Ideas for Developing the Surface War

I have been experimenting with modding to produce a more interesting surface war in terms of unit interactions. Obviously this is not the only possible set of mechanics that would lead to a strategically and tactically interesting land war. But after testing, I think it is pretty clear that these changes (which are very straightforward) will result in a much more dynamic and interesting land war.


1) Increase Scale

This is the single biggest improvement that can be made to PA, in my opinion. Reduce unit movement speeds, reduce unit vision ranges, and reduce weapon ranges and projectile speeds. Alternatively, we could just play on much larger planets, which would have a similar effect of making the map larger.

The reason for this is that time is the ultimate currency of the realm in any real time strategy game. When a player makes a move, there should be a commitment that will take time to reverse. And as a result of that move, it should be advantageous or disadvantageous in a way that will take time for the enemy to change. If you manage to catch an enemy out of position, that should be a mistake that costs more than a few seconds to rectify.

In a real time strategy game, time is the measurement of a player's commitment to a strategic choice. We need more scale or space on the map in order to commit players to decisions that cannot be swiftly reversed.

This ties directly into the idea of "big battles." A battle is not 'big' because it contains a large number of units. A battle is 'big' because it takes place across a large region of space. This is completely different from having two large groups of units fight. Across a big map, the units spread out. The battle begins in the early game with small raids, growing into light skirmishes all over the map, and continuing to escalate more and more in intensity across the shifting front until someone breaks. The "battle" is the entire affair, not just when two blobs collide.

Across a big map, the entire map should be in play all the time, beginning with low unit density, and increasing in density until the endgame, where superweapons are annihilating massive armies.


2) Weapon Inaccuracy as Common Feature

With a few exceptions, almost every weapon in the game should have a substantial value for its "firing_standard_deviation" property.

Particularly for units like Dox, which are rapid fire. Their rate of fire can be increased along with a large inaccuracy, coupled with a projectile that persists past its target for a few tenths of a second. A group of units will actually take quite a lot more damage than a lone individual as a result of shots that missed the original target.

Artillery should have very large values for "firing_standard_deviation." I can attest to the fact that artillery shells can be made positively devastating in terms of damage and splash radius, along with quite cheap, and still be reasonably well balanced in terms of cost of the artillery to the cost of units they kill. Large inaccuracy for long-range artillery both looks and feels awesome, and makes it powerful as a long-term, strategic weapon, but not reliable as an immediate-action tactical response to a specific enemy action.

Having inaccurate weapons also makes fights go on for quite a lot longer, especially between well organized armies, since a significant portion of their DPS misses, and a portion of the DPS that "misses" actually hits a different target.

This has two effects. First, it makes enormous, tightly packed blobs considerably less efficient because they will take much more damage from misses. And second, units struck by "misses" that were aimed at a different unit will be damaged, but if the enemy's aim were excellent that damage would have been efficiently allocated by focus fire. Instead, many units in the group are taking some damage, which is a completely different dynamic from perfectly efficient focus fire at your units one at a time.

It also means there will be many more big explosions, but many of them will miss or land close by to units rather than destroy them. And a lot more gunfire in a close combat situation, with many shots missing.


3) Order of Battle

I am not sure I agree with this assessment, but it seems to be a common opinion that PA feels "bland" or lacking in 'character' of gameplay, or strategic novelty. I believe the core of this complaint stems from relatively bland surface warfare, because the outcome is totally determined by the number of units in each force. The larger group wins, and wins by a predictable amount every time.

What PA needs is ways for players to make tactical choices that will change the effectiveness of their forces. Units that let a smart, savvy player win a fight where they are numerically outmatched. To this end, I think PA needs "support" weapon systems in addition to your conventional front line attack units that collide and destroy each other. Weapons that the player has to make decisions about how to deploy and how to use that will have very different effectiveness depending on many different battlefield conditions.

For example, suppose the missile bot model carried an ammo count of two large, powerful, manually fired cruise missiles that cost metal to construct, and have a lengthy build time. But they deal a large amount of damage and also splash, enabling a missile to destroy a small group outright, or allowing a volley of several missiles to destroy a few buildings. An army in the field with ten of these bots would have a grand total of twenty such missiles at its disposal, and the player must decide carefully how to use each of these missiles. Careful and intelligent use of these missiles could be game-changingly powerful, but losing these bots before they can fire would make them a significant waste of metal. The enemy wants to silence your support guns, while you want to hide them, protect them, and maximize their usefulness.

Put succinctly, "Artillery adds dignity to what would otherwise be an ugly brawl."

What we currently have in PA is an ugly brawl, where the core of a battle is raw numerical superiority, and thus the main task of playing the game is unchecked economic growth.

By contrast, in TA unchecked economic growth will rarely succeed, because you will get run over by someone being more aggressive. You have to strike a very difficult and evolving balance between growth, defense, and aggression, using many different tools for each of those three purposes.

PA needs to find that tension between growth, defense, and aggression. To do this, PA needs combat mechanics that depend on more than just quantity of forces. Otherwise the obvious dominant strategy is unchecked economic growth coupled with mass unit production to try to achieve numerical superiority in a single decisive battle.


Conclusion

We need more scale in terms of battlefield space, either by simply increasing planet size, or by reducing unit speed, vision, weapons range, etc. This will mean you have to spread your forces out in order to control the map, and protect your mexes while destroying the enemy's. Weapon inaccuracy will contribute to this by making smaller groups perform better than larger groups, and making long range weapons more strategic, probabilistic weapons rather than a direct, instant-action tactical response.

And third, PA needs player agency in battles that can affect the outcome of a fight. Such as by choosing to use a limited resource that the player has constructed, or by deploying or moving assets from one area to another to support a weakening front or bolster a push. The blobs of units colliding and mutually annihilating is an acceptable starting point for the surface war, but adding more tools and developing mechanics that allow players to seize an advantage through clever manipulation, positioning, tactics, and planning is essential.

Excellent post. Uber, listen to this guy!

I doubt anyone will disagree with you on scale; some have suggested that larger maps with more dynamic terrain elements can go a long way towards fixing the Dox meta. I mean, whoda thunkit that tiny featureless moons would promote using cheap, fast units? /s

Innacuracy wouldn't hurt either. However, player agency in the form of activated abilities runs contrary to PA's design; units are made and destroyed quickly, with no reliable lifespan in which to perform any behavior besides their automatic attacks. The player's attention, likewise, can rarely be counted on to be present on a given planet at a given time to use an ability reactively.

Player agency may be better expressed by simply making army battles better.

I don't always agree with an Idea on the forum, espically from @ledarsi but when I do




It's the most interesting post on the forum.

"Build dox my friends."

I like long texts like this, though I partly have doubts about your concepts.

Scale:
Personally I think the scale in term of unit speeds, vision and stuff like we have now is fine. The only scale I dislike is the visible unit size. That could be reduced by 50%. That obviously has been discussed before, and if anybody wants to see how 0.4 looks like there is that mod I generated on pamm

I think currently people -for some reason- are playing on rather small maps, so a "little" bigger maps are probably interesting. So instead of an avg of maybe 600, more like 800 with peaks at 1000.
My main concern with that however is more about expansion play.
I think the "time commitments" we have right now are fine, even on 600 you can easily be caught out of position and die to a smaller army.

Unreliable weapons:
This reads like a quite luck based game to me. Yes with bigger armies it will even out between two armies, but the -quite important- early game, where you have rather small groups of units fighting, would easily see very luck based fighting.
I don't like that thought at all. Might be wrong about my feeling. Probably depends on the rate of fire. If you make dox miss 50% but have them shoot 10 bullets per second, then yeah that 50% miss rate probably is fine even in a dox 1vs1.
So you'd need pretty high fire rates to make sure even small scale fights are not dice roles.

Order of Battle:
I don't think we need special abilities for that at all. Heavily disagree.
Right now you can win fights with at least slightly smaller forces, in some situations even considerably smaller forces.
The base mechanics here are: positioning of units (like having a concave) and terrain resulting in armies moving in unfavorable ways. Like a small line that walks into an army, one by one as the most extreme case.
Better map design could improve this even further.
Small scale dox engagments ofc have this concept not as much, as dox are very fast and can therefore fix there positioning pretty quickly, though even with dox I've often seen dox groups split in 2 for some "stupid" pathing reason and have half of them die. A good player should consider the terrain when moving his army to prevent that.
On the other hand with dox you can micro them to get an advantage, though I think that is undesirable and should be fixed. Better make dox a bit slower or something or just so much weaker that people stop building large dox armies for direct fights and instead use small dox groups to raid and not more than that.

Also to get a tension between growth, defense and aggression we need to fix the economy to actually allow for growth and successful defense of growth again

I feel like the unreliable weapons can be compensated for by making sure that it isn't the results of one dice roll that determines a battle, but the results of several thousands.

I am not in favor of special abilities/activated abilities. I was primarily thinking of artillery, long range anti-air, air support, and various other tools that are variable in their effectiveness based on usage. Whether from positional aspects, timing aspects, limited quantity (as in the cruise missile example), or using some other mechanic. I did not mean to be overly restrictive with the cruise missile example, that is only a particularly specific possible case.

By "order of battle" I primarily meant to give a variety of different unit mechanics that create a more complex war machine than a blob of units.

I have tested this, and it does not play like a luck based game except in a fight with a single Dox against a single Dox.

If you multiply the Dox firing rate and give it a substantial inaccuracy, the result is still a very consistent DPS against a single target. However if there are other targets nearby they will also take a smaller, but still fairly consistent splash. The effect of this design is that the smaller group of Dox actually does much better against the bigger group of Dox than they do now. Such as 3 against 6.

For artillery, you have a much stronger case that it is "luck" because a lucky artillery shot in the middle of a group of Dox can wipe them out. But if you have many such groups spread out across the map, it's not terminal, and it's not a statistically likely event unless your groups are huge.

The idea for highly inaccurate artillery is that they don't kill any particular target consistently, but they hit something, especially when firing at a large, dense concentration of units and/or structures. Over a long period of sustained bombardment, it adds up to fairly consistent damage over time.

One word: variety

We can haggle over the details of how we accomplish that variety, but that is what we desperately need and don't have.

Good post like always @ledarsi. I am extremely intrigued by decreasing the speed, vision, and range of all units. It'd be a great way to instantly increase the scale of the game. I really want to try it out.

Someone needs to make a balance mod of this! Could be as simple as roughly halving the vision and range of everything and then speed by... maybe 20-40%?

Scaling vision distances, movement speeds, weapon ranges, projectile speeds, and walk animation speeds down by half across the board works. This is easily tested, I have done it, and it plays well. The walk animations are a pain, however. The only real problem I ran into was walking off the factory floor sometimes creates an annoying glitch where the nanoframe is being pushed by the previously completed unit which didn't leave quickly enough. However, it still builds and is perfectly playable despite the glitch.

The one gameplay problem is that it makes early game quite passive. However this is easily addressed directly by having a specific early game unit that is speedy, but a weaker fighter. This creates an enormous defender's advantage against raids, using either turrets or heavier basic units. But this early game raider creates the possibility of early game interaction even when all the other units need more time to travel.

Essentially this cuts the Dox into two units- a light raider that is speedy but squishy, and a heavier close combat variant that is very slow, but both packs a punch and has a nice chunk of HP for a very low cost. The light variant could easily be the amphibious one (or usable for early game interplanetary attack using a disposable launch vehicle), but would mount lighter weapons and have much less HP.

In addition, this arrangement creates the possible role for a rifle type skirmisher with a fast, accurate projectile that is very effective against low-HP enemies at range, and good at picking off enemy light units. But unlike a more mobile skirmisher, it would be vulnerable to being overrun, either by fast units or by sheer numbers. This unit would be a very effective defender against a handful of light raiders. Against large numbers you're going to need antiswarm of some type, or else a group of your own forces.

I've actually tried a base movement as low as 2.0 (1/6th), and while the units move extremely slowly, it creates a much more methodical battlefield. You just need some units that move faster to get light raids/skirmishers earlier, even though big armies composed of more HP and DPS-efficient units players build later will move slowly.

I agree. Though everything should work passive. I don't want an ability where you have to decide to shoot some weapon right now at some specific location to be part of regular army fights.

For units with high firing rates I think it would probably look cool and have no negative affect.
For artillery I think it needs to be balanced a lot more carefully though.

I've never had the feeling that such units work very well in other games. Too unreliable.

I don't think a slower game adds variety, the ideas about more unit types above can add variety.
I don't think 40% slowdown would be fun. At least not unless we fix the economy to massively buff expansion on top of it. If the units are super slow then the whole game really NEEDs to happen all over the map or players will get bored.

You can create a scale mod right now btw, just build a 0.6x rescale (so the units don't look totally funny in relation to their ranges, a smaller factory would also fix the rolloff glitch) and use TransformPA to scale down vision, etc. Should not be much work to do. I would even play a game with such a mod to see how it actually plays, maybe I am wrong and 40% slower/lower ranged doesnt feel bad. Might actually really be the case. Hmm...

I'm actually quite intrigued by @ledarsi's proposition, especially about the inaccuracy theory.

Even in real life, not all weapons behave the same. If we even look at small arms, and then even rifles and automated weapons: a pistol shoots a lot steadier then an SMG, for instance, but the amount of bullets an SMG can spew for the same amount of time that a pistol fired all it's shots is just outright insane, BUT you have collateral damage as the SMG handles a lot worse then the steady pistol.

Same with rifles and automated heavy weapons.

----------------------------------------------------------------------------------------------------------------------------------------------------

If we translate this to PA we can have some things the rely on their accuracy more then their ROF. What we can do is leverage the range into the inaccuracy percentage... Let me elaborate:

* Bots tend to handle like the SMG. They have an insane ROF, but shoot a small "area" wherein their bullets hit. At max range this would be for example a circle of dia. 2. (lets say a tank has an area of 1, and use this as a standard). At 50% of their max range this would be 1.5, and at 25% this could be 1 (always hitting the tank at max ROF and by effect max DPS). The grenadier would handle in the same fashion, but with more AoE damage, so the grenades would offset this effect, by doing more collateral damage to their other targets. Buildings can be targeted at max effeciency as they tend to, well... not move

* Tanks on the other hand are slow steady vehicles, that handle like a normal rifle. Not an insane range, but reliable and deadly. Tanks hit in a circle of 0.5 whatever they shoot (always hitting a dox, unless it moves suddenly). Slower DPS and ROF, but deadlier in the long run!

* T2 bots is where the fun stuff starts though, as they have increased damage, but lower rate of fire. Mounted machinegun mechanic. Relatively steady but heavily punishing rounds in an area of 1.5 at max range and 1 at 50% range as they are more sturdily fixed (they also look a lot heavier then Doxen). The sniperbot handles like it should though. Very high damage, very long range, but the only thing i'd change is this: give them MORE range, but decreased vision, so they need a spotter!

*T2 Vehicles: The leveler would handle like an increased Bolo (like it does now, but with more deviation at max range). Area 1.5 at max range (can miss a dox, but will almost never miss a Bolo or a building!). The Sheller needs the same deviation like artillery, but with increased AoE damage to offset this. At max range an area of 3.5 but it also does damage in an area of 3 (so it wouldn't necessarily kill it's primary target, but at least take care of everything even near it).

I did leave the vanguard and the inferno out, as they are practically melee range, so this should not affect them... And since the vanguard already does a lot of AoE damage, this unit is in the right place as it is...

That design for bots and tanks could definitely work. I think I can do better, though.

Suppose tanks as a class are designed around a higher price point than bots (unarmored/light vehicles could be cheaper). They have higher stats, but are less cost-efficient; meaning if you spent the entire tank's cost on bots, you would get a lot more total HP and DPS, but spread across multiple individuals.

Suppose your main combat bots were generally priced at around 50-100 metal, and main battle tanks were around 500-1000 metal. You could certainly also have powerful (perhaps more specialized or support-oriented) bots that cost more than those bots. And you could also have light vehicles or support vehicles that cost less than tanks.


Tank Features

Tanks would feature a significantly higher maximum speed than bots, but would handle poorly in terms of turn rate, acceleration, and turret turn rate. This makes them unwieldy at close quarters unless they have a secondary weapon meant to protect the tank against nearby enemies, such as a machine gun.

However, tanks' most distinctive feature would be the main gun. Tank main guns would have excellent range, fire fast-moving shells for high damage, and have a significant reload time of several seconds for each shot. By tweaking single-target damage, splash damage/radius, and cannon inaccuracy, these guns can be made to be versatile against bots and tanks, or specialized against one or the other.

Tank cannon inaccuracy should be enough that misses are expected, especially at range (standard deviation of between 0.8 and 1.2 perhaps), but with a high projectile speed. Missing the intended target is highly likely to still damage it from splash after the shell hits something next to it.

Furthermore, tanks would have considerably longer cannon range than they have vision range, meaning another unit will need to stand in front in order to fully utilize the tank's capabilities. Cheaper, more efficient, and more expendable bots with better vision and shorter range are perfect for this role, leading to combined arms once multiple factories are an option. Slow air recon would also work well in this role, but is vulnerable to AA instead.

The basic idea is that bots behave like a group of small entities, and are thus more vulnerable to splash damage like machine guns, artillery, and flame weapons. Tanks, due to their higher HP and lower cost-efficiency, are very vulnerable to large single-target damage sources.


Tank Unit Designs

Tank main guns would also be a major point of differentiation between tanks. Light tanks would be more versatile, such as firing HE shells that deal more splash damage, making them effective against smaller targets. Other tanks might have shells that do not splash, but which deal high single-target damage, making them stronger against other tanks.

The basic workhorse tank would mount a versatile tank gun with decent single target damage, and a large splash radius for moderate damage, making it useful in many circumstances. Other units might be designed to operate in an environment with such tanks. Such as a basic light vehicle with a high single-target damage cannon. And an advanced tank that is a heavily armored tank hunter, specialized in eliminating enemy armor and not as useful against bots or light vehicles. Or an advanced light tank hunter using long range missiles with high single target damage. These high-power, single-target weapons would obviously work against a bot. But a large group of bots doesn't really care.


Tactics

Under this setup, in a perfectly flat, open field, tanks could theoretically kite bots forever due to their superior speed and range. However, in practice, a group of bots can advance, forcing tanks to retreat until the bots have pushed them back to the enemy base. Or, the bots can retreat to a protected area, using other units to keep the enemy tanks back. Bot armies will be large, slow, and hardy. While a tank group is smaller, speedier, and reliant on maneuver and aggression, as well as retreating instead of becoming stuck in a pitched battle they would lose.

Tanks would serve the highly realistic dual purpose of fire support from behind, and will also switch in a moment into the core assault element at the player's choice.

Tanks are expensive, long-range, valuable fire support units, and you want to preserve their quantity as much as possible while your bots acquire vision and trade evenly on the front lines. However, tanks are also your high speed, high-HP assault units, and are necessary to break a line of turrets where you would otherwise need a truly ridiculous number of bots to crash against their defenses, with massive casualties.

These properties should lead to a preference for combined arms over just mass-producing either bots or tanks. A group of purely bots will lack mobility and punch, and will need a lot of support. However a group of purely tanks will lack the ability to hold ground. And, despite its high shock power, it will bleed strength badly whenever it is used.


If you're interested, I can pare down some of my sprawling mod files and put together a minimalistic proof of concept mod to demonstrate these mechanics and interactions using just a couple unit types.

While reading this it sounded like you were describing RCBM. You may want to give that a try.
https://forums.uberent.com/threads/wip-the-realm-community-balance-mod.58942/