Some ground combat mechanic reflections!

[Jorgen_CAB (OP)]

I have looked at Steves AAR and the mechanic for ground combat and for the most part it seem to play out very nice...

There are two things that I do find a bit perplexing and strange.

The first is the easiest one and that is how bloody and fast paced it can become, especially when we talk about ground combat that might involve an entire planetary surface. The larger the forces the longer a war should take, so some mechanic that make the forces less killy the larger they are should be a feature somehow.

I also would like to see something that reign in the square law of destroying something, that is often not how things actually work out in reality. Resistance tend to be like a rubber band and be more and more resistant as time goes on until it eventually just break and you would not really know when that happen. Right now it is for the most part easy to figure out who is winning and how long it will take after a few rounds as the winner just kill more and more enemy troops for everyone he himself looses.


The other issue I have after some deliberation is the front line mechanic... it sort of make little sense that two entrenched units fight each other, they would have to leave their entrenchment to do that in the first place.

I would instead only allow the offensive front to be the ones that can actually inflict casualties on the enemy while the defensive line still have to randomly target an enemy unit in either defensive or offensive line. If a unit in the defensive line target a unit in the enemy defensive (or support, echelon) line they do no damage. Only support weapon in the defensive line can do damage to the enemy defensive line and artillery in the support line etc. So you would still see attrition over time on both sides, more like skirmishing and bombardment of enemy positions.

The offensive line should also increase its size so attacks are a bit more likely to hit them and not the defensive line, perhaps 150% of their original size or something.

The current system can become a bit weird as an attack pull back all their forces in support line the defender can send in their armoured reserve with no retaliation what so ever as none of those unit in support can shoot back in the 8 hour ground combat round that follows.

It also feels a bit weird that you can entrench all your forces and THEN attack with full entrenchment values. Two entrenched enemies should just be staring at each other and lobbing artillery rounds and making air-strikes etc...

My suggestion above of changing the front line mechanic a bit would also to some degree solve the overly brutal nature of the games ground combat. Also, in many real wars between two relatively equally qualitative standard you would still loose about as much manpower as the enemy even if you had roughly a 3:1 advantage in strength. The way the game is setup you would probably not reach that (unless you attack with no entrenchment?!?), but I might be wrong though.

Another thing I also would like to see is that population play a role in ground combat... we should be able to quickly form very light militia formations from large population planets. They might not be able to destroy a qualitative opponent but they would drag out the war for a long time and make it very costly. Militia formations should be some sort of readiness level which cost you some wealth in maintenance with a minimal level that cost nothing but only give you a very small number. But in a world with 10.000.000 people you should be able to get a considerable number of militia forces even at minimal level. Once the militia have fought and hostilities are over it would replenish its number after a while, not immediately. Raising the militia level also should take time, so you could not do it just before an invasion. The militia should basically be like half strength of the lowest normal troops, but they are many so killing them with tanks would take a very long time.

Some of my quick reflections of what I have seen so far...

[Bremen]

The speed I'm okay with. I mean, if I have ten divisions and the enemy has ten divisions, I don't see any reason it shouldn't work out the same as ten one division on one division fights. It's not like chokepoints are likely to be an issue on a planetary scale with both sides having the tech for rapid transport.

I do kind of like the idea that if boths sides are on defensive there should only be artillery fire (including FFD supported orbital bombardment), personally, and I agree that "fortify in place and then attack" is an odd behavior but seemingly incentivized by the current system. Maybe this is countered by the fact that having no units on defense means you can easily be overrun by attackers, I'm not 100% on how the ground combat works in that scenario, but if so it would seem tricky if the AI doesn't know how to make use of that.

I think the idea of having defensive units randomly target and inflict no damage if they "pick" a defensive one is massively exploitable, though (for instance, having 1 attacker and a thousand 1 man formations on defensive). Rather I would say that all defensive formations should only target attacking formations, while attacking formations will target both defensive and attacking formations - it still works out fine.

That said it may be a little late in this version cycle to make that change, even if I like the idea. Once C# comes out I'm sure we'll all have a better idea of the ground combat and can make suggestions from there.

[Jorgen_CAB (OP)]

The speed I'm okay with. I mean, if I have ten divisions and the enemy has ten divisions, I don't see any reason it shouldn't work out the same as ten one division on one division fights. It's not like chokepoints are likely to be an issue on a planetary scale with both sides having the tech for rapid transport.

I do kind of like the idea that if boths sides are on defensive there should only be artillery fire (including FFD supported orbital bombardment), personally, and I agree that "fortify in place and then attack" is an odd behavior but seemingly incentivized by the current system. Maybe this is countered by the fact that having no units on defense means you can easily be overrun by attackers, I'm not 100% on how the ground combat works in that scenario, but if so it would seem tricky if the AI doesn't know how to make use of that.

I think the idea of having defensive units randomly target and inflict no damage if they "pick" a defensive one is massively exploitable, though (for instance, having 1 attacker and a thousand 1 man formations on defensive). Rather I would say that all defensive formations should only target attacking formations, while attacking formations will target both defensive and attacking formations - it still works out fine.

That said it may be a little late in this version cycle to make that change, even if I like the idea. Once C# comes out I'm sure we'll all have a better idea of the ground combat and can make suggestions from there.

In terms of scale what I meant was that an engagement between 2000 or 200.000 on each side should be vastly different in time scale, right now it scale rather linear while in reality it would not as the larger an operation is the more complicated it is and organisation and logistics take more time. There should just be a reduction on the total allowable width that actually fight the larger the battle is during each 8 hour cycle or there should be a % chance that no combat occur in that phase might be easier to implement as the armies scale in size.

It also is not realistic that 10:1 odds means all 10 divisions get to fire at the same time in reality... a 10:1 odds in reality will never mean 10:1 odds in the field... there are only so many soldiers and material you can throw in... even in terms of artillery to some degree at one point in time.

The exploit you mention does not happen as each of you targets is a certain width, the same way that damage in ship components work. If you kill that one soldier you randomise a new one until all your hits hit something. I'm sure this is how it currently work, otherwise you could always exploit this anyway.

The effect of scaling would then mean if you dump 1.000.000 soldier to beat 2000 soldiers (width should be the factor) then there would be a much higher chance several 8 hour ticks go by with nothing
happening representing the extremely large forces organising rather than attacking. It would be an abstraction to the fact that such large forces are involved. If you instead drop 20.000 men it is a much faster combat and the chance that an 8 hour cycle is skipped is much lower. It has to scale none linear to work. So a larger force is much more likely to skip a turn than a smaller one. The factor should then be both forces added together or something. It would be an abstracted mechanic to show the lulls and organisational hassle of managing really large armies.
This would also make smaller more powerful armies way more effective time-wise than a more numerous unwieldy army which is realistic. The end result would still be the same, just the time it takes would differ in this instance.

In addition there should also be a limit of how much width could ever target another width, it would probably be realistic to restrict it at someplace around 10 times the numbers.

[Bremen]

In terms of scale what I meant was that an engagement between 2000 or 200.000 on each side should be vastly different in time scale, right now it scale rather linear while in reality it would not as the larger an operation is the more complicated it is and organisation and logistics take more time. There should just be a reduction on the total allowable width that actually fight the larger the battle is during each 8 hour cycle or there should be a % chance that no combat occur in that phase might be easier to implement as the armies scale in size.

It also is not realistic that 10:1 odds means all 10 divisions get to fire at the same time in reality... a 10:1 odds in reality will never mean 10:1 odds in the field... there are only so many soldiers and material you can throw in... even in terms of artillery to some degree at one point in time.

You're thinking of it like a war, though, and I'm thinking of it like a battle. A battle with 10,000 men over 100 miles wouldn't be 100 times as long as a battle with 100 men over 1 mile, it would be pretty similar. You might wait weeks before launching a major attack, but once you do it will probably be over quickly one way or another. Since we're talking about planetary scales with armies of probably 10s of thousands, maybe slightly more, I don't think width or maneuver is likely to be a big issue issue.

Making it so 10 divisions can't simultaneously engage one division is more of a balance consideration, and I have no strong feelings on it. Again, I'd say it might be best to wait until we've played the first version before actually worrying about balance.

The exploit you mention does not happen as each of you targets is a certain width, the same way that damage in ship components work. If you kill that one soldier you randomise a new one until all your hits hit something. I'm sure this is how it currently work, otherwise you could always exploit this anyway.

Except you don't kill the soldier, because your suggestion was if a formation on defense picks a target that's also on the defense it deals no damage:

If a unit in the defensive line target a unit in the enemy defensive (or support, echelon) line they do no damage.

[Jorgen_CAB (OP)]

In terms of scale what I meant was that an engagement between 2000 or 200.000 on each side should be vastly different in time scale, right now it scale rather linear while in reality it would not as the larger an operation is the more complicated it is and organisation and logistics take more time. There should just be a reduction on the total allowable width that actually fight the larger the battle is during each 8 hour cycle or there should be a % chance that no combat occur in that phase might be easier to implement as the armies scale in size.

It also is not realistic that 10:1 odds means all 10 divisions get to fire at the same time in reality... a 10:1 odds in reality will never mean 10:1 odds in the field... there are only so many soldiers and material you can throw in... even in terms of artillery to some degree at one point in time.

You're thinking of it like a war, though, and I'm thinking of it like a battle. A battle with 10,000 men over 100 miles wouldn't be 100 times as long as a battle with 100 men over 1 mile, it would be pretty similar. You might wait weeks before launching a major attack, but once you do it will probably be over quickly one way or another. Since we're talking about planetary scales with armies of probably 10s of thousands, maybe slightly more, I don't think width or maneuver is likely to be a big issue issue.

Making it so 10 divisions can't simultaneously engage one division is more of a balance consideration, and I have no strong feelings on it. Again, I'd say it might be best to wait until we've played the first version before actually worrying about balance.

The exploit you mention does not happen as each of you targets is a certain width, the same way that damage in ship components work. If you kill that one soldier you randomise a new one until all your hits hit something. I'm sure this is how it currently work, otherwise you could always exploit this anyway.

Except you don't kill the soldier, because your suggestion was if a formation on defense picks a target that's also on the defense it deals no damage:

If a unit in the defensive line target a unit in the enemy defensive (or support, echelon) line they do no damage.

On the picking target thing you are somewhat right... the solution is then that you first weight the likelihood that you either pick a defensive or offensive part of the line and THEN you randomise where you hit.

So if the enemy has 30% of his line in the offensive front it would mean that each of your defensive units randomly pick that force to fire at in 30% of the time and 70% of the time they do nothing but fire there artillery and mortars at a defensive front line unit or nothing at all if they happen to pick the support or echelon line as that is how it currently work.         

I also agree that we need to wait and see... but I think I have a good grasp of how the mechanics work over all.

I also agree there is a difference between a battle and a war, the game should simulate both... that is why I suggest that really large conflicts take MUCH longer as an abstraction mechanic and that there is a limit on engagement possibilities.

I also find it very strange that same level of technology would not make it still hard to find and engage each other, there are things like disrupting sensors and stuff like that to hide ones movement etc. Life is never really so black and white...

[Bremen]

On the picking target thing you are somewhat right... the solution is then that you first weight the likelihood that you either pick a defensive or offensive part of the line and THEN you randomise where you hit.

So if the enemy has 30% of his line in the offensive front it would mean that each of your defensive units randomly pick that force to fire at in 30% of the time and 70% of the time they do nothing but fire there artillery and mortars at a defensive front line unit or nothing at all if they happen to pick the support or echelon line as that is how it currently work.         

It's still easily exploitable; you could fill a defensive line with cheap units never meant for combat while your actual combat units are on offensive, and thus the defender would hardly ever get a shot off. It also runs the opposite of your suggestion about penalizing ten units attacking one, because it would mean one defending formation against many attackers would spend most of its time being pounded while unable to fire back.

[Rabid_Cog]

The mention of "battles" and "war" sparked an idea. They really should be split up. Even an outnumbered defender can win a battle or gain local numerical superiority through superior mobility or strategy.

Instead of each side just blasting the other, all offensives are done in "battles". There may be 1 or more battles each ground combat tick with larger forces likely to have more battles but with diminishing returns. It would basically work as follows:
* All units on both sides start in the Reserve. Units move from Reserve to Battle and back to Reserve once the battle concludes.
* All offensive (in reserve, good health and in the front line) units are put in a randomized list. That is a single list including units from BOTH sides.
* Looping through the list from the top, a check is performed to see if each unit attempts to start a battle. This check is a, say, 20% chance modified by commander stats (perhaps something called "aggression"?) and the number of ongoing battles. If the check is passed, a potential battle is created for this unit.
* The unit picks a target using the current rules.
* Both sides now randomly assign reinforcements from the list of all Reserve units. Every unit has a chance to be assigned to this battle related to how many reinforcements have already been assigned by this side on this battle line (front line or support line). This is all the units that could potentially be involved in this battle.
* Once all forces have been assigned, the attacking commander makes a final Balance-Of-Forces comparison on the relative strengths of the two sides of the battle. If it is sufficiently in his favour (again, perhaps modified by aggression or something) he pulls the trigger and the battle commences. If he feels the odds are against him, the attack is cancelled and all units are returned to the Reserve.
* When battle is joined, it works exactly like combat works at the moment, except that no other units except those in the battle can be targetted.
* At the end of a ground combat tick, the battle usually ends, but if the attackers are doing well (again the attacker makes the Balance-Of-Forces check) it has a chance to continue in the next tick. Badly damaged units have a chance to withdraw from the battle, then the reinforcement step happens again as each side brings new units into the battle.

If one side has such a numerical superiority that it starts running out of targets, then too bad. Those units, while wanting to fight, simply don't have the space to do so. So whether you outnumber your opponent 100x or 10000x has much less of an impact. Army size has a much greater impact on how LONG you can keep fighting and what casualties you can sustain, rather than on how strong you are in a single engagement. You cant instagib a Chaos God by just chucking 42 trillion guardsmen at it. It would take a while to chew through them all, but eventually it (likely) would.

It also offers some protection for support units. Even if they get targeted the first time, frontline units have a chance to rush in and defend it via the reinforcement mechanic.

This also enables further mechanics like allowing individual battles to have different terrain modifiers depending on what terrain is available on the planet and even allow fighting over objectives down the line.

[Jorgen_CAB (OP)]

The mention of "battles" and "war" sparked an idea. They really should be split up. Even an outnumbered defender can win a battle or gain local numerical superiority through superior mobility or strategy.

Instead of each side just blasting the other, all offensives are done in "battles". There may be 1 or more battles each ground combat tick with larger forces likely to have more battles but with diminishing returns. It would basically work as follows:
* All units on both sides start in the Reserve. Units move from Reserve to Battle and back to Reserve once the battle concludes.
* All offensive (in reserve, good health and in the front line) units are put in a randomized list. That is a single list including units from BOTH sides.
* Looping through the list from the top, a check is performed to see if each unit attempts to start a battle. This check is a, say, 20% chance modified by commander stats (perhaps something called "aggression"?) and the number of ongoing battles. If the check is passed, a potential battle is created for this unit.
* The unit picks a target using the current rules.
* Both sides now randomly assign reinforcements from the list of all Reserve units. Every unit has a chance to be assigned to this battle related to how many reinforcements have already been assigned by this side on this battle line (front line or support line). This is all the units that could potentially be involved in this battle.
* Once all forces have been assigned, the attacking commander makes a final Balance-Of-Forces comparison on the relative strengths of the two sides of the battle. If it is sufficiently in his favour (again, perhaps modified by aggression or something) he pulls the trigger and the battle commences. If he feels the odds are against him, the attack is cancelled and all units are returned to the Reserve.
* When battle is joined, it works exactly like combat works at the moment, except that no other units except those in the battle can be targetted.
* At the end of a ground combat tick, the battle usually ends, but if the attackers are doing well (again the attacker makes the Balance-Of-Forces check) it has a chance to continue in the next tick. Badly damaged units have a chance to withdraw from the battle, then the reinforcement step happens again as each side brings new units into the battle.

If one side has such a numerical superiority that it starts running out of targets, then too bad. Those units, while wanting to fight, simply don't have the space to do so. So whether you outnumber your opponent 100x or 10000x has much less of an impact. Army size has a much greater impact on how LONG you can keep fighting and what casualties you can sustain, rather than on how strong you are in a single engagement. You cant instagib a Chaos God by just chucking 42 trillion guardsmen at it. It would take a while to chew through them all, but eventually it (likely) would.

It also offers some protection for support units. Even if they get targeted the first time, frontline units have a chance to rush in and defend it via the reinforcement mechanic.

This also enables further mechanics like allowing individual battles to have different terrain modifiers depending on what terrain is available on the planet and even allow fighting over objectives down the line.

This is roughly what I would like to see in general, it would be way more realistic and also quite fun to see which units was involved in battles and how things went. It would also give a somewhat benefit to qualitative armies that would probably win more but a low quality army could still hold on to a battle field for quite some time.  It would also scale better as battles/wars grow in size.

If terrain and structures was abstracted into the battles then we could see infantry trying to stay in built up areas and difficult terrain and tanks formations battling it out on more open ground. If you brought a lot of tanks into a mainly city brawl then those tanks would not be so powerful anymore. You might receive some bonuses for winning the battle on the open but eventually you run out of open space to conquer and most concentrate on the built up areas, without infantry that will be a nightmare.

You might even be able to strategically just give up the fight on open terrain and cling to the cities or mountains or something. Immediately giving up bonuses to the enemy but also saving valuable resources to protecting that allot more easily defended, especially if you are outnumbered and outgunned.

There might also be an option of disbanding your troops to fight a guerrilla war, that is often quite realistic and dangerous in the real world.

I also think there could be some differences on battle space if you are fighting on a small asteroid or on a vast planet surface in hundreds of cities and with billions of people involved.

[Jorgen_CAB (OP)]

On the picking target thing you are somewhat right... the solution is then that you first weight the likelihood that you either pick a defensive or offensive part of the line and THEN you randomise where you hit.

So if the enemy has 30% of his line in the offensive front it would mean that each of your defensive units randomly pick that force to fire at in 30% of the time and 70% of the time they do nothing but fire there artillery and mortars at a defensive front line unit or nothing at all if they happen to pick the support or echelon line as that is how it currently work.         

It's still easily exploitable; you could fill a defensive line with cheap units never meant for combat while your actual combat units are on offensive, and thus the defender would hardly ever get a shot off. It also runs the opposite of your suggestion about penalizing ten units attacking one, because it would mean one defending formation against many attackers would spend most of its time being pounded while unable to fire back.

Sure... could still work if you allow a certain amount of defenders to attack the offensive units though, sort of a mobile reserve. Sure it could be solved to satisfaction.

[Hazard]

While I'd like further development in ground combat mechanics, the current system is functional. I'd say let Steve bring C# out first.

[Bremen]

While I'd like further development in ground combat mechanics, the current system is functional. I'd say let Steve bring C# out first.

Yeah. While I like the idea of fights where both sides are on the defensive just exchanging artillery fire (which would also effectively be the difference between a battle and a war, IMHO), I'm happy to wait and see on that. The rest sounds to me like it's complicating the process for no actual gain, though again it may be best to wait and see.

[Garfunkel]

Some of this stuff that you're proposing isn't actually a problem, or is wholly unnecessary. Because you're still thinking of it in terms of Earth based wars at localised level and not interstellar future war on a planetary level.

So things like:

It also is not realistic that 10:1 odds means all 10 divisions get to fire at the same time in reality... a 10:1 odds in reality will never mean 10:1 odds in the field... there are only so many soldiers and material you can throw in... even in terms of artillery to some degree at one point in time.

is not as important as you think. I used to think the same way and this is of course obviously true for our own military history, because ever human war ever has always been constrained and limited by geography. Even the largest human conflict, the Eastern Front of WW2, was limited and only took place over some minuscule 0.0001% of Earth surface.

Once your sensors are good enough that you can pinpoint 1000 power armoured soldiers on the surface of a planet, you have in essence unlimited real estate to encircle and envelope them in three dimensions with your 10,000 power armoured soldiers. We're not doing tactical or even operational level combat here, purely strategic level, so having those 10,000 guys all be able to fire on the 1000 guys is reasonable and believable.

Even asteroids are generally big enough (when modelled in Aurora) to have room for thousands of soldiers and hundreds of vehicles.

Similarly:

in terms of scale what I meant was that an engagement between 2000 or 200.000 on each side should be vastly different in time scale, right now it scale rather linear while in reality it would not as the larger an operation is the more complicated it is and organisation and logistics take more time.

Well yes and no. The planning isn't that different, because it's not one person or one staff doing all of that as the more troops you have, the more headquarters you would have as well, and they can (and would!) simultaneously plan on multiple levels. Logistics, yeah for sure stockpiling would be a massively bigger undertaking but Aurora doesn't really simulate that aside from bringing in a supply formation big enough to cover consumption.

But most importantly,

While I'd like further development in ground combat mechanics, the current system is functional. I'd say let Steve bring C# out first.

Yeah. While I like the idea of fights where both sides are on the defensive just exchanging artillery fire (which would also effectively be the difference between a battle and a war, IMHO), I'm happy to wait and see on that. The rest sounds to me like it's complicating the process for no actual gain, though again it may be best to wait and see.

it's probably best to get C# out in some form and start playing it, to see what sort of shenanigans players get up to. We had that massive argument about ground combat six months ago or so, in which lot of the debate just stemmed from misunderstandings and so on. Especially as Steve has only fought one, special, enemy so far.

[Jorgen_CAB (OP)]

Some of this stuff that you're proposing isn't actually a problem, or is wholly unnecessary. Because you're still thinking of it in terms of Earth based wars at localised level and not interstellar future war on a planetary level.

So things like:

It also is not realistic that 10:1 odds means all 10 divisions get to fire at the same time in reality... a 10:1 odds in reality will never mean 10:1 odds in the field... there are only so many soldiers and material you can throw in... even in terms of artillery to some degree at one point in time.

is not as important as you think. I used to think the same way and this is of course obviously true for our own military history, because ever human war ever has always been constrained and limited by geography. Even the largest human conflict, the Eastern Front of WW2, was limited and only took place over some minuscule 0.0001% of Earth surface.

Once your sensors are good enough that you can pinpoint 1000 power armoured soldiers on the surface of a planet, you have in essence unlimited real estate to encircle and envelope them in three dimensions with your 10,000 power armoured soldiers. We're not doing tactical or even operational level combat here, purely strategic level, so having those 10,000 guys all be able to fire on the 1000 guys is reasonable and believable.

Even asteroids are generally big enough (when modelled in Aurora) to have room for thousands of soldiers and hundreds of vehicles.

Similarly:

in terms of scale what I meant was that an engagement between 2000 or 200.000 on each side should be vastly different in time scale, right now it scale rather linear while in reality it would not as the larger an operation is the more complicated it is and organisation and logistics take more time.

Well yes and no. The planning isn't that different, because it's not one person or one staff doing all of that as the more troops you have, the more headquarters you would have as well, and they can (and would!) simultaneously plan on multiple levels. Logistics, yeah for sure stockpiling would be a massively bigger undertaking but Aurora doesn't really simulate that aside from bringing in a supply formation big enough to cover consumption.

But most importantly,

While I'd like further development in ground combat mechanics, the current system is functional. I'd say let Steve bring C# out first.

Yeah. While I like the idea of fights where both sides are on the defensive just exchanging artillery fire (which would also effectively be the difference between a battle and a war, IMHO), I'm happy to wait and see on that. The rest sounds to me like it's complicating the process for no actual gain, though again it may be best to wait and see.

it's probably best to get C# out in some form and start playing it, to see what sort of shenanigans players get up to. We had that massive argument about ground combat six months ago or so, in which lot of the debate just stemmed from misunderstandings and so on. Especially as Steve has only fought one, special, enemy so far.

Logistics is a universal problem that you will not solve as long as you have individually independent minds around...

As for the sensors I simply disagree life has repeatedly shown that development in one area always develop a counter and then a counter to that counter and so forth, more or less. Weapons versus armour or sensors versus countermeasures and concealment.

I also agree that nothing should be changed at this time which is why I put it into a separate thread and not into the suggestion thread. 

[Rabid_Cog]

I also agree nothing should be changed yet. There is simply not enough data out there to come to conclusions like "collateral damage is too much" and "ground combat are too fast and brutal".

Nothing wrong with thinking of ideas, however (as long as we keep anything but a simple blurb out of the suggestion/discussion threads to prevent clutter). Ideas give alternate viewpoints and may lead to paying attention to things that we/Steve would otherwise not even consider.

[Jorgen_CAB (OP)]

I would say that the biggest pet peeve I have with the new system is that the "defender" have no real advantage when both sides have max entrenchment. There are little incentive for the "attacker" to put any troops in offensive line until they whittled down the defender from the entrenched position which give the "attacker" as much of an advantage as the "defender". Drop the troops and keep them in support line until max entrenchment kicks in and then fight. The only thing you sacrifice is time, which of course can be a problem.

I think there at least should be some restriction on the availability of the entrenchment bonus on the one that want to attack rather than defend. Right now there are no real difference in stances, just fighting.

There should perhaps at least be a stance on the troops such as defending or attacking, if both forces are on defensive stance only artillery and airstrikes can be performed. If you elect to attack you will suffer a big penalty to the entrenchment bonus, perhaps as much as 75% so armoured forces rather be set at offensive to at least be able to do breakthroughs and give up the entrenchment bonus.

There should always be some bonus to defending. Terrain seem to effect all sides equally so it give no direct advantage to the "defending" side as there are no defending or attacking position except for offensive and defensive line. No real home advantage bonus so to say.