Semper Fi - Development Diary #3

[unmerged(162341)]

You guys from Paradox are realy trying hard and for that thank you!

However please note my observation and the key-word here is BALANCE.

Your work with defensive AI is outstanding.
However the atack Ai is not following this, if Atacking AI will stay the same as currently.

Please undestand this simple words :If you make Ai in defence capable to pull back ,shorten frontline and digg in on strategic spots,than you must improve AI atack to--Note: to counter this you must make AI,especialy spearhead formations,being capable of encirclements(large ones),exploitation, cutt-offs (intentional) behind the lines..and chasing-yes its time to introduce term chasing retreating enemy!

So to simplify:Good work, however will again push gameplay from balance to the other side if you dont make Atacking AI more effective.

Please make sure to realise this my words to prevent Semper-Fi being vicitim of to much accent just on defence-which therethicaly and practicaly could leed to "trench warfare",becouse defence will be much more cost effective,atractive and rewarding compared to atack,which of course sounds and it is more like ww 1 ,not ww2.

So I urge you to improve atack AI to.

If you allready have this in mind I opologize and thank you.

 

[joeenochs]

The tooltip of the withdraw stance says: "units won't attack".

Getting a good defense line can sometimes also mean to attack. For example if there is a river and the opponent has managed to cross it only at one place.

Are you guys considering this or will I need the defensive stance for this?

 

[telesien]

You guys from Paradox are realy trying hard and for that thank you!

However please note my observation and the key-word here is BALANCE.

Your work with defensive AI is outstanding.
However the atack Ai is not following this, if Atacking AI will stay the same as currently.

Please undestand this simple words :If you make Ai in defence capable to pull back ,shorten frontline and digg in on strategic spots,than you must improve AI atack to--Note: to counter this you must make AI,especialy spearhead formations,being capable of encirclements(large ones),exploitation, cutt-offs (intentional) behind the lines..

So to simplify:Good work, however will again push gameplay from balance to the other side if you dont make Atacking AI more efective

Please make sure to realise this my words to prevent Semper-Fi being vicitim of to much accent just on defence-which therethicaly and practicaly could leed to "trench warfare",becouse defence will be much more "cost effective",atractive and rewarding compared to atack.

So I urge you to improve atack AI to.

If you allready have this in mind I opologize and thank you.

That may be an issue. If this won't be checked properely we might get from situation when Germany always beats Russia to situation where Germany never even get them hard time. I think I speak for majority of players when I state, that the 50-50 outcome in case of otherwise roughly historical development and neither Germany or SU being controlled by human player would be most entertaining and not that far from reality.

 

[blue emu]

I think Rommel would disagree, especially since we're talking about combat delay, so the tactical level.

You missed a word in my post:

... no sane real life commander...

Agreed. In reality trying to leapfrog your units like HOI3 encourages would be a logistical and transportation nightmare at the tactical level.

... but HOI doesn't pretend to simulate the mechanics of warfare at the tactical level. Did real-life (WW-II) Infantry really get a random chance to "block" incoming Artillery rounds? I can't recall that happening.

It tries to simulate the results.

 

[The Starfox]

It tries to simulate the results.

The issue is though that the attack delay as currently implemented does not simulate as well as the previous solution. The current system has units stopping in all situations, no matter how trivial the resistance, for several days at a time, every time.

The previous system better simulated results because it allowed for pre-planning of attacks, and marching routes, commanders had been given their objectives ahead of time and pushed on non stop if possible until they were achieved. Should those objectives be changed or routes of march altered while they are happening then naturally a delay is applicable. Guderian crossed Luxembourg in one day, and southern Belgium the next, brushing aside light resistance.

In the current game however, he would have stopped for at least 3 days in Luxembourg and then 3 more days in southern Belgium, and then again at Sedan. How is such a breakthrough as occurred several times in that war to be simulated if the units doing the attacking must stopped constantly for several days, allowing the enemy ample opportunity to rush in reserves that are not hindered by any delay at all.

I believe the previous system better simulated results, and do not know of the purpose for the change to the new system, as I see only a worse system not a better one.

 

[unmerged(131628)]

You missed a word in my post:

Originally Posted by blue emu
... no sane real life commander...

And you are saying he wasnt sane?????? He knew very well what he is doing and what are potential risks and opportunities.

 

[unmerged(148761)]

... but HOI doesn't pretend to simulate the mechanics of warfare at the tactical level. Did real-life (WW-II) Infantry really get a random chance to "block" incoming Artillery rounds? I can't recall that happening.

It tries to simulate the results.

Oh I agree that the system abstracts everything that takes place on the tactical level and I am completely ok with that. But in this case the abstraction leads to a strategic simulation that seems unrealistic to me. It is impossible for a division to really drive quickly into enemy territory if it achieves a breakthrough, this has to be done at the corps level which wasn't necessarily true IRL.

For example, in WWII combat the advantage of armor was that armored units had the power and toughness to create a breakthrough and had the speed to exploit it. In game you have to choose one or the other. Often I have my infantry fight and win the battle while my armor sits and waits around, and then after the battle the armor speeds through the broken line and forms an encirclement. Honestly I could probably use this strategy with an all Infantry/Calvalry army and skip on armor altogether. Again, this doesn't really have anything to do with the wisdom of having reserves-- the division that creates a breakthrough (as long as it hasn't taken massive casulties or equipment damage) is the best suited to push forward and exploit the breakthrough.

I just think organization and the attack delay are a little redundant-- if my unit is hit hard, then of course it should have to wait and recover and I should bring a reserve division forward. But this is already covered with the concept of organization, which has the advantage of being related to the scale of the actual combat that took place. Nobody wants to see an armored offensive stalled because an MP brigade or HQ got in the way of your tanks and the drivers had to get out and clean the MPs and staff officers off of their tank treads.

(Finally, I don't expect this to change, and I don't actually care that much, I am just explaining my position.)

 

[jackda]

Yes but the price is that the attack delay is kind of arbitrary. It makes no difference if you attacked a couple of armoured divisions or a single militia unit with your elite forces. That is pretty unrealistic.

I think the critical importance of reserves could just come with a good combat model in general.

For example if your initial breakthrough forces lose org, you need exploitation units as well.

Blue is right, but i think i understand what is your problem.

Attack delay should be scaled with the length of the battle the unit delayed won.

For exemple, if a unit is able to beat a milicia in 15 hours it should be able to start another fight far sooner than if it was fighting 2 Arm divs for several days.

 

[unmerged(76022)]

Oh God those HQ units are so oversized! Just so you know there is a small minority of players who use HQ units as front line brigades attached and doing combat line all other units, now this super sized HQ units are going to mess things up something aweful.

for example my army of panzer's are organized like this:

Army HQ + HARM + HARM + SPART + ENG (speed 5 to 6)

under this army are five (5) division HQ with the same units attached then under each Division HQ are four (4) ARM+ARM+SPART+SPART+ENG and one (1) LARM+LARM+LARM+SPART+ENG.

My HQ's are used in combat just like any unit in the division and they are actually stronger because they use HARM but a little slower so they make great "hinges" on my encircling thrusts.

Now already many providences are small in size and the unit chit covers up valuable infomation displayed on the map and now you guys are going to make supersized chits for the HQs? that is just nasty and going to make manual control much harder... I know you want the AI to do all the fighting in the game but some of us like to micromanage and dont need supersized HQ chits all over the place covering up everything else. I sure hope this is moddable cuz i wont even consider the expansion worthy if all it does is cater to the AI lovers.

 

[LeeDub]

Oh God those HQ units are so oversized!

Disagreed. Any smaller and you might as well throw this feature out the window. They need to be recognizable quickly.

 

[Birken]

Oh God those HQ units are so oversized! ...

Didn't I state in a previous DD (one actually related to HQ sizes), that HQ-size is an option?
The same goes for the hierarchy lines.

Disagreed. Any smaller and you might as well throw this feature out the window. They need to be recognizable quickly.

Yes!

 

[unmerged(53215)]

It looks like the HQ units are bigger only when the view is zoomed out, check the different screenshots. I like this feature as is and would not want color coded units.

 

[Modo]

You missed a word in my post:

Please. Do I have to explain to you the difference between the tactical level, where Rommel did consistently well, and the strategic level, which has nothing to do with the combat delay...

 

[Guibod]

Thanks!
<SNIP>

Did they follow the AARs Kirth Gersen and I wrote about AI? ? I remember you did and you encouraged me. Did they as well? That would be, at some degree, a way to have some more clear ideas about how to give players better interaction tools with the military AI.

No problem Gladiator, both you and Kirth did great for mundanes to tame down the AI system. You both pointed weak points in the AI mechanics that I am sure Paradox Interactive people already knew.
I don't expect Doomdark to spoil an enthousiastic Dev Diary with deficiency aknowledgement. Especially if there's no immediate answer in Semper Fi featured fixes. Those are the fundamentals of communication in IT & video game industry. What I expect, as I trust PI involvement, is that our remarks and feature requests are now part of paradox' private roadmap. We can expect some feedback somewhat medium-term, but not right now.

 

[Jorgen_CAB]

I agree with the camp that find the combat delay a problem. It makes the game on the tactical level feel more like WWI than WWII in many instances. They have a perfect tool such as Organization and if fully used could easily substitute the delay system.

One solution could be that any division that enter into enemy held territory would loose some ORG on movement as well as in combat. Another would be that divisions standing in a province next to an enemy province with military units receive much less ORG regain (and reinforcement as well i presume).

I'm not totally against the combat delay because it needs to be simulated somehow. The current solution is a poor simulation, it forces you to decide if you want to use your armoured divisions to punch a hole or to drive through it. You can't do both unless you use an unhistorical amount of armoured divisions to simulate your armoured corps and armies.

 

[GrimReaper]

Excellent! I will enjoy the new AI for sure

 

[unmerged(52476)]

Would it be possible to make a SeFi Development Diary Archive and Sticky it?

Thanks!

 

[Surt]

I'm really missing some way of controlling garrisons and anti-partisan groups.

Garrisons units should never leave their objectives (until a change in the strategic situation makes it safe), if a crisis occurs the groups mobile units should be reserves to support them. Far too often I see the AI leave some port open because it sends the garrison off to some crisis, thereby opening for invasion.

Anti-partisan should spread out in the interior of the partisan area and hunt any partisan that pops up, when no partisans are near they should protect the supply routes and take back partisan provinces. Partisans far too often causes the AI to get units from far away, from important garrisons or even from the front line and the AI uses too many units at that.

Also all HQ's should have an optional area of responsibility, Theatre already implicitly have it for the theatre, but lesser HQ's should have options like the planes. I could think that you could select one or more of provinces, regions, countries, subcontinents (continuous occupied (or allied) areas limited by Straights, neutrals, own border) or by an radius or cone from some point or by command radius (where axis of attack makes it move else its a static position).

 

[Gankenstein]

I don't expect Doomdark to spoil an enthousiastic Dev Diary with deficiency aknowledgement. ... Those are the fundamentals of communication in IT & video game industry.

No they are not. It is widely accepted that being honest about the state of a product is more profitable in the long term. And why spend time & effort reading stuff if it's all bull anyway? Speaking for myself, I'm much more likely to be a loyal customer to someone who goes "we admit this sucks & expect to adress it at such and such a time" than to someone who stonewalls obvious issues.

<-- aborts rant, enough said already

 

[Gladiator]

I'm really missing some way of controlling garrisons and anti-partisan groups.

Garrisons units should never leave their objectives (until a change in the strategic situation makes it safe), if a crisis occurs the groups mobile units should be reserves to support them. Far too often I see the AI leave some port open because it sends the garrison off to some crisis, thereby opening for invasion.

Anti-partisan should spread out in the interior of the partisan area and hunt any partisan that pops up, when no partisans are near they should protect the supply routes and take back partisan provinces. Partisans far too often causes the AI to get units from far away, from important garrisons or even from the front line and the AI uses too many units at that.

Also all HQ's should have an optional area of responsibility, Theatre already implicitly have it for the theatre, but lesser HQ's should have options like the planes. I could think that you could select one or more of provinces, regions, countries, subcontinents (continuous occupied (or allied) areas limited by Straights, neutrals, own border) or by an radius or cone from some point or by command radius (where axis of attack makes it move else its a static position).

All of that fully reflects my viewpoint, my experiments, and my wishes for future AI developements!

Great post!