Home Trending Latest threads New posts Developer diaries

AI

Jump to latest Follow Reply
  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
After extensive testing of the AI I confirm they work - as in the game doesnt crash.

There is no way we can get the AI countries to focus on taking a particular province. The keypoint settings only tell the AI where to GARRISON, and they'd have to actually control those provinces for those lines to be of any use. Hence, setting Brussels to 1000 doesnt do any good unless Germany has already taken Brussels.

The biggest impediment to a historical Western Front forming is Belgium. Simply put they are too good, as the Germans have to cross the Maas to get to Brussels. I've dumbed down Belgium by stripping them of a few infantry and artillery techs to no avail. This suggests there is only one solution left - do what Paradox did with HOI. Everyone knows that France in 1940 wasn't that pathetic but to simulate the easy time the Germans had, France was arbitrarily dumbed down. I suggest we do the same to Belgium in terms of tech and OOB - drastically. This will improve gameplay and have minimal detrimental effect as I doubt many people will be dying to play Belgium anyway.

The fortress infantry units in the OOB should be eliminated. Infantry divisions 4, 5, 6 have been removed and 1, 2, 3 converted into militia.
The result: NO AVAIL!! The French army charges into Luxembourg and takes Koln. The Western Front stabilises at the Rhine, with Germans lacking in strength to retake Koln and French lacking the numbers to breakthrough.

Somehow we will have to make sure this deadlock happens at the Marne, not the Rhine.
 
The trouble with dumbing down countries is that while it makes life easier for the AI, it can make it too easy for a player.

Incidentally, a lot of board wargames of WW1 also seem to have trouble making the first couple of months of the war work out right.

Which does lend more weight to the suggestion of creating a November 1914 scenario, after the Germans have occupied Belgium, the front lines have been stabilised in the West, the Russians driven back out of Prussia in the East, and Turkey has just entered the war. The fact that most major countries are already at war should also make decision-making easier for the AI, and we won't see problems like Britain dissolving its garrisons in Egypt...

Of course, I'd suggest we get the game-critical parts of the mod (like the tech tree) running properly first before we make a start on this.
 
I think the fundamental conundrum is that if Germany is strong enough to break through to the Marne, it is strong enough to win the war outright in 1914.

If they only get a stalemate, then it means they weren't strong enough to break through Belgium........

So the historical situation was a bit of a fluke really - the Germans shouldnt have got that far to start with, and once they did, shouldnt have thrown it away.

UPDATE: I seem to have found a possible solution. Usually the front stabilises at Arlon-Metz, which is an improvement over the Battle of the Rhine situations we've been getting up till now. On one occasion I've even seen Germany occupy all of Belgium. What I did is no doubt very controversial - I gave Metz to Germany.

Justification: France typically charges into Luxembourg, and from there threatens Westphalia. Giving Metz to Germany will prevent this from happening. In terms of gameplay this gives Germany the ahistorical advantage of attacking Belgium from the south. I will argue that Belgium didnt put up much of a fight, and they couldnt have, so this is not a gamebreaking issue.

Allenby will argue that this makes the Schlieffen Plan unnecessary. I've given the French a level 5 fort in Chaumont - which is the only extra French border province as a result of my change . The Schlieffen Plan is however still a good idea. Metz can reach Chaumont only by crossing a river, and with the level 5 fort, a direct assault on France is still unfeasible. If we implement this plan we will accordingly compensate France for the lost ICs and resources and deduct Strasbourg and Colmar to make up for the gains in Metz.

UPDATE AGAIN: Germany seems to break Nancy too easily. I'm upgrading all French border forts to level 5.
 
Last edited:
As far as I'm concerned taking such a vast step as to make a major territorial change like that would be too much. Solving this matter by changing the political map of Europe to a situation which as ahistoric is not the way in which this problem should be resolved. :)
 
I concur with Allenby, the ends do not justify the means in this case.

What about adjusting the garrision AI for Belgium? Where only minimal troops are in Brussels and Arlon, and have the bulk of their forces in Antwerp. It should simulate the situation in 1914, where the Germans rolled in and demolished the Belgian forts, and left Antwerp with the Belgian army in it as the German's turned to continue the Schlieffen Plan's right arm.
 
StephenT said:
Which does lend more weight to the suggestion of creating a November 1914 scenario, after the Germans have occupied Belgium, the front lines have been stabilised in the West, the Russians driven back out of Prussia in the East, and Turkey has just entered the war. The fact that most major countries are already at war should also make decision-making easier for the AI, and we won't see problems like Britain dissolving its garrisons in Egypt...

Of course, I'd suggest we get the game-critical parts of the mod (like the tech tree) running properly first before we make a start on this.
Click to expand...

Might I suggest that such a scenario is moved to late February/early March 1915, after the Battle of Champagne and before the Battle of Neuve Chapelle? I think a scenario in November 1914 might make the August one less relevant, whereas if it is moved to the next year, then the same principle that you are essentially looking for - a stabilised situation in the west, the Russians on the run in the East and Turkey in the war can be maintained. :)
 
An August scenario would run into exactly the same problems we've been discussing; the German AI will never get past Belgium - unless we weaken Belgium or strengthen Germany too much, in which case the French will never hold Paris...

I suggested November '14 because it's the earliest possible moment at which the war settled into the familiar pattern it would keep for the next 4 years, so the AI doesn't have to do anything complicated (like launch an invasion, or move an expeditionary force to Flanders.) Thus giving you more time to fight the war.

In other words, my suggestions would be:

The Great War: January 1914- (completely open; make your own alternative history)

Trench Warfare: November 1914- (fight the historical war, more or less)

The Final Offensives: Late November 1917- (The USA is fighting to make the world safe for democracy, Soviet Russia has proclaimed world revolution, Germany is training its stormtrooper divisions for a final crack at victory, and in the pre-dawn hush near Cambrai a strange rumbling and squealing noise breaks the silence... Oh, and General Allenby is just outside Jerusalem.)
 
Stephen's got the right of it. November makes a good time to start a historical western front style.

Although a 1915/16 scneario where all the minors and other powers start to pick sides would also be nice, for a starting historical setup but possible variations with who sides with who, and who stays neutral.
 
I think I've solved the problem of Belgium - or at least mitigated it.

I've stripped them of some techs:
Bolt Action Rifle 1101 (soft att +1)
Field Fortifications 1102 (ground def +1)
Quick Firing Artillery 14201 (hard att +1)
Infantry Gun 30mm+ 14202 (ground def +1)
Field Artillery Gun 14203 (soft att +1)
Pack Artillery Gun 14205 (mountain, cavalry soft att +1)
Strong Point Warfare 11103 (ground defence +1)

And removed infantry units 4, 5, 6 from the OOB.

However this STILL results in Battle of the Rhine.

But I came up with a masterstroke - the main problem was crossing the river into Brussels and the level 2 fort. I changed Brussels and Antwerp to level 1 forts (after all, a few Krupp guns did the trick). Then, altered the ai to set the river parameter in the front section to 4.5 . Any river attack is hardcoded to be about a quarter or a fifth of the effectivity of a normal attack. Setting this to 4.5 effectively removes the detrimental effects of this. It took a while, but Belgium was finally totally occupied by German forces in May 1915.

I will apply this change to all AIs. Players will not be affected because these changes only affect AI.
 
The Great War: January 1914- (completely open; make your own alternative history)

Trench Warfare: November 1914- (fight the historical war, more or less)

The Final Offensives: Late November 1917- (The USA is fighting to make the world safe for democracy, Soviet Russia has proclaimed world revolution, Germany is training its stormtrooper divisions for a final crack at victory, and in the pre-dawn hush near Cambrai a strange rumbling and squealing noise breaks the silence... Oh, and General Allenby is just outside Jerusalem.)[/WThe Great War: January 1914- (completely open; make your own alternative history)

Trench Warfare: November 1914- (fight the historical war, more or less)

The Final Offensives: Late November 1917- (The USA is fighting to make the world safe for democracy, Soviet Russia has proclaimed world revolution, Germany is training its stormtrooper divisions for a final crack at victory, and in the pre-dawn hush near Cambrai a strange rumbling and squealing noise breaks the silence... Oh, and General Allenby is just outside Jerusalem.)
Click to expand...

We'd have February 1916 (Verdun, Somme, Jutland year), November 1917 (USSR) and March 1918 (just after Brest Litovsk).

Might want to add a "what if" at the end - if Ludendorff wasnt sacked in late 1918 and they fought on. Needless to say it will be practically impossible for Germany to win.......
 
StephenT said:
An August scenario would run into exactly the same problems we've been discussing; the German AI will never get past Belgium - unless we weaken Belgium or strengthen Germany too much, in which case the French will never hold Paris...

I suggested November '14 because it's the earliest possible moment at which the war settled into the familiar pattern it would keep for the next 4 years, so the AI doesn't have to do anything complicated (like launch an invasion, or move an expeditionary force to Flanders.) Thus giving you more time to fight the war.

In other words, my suggestions would be:

The Great War: January 1914- (completely open; make your own alternative history)

Trench Warfare: November 1914- (fight the historical war, more or less)

The Final Offensives: Late November 1917- (The USA is fighting to make the world safe for democracy, Soviet Russia has proclaimed world revolution, Germany is training its stormtrooper divisions for a final crack at victory, and in the pre-dawn hush near Cambrai a strange rumbling and squealing noise breaks the silence... Oh, and General Allenby is just outside Jerusalem.)
Click to expand...

Forgive me good sir, I have misled you. The August 1914 scenario would infact start on 22nd August 1914 with the war well underway: the battles of the Frontiers would have been and gone, Germany would have virtually all of Belgium, and the Battle of Tannenberg would be in progress. Therefore, the German player is still given the chance to have lunch in Paris and dinner in St. Petersburg.

If you want to go ahead with having the 'Trench War' scenario begin in November, then do so by all means, but I want to make it clear that there shall be an August 1914 scenario like the one I've just briefly outlined above. :)
 
Take a look at this!

SuperBritain.JPG


I've got the British AI to become gung ho outside the British isles! However they did NOTHING in France from 1914-1918, and then when France became a German puppet they started shipping men across the Channel en masse!!

Hmm.......this might suggest an absurd solution.......
 
Take a train on the new British Imperial Railroad, tickets for a non-stop train from Constantinople to Capetown now available. :)

Yep that is one large invasion force from Britain...perhaps ptan you should take a moment and load that AI into a regular game of HOI and see how it works there.
 
Good job with the AI, Patch. This is your department now, especially as you're the only one who seems to have half an idea as to how AI works :wacko: :)

Seeing as I took the tech tree from you, it seems only fair - I'll keep that for now, and if/when Johan comes back, I'll hand it to him as he was reluctant to let it out of his sight last time.
 
Hi! I tried the TGW mod for the first time today 1.03b. I took Germany at normal diff for a spin and European France was all occupied in less than 2 months and I had an offensive going against Russia at the same time. I guess the French AI should be a bit better at defending itself. Probably is a lot more even in MP though.

Other than that looks like a tremendous mod with lots and lots of work behind it. I'm really impressed!
 
Last edited:
You must log in or register to reply here.
Top Bottom
Back