Slitherine

if it's been mentioned I'm mentioning it again!

spamming! I have no problem with spamming by the AI. it is just the AI after all. what I don't like is the all of one type spam just finished Kkarkov42 and the AI was spamming more KV-1's than the USA spammed tanks of the whole war how about the spam be more of a mix of equipment around a city?

huertgenwald wrote:The one thing which needs changing is that elite repair only increments in single steps, as it does from strength 10 on upwards.

This is a good idea. It would add very interesting tradeoffs to the game play, especially if between-scenario elite replacements were less expensive than in-scenario elite replacements (which would be more "realistic"). If they all cost the same, it wouldn't matter so much. Maybe best implemented only at higher difficulty levels, though.

I forgot to mention some important things which I think should be considered for the new mechanics

1. Plane Spotting
Planes should reveal the terrain for the current turn when overflying an area...as by now they only reveal their spotting range hexes on arriving at their destiny hex...this is pretty unrealistic and may be considered for redesign
maybe a possibility for ON/OFF in gamerules.pzdat

2. New Trait
outdated or noproduction or something similar
This is the absent trait between nopurchase noupgrade and noreplace
It means that this equipment can not be replaced anymore when its production date has ended and has mandadorty to be upgraded in
order to conserve it...
This mainly will improove realism in user campaigns and scenarios but opens interesting new tactical considerations in MP too.
It feels weird to replace a PzIII or a 39 Infantry Model in 1944 since the equipment was not longer in production since a while

3. Annoying AI blunders
I run a test to observe the German AI at the Kursk scenario and found that Tanks attack entrenched AT guns although an Infantry unit is available too..that is as much as silly as placing Artillery in a Truck next to the enemy and should be adressed if possible

What also may worth consideration:

REINFORCEMENT/UPGRADE ZONES
to be placed on ground via Editor (maybe with some indication...a Depot Hex ?)
R= Reinforcement
U= Upgrade
to allow only in specified areas to place new purchase or upgrade equipment

Checkbox Editor (Reinforcement/Upgrade Hex ON/OFF)
If checked city upgrade and city unit placement deactivated

This would greatly increase tactical depth mainly for user scenarios/campaigns but as well for MP and maybe even for Afrika Corps when taken into account of sceanrio design from the beginning. The Africa campaign failed by a fair share due to a lack of constant supply ..this may be interesting to exploit

EXPORT MORE GAMERULES TO gamerules.pzdat
examples:
probability of evade to take effect
how many turn after capturing city to upgrade/place new units

NEW TRAIT
Class: Recon
retreat = Evade an attack and retreat > recon (user made content) > ability to assign % in gamerules > applys to Ai units too
Class: Fighter,Tac Bomber,Strat Bomber
outmaneuver = Evade ability for planes > recon (user made content) < ability to assign % in gamerules > applys to Ai units too

EDITOR SCENARIO PARAMETERS

Ability to determine how many max units per unit class are allowed in scenario
Only apply if number > 0 otherwise ignored
Only for main unit classes from 0 to 9

Together with max core/aux/elite slots this could be a great tool for vanilla campaign design as much as user made campaigns and MP maps as it opens more strategic options.
As well this could influence AI purchase and avoid spamming of the same unit type all over the place

robman wrote:

huertgenwald wrote:The one thing which needs changing is that elite repair only increments in single steps, as it does from strength 10 on upwards.

This is a good idea. It would add very interesting tradeoffs to the game play, especially if between-scenario elite replacements were less expensive than in-scenario elite replacements (which would be more "realistic"). If they all cost the same, it wouldn't matter so much. Maybe best implemented only at higher difficulty levels, though.

Elite replacements already are less expensive in pre-deploy, are they not?

4kEY wrote:

robman wrote:

huertgenwald wrote:The one thing which needs changing is that elite repair only increments in single steps, as it does from strength 10 on upwards.

This is a good idea. It would add very interesting tradeoffs to the game play, especially if between-scenario elite replacements were less expensive than in-scenario elite replacements (which would be more "realistic"). If they all cost the same, it wouldn't matter so much. Maybe best implemented only at higher difficulty levels, though.

Elite replacements already are less expensive in pre-deploy, are they not?

Yes. That disparity would need to be continued (or even increased) if this suggested change were to have any significant effect on game play.

Rudankort wrote: A separate question is, what new options for configuring the AI are needed in the Editor to enable better user-created single player scenarios.

I know its not directly related to AI configuration but it would help infinitely if there could be 2 types of triggers..
those who run only once and expire and others who run every turn in the time specified: example(1/-1) would run every turn from start to end...this important to introduce getting prestige for holding objects or reinforcements or or or...makes great difference if a player has to go for certain objective in order to have sufficient prestige in order to win scenario or facing a constant flow of reinforcements giving certain conditions...this could avoid the annoying scripting of single triggers for each turn

dks wrote:if it's been mentioned I'm mentioning it again!

spamming! I have no problem with spamming by the AI. it is just the AI after all. what I don't like is the all of one type spam just finished Kkarkov42 and the AI was spamming more KV-1's than the USA spammed tanks of the whole war how about the spam be more of a mix of equipment around a city?

Thanks for bringing this up! This is the only AI behavior that really raises my blood pressure. Proper use of artillery as so many others have mentioned would be my only other request.

I'll second huertgenwald's request for the ability to re-enforce in increments of 1.
As some one who plays casually for fun and enjoyment, if the AI improves its kill rates this feature will be vital for new/casual players like myself, especially in the post 42 DLC's.
Great to see the Africa Corps version is on the horizon to add even more to a fantastic game.

Some wishes are off topic - not only for AI

* show hero bonus for units in the unitlist (tooltip over the small hero-heads)

* reload equipment.pzeqp (cheat code) or cheat codes for heroes and kills (for testig purposes)
* add trigger-condition "select unit <unitname>" to explain new units
* add trigger-action "open file <filename> in a window" to explain new units
*add trigger-action to remove a unit from the map (and to change a unit to an other unit placed in scenarioeditor on the same hex with transfer of heros, awards, experience, kills...)
*add trigger-action to add experience (and other bonus) to a unit
*give us the left ID's (example: HoverUnitStar6... for the main.htm) to show experience up to 999

*define buy-list for the computer-opponent (in scenario editor)
*define targets for units (enemy units or hexfield) for the computer-opponent

AI-helping traits
*avoid direct contact (most artillery-units)
*force enemy contact (KV2)


later.....

*remove all implicit trait-restrictions (para only for Infantery..)
*use only traits for unit-characteristics. no implicit traits. You have the traits-column in the unitclasses file. Use it, please!
*allow definition of bonus for awards (prestige at the beginnig of a scenario based on the medals: 3 units with ironcross1 -> 3 x <prestige column in awards-file> at the beginning of a scenario + x units with ironcross2....)

Rudankort wrote: A separate question is, what new options are needed in the Editor to enable better user-created single player scenarios.

A very important issue is the ability to tie weather actions to action zones or new "weather zones" in order to generate different weather on different map sections.

A max of 6 weather zones would make a great difference...they could be placed and work exactly like the action zones so its almost the same code just tied to a different string

The weather display in the UI would have to be amended accordingly to show weather by "hoover with mouse over hex" and the minimap could show darker groundtiles for rain/clouds and brighter ones for snow for their corresponding areas.

So with little effort and ressources there could be created a huge improvement

Air units again:

Dunno if that is just my experience, but it seems that AI air units set to attack sometimes don't fly back to base to rearm even when out of ammo. It seems that AI is keen to use the hex right on top of the air base, but not those adjacent to it, which allow to refuel and rearm (but not repair). So if the airbase is occupied, other AI planes don't return (at least that is my impression)....maybe this has been introduced to make the AI use carriers (which are one-hex only)?

Anyway I had several instances in which planes out of ammo flew around over my units for one or more turns.

If that is a common issue experienced by others as well it would be nice be improved, so that airfields are used better...

After reading airborne's thread about 1.06 I just want to add one thing that would IMO help modders quite a bit, but it is not AI related:

The current campaign system described here viewtopic.php?f=147&t=29103 works reasonably well for mods that don't change certain game files.

However IMO one of the best things in PzC from the modding perspective is that nearly everything can be changed, be it gfx, dat or even html files.

The current workaround for bigger mods changing these files is either using GME (which seems to be not developed further anymore, but remains free), or even use another installation. Both are not ideal, esp. the installation and usage with GME can be confusing and error-prone. If it is not already on the dev team's radar I'd like to ask for a major improvement in this field, like a "moddir" function that covers replacements for all moddable game files.

I can't comment whether this is technically easy to do or even possible, but if it could be provided it IMO would help a lot.

bebro wrote: However IMO one of the best things in PzC from the modding perspective is that nearly everything can be changed, be it gfx, dat or even html files.
I can't comment whether this is technically easy to do or even possible, but if it could be provided it IMO would help a lot.

chris10 wrote:

Rudankort wrote: I agree that these two tools are absolutely great, but what help exactly can we give in relation to these? I was talking more about making the mods more exposed to the player community in general, to people who never visit this and other forums in the first place (and such players are a majority).

saying that the mod support should be increased making the exe always looking in the mod folder first and only drag from the main game what it doesnt find there..this would enable complete conversion,user ui customization,overlays etc etc and a generic mod enabler would be completely obsolete for PzC which could boost the amount of people willing to look into mods.

I support many of the suggestions that have already been made, about AI attack sequencing for example.

I'm not too sure the AI should prioritize attacks on fighters with its own fighters in the first few years when the Germans have air superiority. The AI's planes now generally inflict more damage to my ground forces in the couple of turns they're on the map than they would inflict when attacking an experience fighter unit. In late war scenarios, maybe the AI could indeed focus more on achieving air superiority.

The scenarios where the AI should be attacking could also be more challenging by making sure the AI will attack from as many sides as possible (the AI should get enough units to do this). In Demyansk Pocket, the AI gave me all the time I needed to move out and fight it on my terms as its attacks seemed to be linked to a supply unit arriving at an airfield. As such, the scenarios was a lot easier than it could've been if the AI had kept on attacking.

To me, that's the primary current weakness of the AI on the operational level: it doesn't really time its attacks with other attacks along the line and there are often a number of units at the edge of the map that don't really do anything in the scenarios where the AI is defending or a large number of units at the edge of the map where the AI should be attacking, but is constrained by having to wait for triggers (it seems). Making the AI attack like a human player is more or less impossible, but the suggestion about giving the AI a balanced force at various locations on the map is a good one. It would hopefully also decrease the chance that the AI will buy tanks, such as KV-1's early on in the Soviet campaign and IS-2's later on.

The AI's use of artillery is fairly underwhelming in general: it will hardly ever, or simply never attack a non-artillery unit that isn't also being attacked by a ground unit. I have seen the AI fire at infantry units and my artillery, but that's about it if a unit wasn't also attacked by a ground unit earlier on (as already mentioned, the habit of the AI to use artillery after an enemy unit has been attacked by a ground unit is also something that could be improved).

Perhaps the AI could also get some help when it comes to determining where it should place newly purchased units. Currently, the AI places units at various hexes without much regard for the chance the units will survive for a single turn or if the objectives is being assaulted by a large number or a small number of my units. I'm not sure if it's possible to assign priority objectives to the AI that need to be held at all costs, but that could help. It would also help prevent a situation where the AI spreads its units over multiple objectives, decreasing the overall efficiency of the defending forces.

Another issue coming to mind regarding the AI:

As it seems to be driven by victory hexes, it can act a bit strange when it tries to make a naval invasion. Depending on map layout and location of VHs it sometimes rather tries to approach them via sea staying embarked, even when there are nearby beach hexes where it could land to progress from there. That makes these units of course quite vulnerable.

Not sure how this could be improved in the best way, but if there's work to have the AI handle land transportation better there's maybe some fallout from it that helps with this issue too.

That would be corrected by allowing the designer to specify target hexes for individual units. This has been requested further up the thread. It is especially useful for co-ordinating attacks between air and ground units. It would be great if it could be done like triggers are now - by unit name etc. That way a division can be assigned a target/objective, another will go to another target etc. If it could be done progressively that would be brilliant - take A then B then C.

I'm not sure how hard this would be to do from a coding POV, though. Probably a pain in the end.

right

My biggest gripe about the AI is the retreat choices it makes for both my units and the enemy. Defeated units should never run in the direction of the attack, they should run away, like Sir Robin.

All too often I see units with perfectly safe paths away from harm instead plunge a hex deeper into enemy territory. Like most board war games, the first choice for a retreat hex should be one not touching an enemy unit. That would eliminate all these absurd suicidal charges.

We aren't playing with the Imperial Japanese army, after all.