Development Diary #8 - 3rd of December 2008

[Sangeli]

If a chess computer in the 1980's was nearly as the best human player, I see no reason why the AI in HOI 3 can't be at least adequate? The AI in itself has advantages that the human player doesn't; the AI lacks (or should lack) fog of war giving it complete awareness of the current situation. Additionally, the AI can perform caculations whereas the human can only make educated guesses. And as a result of knowing the exact situation and having the ability to perform battlefield calculations perfectly, the AI should have the advantage of being to plan a perfect campaign. That means anytime the AI is on the strategic offenive it can calculate the best path to encircle the maximum amount of enemies taking into account the exact time it would take the human to counterattack, retreat, etc. The AI has the ability to outplay the human player but the right now the system is garbage. All that is needed to fix it though is tact programmig and trial and error.

[Jamie550]

If a chess computer in the 1980's was nearly as the best human player, I see no reason why the AI in HOI 3 can't be at least adequate? The AI in itself has advantages that the human player doesn't; the AI lacks (or should lack) fog of war giving it complete awareness of the current situation. Additionally, the AI can perform caculations whereas the human can only make educated guesses. And as a result of knowing the exact situation and having the ability to perform battlefield calculations perfectly, the AI should have the advantage of being to plan a perfect campaign. That means anytime the AI is on the strategic offenive it can calculate the best path to encircle the maximum amount of enemies taking into account the exact time it would take the human to counterattack, retreat, etc. The AI has the ability to outplay the human player but the right now the system is garbage.

You can more or less test every possibility in chess without too much work.

With complex games like HoI, not really

[peo]

If a chess computer in the 1980's was nearly as the best human player, I see no reason why the AI in HOI 3 can't be at least adequate? The AI in itself has advantages that the human player doesn't; the AI lacks (or should lack) fog of war giving it complete awareness of the current situation. Additionally, the AI can perform caculations whereas the human can only make educated guesses. And as a result of knowing the exact situation and having the ability to perform battlefield calculations perfectly, the AI should have the advantage of being to plan a perfect campaign. That means anytime the AI is on the strategic offenive it can calculate the best path to encircle the maximum amount of enemies taking into account the exact time it would take the human to counterattack, retreat, etc. The AI has the ability to outplay the human player but the right now the system is garbage. All that is needed to fix it though is tact programmig and trial and error.

Not so much no.
Chess is a very simple game in regards to moves. Had HoI been a turn based game with very little possibility for innovation by the player it could be simpler since then you could reduce it to calculations.
However that isn't HoI.
In theory any computer can given enough time to count beat any player in chess since it has simple rules and fixed moves.

[Johan Jung]

If a chess computer in the 1980's was nearly as the best human player, I see no reason why the AI in HOI 3 can't be at least adequate?.

Depends on if you consider war games to resemble warfare (chess) or simulate warfare (war games). I prefer the later. How are you going to represent "stupid/irrational/non-cooperational" human behavour in a game-AI? Of course, I prefer playing against human opponents rather than any AI:-) Also, when it comes to chess, it doesn't use probability outcomes of your moves/battles (ie, always 100% success, and warfare is a lot more about uncertainty/chaos etc). IMHO.

[Bobb4]

If a chess computer in the 1980's was nearly as the best human player, I see no reason why the AI in HOI 3 can't be at least adequate? The AI in itself has advantages that the human player doesn't; the AI lacks (or should lack) fog of war giving it complete awareness of the current situation. Additionally, the AI can perform caculations whereas the human can only make educated guesses. And as a result of knowing the exact situation and having the ability to perform battlefield calculations perfectly, the AI should have the advantage of being to plan a perfect campaign. That means anytime the AI is on the strategic offenive it can calculate the best path to encircle the maximum amount of enemies taking into account the exact time it would take the human to counterattack, retreat, etc. The AI has the ability to outplay the human player but the right now the system is garbage. All that is needed to fix it though is tact programmig and trial and error.

The fact that human's beat chess playing computer all the time... well enough said.

The computer can fight battles but it has a hard problem sticking to a plan, it is forever evaluating your moves and countering them.
You retreat it attacks. You fall back it follows. You slam the door and its army is encircled. Did the AI do anything wrong? No.
Would a human player have seen the bigger picture, maybe? So getting the AI to do the right thing is a lot harder than you think.
Now add in to the equation that there is no Comtern, Allies or even Axis for the AI.
Each country is fighting by itself and you can understand just one of the problems the AI programmers’ face.
Now lets throw in research, logistics, supply, diplomacy leadership and another 10 000 odd statistics and the movement of just one division by you can cause the AI to change literally thousands of variables.
Yes it theoretically should be able to fight better but fighting is only a minuscule part of the game.
My hats go off to the game designers, their task is thankless and despite having two highly successful games under their belt, no one will be totally happy with what HO3 eventually is.
Modders will go to work, forums will spring up and thread after thread will complain and later a much awaited patch/expansion will fix it… that is until there is talk of HO4

[Paul_M]

This is really for the paradox people.

When you look at division in WW2, and compare them from the various countries for the most part they are fairly similier.

Almost all infantry divisions had 3 regiments (or brigades) of infantry with 3 battalions, 1 regiment of artillary, and then one extra "regiment" of support troop battalions (Anti-air, Anti-tank, engineers, and recon). If you count this up you see you need 5 "brigade" slots to make a standard division.

I don't know what is planned but a way to differentiate divisions is by making various kinds of brigades of a type. So for the support brigade (which is a rather important difference between the countries) you could have:

Support Battalions Poor (this would represent the Russian infantry divisions as they had only companies rather than battalions for their support troops).
Support Battalions Avg (this would represent the commonwealth)
Support Battalions Strong (this would represent the Germans/SS/USA divisions who had generous alotments of AT, AA, engineers, and recon troops)

The difference is in time and cost, so you make a choice between quality and quantity and cost.

There are other things such as over time divisions changed. Some concrete examples are:

Germany infantry divisions started with 3 regiments of 3 battalions and in 42 switched to 3 regiments of 2 battalions (could be the date is later) so that a "Infantry 41 division" is considerably more robust then a late war infantry division. There are differences as well as the later war division had a "heavy weapons/recon" battalion I think was lacking in the early division. Also the battalions themselves were smaller but had far more automatic weapons/support weapons per man.

German Panzer divisions started with 2 mech/motorized infantry regiments of 2 battalions, 2 panzer regiments with 3 battalions, an artillary regiment (a fair number were self propelled), engineer battalion, AT battalion (often self propelled), AA battalion, armoured car/motorcycle infantry battalion (or two) but in 41 essentially all panzer divisions lost the second regiment and the 3 battalion when they expanded the number of divisons leaving them with only a single panzer regiment with 2 battalions of tanks.

US/UK armoured divisions started tank heavy and infantry lean and over time lost a fair number of tanks and gained more infantry, the exact numbers of which is a tad variable so I don't want to run off memory here.

The trouble with most of the brigades in HOI2 is that they aren't normally brigade size formations except for artillary most of the time they would represent a battalion sized formation (recon, AA, heavy tanks, SP-AT).

Some of this probably absurd details that don't need to be in the game.

However, I'd suggest looking into how many brigade slots you allow and making a "support" brigade rather than the specific AT/AA/Recon and so forth brigades. How you want to treat those particular elements is harder to comment on as it depends on a lot of things but making them nominally battalion sized formations added into an extra slot is reasonable.

Just about every country had these sort of extra units, it comes from the fact a lot of units were raised as regiments and then stuck together to form divisions. This was true for both the commonwealth and germany, and was very true for Russia in 42 when it was rebuilding (they had very few officers that had the training to handle a division due to both the purge and losses of front line troops and so had a lot of brigades). They normally existed as "corps troops" which were mixed and matched to the needs of the moment. The US also had them by the bucket load which is why most of their infantry divisions would end up with a tank battalion attached plus some AAA etc. But these types of units show up a lot in any operational game/simulation and amount to a fair amount of combat power in total as a lot of artillary existed in this form (large calibre guns for counter battery fire for example).

I'll also comment that you might want to look into making types of the various brigades: so infantry brigades that have smaller manpower and IC costs for less combat power to more closely approximate a lot of the minor countries militaries. How much "flexibility" comes at a cost of programing difficulty and fiddle faddle is hard to say from the outside but it might be at least worth looking into.

Again in summary my main point is could you look into the exact number of slots available again? The vast bulk of real world divisions had 5 slots, depending on if you consider the support units "folded into" the infantry or not.

[Paul_M]

The fact that human's beat chess playing computer all the time... well enough said.

The computer can fight battles but it has a hard problem sticking to a plan, it is forever evaluating your moves and countering them.
You retreat it attacks. You fall back it follows. You slam the door and its army is encircled. Did the AI do anything wrong? No.
Would a human player have seen the bigger picture, maybe? So getting the AI to do the right thing is a lot harder than you think.

Chess is a very straightforward game to program as there exists large bodies of manuals on what counter to make, there is no terrain, and a strickly limited board. Where a computer beats a human is that it has the ability to evaluate moves in huge numbers of steps ahead. I used to play chess daily and even after years of doing so seeing the board more than 3-4 steps ahead took a "good" day. Also playing against a computer is very "dry" as it takes out the human element.

You example is a bad one because yes the AI goofed as no a human would not have missed that as the strength on your flanks is a particular concern for anyone with a lick of sense. The proper evaluation of enemy strength and intention is paramount of planning. I've called a halt to offensive operations for exactly this reason. I could go forward more but I would be exposing my flanks and leaving myself dangling in the wind which is not something that makes sense to do (normally, as always, particular circumstances could change that).

The task faced by the AI programers is brutally difficult as there are, unlike in chess, very few hard and fast rules. But the largest problem I see with the AI is it does not make long term plans, and it tends to react too much sending units scurrying back and forth endlessly. Watching a unit you have lent to the AI sometime its behavior often comes close to nonsenical. I have seen air units shipped all over the world and rarely even begin to understand why they are there and when they are there its rare they actually seem to do something.

How to program in a long term planning loop is hard to say. I would hate to try and do it (and yes I do program). Just putting down why and how I plan an assault I see a large number of IF-THEN-ELSE conditions showing up. Then once the attack begins and the situation becomes fluid this is when you need quick reactions to developing conditions. Plus the proper utilization of reserves and the reformation of a reserve, stop lines etc. All of this is a complex interaction of short term goals within an overall framwork of a long term objective. But first and formost you need that long term objective as "being only reactive" means basically you have lost in some situations. Reactiing to opportunities is different then always "responding" to the enemy...I hope this makes sense.

As I have a goal with specific objectives the AI is hard pressed to counter this in general. The same it true in Chess, if I have a plan and you don't you may complicate my life but its unlikely in the extreme that you are going to defeat me. As the AI lacks a plan its only chance at sucess largely lies in overwhelming numbers and when that fails it has nothing to fall back on. This isn't to say that even the current AI can't give you a "hard time" I can think of several examples from my last game but for the most part the lack of an overall operational plan is crippling.

As a concrete example from my last game "Overlord" in fall of 44 had the following objectives: Land at Caen, drive through Paris to the Vichy Border as a primary attack goal to isolate the western battlefield from supply, mop up the divisions trapped. Stage 2 establish a secure boundry using the minimum number of required provinces (somewhat reactive). Changes to the original plan were the re-use of the invasion Corps to secure a province by naval invasion after Caen (coast hopping) and that they were used to secure the coastal flank, plus occupation of Djion due to response to enemy moves. And lets not talk about all the naval movements required to assemble the invasion force, use of naval gunfire, plus airpower used to both prepare the battlespace and during the combat itself. My mind boggles at trying to write a conditional sequence that could replicate that.

You can only program something like this if you have first a routine that establishes theatre level objectives plus evaluates them for plausability since piling up troops for an invasion makes no sense when you have no transports for example.

Another concrete example. Japan controls India China and Persia. Where/how to proceed?
Option 1: continue drive to home islands, by island hoping and closing in until bases which allow strat bombing to commence.
Option 2: land in occupied India and cut the japanese forces in half making use of the better terrain to exploit an armoured/mechanized force advantage.
Option 3: land in annexed Persia to allow for a US controlled province to ensure supply without requiring British involvement.
Option 4: declare war on Iraq and attack overland through Iraq into Persia ensuring a land route to the powerful commonwealth land forces.
Option 5+: ???

Which is best? Where to assemble the troops given a particular option? How many divisons are needed? etc etc etc

This is a horse of a different colour to programing a chess game AI.

[Bobb4]

Chess is a very straightforward game to program as there exists large bodies of manuals on what counter to make, there is no terrain, and a strickly limited board. Where a computer beats a human is that it has the ability to evaluate moves in huge numbers of steps ahead. I used to play chess daily and even after years of doing so seeing the board more than 3-4 steps ahead took a "good" day. Also playing against a computer is very "dry" as it takes out the human element.

You example is a bad one because yes the AI goofed as no a human would not have missed that as the strength on your flanks is a particular concern for anyone with a lick of sense. The proper evaluation of enemy strength and intention is paramount of planning. I've called a halt to offensive operations for exactly this reason. I could go forward more but I would be exposing my flanks and leaving myself dangling in the wind which is not something that makes sense to do (normally, as always, particular circumstances could change that).

The task faced by the AI programers is brutally difficult as there are, unlike in chess, very few hard and fast rules. But the largest problem I see with the AI is it does not make long term plans, and it tends to react too much sending units scurrying back and forth endlessly. Watching a unit you have lent to the AI sometime its behavior often comes close to nonsenical. I have seen air units shipped all over the world and rarely even begin to understand why they are there and when they are there its rare they actually seem to do something.

How to program in a long term planning loop is hard to say. I would hate to try and do it (and yes I do program). Just putting down why and how I plan an assault I see a large number of IF-THEN-ELSE conditions showing up. Then once the attack begins and the situation becomes fluid this is when you need quick reactions to developing conditions. Plus the proper utilization of reserves and the reformation of a reserve, stop lines etc. All of this is a complex interaction of short term goals within an overall framwork of a long term objective. But first and formost you need that long term objective as "being only reactive" means basically you have lost in some situations. Reactiing to opportunities is different then always "responding" to the enemy...I hope this makes sense.

As I have a goal with specific objectives the AI is hard pressed to counter this in general. The same it true in Chess, if I have a plan and you don't you may complicate my life but its unlikely in the extreme that you are going to defeat me. As the AI lacks a plan its only chance at sucess largely lies in overwhelming numbers and when that fails it has nothing to fall back on. This isn't to say that even the current AI can't give you a "hard time" I can think of several examples from my last game but for the most part the lack of an overall operational plan is crippling.

As a concrete example from my last game "Overlord" in fall of 44 had the following objectives: Land at Caen, drive through Paris to the Vichy Border as a primary attack goal to isolate the western battlefield from supply, mop up the divisions trapped. Stage 2 establish a secure boundry using the minimum number of required provinces (somewhat reactive). Changes to the original plan were the re-use of the invasion Corps to secure a province by naval invasion after Caen (coast hopping) and that they were used to secure the coastal flank, plus occupation of Djion due to response to enemy moves. And lets not talk about all the naval movements required to assemble the invasion force, use of naval gunfire, plus airpower used to both prepare the battlespace and during the combat itself. My mind boggles at trying to write a conditional sequence that could replicate that.

You can only program something like this if you have first a routine that establishes theatre level objectives plus evaluates them for plausability since piling up troops for an invasion makes no sense when you have no transports for example.

Another concrete example. Japan controls India China and Persia. Where/how to proceed?
Option 1: continue drive to home islands, by island hoping and closing in until bases which allow strat bombing to commence.
Option 2: land in occupied India and cut the japanese forces in half making use of the better terrain to exploit an armoured/mechanized force advantage.
Option 3: land in annexed Persia to allow for a US controlled province to ensure supply without requiring British involvement.
Option 4: declare war on Iraq and attack overland through Iraq into Persia ensuring a land route to the powerful commonwealth land forces.
Option 5+: ???

Which is best? Where to assemble the troops given a particular option? How many divisons are needed? etc etc etc

This is a horse of a different colour to programing a chess game AI.

Actually if you read my thread properly you will realise I was saying exactly what you are saying. Just using a lot less words.
To sum up what I think. "It is easy to programme a computer to play chess, very hard to programme a computer to beat a human in a war game!"
Using your examples further reinforces this arguement

[Alexander Seil]

I think I mentioned it in this thread before- with the command chain, you could presumably break down all AI planning (the military party of it, at least), into trivial "tasks" carried out each level of command, in turn, starting with the Theater AI and moving down, relying on crude min/max problems to pick the optimal offensive route. The advantage of this is that the programmer does not have to concern himself with how the AI "wages the war," but with how a Corps chooses which province out of 2-3 to attack. It effectively scales the problem down.

[Huncowboy]

Amazing. I wish you guys make a ton of money with this because you deserve it. It is so rare to see a developer who is going for the core fan base. Keep it up! My most played game ever was HOI2. This will be next in line.

[Alex_brunius]

I think I mentioned it in this thread before- with the command chain, you could presumably break down all AI planning (the military party of it, at least), into trivial "tasks" carried out each level of command, in turn, starting with the Theater AI and moving down, relying on crude min/max problems to pick the optimal offensive route. The advantage of this is that the programmer does not have to concern himself with how the AI "wages the war," but with how a Corps chooses which province out of 2-3 to attack. It effectively scales the problem down.

Exactlly, Corpse AI takes care of 2-3 provinces to attack, Army Group AI cares about 2-3Regions, and HQ AI cares about 2-3fronts. Brilliant thinking.

I really hope we will se individual personality traits of leaders affecting AI behaviour (offensive, defensive, logistics wizard, exc). That would be really really awesome. Make our quad core processors work!

[Paul_M]

Actually if you read my thread properly you will realise I was saying exactly what you are saying. Just using a lot less words.
To sum up what I think. "It is easy to programme a computer to play chess, very hard to programme a computer to beat a human in a war game!"
Using your examples further reinforces this arguement

Well I didn't disagree with you in general so I'm not sure about the reading it properly bit but I did disagree with you over if the AI had made a mistake or not. Yes it was wordy but its hard to give concrete examples without using some words, and without concrete examples it's hard to make the point to someone who may not be a programer (not directed at you specifically in this case).

With regards to the other comments...

I'm not sure if the breaking it down will work or not. It depends on if there is a plan behind what the AI is doing that lasts for any length of time. Part of the problem is that "strategy" is confused in the minds of most people with mangement and in many "strategy" games with micromangement. Strategic thinking is something you do while the game is booting or in the morning over cearel its not something you change every 5 minutes without a disasterous consiquence. Operational planning is what happens after strategy decides the type of war and management/efficiency produces the means to fight it and long before tactics. The scrips likely pre-define strategy and production (or production defines strategy whichever), and I've seen litte evidence in the past of Operational planning. It's not like if you have a weak point in the line the computer won't find it and attack it but it attacks it because it is the weak point not as comprehensive part of a plan. This allows people to exploit this by leaving a weak point knowing the computer will attack it. It's possible the AI has some sort of objective but it is hard to see what it might be.

I've seen the AI launch attacks, even annoying attacks, which were rather disruptive of my plans but they lacked that sense of being overall a part of a plan. Of course as the attacks failed (in the current game not as a general rule) its hard to see what the AI would have done had they succeeded (again I am speaking about specific attacks in my current game not in general) so again it hard to evalute it.

I have noticed that the AI/AI script has improved every time either a new game comes out or a patch. I anticipate the AI with HOI3 will be better than the current one and will improve as well.

I guess in the end what I don't see is a focused attack to gain specific objectives but instead a scattered effort which is largely unfocused and lacks an evaluation feedback loop. I'm not sure I'm making any sense though. But I have watched the AI fighting the AI and it is just bizzarre. Units move around for no obvious purpose...or if they had a purpose it was contermanded the next day. Units are tossed into combat willy nilly with no effort made to exploit capabilities or achieve anything. Naval invasions that are suicide missons and make no effort to achieve anything. Not covering invasion beaches. Not maintaining a reserve, though most human game players also fail this. The British not ensuring naval supremecy in the Mediteraenian. Unescorted transports.

I could summerise by saying I am often left with the question "Why did the AI do that?" and the usual response is "Damned if I know but it sure didn't make a whole lot of sense." I am probably not being completely fair as the AI does manage to do some things...Japan overrunning China and India for example. Where the AI script falls appart is when the means are marginal is likely a fair assessment, but its also true that when you have overwhelming superiourity you can have a bad plan and still win, probably a fair assesment is a human is more like to snatch defeat from the jaws of victory then the AI script.

But I don't know how much of this is inherent to the script, and how much is a limitation of the script and how much is just bad programming logic. I do know that a good script to do this sort of thing is very hard to produce.

[Alexander Seil]

As I said, the reason the AI does that in HoI2 is because it is awfully fuzzy and does not really have any way to reasonably formulate goals - it's just a "Front AI" that behaves like an amoeba, trying to seep through the cracks in your line. But it's fundamentally impossible to "fix" with any amount of scripting, unless they adopt a new paradigm for the AI. I suspect that what I suggested will end up being reasonably close to what Paradox intends - otherwise I don't see how the delegation-based command system can possibly work.

[Paul_M]

As I said, the reason the AI does that in HoI2 is because it is awfully fuzzy and does not really have any way to reasonably formulate goals - it's just a "Front AI" that behaves like an amoeba, trying to seep through the cracks in your line. But it's fundamentally impossible to "fix" with any amount of scripting, unless they adopt a new paradigm for the AI. I suspect that what I suggested will end up being reasonably close to what Paradox intends - otherwise I don't see how the delegation-based command system can possibly work.

I won't argue with you as I don't know myself! But without a way for the overall top level of the command chain to form coherent goals then I don't see what the lower levels will do that is fundamentally any different then the amoeba and potentially worse. Attacking this province may be sensible to the lower echelon leaders and utterly stupid to the overall "big picture." Does what I just wrote make sense?

I guess in programming lingo there needs to be a top down structure not a bottom up one or worse a top disconnected to the bottom structure.

[Nats]

I am most interested in the progression of the graphics - they dont look too bad, well, theyre getting better. Still very mundane graphics though on the whole. Could do with some ... pizzaz! At the moment I almost fall asleep looking at them. I really dont like the single unit soldier graphics. Just doesnt say 'army' to me. what was wrong with the HOI2 soldier graphics? I know it would be more work to make three soldiers suit each faction but it would look so much better. Same for the tanks - why only one tank. Why not show three to depict a whole division? I think considering there are so many troops in one area you really need an icon that will depict that such as loads of tanks with smoke coming off their tracks, or a fleet of ships or a whole army of troops. A sigle sprite just doesnt seem right. I havent liked this in any of the other games you have done recently either.

As for the map - well its pretty dull. Earth from close orbit (which this seems to depict) is far more exciting than this depicts - roads, railways, cities, bridges, ports, etc. I would really love to see this kind of immersion on the maps so that they depict the man made world not just a load of barren scenery which just doesnt look the part.

And what happened to snow on the mountains, rivers, lakes, fog, mist, mud, rain, snow effects. You could protray so much with the map graphics that would really inform the game rather than relying on icon after icon, overlay after overlay. All this fiddling around really spoils the gameplay. I hardly ever remember look at the weather or whether its night or day etc before I attack an army. Its just so unnatural. IMO you really need to make the game more intuitive for those of us (and Im sure there are many like me) who hate going through stacks of overlays and turning icons on and off to fight each battle. I want to see immediately on the map where my new battles are by small explosions and smoke. I want to see bombers dropping bombs on my cities and the cities smoking and ruined. I want to see ack ack and searchlights.

Am I expecting too much - I dont think so. Computers are well capable of all this these days. The world is waiting for a great wargame like this. You have produced loads of games along the lines you are showing at the start of this thread - please show you are willing to break off the shackles and be an imaginative developer. Not yet another who finds a formulae and just sticks with it for each and every new game. Lets get back to the hayday of PC games where new games actually broke new ground rather than just being yet another clone. PC gaming - the clone wars thats what the noughties have become!

[Alexander Seil]

I won't argue with you as I don't know myself! But without a way for the overall top level of the command chain to form coherent goals then I don't see what the lower levels will do that is fundamentally any different then the amoeba and potentially worse. Attacking this province may be sensible to the lower echelon leaders and utterly stupid to the overall "big picture." Does what I just wrote make sense?

I guess in programming lingo there needs to be a top down structure not a bottom up one or worse a top disconnected to the bottom structure.

You can't code creativity into an AI. Not yet, anyway, and not into a game AI. But the idea is that the Theater level AI only ever formulates very simple goals - such as, "broad offensive against country X," and then the rest is handled by lower echelons. You can make it more complicated by having it prioritize certain strategies already at the Theater level (say, emphasize taking province Z instead of Y, and have Army Group commands take note of that), but the point is to get it down to a problem of a Corps choosing to attack 1 province out of 3. If you don't start the real decision-making at that level, anything on top of that will collapse like a house of cards

[el alamein]

I agree with most of the threads to this topic. Most of the successful wargames I have played have some sort of objective setting, whereby the player (or computer) sets and objective for a theatre and then the computer organises the units to achieve that objective with the land/air and sea units at its disposal. An early example of this is the SSG wargame Battlefront. If there is an option to lend a players units to the AI then this would be the best possible way to control them, by setting a city/port/airbase as an objective to capture, and the AI uses the units under its command (Army Group/Army/Corps etc.) to achieve this. Another example is the War in the Pacific game - you can delegate commands ie. SW Pacific Command or the China Theatre to the computer to handle for you so you can concentrate on a specific theatre(s). I would assume that cities/provinces with victory points would be the more obvious objectives for the AI to capture.

[Alexander Seil]

Well, War in the Pacific is a pretty bad example, because the AI is widely regarded to be broken. I'm with you on the rest, though.

[datupuki]

I don't like the fact we can't use counter in 3d mode

i don't like the almost hex like boarders

[Paul_M]

You can't code creativity into an AI. Not yet, anyway, and not into a game AI. But the idea is that the Theater level AI only ever formulates very simple goals - such as, "broad offensive against country X," and then the rest is handled by lower echelons. You can make it more complicated by having it prioritize certain strategies already at the Theater level (say, emphasize taking province Z instead of Y, and have Army Group commands take note of that), but the point is to get it down to a problem of a Corps choosing to attack 1 province out of 3. If you don't start the real decision-making at that level, anything on top of that will collapse like a house of cards

The problem is that if say I have at whatever level the decision made to surround and isolate the units in Province A the last thing I want to have happen is before the overall forces are assembled the I Corps commander to attack the weak link province I am planning on ending the encirclement on and starting a chain of reactions by the enemy which likely blows the plan out of the water.

Information needs to pass in both directions up and down the link. However, speaking only for myself I don't start the planning process by worrying about which province I Corps will attack but by setting the overall objective. Only once I have set the objective "Isolate and destroy all German Army Units west of the line running north of province eksupsolon" do I start worrying about what I Corps, I Armoured Corps, etc will do to implement the plan. And only once I have decided that I Corps will attack in that direction and I Armoured Corps will attack in that other direction do I worry about exactly which provinces will I Corps attack in what order. Finally at the end I worry about which division will do what.

I may modify start dates, change the orders and so forth of the various divisions depending on organization, type, terrain, leadership, enemy forces etc in a "refinement phase" of the planning process.

Again this is without considering airpower, naval lift, or naval support since all three add a lot of complication.

For me the process is fully top down. You are correct that for the plan to accually work what province(s) is(are) attacked by what unit(s) is the make or break decision for the plan. But without an objective to the action it is just attacking randomly. It is not much different then playing a poor chess player who basically moves pieces at random. That can, infact, be maddening but is unlikely to result in consistant victories.

But I agree creativity won't exist as in reality I am not sure that the term AI is fully applicable. AO (Artificial Opponent) is likely far more accurate as I am dubious about "intellegence" in the extreme.

I guess we will have to see what is done and how it works in reality.