What are the planned unit roles?

I think it's possible to carve out many distinct unit roles in a complex ecosystem if you have e.g.

1. Different movement types affected by terrain e.g. bots, vehicles, hovers, amphibs

2. Different weapon types (e.g. lightning guns with temporary stun, heat rays that do more damage up close, weapons that do little damage but push the enemy away, etc.)

3. Special unit characteristics, e.g. ability to cloak, all terrain movement, ability to deploy to increase HP.

4. Unit synergies e.g. linkable shields (see below), units that do no damage but instead stun enemy units or slow their movement/rate of fire... must be more possibilities here?

I agree that if there are lots of units they should all have a unique niche in the ecosystem. They should also be designed to work together.

There's one really nice unit synergy I'd like to mention from Zero K. The shield bot factory builds bots with small personal shields. When they stand close to each other the shields link together. One unit (felon) depletes shield charge to fire a powerful lightning weapon. Felons are weak on their own as they quickly deplete their own shield and stop firing, but in a shieldball where they have access to a large shared shield charge they can be devastating. I think imaginative unit interactions like that are really cool.

Also one way to cut down on unit numbers - just make basic structures buildable on multiple terrain types. For example in ZK wind turbines can be built on land or water, fusion plants can be built on land or underwater.

I prefer if you act on the ideas instead of just listening

I wasn't advocating do-everything units, I was advocating do-more-than-one-thing-units. Of course there is also a point at which more roles for one unit start reducing the overall complexity of the game.

Using Anti-Air as an example probably was not that good of a choice, but having only one role makes it very boring to use. It is basically like this: Opponent builds a lot of air units -> build Anti-Air
Opponent stops building air units and you have enough Anti-Air to be safe from the air units he currently has -> don't build Anti-Air

In a more generalized fashion:
Do I need more units with role X?
yes -> build the corresponding unit
no -> don't build the corresponding unit

Multi-role units allow you to fill in the role you need, but still build a unit with a role which helps you carry out your game-plan.
Like a unit, which provides ground support fire from the back of your troops, but also has short range anti-air-missiles or a fast machine-gun unit, which can attack both ground and air units. Both provide anti air capabilities but which one you build depends on the situation. The first one can support a ground push better, while the second can also function as a raiding unit.

You can't shoot for awesome without awesome units
(On a sidenote: One of the reasons why people love the Starcraft games so much is because they have really interesting units. Especially those from the first one and its expansion.)

What I am detecting in your reply is that you find it dull when you only have one choice of unit to do a particular job.

Case in point- anti-air. If you only have "the anti air unit" and if the enemy is building air, then you must purchase this one type of anti air unit, and nothing else will actually do anything to help, then you rightfully complain about bad design.

Now imagine you have several choices of anti-air unit. You have a cheap, short-range flak vehicle unit that deals small amounts of splash damage, and is effective in groups. You have a mobile SAM launcher unit that is expensive and fragile, but has excellent range and stopping power, and a large minimum range within which it cannot target enemies. This unit's drawbacks would be cost, and a slow rate of fire. And you could also elect to use a laser anti-air unit that deals constant damage to its target. Or another one with a short-range, giant splash damage missile with a tremendous cooldown that obliterates entire clouds of air units in one shot. There are endless possibilities, even among dedicated anti-air alone.

Suppose that your opponent is using fast bombers in small numbers to conduct raiding missions on your outlying economic structures. A few strategically positioned SAM vehicles can cover a large area effectively, and will eliminate a few aircraft each time they enter its range.

Suppose your opponent is constructing a giant armada of gunships and you think he intends to use them to smash your ground army he has scouted advanding on his base. Now would be a good time to start adding those splash-damage flak units to that ground army, because they are numerous and not a point of failure (like a few SAMs would be) and they deal splash damage, which becomes increasingly effective the more planes he has.

I think it is dull, if there is only one unit choice, because the units are designed that way.

What I want is, that the unit choice isn't forced through the design, but through the gameplan you chose to follow.

If I plan to do a big ground push and need Anti-Air, I am of course taking the Anti-Air unit, which can also provide ground support fire if there are no air units around.
If I plan to turtle in my half and win through attrition and economic harrass, using the Anti-Air unit with raiding capabilities is probably the better choice.

I don't want to have to put back my own plans, just because my opponent happens to be building air.
It's like chess, if you only react, you lose. To be good, you have to be able to both react and follow your own plan at the same time.

Then it's going to take that one gunship a **** of a long time to kill your land army.

In my opinion, neither SupCom1 nor SupCom2 really got this right. In SupCom1 they spent far longer than they needed to on many units, because they had to have "Tank 1" "Tank 2" "Tank 3". This was *slightly* less bad in FA but still very bad. SupCom2 had the "My army magically gained completely new capabilities from nowhere!" aspect.

Ultimately, I'm backing FunkOff on this one. Relatively few units, with combined roles, and no overlaps, because then you're just wasting your time. It's OK to have a few units with singular roles if they are somewhat more powerful and investment-heavy- see, say, Siege Tanks or Colossi in Starcraft 2.

So my proposed units are something like this:

Land:
Peewee- relatively high DPS for it's low cost but not much health, moves quickly, can target land, air, and submerged with single-target weapon. Serves well against light harass from both air and skies, and can harass undefended enemy areas. Short range means not very effective in large groups. It's also amphibious.

Challenger- primary tank. Good range, high health, high damage, land only. Can also build mass extractors and repair themselves slowly- can't repair anything except self.

Sumo- very heavy tank. Quite expensive and slow-moving, but extremely high damage at a good range and high health. Sumos can be countered without too much trouble if you take them out before they get to you, but if you don't, you're in for a mess. Sumos can also cheaply and quickly create a weaker version of the static PD/AA. Is amphibious but defenceless in that state.

Flak- Not particularly accurate but carpets the sky in high-damaging explosives. Can also fire depth charges. Flak has medium range. Flak forces your opponent to send his aircraft in in waves.

Engineer- builds things. Also projects a stealth field and can clear landmines.

Commander- good health, good damage and range, plus DGun, but can never be repaired. If you use him early in the game and he becomes damaged, he will be a big opportunity for the opponent later. Can also construct things quickly and is amphibious. Is defenceless when underwater.

Buildings:
PD- simple turret. Relatively cheap with good range and damage but no splash. FA's T1 PD was fine, really. Can also build and repair wall segments in a fairly large area when not firing, which are cleverly placed to be at the right height (unlike those irritating walls in SupCom which never worked properly).

AA- High damage and range with a good muzzle velocity but no splash. Much cheaper than a few Flak and relatively high health. Effective against light air harass but heavy air will take it out with no trouble. Can also repair other structures around it when not firing.

Artillery- Not cheap, but mean. High splash, range, and muzzle velocity. If your opponent has unstealthed groups of units in the range of some Artillery emplacements, they won't survive long. Think deployed Siege Tank. Static artillery has decent health and a significantly longer range than mobile artillery, but is a lot more expensive. Artillery emplacements can also fire torpedoes against submerged targets.

Missile launcher- doubles as both tactical and strategic missile launcher. The launcher itself is relatively cheap, but missiles can be expensive. Tactical missiles do high damage but aren't very strong and weak splash, whereas strategic missiles are very expensive but high damage, high splash, and high resistance to being intercepted. Tactical missiles can be stored up and fired in bursts. Tactical missiles have a longer range than Artillery by a non-trivial margin.

Defender- can shoot down enemy missiles in a medium area. It's weapon can also do light damage to enemy aircraft and tanks with high accuracy. It takes many Defenders to shoot down a strategic missile, but one can do a few tactical missiles (4-5). Defenders can also shoot down strategic missiles going over their heads.

Radar- can detect enemy units in a fairly wide area, but less area than you'd find in static Artillery or TML. Is also a sonar.

Big Bertha- expensive, but can fire pre-built land units or artillery shells to any destination.

Air:
Drone- cheap, light air harassment. Medium damage to a single target, low health and range. Has a small radar and doubles as a scouting unit.

Bomber- Large and in charge, the bomber does high splash damage to pretty much everything in a line, and a few bombers can easily annihilate a contingent of tanks, but it's not cheap and won't hold up long under enemy AA fire. It's range is also low, so non-trivial AA has a habit of destroying it before it drops a single bomb. Can also drop Torpedoes to destroy submerged targets.

Gunships- Very high health and some medium single-target damage, the gunship can seek and destroy enemy AA units. You'll need many Flak, or some static AA turrets, to defend against Gunships, but if you mass Flak, you can kill any number of Gunships. Gunships can also transport a couple units, but they have no protection from enemy Flak.

Transport- can transport many land units, all functions are disabled during transport, and transports are weak. Units in a Transport are not vulnerable to Flak.

Flying Engineer- expensive, but you can stack as many as you like in a single space. Hope there's no enemy flak around.

Interceptor- very weak, but relatively cheap (still quite a bit more expensive than Drone). They move faster than any non-Drone air unit and pack a medium single-target anti-air punch. Interceptors can, when not shooting, lightly repair other air units and themselves.

Comments?

A tight unit list is a good idea, since it allows each unit to serve a distinct role and keeps the amount of information that the player needs to keep in their head down.

This will also help to keep development times down, and give some room for expansion if there's time/money later on.

Also, deadmg's list looks good, though I would add Jammers to his list, of the mobile and stationary variety, so that radar has a counter.

Let's not forget one of the most-loved features of TA was the extensive unit list. This is something every spiritual successor to TA has gone to great lengths to preserve, including 3rd party mods for TA, Supreme Commander 1 (not 2), and of course Zero K.

Land set needs mobile missile launcher like supcom2 imo .

Also, refer to some interesting air ideas to prevent one-dimensional "spam fighters and smash them against enemy fighters" here:

viewtopic.php?p=527592#p527592

I hope all land, air, and naval combat is diversified enough to actually be more interesting than "spam x unit, smash x unit blob against enemy blob." That's kinda what naval and air is in supcom2 (uef subs vs subs and uef fighters against fighters...). It's okay, but gets boring.

There's no point in MML when you have mobile artillery.

Oops, apparently I edited out the mobile artillery and forgot to edit it back in.

As for air warfare, then you could simply transport in flak units and lead the enemy over them. That would be le painful.

One thing I really want to be a large part of the game is utility things. Both buildings and Units. You mention some in the OP. Not every unit/building doesn't need to do the standard stuff. By that I mean generating resources, building things or shooting stuff.
Things like different radar buildings, cloak and other simillar things. I think units/buildings with abilities are more fun than units with big guns.

While I'm a fan of TA's massive unit list, I must point out that are significant drawbacks to that model, the biggest one being unit redundancy. For ground forces, you had T1 Kbots, T2 Kbots, T1 Vehicles, T2 Vehicles, and T1 Hover Units, each of which had at least five different models of unit available. Then there was T1 and T2 Air Planes and Sea Planes, which is another fifteen. Add in the T1 and T2 Naval Yards for another five, and you've got fifty different units right out of the starting gate. Many of those factories had more than five units in them, and in some cases significantly more than five.

However, this leads considerable unit role overlap, which is inevitable given the sheer numbers involved. Just looking at light ground forces for the Arm, there's the PeeWee and Warrior Kbots, and the Flash and Stumpy Tanks. That's four different units, all competing for the light ground unit niche right there.

This is a trend that's repeated throughout TA, where there are often two or more units competing to fill the exact same role in slightly different ways. While this gives the player lots of choice to work, the reality is that many of those choices were not good ones. Only a relatively small subset of units got used....

Also, the balance between units becomes increasingly more complex as the number of units in the game grows. You have to consider how each unit fares on their own against the others, how they fare in small groups, how they fare in mixed groups, and so on. The number of variables that has to be considered grows much, much faster than the size of the unit list.

In addition to the balancing headaches, each new unit model that the developers want means a significant amount of time spent creating and texturing it, which is time and effort that they can't put into other areas of the game.

Also, the amount of information that the player has to keep in mind is at least proportional to the size of the unit list, and there's only so much a player can keep in mind at one time. The more units there are, and the more complex their interplay, the worse this problem gets.

So, I'll say it again: Right now I'd like to see a nice, tight unit list. Just the core set needed for a good gameplay experience. If that can be done, and the game does well with it, then the devs can go back and see about adding new units.

I do like support units. In a 40 player game, it would be nice to be the ally providing global radar coverage and transport aircraft.

I don't remember how that was implemented in TA/SC, but in SC2 mobile arty has a smaller range compared to point defenses, and is pretty much terrible to aggressively attack with due to the incredible delay between shot and shell landing and generally just being inferior to the "punch" that tanks deliver. MML are meant to defeat structures in early-game (and even later in the game with good micro), and they are great at that. Mobile arty is for splash damage against land units, mostly, and pressure to micro, as well as for defense. Not much good for anything else which is done better by tanks/assault bots or MML.

From what I remember TA had missile launchers and they worked great against structures (which is arguably their purpose)... so I would like to see them personally.