell, this might make more sense)

If the bricks had the advantage of hitting me while I can't hit them I would retreat anyway and most likely would be in range of the loyalists instead (and would them fire on them).

Well actually i'm much more looking for avoiding single unit micro management, especially when many squads have to battle against each other. This of course requires a very good AI and possibly being able to learn from its mistakes.
Just because playing with tons of units looks to me a bit incompatible with unit micromanagement.
I see this either as an additional game mode which would much more strategical instead of being tactical or as an option you can choose to play with if you know you're not good at micromanagement and if the game is really strategy oriented, meaning that even if you're not good at micromanagement, the decisions you take make you win.

I don't believe anything needs to be taken *away* from the type of unit control that Supcom and TA always had.

I DO think it would be good to take Supcom's UI and interface, polish it up just a tad (Orders as a first class entity for example), and then also on top of that start putting in place some higher level AI helpers - more complex orders, squad / battalion AI, that sort of thing.

That way you still have your intense, 1 tank vs another tank kind of fun, while allowing the game to scale immensely beyond that.

Supcom, once it gets past that beginning stage, really starts to fall apart - units react slower, the pathfinding tanks, and so on. Not only that, but silly stuff like a unit shooting the ground and killing its buddies is ok if you're there to look at it, but the more units you've got on the field the more that just becomes an irritant, not good gameplay. I really hope Uber spends a good amount of time tackling that mid-end game stage.

If units shot into straight into a cliff then that is a bad priority. In what situation would you wan't your units to fire into a cliff?

And boy is this micro heavy to avoid.

Imagine standing on top of a cliff a bit away from the edge. The enemy is below the cliff. When they come in range your units will start to fire into the ground in front of them. If they do friendly fire they are even likely to hit the ground and do damage to your own units. You have to turn them to hold fire, order them to the edge of the cliff and turn them to fire at will. Then you have to check to see if every unit is actually close enough to the edge so that they don't end up hitting the ground anyway.

What if units are smart enough to avoid firing straight into the ground?
They don't end up doing damage to your own units. You don't have to excessively micro the units to use the hills to your advantage.
Now that you have control of the cliffs you can go near the edge and shoot the enemy and retreat if you are taking too much damage and your units will be in cover in an instant.
That is much more tactical allowing hills to have a natural strategic advantage.

And what if you could actually give an attack-move command to the edge and units will automatically stop to fire when the enemy is in Line of Fire?
You could even use hills as cover from certain angles without exposing your units unnecessarily . Cool huh?

How do you people draw the parallel between RTSs like TA and MOBAs?
The only thing the have in common is a top-down view, that's it! No matter how much Uber doesn't give a **** about this game, they can never make it into a MOBA

DOTA started as an RTS itself, Warcraft 3. Had a custom mode called footman wars which was based around the idea of units spawning and managing themselves. Led to a lot of variations such as tower defense and DOTA notably. DOTA in particular being focused on managing your hero unit at the expense of all other aspects of War3. So, a one-unit RTS.

MOBA, ARTS, whatever funny term you want to use it's really Warcraft 3.

Ok, so we are talking about an in-addition-AI-system. This thread is getting better.

In TA there was a super computer unit. Now, I am not going to argue that this unit would be viable in today's RTS, because the way the super computer worked is it allowed units to fire at radar blips out of line of sight. Obviously requiring this unit to have a function that we currently take for granted may not be popular.

Ergo, when FA was created the idea was you could have a base assistant instead. The base assistant would help you grow the base. However, as we well know, base planning is crucial to winning annihilation games. Automating that could take the strategy out of the game. GPG had the compromise that support commanders could assist economically by being resource producers. This was relatively cool. By the end of a scenario in FA I generally have more support commanders than any other unit, and ironically the micro to get them all producing metal and energy is atrocious.

Doud a fine addition to PA that Nuetrino has already confirmed is Area Commands. In Zero-K you can grab some units and tell them to reclaim, attack, build mexes, etc in an area. Area commands are actually extremely useful because you pick the area and see the boundary provided. I suppose if an area fully encompassed a planet, even the back side of the planet would be wrapped as well. It's a very accurate macro tool thus satisfying high end game play.

I am open to the idea of AI units in the game that can be toggled on and off including the Super Computer and/or the Support Commander. In general, if you have multiple operations, being able to provide yourself with temporary AI support on rocks you are not currently concentrating (planet, moon, asteroid, roboconstruct) could be cool and if there was some way for the enemy to destroy your AI (support commander) this could also be cool.

Why did my bots stop building up on one of the moons of Jupiter? Oh my support commander got fragged by a small strike force and I haven't qued any new commands there recently.

I am leaning heavily for this to be a mod though. For example, if someone stops my nukes because his AI built him anti-nukes he wasn't even aware of getting placed I would be frustrated.

As a resource producer support commanders are pointless spam units and need to go.

As an AI companion they simply would need to be predictable in their behaviour.

Why?

Because they're pointless spam units.

I just use them as mini commanders. They have formidable weapons, very high HP, and good engineering capabilities. And on seraphim and aeon commanders you could teleport them around the map. This gave you the ability to VERY quickly set up firing bases to attack your opponent from unprotected angles (or to spam up a quick experimental).

They are almost NEVER used as resource generators or "base assistants".

All I see people do is amass them in ridiculous numbers to generate resources.

I like all of this.
But as i read my post i realize i may have splitted it. It's not only about micromanagment, but also the reporting stuff in real time. Having to check many positions on a regular basis to ensure nothing wrong is occuring or everything is working as expected and according to assignments is quite a bit difficult. I know many players consider this as being part of a classical standard RPS. So see this also as an option :

I would like to be notified when a specific squad / Area is under Threat and requires an immediate action from my command. An this because i'm not expecting the AI to anticipate and decide what to do when an assigned objective is about to be failed. I want to be notified that there's something to do right now.

I see this as a command center where a flow of information is comming in.

Now, the question is : Should any unit be able to report, or does it require a specific unit to be included and protected in each squad / zone ? Having such unit may introduce the possibility of losing contact if it's destroyed. Meaning that AI should be smart enough to perform both the following things : complete achievement which has been assigned to a unit squad and having the unit squad protect the unit which is here to report information.
Losing contact would mean something is definitly going wrong.

If you give UI / Game improvements to the gameplay, it should not be linked to a specific unit.

A more reasonable way to implement it is to make a higher level organisational unit - something like a squad. When formed, one of the units get automatically 'promoted' to squad leader - this unit is then the source of information on the squad.

Squad under attack? A window pops up zoomed in on the squad commander. Squad idle? An icon appears in the 'idle units' tab. Squad leader is destroyed? New unit is automatically promoted and alert is sent to the player. Select a member of the squad, and you automatically select the whole squad. And so on.

The idea is that you want to reduce the number of orders required to get large amounts of units moving, and you want those units to be as smart as possible in terms of executing those orders.