On Being A Commander

One thing that has irritated me throughout every single RTS game that I've played is that you're supposed to be a 'commander' of an entire army, and yet for some reason the tools at your disposal are dismal at best and non-existent at worst.

For a game like PA with several combat arenas spread over different planets, wouldn't it make sense to give the players the tools at their disposal that a region commander would have?

Make things as 'Smart' as possible

Some people like to argue that it's important that units don't become too automated, that if units fought in the most optimal manner it would become a game of AI wars with little to no user involvement.

I strongly disagree. The biggest problem Supreme Commander ever had was that it was designed as a grand scale RTS, and from there it went downhill into the tactical RTS trap. I'm not against tactical games, but frankly, there are lots of tactical RTS games that do it just that much better than Supcom simply because they were designed around tactics.

As evidence, look how terribly bad the gameplay is in Supcom when you start getting to the level where you're smashing large armies together. You get path finding issues, your units act stupid, the power multipliers of several units (shields, artillery) make bases impregnable, and unit control is at best relegated to designating your main unit types into 'Group 1, 2 and 3'. 99% of all the battles with 100+ units is simply 'select group, click in middle of enemy base, occasionally tell entire group to shoot something.
Yes, it's still amazingly fun despite the issues, but that's where PA has tons of room to improve things.

A commander has no business giving individual orders beyond a certain scale. I'm not suggesting removing individual unit orders, but I do believe it is very important to have the ability to PLAN AHEAD:

1. Have an order/plan system that is completely separate from the units, so that you can plan & set up attacks in down time while you're building your troops. Make it possible to modify the orders, specify units to carry out the orders, and coordinate movements between units. Give a listing of the units required to execute the plan, and show a real-time summary of which units you have assigned to the plan compared to how many you need. Allow for multiple plans, and track the success rate of the plan. Assign factories to the plan so they can automatically produce the correct # of units required. Hold down a key, click & drag to shift the entire plan around.

2. Automate individual unit combat to be as smart as possible. Allow players to designate retreat zones (part of the plan) and specify targets that must be taken out at all costs. Artillery should retreat if being attacked by ground troops, long range skirmishers should kite oncoming enemies, but let the player make those decisions in his planning.

3. Economics / Base management. Automated building replacement. Base templates. Same idea, set up an entire base in your plan, assign some construction units to the plan, set the economic priority of items and leave them alone to get their thing done. Use an existing base or stored template to quickly create a required layout. Since a plan isn't unit-dependent, your base can be tasked with the job and automatically replace engineers that may get destroyed.

4. Record the statistics for each plan. How many resources were used to build it, how long did it take to complete, how many enemy units did it destroy, how many friendly units were destroyed, how much is it producing etc.


Put the information a commander needs right at his fingertips.
- You should be getting reports on the capabilities of the enemy, estimates of his numbers, his income, his build potential, what he was building when you last saw his base, any troop movements and where they were headed & how fast since the last spy report.
- You should be getting reports on the enemy's tech-tree progression, whether it be research, or tech level, or infrastructure upgrades. Identifying an enemy's progression by visually watching a spy plane fly over the enemy may be good for a small scale game, but managing that with multiple players on several different combat zones can quickly become unmanageable.
- Be able to reverse-engineer wrecks to identify what attacked you and why they were kicking your ***.
- Be able to select and compare your progress to what you know of the enemy - he might be in a non-threatening huddle on another planet, but if you know he's out producing you 3:1 you might not feel so confident about leaving him alone.
- Spoofing intelligence reports should be possible - physical camouflage, decoy heat signatures, whatever works.
- Radar should identify incoming aircraft / spacecraft, but line of sight should be more important for ground movement. Give players the ability to punch holes in an enemy's intelligence gathering (take out patrols / watch towers) before an assault to make it harder for the enemy to return fire. One big circle that detects everything makes it real hard to maneuver or surprise the enemy.

And so on. The key to making this actually work across multiple combat locations is going to be to give the player the true commander role, and reduce the dependence on both unit and intelligence micromanagement.

Along these lines, something I always thought was weird is that you have no way of tracking progression of your own projects aside form visually checking on them. If you've watched and Starcraft2 casting you'll know the replay viewer as whats called "The Pruduction Tab", it basically shows everything the players have in productions, shows what(unit icon), how many and how far long it is. It also stacks multiples of the same unit and shows the progress of the one that's furthest along, switching to teh next furthest along when the first finishes.

I've always thought it was weird to not have this as a player, and I think it fits in pretty well into having the right tools.

Mike

I definitely support more tools for the player to make strategic decisions (e.g., production tab), and I would also support the reduction of stupidity (I'm using this instead of 'smarter units') -- I mean, units should be able to pathfind well, and I hope this is improved in PA (over SupCom and TA).

I'm not sure that all units should be smarter, but it would be nice if there were some smarter units: higher tech-level artillery targeting the center of a mass of enemies rather than individuals, etc..

Overall, as long as the planning system is done by the player (and as long as players can manually override plans or make them as they go along...) I think this could be a nice system.

In general, I agree with all your points (first and second posters)

Well except for a screen that shows me special project progress, that could be nice, i liked the interface so far. Not overpact, a good overview, with a simple style to use. And for quick decisions the keyboard shortcuts.

If you bring in that Tabs as optional, okay. But pls not fixed. Couse i for my part wouldnt need this tabs that show me all informations around. It would only block my sight for the important thing. The Action on the Battlefield.

Having a construction template just takes away some of the fun! Being active with your decisions and continually building what you need keeps the player focused and makes the experience more fun!

Having templates(player created ones) takes nothing away from the fun, in fact it lets you focus more on the fun aspects of the game. I'd say the majority of players that building bases is one of the least fun aspects, why not streamline that process to allow for more time to be devoted to the better parts of the game?

Mike

I like the idea of the visual overlays describing information about yours and your enemy's base, but only in the later game when things become inter-planetary and harder to manage. Maybe have a buildable intel unit that allows you to access this information, like the "Targeting Facility" in TA, that takes a significant amount of resources to build?

People like me, who are into games like Tropico, Anno and simcity normally like base-building... and since those games are immensly popular I'd say quite a large amount of players likes to build their bases manually.

And even if templates were in the game there is nothing stopping you from doing so.

Mike

For some reason I keep referencing Sins in these topics... SoaSE had a way of letting you keep an eye on everything at a glance by keeping everything organized into dynamic icons arranged in a big scrolling sidebar on the left. Anyone who's played SoaSE knows what I'm talking about - that game was very good at making it easy for you to keep track of the insane amount of stuff that was going on. It allowed you to keep track of what units were where and what their status was. Very nifty for organizing all of your info and allowing you to keep an eye on one location while being in another.

I never used it! xD

I sort of agree on some levels with the direction you're going with this for one simple reason, scale. PA is promising to be bigger and badder than anything else out there. It's going to have more units, more terrain, more everything, than anything else out there which makes even the idea of trying to wrap your poor mortal fingers around the controls to manage all of that as you might in a more traditional RTS a bit ludicrous.

But it also raises a tough question. What, and how much, should you automate?

What can you realistically automate without rendering the game unplayable on the average home users hardware?

Do you automate the Macro side of things more making it easier for players to focus more on the battles, because battles are naturally more fun?

Or do you automate the micro side of things and allow players to focus more on the sim city side of things ?

Just where should the line be drawn between automation, and interaction? What should we be allowed to automate, and what should we not?

It's not an easy question to answer is it?

I remember during the production of supreme commander there was talk of a base AI that would do things such as automaticaly rebuild destroyed structures.

I'm not sure how i feel about that but it seemed like somehting that should be brought back up.

Scale is, indeed, the biggest challenge that PA will have. Supcom reached so close to the goal, but it fell apart at the end due to what seemed like a refocus on what the 'general public' might like, ie small tactical battles.

I believe that the only 'bad' sort of automation is where you force the user to relinquish control. To use a SOASE example, the 'automatically place buildings' button is a crappy sort of automation - it made things a bit easier yes, but it was basically a 'we're going to randomly dump buildings all over and you have no control over it'.

A good sort of automation would be something like setting engineers to 'expand resources' and they automatically go out and claim all resources near your base in the shortest route possible.

A better sort of automation would be to allow for planning your base on the moon first, and then launching your commander up there and being able to ignore him as he sets up and completes your plan.

I don't like the idea of paying for intelligent units or UI upgrades (either limited to some units or enabled with some structure). I think there is something to be said for consistency in unit intelligence and the UI. If I have a unit I want to know how it acts without checking if some structure somewhere has been destroyed.

In general I agree with OP although there is always the question of what should be automated. Clearly maximum automation is bad because that would be an AI that executes perfect play. Everyone's opinion will lie somewhere between no automation and full automation. I think I am fairly high on that scale but as an example I am not sure about 1.

My feeling is that all the automation described here would either be far too smart or useless. Specifically if you have units that execute an attack plan then they will either have to make important decisions by themselves or the units will act too stupidly for this to be entirely automated because no plan survives contact with the enemy. But most of the automation in this suggestion looks fine to me.

I don't want to reduce meaningful player choices with automation. So my automation limit is when the effort required for a player to enter all the contingency plans outweighs the complexity of the task avoided with automation.

I just thought up and example for this. Say you have a system which allows a player to configure exactly when their army should do a full scale retreat. I think this is is bad because there are so many complex factors to take in to account. I would not use that automation system because simply selecting everything and right clicking away from the battle takes a lot less time.

Honestly Googlefrog I think Zero-K has a lot of it right, I just think that it would be ideal to be able to break the link between unit & orders entirely, and treat plans as entities in their own right which can be modified in an easy manner. Zero-K makes it easy to issue orders effectively, which goes a long way towards reducing required micro, but it doesn't solve the problem of assigning the similar orders to multiple successive waves of units.

I wish I knew some way of making some animated illustration of how it should work. It has to be simple enough to work with, and yet powerful enough to make it worth while to spend the time making those plans.

Base building would definitely benefit. Also, being able to plan an assault before you even have any units ready would be good. Estimates of the plan's cost & time to completion would be necessary.

This reminds me of the Ferry system in SupCom, where you could tell a transport to ferry to a destination, rally units to the ferry point, and they end up at the destination via transport. I suppose if you could set an order queue to be tied to a point on the map, units could rally to it and then execute that order queue.

This is easier, as base template systems have been done in FA, so it isn't as much to do originally.

Love the idea of defining plans independent of units and assigning them to it. Especially for things like synchronized attacks. Here's an example of how synchronized attack routes could work:

1. Define attack paths.
   • Click at staging area, drag along route and release at destination.
   • (repeat as needed)
2. Assign units/factories.
   • Select units/factories and right-click attack path to assign.
3. Group attack paths.
   • Select paths and use shortcut to group them.
4. Watch the destruction!
   • Attack waves now arrive in sync along attack routes at timed intervals at an attack destination.

(Disclaimer: I might re-post how synchronized attack routes could work in a more topical location )

*for more UI-power*

There already have been a couple of threads on different aspects of more UI power and the general consensus was always in favor of more UI power. I hope for an UI that is as powerful as possible while being as simple, intuitive and quick in usage as possible.


Here are some of the discussions on improved UI power:

Ideas from Zero-K

Global Construction Orders

Powerful UI or Not?

On tactical formations and movement

On production queues (advanced factory UI)

Suggestion: Add WiC-like communication system

Advanced construction templates


The circumstance, that most of those threads are created by me is because, as I said, I am *for more UI-power* and the discussions I started are the ones most present to me... this there might still be some other threads on more UI-power, that I have missed... so I am not intentioanlly and specifically advertising the threads I started.

More UI-power is fine, but a powerfull UI is still not enough when your units get smashed once you are not looking or if there are some very boring yet game deciding phases like initial base construction.

What i agree on:
There should be persistent move / patrol paths which you can assign units to like you could do with ferries. You could do something similiar in SupCom and FA by setting the path for the factory and if you changed the path, all units which were still moving on the path adapted to the changed waypoints, but it always lacked the possibility to combine that with "gather" commands or to make an attack plan in advance, gather your troups and then just execute the attack. Using the paths on factories was quite stupid since it would just have produced a steady flow of units which got wiped out by enemie defenses.

But:
Units should also act less stupid when NOT executing an order or when on a patrol. You already made a tactical decision for this units, so why force the player to micro? He sure has more important tasks at hand and it's just not fair if his units get outsmarted by someone who simply has the budget to afford multiple monitors and is capable of microing multiple frontlines at the same time (or just beeing able to react faster, thanks to more viewports). Having the AI playing PERFECT at that level is not much of a problem, it only would be if this feature would only have been accesible for one side.

Preserve those features for the competive part which are acessible in the same way to every player, independent from PC power or effective screen size, but don't even try to preserve features which require fast reaction or which would possibly require attention on multiple sites at the same time.

Having multiple or larger monitors available may not become equal to cheating. That's by the way the reason why the maximum viewport in WC3 and Stracraft 1/2 was always the same, independend from monitor size, aspect ratio and screen resultion. They had to put artificial restrictions on the viewport to prevent users with large screens to get an unfair advantage in PvP because beeing able to micro at multiple sites at once meant an huge advantage.