Defensive Battles

[AgentTBC]

I've finished up through DLC '43 (on Rommel) and so far and while some of the offensive scenarios are becoming appropriately difficult to obtain a DV and can really grind away at your unit strength (Prohkorovka ugh), the defensive scenarios are still mostly trivially easy. I think some of this is how poor the AI is at making use of artillery what with generally launching its ground attacks before firing the arty. I'm not sure why the AI can't be coded to always perform the suppressing artillery fire before ground attacks because the one time I lost a unit on a defensive battle was the one time the AI suppressed a grenadier unit with two arty shots before launching an infantry assault and caused it to surrender since I had not left a retreat path for it, never having seen the AI suppress with arty before and not expecting it.

Even apart from that I think the defensive battles need some extra oomph. Given my experience on Rommel I think the only question on any of the easier difficulty levels would be not whether you get a DV or not but whether you can completely wipe all Soviet units off the map before the scenario ends.

Like I said, proper use of AI artillery is probably the single thing which would help the defensive scenarios out which doesn't involve simply throwing in extra units.

[4kEY]

Yup. It'd open a whole new level of strategy.

[El_Condoro]

This was predicted a long time ago when the DLCs first appeared. Until Rudankort returns there won't be any news on an upgrade to the AI, especially in its (in-)ability to launch effective attacks without swarms of units (and even then failing miserably!) Artillery suppression is certainly one area that needs to be improved. Another is the AI's habit of sending powerful units into close defense terrain such as forests and hills; if it stayed clear of those terrain types with its big guns it would be more effective, too, and less AI units would be needed in each scenario to give the player a headache.

[Zhivago]

This was predicted a long time ago when the DLCs first appeared. Until Rudankort returns there won't be any news on an upgrade to the AI, especially in its (in-)ability to launch effective attacks without swarms of units (and even then failing miserably!) Artillery suppression is certainly one area that needs to be improved. Another is the AI's habit of sending powerful units into close defense terrain such as forests and hills; if it stayed clear of those terrain types with its big guns it would be more effective, too, and less AI units would be needed in each scenario to give the player a headache.

Can't Lordz/Slitherine find someone else who knows how to work with this PzC computer code (or can figure it out) besides Rudankort? Panzer Corps has to be one of Lordz/Slitherine's best and most profitable game titles now. It seems pretty risky to have the entire franchise dependent on whether or not one person decides if he wants to work on the game or not anymore.

[Kerensky]

Unfortunately it's not that simple.

First of all, defensive scenarios are potentially lethal so we intentionally take the teeth out of them. Who could forget the horror stories of the Bagration scenario of the Vanilla game? Putting the AI on the offensive against an unprepared player can absolutely gut their CORE, leaving them unprepared for the rest of the campaign. Considering the complete Grand Campaign currently has something like 80 scenarios, and will be complete at almost 100, a battle that decimates the player isn't really an option. At least not until we get to late 1944 and 1945.

Also, offensive scenarios take the initiative away from the player, which has to be carefully managed. When the player is attacking, other than AI counter-attacks, they advance at their own pace. If a unit is beat up, they can leave it behind their advance to refit and repair. In a defensive situation, moving behind the front line may not always be an option, especially when the AI is pressing your front and flanks with high speed armor such as T-34/43s and T-34/85s.

Bottom line, it's a fairly complicated situation, but we feel that the solutions we have are the best for the vast majority of our players. We've seen that most of our forum going players are typically our more advanced players so we have to be careful not to improve and buff the AI too much.

[El_Condoro]

There are 5 difficulty levels and 3 more 'advanced' difficulty levels, so perhaps one or more of those could be buffed for those players who would like the AI to play more intelligently?

[Amaranthus]

Thanks Kerensky, I appreciate those thoughts - but couldn't you reserve a souped up AI for the General+ difficulties settings that bars no holds? Most forum players use FM or higher difficult settings, and would like some tough defensive battles. Most causal players probably use Colonel or lower, and wouldn't therefore be effected.

I know I personally dislike giving the AI too many artificial advantages (e.g. Manstein overstrength), and would much prefer to have a focus on a human-like AI thinking (make it as tough as you can - a bit like a strong chess AI). This is the thrill I get out of PBEM (the unexpected moves, and lack of predictability of long-term plans [the AI has little]) that the SP game (which I otherwise love) currently lacks - sorely.

[Kerensky]

That is always a possibility, but it's also not my area of expertise.

[4kEY]

Who could forget the horror stories of the Bagration scenario of the Vanilla game? Putting the AI on the offensive against an unprepared player can absolutely gut their CORE, leaving them unprepared for the rest of the campaign.

We've seen that most of our forum going players are typically our more advanced players so we have to be careful not to improve and buff the AI too much.

Hehe..I remember Bagration made me sweat. I became a better player there. Not losing any units was a challenge even with oodles of Tiger IIs.

Nowadays I'll just go into the scenario editor and raise the Exp of the enemy. I started 43 East with computer Exp raised by 200 points.

Supression is such a big deal that if the AI was programmed for it, all but the most recon-stocked would be gutted.

[VPaulus]

Can't Lordz/Slitherine find someone else who knows how to work with this PzC computer code (or can figure it out) besides Rudankort? Panzer Corps has to be one of Lordz/Slitherine's best and most profitable game titles now. It seems pretty risky to have the entire franchise dependent on whether or not one person decides if he wants to work on the game or not anymore.

It's not that easy. Please understand that the code was all done by Rudankort, and under these circumstances it's always very hard to find and fit someone for that position.
Besides Rudankort is the Alma Mater of Panzer Corps.
Rest assure, that the AI will be fine tuned sooner or later, and that Panzer Corps is still a living project and it's not resumed just to new DLC. However we need a little patience and please remember that this is not a AAA developer nor a AAA title.

[dks]

just a few thoughts. first we would not have a PzC as we know it without Rudankort. he took PGF so we could play PG again on up to today's date OS's. from PGF we got PzC with the help of some die hard PG fans and many others. soon I suspect a PzC 2 will come out after PzC 1 with much more goodies. this time unlike PG, one can mod the heck out of this PzC if one wants to without hex editors and special made editors after the fact.

an old PG fanatic here and getting my rump stomped again and again playing PzC if I get impatience and try to move toooooooooooooo fast thanks to Rudankort and company

[Zhivago]

Can't Lordz/Slitherine find someone else who knows how to work with this PzC computer code (or can figure it out) besides Rudankort? Panzer Corps has to be one of Lordz/Slitherine's best and most profitable game titles now. It seems pretty risky to have the entire franchise dependent on whether or not one person decides if he wants to work on the game or not anymore.

It's not that easy. Please understand that the code was all done by Rudankort, and under these circumstances it's always very hard to find and fit someone for that position.
Besides Rudankort is the Alma Mater of Panzer Corps.
Rest assure, that the AI will be fine tuned sooner or later, and that Panzer Corps is still a living project and it's not resumed just to new DLC. However we need a little patience and please remember that this is not a AAA developer nor a AAA title.

Of course it would be preferable for Rudankort to come back--he helped raise Panzer General from the dead. However, no matter what product a company is making or selling, it would seem logical to have more than one person who knows how to make or sell the product. If Rudankort decided never to come back, Lordz should have some plan B to continue the franchise. And I disagree that this is an AAA game--it could be an A game if it was developed for more platforms like Mac and the ipad.

[VPaulus]

Alex just returned and is doing fine under circumstances, we cannot give and information about a Mac or iPad version of Panzer Corps, just that we still have it on the planning, although delayed as Alex took a very well deserved and much needed break.

It is a private matter though.

[Gwaylare]

The skill of an AI is a critical part of any round based game like PzC. Realtime strategy is dominating the market, because they are able to deal better with a bad AI. Round based strategy cannot. So to develop PzC further on for me there two ways: Focus on a PvP part and get a world wide multiplayer community or develop a real AI, able to deal with all features of the game and able to launch counterattacks.

So I tried to play DLC 39 but did not play beyond Piatek, because it just was not a challenge at all.

Do not get me wrong, I like PzC very much and I play this game daily. But I use the multiplayer mode only. So the game is really great and the scenarios are well desinged and quite good balanced. But to buy further DLCs for me this game needs a serious AI.

Best Regards
Gwaylare

P.S.
It would be a great idea to have a modable interface to modify AI behavior. A simple list of DOs and DONTs will strengthen the AI a lot.
Another simple trick would be to go through the units more than one time, to give them the chance to move some units first and attack after that or to fire with artillery before an attack.

[AgentTBC]

The reasons for a developers absence are indeed a private matter which we should have no interest in; the effect that absence has on the quality of a game available for purchase is very much a public matter and is certainly open to discussion. It doesn't really matter why the AI hasn't been tweaked in some very basic and important ways, only that so far it has not. The primary example being, of course, the inability of the AI to use artillery. It would be very difficult or impossible for the AI to use it as well as human... but it would be trivial to make it use artillery much better than it does now. Consider the following changes:

1) The AI should never under any circumstances move an artillery unit next to an enemy unit deliberately.
2) The AI should move and fire artillery before any other units.
3) The AI should prioritize suppressed enemy units for ground attack.

BOOM, you've just vastly improved the play of the AI player with three extremely basic and easy changes. Yes, we can come up with examples of times it is good to move self propelled artillery next to an enemy unit, as in when you wish to cause a surrender. HOWEVER, the AI is incapable of those maneuvers except purely by accident so preventing it from ever moving artillery next to enemy units will be an improvement in 98% of cases. And so on.

I think the changes I list above would make for a much more interesting and fluid game.

[soldier]

Unless the artillery units need to move into position to fire, it would be good if they could fire first but actually move last. That way (hopefully) other units would be in front of it at turns end and the tank attacks would be much more effective.
I remember some people complaining about the first incarnation of Bagration being too tough, well it would have been a lot tougher if all the German units had been suppressed before the IS 2's let loose. Seems like a good method to make defensive battles tougher and the AI more dangerous.

[MartyWard]

Unless the artillery units need to move into position to fire, it would be good if they could fire first but actually move last. That way (hopefully) other units would be in front of it at turns end and the tank attacks would be much more effective.

I think just doing that would make the AI so much better. It seems like it should be able to be done.

[brettz123]

I agree the artillery AI really needs to be tweaked as right now it is only any good in defensive situations. My concern is that this keeps the game playable 1939-1943 but once the 44 and 45 DLCs roll out where the AI will be attacking more often it should be a concern to everyone who plays the game how inept the AI really is at using artillery offensively.

[boredatwork]

Unfortunately it's not that simple.

...Putting the AI on the offensive against an unprepared player can absolutely gut their CORE, leaving them unprepared for the rest of the campaign.

^This

The problem with asking/expecting changes is that not everyone enjoys the game for the same reason and improvements to one area for some people may negatively impact the enjoyment in others.

PG was one of the most commercially successful turn based PC games not on the strength of it's AI - which was in many respects much worse than PzC - but because of the RPG nature of it's campaign. Making PzC's AI more artillery savy would indeed improve the challenge on a macro level by making defensive scenarios more challenging. The problem is with the high casualties in PzC it would also *greatly* increase the probability of total kills of CORE units - which runs counter to the RPG element which is the game's strength over the thousands of other wargames out there.

I'm not saying that it shouldn't be done - but IMO it shouldn't be done unless it is done as part of an overall plan that will add challenge without losing the RPG appeal. The simplest course IMO would be adding the ability to reform destroyed units so that histories and medals remain intact even if experience is continually reset to 0.

[brettz123]

Unfortunately it's not that simple.

...Putting the AI on the offensive against an unprepared player can absolutely gut their CORE, leaving them unprepared for the rest of the campaign.

^This

The problem with asking/expecting changes is that not everyone enjoys the game for the same reason and improvements to one area for some people may negatively impact the enjoyment in others.

PG was one of the most commercially successful turn based PC games not on the strength of it's AI - which was in many respects much worse than PzC - but because of the RPG nature of it's campaign. Making PzC's AI more artillery savy would indeed improve the challenge on a macro level by making defensive scenarios more challenging. The problem is with the high casualties in PzC it would also *greatly* increase the probability of total kills of CORE units - which runs counter to the RPG element which is the game's strength over the thousands of other wargames out there.

I'm not saying that it shouldn't be done - but IMO it shouldn't be done unless it is done as part of an overall plan that will add challenge without losing the RPG appeal. The simplest course IMO would be adding the ability to reform destroyed units so that histories and medals remain intact even if experience is continually reset to 0.

Good points but this could be implemented as part of the harder difficulty levels. So you can satisfy both parties I think.