RE: How is the AI? (Full Version)

All Forums >> [New Releases from Matrix Games] >> Strategic Command Series >> Strategic Command WWII War in Europe



Message

Ironclad -> RE: How is the AI? (11/22/2016 6:54:47 PM)

Splendid work by the AI - it has just caught me in early 43 with a deadly and rapid D Day taking full advantage of my light defences in the west as the Reich was committed to a yet another major (and over ambitious) offensive in the east. A fun game - its a long time ago since I can recall such an enjoyable non-multiplayer experience.



xwormwood -> RE: How is the AI? (11/22/2016 7:05:50 PM)

I've been through the same experience, and share the feeling (the pain for losing the west, and the joy about the AI).
:)



GBoggon -> RE: How is the AI? (11/22/2016 7:11:14 PM)

The AI Is not bad, but when I find a AI that out plays me at this sort of game with out cheating I will ask for my life support to be switched off[;)]. It is good for learning the game though. I look forward to playing a human as I think the game balance is better than in the old game.



xwormwood -> RE: How is the AI? (11/22/2016 7:21:19 PM)

Famous last words...
[:D][;)]



n0kn0k -> RE: How is the AI? (11/23/2016 4:01:54 AM)


quote:

ORIGINAL: Gordon Boggon

The AI Is not bad, but when I find a AI that out plays me at this sort of game with out cheating I will ask for my life support to be switched off[;)]. It is good for learning the game though. I look forward to playing a human as I think the game balance is better than in the old game.


It's not cheating afaik. Huberts A.I. blog is a good read if you're up for it. It's in the dev blog section.



Petiloup -> RE: How is the AI? (11/23/2016 11:10:37 AM)

The AI is just terrible on a strategic level but does a good job on a tactical one.

Just finished my first game with an Axis major victory not even sweating a little.

Just did the following:

- Poland [evidently]
- Accept to take Norway and Denmark.
- Crush Holland [just move against the Capital and leave the reast as it will surrender]
- Same for Belgium.
- Take the whole of France and move German Troops to take Algiers then you get it all.

In the meantime invest 5 Diplomacy ****s to Turkey with the German and 3 to Spain with Italy.

Once Turkey is Axis then switch to Spain with the German as well.

After the Fall of France you get Spain easily then you'll get Hungary, Romania and Bulgaria [assuming you gave to Stalin what he wanted]

Then move all your Diplomacy points to keep Russia out of the war as long as possible.

- From Spain take Gibraltar.
- From Turkey take Palestine, Jordania, and help Iraq then crush Egypt between the Afrika Korps and Palestine troops [not forgetting to get Malta in the meantime].
- Once done move troops from Egypt close to Persia and make sure the whole Turkish army is on the border with Russia.

During that time invest heavily in Infantry, Armor, Planes, Long Range Planes, Production research. Leave out Rockets, Hvy Bombers, Naval, Subs, Anti-Subs.

That should give you the edge once Russia enters the war around 1942 as they will have more troops but not as good as the German, keep upgrading troops as much as you can [Infantry, Planes, Tanks]

Once Russia enters the wars move Diplomacy points against the US [it's expensive but worth it] to delay the US entry as long as possible.

Dont build any Naval troops to get enough troops to start moving into Russia from the West and from Middle East [Get back Persia as well].

In 1942 the US should go to war as well and will first invade Marocco, just move 2-3 tanks and 2-3 armies to push them back to the sea. Use Uboats and the Italian Navy and don't care if you loose the Italian Navy in the meantime.

Start when the US enters the war to search Advance UBoats so they are more efficient sinking transports and other ships [don't attack destroyers].

Upgrade the whole Spanish Army and build everything you can for them, leave them in Spain just as a reserve.

The AI surprised me invading France in 1943 while I was busy in Russia but I stopped them with the Spanish army long enough to bring more troops. [I did build a lot of Garrison in France as well to slow down any invasion, especially on the ports].

Once the France invasion is containing then start the push back with tanks and have some fighters in cover. They will get hit heavily each turn but your production is almost 1,500 points at that time so you can afford it.

In 1944 France will be cleared of Allies troop easily.

In Russia just destroy troops at every opportunity with Tanks and planes. I upgrade my Tanks constantly and the plane as well. In 1943 all the German army was in Mobility 2 so you can move quite well in the vastness of Russia.

Winter just upgrade and repair troops, keep your HQ close enough and build a few as needed. Those are a must or your troops starts to suffer penalties and can't be reinforced/upgraded.

Turkey front is hard to move but with 3-4 Italian Tactical and Medium bombers you can move [even slowly] enough to capture oil wells.

Upgrade German and Axis Allies troops as they benefit from the Germany research so the Rumanian/Hungarian/... are as good as German troops almost.

Strategy in Russia is to move against Leningrad to take it first, move against Smolensk and in the direction of Moscow. In the South just destroy troops when possible and move slowly towards Kiev/Odessa.

Besides 1 tank in the South all the tanks are against Leningrad/Moscow. All the Luftwaffe in the North as well. With 2-3 planes you can soften any Russian unit in a city and take it right away then keep pushing forward. Encirle the Russian troops at every opportunity.

It's a slow process as there is a lot of Russian but they take a lot of losses against Tanks so you can push forward. By keeping the Luftwaffe together you can clear any hex even heavily defended.

Just keep fighters all over the front line to protect your own troops.

Once Leningrad is taken, join with the Finns and move to Moscow.

The South is getting easier as Russian troops are moving North to protect Moscow.

By 1944 the whole front was getting much easier, Tanks upgraded to 5 were plowing through Russian troops like a knife in melted butter.

By 1945 most of Russia should be conquered and Moscow fallen. The rest of Europe free of Allies.

VICTORY.

The AI is really unimaginative strategically, the attacks are following the real WW2 with invading North Africa then Normandy.

Now tactically it does a great job I must say, great use of air power to soften troops then destroy them with infantry, and then use Tanks to create a hole and exploit. So whenever possible you need a second line of troops to prevent a breaktrough.

It never tries to attack somewhere else like Spain or take Portugal as an Ally or invade Denmark, anything that I wouldn't be able to prevent.

I had to keep no troops but a few Garrison here and there just in case.

At the end I was destroying US troop transports with UBoats and surface ships like it would be Turkey shooting.

All the AI was doing is keep sending more and bombing my production in France/Germany.

First game and I won without even trying anything fancy.

So if you are a casual wargamer then it's a great game, looks amazingly nice, the AI is doing well for that level.

If you are an experienced wargamer like me then give it a pass. Not worth the money.

Might be fun trying against a human.



n0kn0k -> RE: How is the AI? (11/23/2016 12:15:23 PM)

quote:

ORIGINAL: Polonthi
The AI is really unimaginative strategically, the attacks are following the real WW2 with invading North Africa then Normandy.


This is based on one game. That doens't mean it will do exactly the same the next game.
It will also invade the UK for example after you've taken it in Sealion.
Many more variations can be added over time.

I suggest to take a look at the following scripts for all the variations:

Strategic Command WWII - War in Europe\Campaigns\_1939 Storm over Europe\Scripts\AI\amphibious.txt
Strategic Command WWII - War in Europe\Campaigns\_1939 Storm over Europe\Scripts\AI\amphibious_minor.txt
Strategic Command WWII - War in Europe\Campaigns\_1939 Storm over Europe\Scripts\AI\offensive.txt

Interesting game you've played though. Keeping the major parties out of the game.
I guess diplomacy needs some more balancing if you can string all the major parties along like that.
Perhaps it should not be possible to use diplomacy on major parties, or make it even more expensive as it is now.
May I ask on what level you were playing? From your report I can make out you've spend a lot.




Hartmann -> RE: How is the AI? (11/23/2016 1:08:11 PM)

I find Polonthi's recommendation

"So if you are a casual wargamer then it's a great game, looks amazingly nice, the AI is doing well for that level. If you are an experienced wargamer like me then give it a pass. Not worth the money."

arrogant. The only recommendation I could make sense of would be that very experienced wargamers should pass ANY wargame except for PBEM or multiplayer. If, on the other hand, I compare the competence of SC WW2's AI with the AI of more or less all other grand strategy wargames, then it fares VERY well. If I would have to mention a recent game where the AI destroys all fun in single player mode, I'd take a long and hard look at HOI4 instead.

Well, that said, the AI certainly could be improved still. The dev blog explains that there's a strategical, tactical and "fuzzy" AI. The stratgical AI, which is at issue here, is not hardcoded anymore, but rather constituted by strategy scripts which fire under certain conditions. It seems from Polonthi's game that the AI hasn't a script yet for dealing with

1) an Axis player trying to secure the whole of the Mediterranean (by means of strong diplo focus too) while keeping Russia out of the war as long as possible
2) an Axis player who actually already has achieved that goal

But as the AI isn't hardcoded, it should be possible to provide scripts for these scenarios, where e.g. in 1) the AI "sees" that it has to do more to prevent the diplomatic fall of the Iberian peninsula, and in 2)that it now should consider landing in Portugal/Spain instead of France. In any case, it should not (oblivious to von Moltke's rule) stick to the plan it once chose regardless of what the enemy does or did achieve.





FelixPraedo -> RE: How is the AI? (11/23/2016 2:20:01 PM)

quote:

I find Polonthi's recommendation

"So if you are a casual wargamer then it's a great game, looks amazingly nice, the AI is doing well for that level. If you are an experienced wargamer like me then give it a pass. Not worth the money."

arrogant.


Polonthi: your victory earns you bragging rights I'd say.
Looking forward to a PBEM game against you as soon as multiplayer is ready!

quote:

Well, that said, the AI certainly could be improved still. The dev blog explains that there's a strategical, tactical and "fuzzy" AI. The stratgical AI, which is at issue here, is not hardcoded anymore, but rather constituted by strategy scripts which fire under certain conditions. It seems from Polonthi's game that the AI hasn't a script yet for dealing with

1) an Axis player trying to secure the whole of the Mediterranean (by means of strong diplo focus too) while keeping Russia out of the war as long as possible
2) an Axis player who actually already has achieved that goal

But as the AI isn't hardcoded, it should be possible to provide scripts for these scenarios, where e.g. in 1) the AI "sees" that it has to do more to prevent the diplomatic fall of the Iberian peninsula, and in 2)that it now should consider landing in Portugal/Spain instead of France. In any case, it should not (oblivious to von Moltke's rule) stick to the plan it once chose regardless of what the enemy does or did achieve.


Hartmann: the strategic AI can certainly be further developed using scripts, one more reason the game is worth the money!
Who knows one day it will be unbeatable....

Cheers!



IainMcNeil -> RE: How is the AI? (11/23/2016 2:30:29 PM)

It sounds like there is more of an issue with Diplomacy than anything else. This seems to have been the decisive impact in the game mentioned. I've been finding it definitely pays to go for Spain and Turkey so it may be we need to raise the cost of those or add a bit more uncertainty to that tactic. Maybe a diplo chit should have a small chance of backfiring too (maybe 1%) - people don't like being pressured! This would reduce the rate at which nations moved towards your alliance and increase the risk of such tactics.

Or maybe we need to make the AI keep more watch on what you are doing and react to it. Once Turkey and Spain are pro-Axis the Allies have an enormous mountain to climb.



Hubert Cater -> RE: How is the AI? (11/23/2016 2:49:45 PM)


quote:

ORIGINAL: fatgreta1066

I'm playing the 1939 Campaign as Allied. It took Germany until November 25 to conquer Poland. It seemed to me that the AI did a good job of cycling units in and out of position when making important attacks (I.e. against Warsaw). It seemed to not do a very good job in figuring out how to bring Poland down. By comparison, I played 5 times as Germany and conquered Poland in 2 turns 4 times, 3 turns once.

Germany was in good position to invade France by late March, 1940. I held on to France until early August. I thought the AI diffused it's initial strikes a bit too much, not creating a breakthrough when it might have. To be fair, I did a pretty gamey thing by placing the entire French navy (including its Med fleet) on raiding duty against the Norway convoys, as well as some RN units. That cost Germany a number of MPP. I also sent an extra British Corps to France, in addition to the BEF.

Now that Germany has French ports, we will see how the AI handles the Battle of the Atlantic, and the war in Africa.


This post caught my eye so as Poland and the invasion of France and the Low Countries were a portion of the AI that I had spent quite a bit of time on fine tuning.

General speaking, any optimizations here for the AI would not only be good for this part of the game but also for every other part of the game when it comes to effective AI combat.

For Poland, the goal on my end was to attempt to recreate the 2 turn victory that most human players can achieve... essentially this was the bar.

The challenge in Poland for the Axis AI was that while there are many targets and many towns between the Axis start lines to Warsaw, how to ensure the Axis AI would focus enough of its assault not only on Polish front line units, but to also focus its drive towards Warsaw as quickly as possible.

This was a tricky one since there are so many targets it is easy for the AI to get distracted but after a lot of watching the AI play and determining what most human players would do, we did at one point have the AI (if the rolls were good and the weather cooperated) capture Warsaw within two turns 50% of the time. Part of this was achievable through more tightly focused scripting but also by adjusting certain behaviour such as having the AI to hold back its Tanks (not always and only under the right conditions) after the initial wave of attacks. This allowed it to breakthrough a bit further with its tanks and position itself closer to Warsaw for the follow up turns.

This one general behavioral change, combined with many many others, seemed to improve AI assaults quite a bit and also played out quite well in other parts of the map such as France and the Low Countries and so on.

Unfortunately we also then found that Poland was too easy to defeat in general so there were changes that led to requiring the Axis to use most of its frontline units to achieve quick victories over Poland, i.e. rather than being able to move half of those units towards France right away, and this then led to the AI also being less likely to capture Poland in 2 turns on a regular basis.

As it is now, the results will vary from game to game, and although it is extremely rare for the AI (I've run countless AI vs AI debug games during development) two turns is still an outside possibility but 3 or 4 turns is more likely. In some cases where the rolls and weather combine to frustrate the AI a longer time frame sometimes occurs as it did in your game here.

That being said, on average by the time of Barbarossa things seem to even out and the fact that the AI was able to capture France by August shows that made up some of that delay in Poland as an August capitulation against a human player in France is still a pretty good showing for the AI... at least from all the tests we've run.

Based on this, my guess is that Barbarossa will likely start near the historical timeline as well for the Axis AI.



Hubert Cater -> RE: How is the AI? (11/23/2016 3:03:00 PM)

quote:

ORIGINAL: fatgreta1066

I've always heard that AI in general is better as a defender than an attacker. I know very little about programming so I can't speak to that other than to say the thought seems to make sense to me.

I'm playing on the medium setting, so to speak. No advantages for the AI in production and such.

And also, I love this game, I certainly don't want any comment to appear critical of it. This is the most fun I've had with a PC game since the first Harpoon way back in the 90's, and High Command as mentioned in another thread.


Thanks again for this feedback and while I can't speak for all games what I've found is that many games tend to give the defensive AI quite a few bonuses, in terms of extra units or otherwise, to help it play a more complete defensive game.

One game in particular that stood out this way was the original Panzer General series from 1994 where if you were ever to play the scenarios in reverse, i.e. as the Allied side you would quickly notice how lopsided some of the scenario were in favor of the Allies and only to ensure that the Axis human player otherwise had a very tough time of it all. The Kursk scenario comes to mind as I have a very distinct memory of that one.

Speaking for the AI in Strategic Command I would actually argue the reverse in that that the AI will probably be more impressive on the OFFENSIVE than on the DEFENSIVE as the defensive portion of the AI is very much limited to even what a human player is able to achieve, and that is just to say try and hold off the attacker for as long as possible by staying in good defensive positions such as behind rivers, inside towns and cities and to remain in as good a supply as possible.

As an aside, and from a development point of view after quite a bit of effort working to improve the tactical combat portion of the AI for this release, hearing that the AI can still manage to capture France on par with the efforts of some players for when they play the Axis, is always nice to hear [:)]





Hubert Cater -> RE: How is the AI? (11/23/2016 3:11:08 PM)

Thanks for the feedback Polonthi and we can certainly add more adjustments to take into account your diplomatic strategy, since as mentioned most of these are fairly straight forward to improve on our end. Taking into account the many different strategies is always a challenge, but not impossible and we always appreciate this type of feedback where we can continue to improve the game.

That being said, and again from a development point of view, seeing the game still last until 1945 tells us that despite any failings, the AI still gave you a pretty good run and that is still good to see from our end. The fact that the AI still managed to surprise you with an early D-Day is also good to hear as that was an area we spent quite a bit of time working on since as Iain also mentioned, most games if not all, tend to fail in this regard.

To answer some of your more specific queries, the AI will indeed vary from game to game and doesn't always necessarily land in North Africa via the Torch landings, and could indeed make landings in Spain and even southern France, although these are of much lower percentage chances for them to occur, but they are possible once Spain joins the Axis. Again each game is different and we are happy to continue to improve the game as the scripting system does allow us to make these types of adjustments quite easily once an area is highlighted and in need of adjustment.

For example, Sealion came up as an issue during Beta (and during the video Beta previews) and I think players will find that much more challenging now in the official release.




BillRunacre -> RE: How is the AI? (11/23/2016 3:44:19 PM)

Thanks for the feedback Polonthi, and as Iain said it might be that some tweaks to diplomacy will help here.



Grotius -> RE: How is the AI? (11/23/2016 3:59:33 PM)

I think the strategic AI is pretty good. Polonthi is not pointing to major flaws in the AI's overall approach, but instead to one small part (diplomacy) that could easily be countered. And even Polonthi didn't win a Decisive Victory. Has anyone won a Decisive Victory as the Axis yet?f

I've been playing wargames as long as anyone here, and the AI is giving me a great game, but I confess I haven't paid much attention to Diplomacy.



FF_1079 -> RE: How is the AI? (11/23/2016 4:25:32 PM)


quote:

ORIGINAL: Iain McNeil

It sounds like there is more of an issue with Diplomacy than anything else. This seems to have been the decisive impact in the game mentioned. I've been finding it definitely pays to go for Spain and Turkey so it may be we need to raise the cost of those or add a bit more uncertainty to that tactic. Maybe a diplo chit should have a small chance of backfiring too (maybe 1%) - people don't like being pressured! This would reduce the rate at which nations moved towards your alliance and increase the risk of such tactics.


This.

Spain and Turkey both had a VERY high resistance to being diplomatically pressured and could easily have taken offense and swung away from Germany. I believe the old CoS diplo worked that the higher a nations favorability of you was, the easier it was to influence them. If they had a low to negative leaning towards you, diplo could easily backfire and have negative outcomes.

I find the AI is a good opponent for me in my casual way of playing the game. The AI will punish any sloppy play I make, and take advantage of any holes it sees. I am very happy with the base game and look forward to playing it with modded scripts after I beat the base game. Good job.



sveint -> RE: How is the AI? (11/23/2016 6:23:53 PM)

It's far too easy to make nations join the Axis (Spain and Turkey).



Capitaine -> RE: How is the AI? (11/23/2016 6:29:34 PM)

You may be ignoring the ability of an Allied player to oppose Axis diplomatic moves. Not sure if it was an event or not, but in one game the Allied AI secured Sweden as an ally early in the war. I'd think a human Allied player could make diplomacy a harder task on the Axis.



Hubert Cater -> RE: How is the AI? (11/23/2016 7:02:01 PM)

Agreed as we'll look into having the Allied AI be a bit more reactive to this type of Diplomacy as even Edwin posted some feedback on this as well. All in all some straight forward adjustments on our end to hopefully improve the game even further.



FF_1079 -> RE: How is the AI? (11/25/2016 1:42:11 AM)

He is an outside idea on a simple fix to the diplo, keep the 50 mpp cost for influencing nations that have a 51% or higher initial/current view of your nation. Increase the FIRST diplomatic overture to a nation whose relationship is 49% or lower favorable to 150 mpp to simulate the initial effort to overcome resistance and then drop back to 50 mpp for each additional point. If you are serious about winning that nation diplomatically, you will have to reduce your research and development or war spending by 100 mpp extra for each unfriendly nation attempted.

This may not slow Germany much - but it will keep an Axis player from emptying Italy's MPP bank as well to get that plus 30% combined chance.



ILCK -> RE: How is the AI? (11/25/2016 2:45:24 AM)

I found the AI countered my diplomatic moves with Spain quite well. They essentially stalled me out on the verge of getting them into the war and moved Spain away from Germany in fact.

As the Allies the Germans launched a VERY early and unexpected Fall Gelb that began in January 1940 and took me rather unprepared.





wodin -> RE: How is the AI? (11/25/2016 5:21:38 AM)

Considering how many people I see post about this games useless AI or that games useless AI I realise I must be absolutely useless at wargames..which is great cos on the whole I've found most AI's a worthy opponent. I've actually never played a wargame and analysed it enough to suss out an AI..



sPzAbt653 -> RE: How is the AI? (11/25/2016 7:49:27 AM)

A comment about the computer's use of Airborne units [I don't think anyone else has commented on this]. Every now and then the computer will launch an Airborne assault behind my lines. This seems to work well in this game, as I can recall in other games the computer is either constantly dropping these units and making a boring nuisance of itself, or the computer doesn't use them as airborne units at all, which is disappointing. SC3 falls somewhere in between - not often enough to become a nuisance, seldom enough to keep us on our toes.

Good job with the Airborne scripts, if such a thing exists [&o]



IainMcNeil -> RE: How is the AI? (11/25/2016 8:54:30 AM)

I think most of the people complaining about the AI are finding a specific loophole that was missed and exploiting it rather than looking at the big picture of how the AI is playing. To some extent a game as complex as this is always going to have loopholes as when you close one you likely open another (just like tax law!) and so you play a cat and mouse game, but for the vast majority of normal players the AI does a more than decent job. It certainly gives me a good game! It lets me play in a realistic way and still achieve goals. Other games that often force you to play to the loop holes on the hard settings to survive and I really don't enjoy that.

Having said that I'm sure there are some bugs we missed or things that can be improved with some more of Hubert's time!



Hartmann -> RE: How is the AI? (11/25/2016 11:36:51 AM)


quote:

ORIGINAL: Iain McNeil

I think most of the people complaining about the AI are finding a specific loophole that was missed and exploiting it rather than looking at the big picture of how the AI is playing.


I think that nails it. If I see how the AI (on both sides) handles Fall Gelb and Barbarossa, and compare this to other AIs of WW2 strategy games within the same category, everyone should see that it is really rather good. OF COURSE it is not as good as an experienced human player(it still sometimes misses opportunities as well as obvious threats), but it "feels" much more like a (casual to mediocre skilled)human than other AIs.

The only time it really shows that it's "just" an AI is situations where it stubbornly sends one single unit after another to its certain doom, like it sometimes does when the African front is stalemated or when the US tries to transport troops after a successful "Sea Lion".



Hubert Cater -> RE: How is the AI? (11/25/2016 1:04:40 PM)


quote:

ORIGINAL: sPzAbt653

A comment about the computer's use of Airborne units [I don't think anyone else has commented on this]. Every now and then the computer will launch an Airborne assault behind my lines. This seems to work well in this game, as I can recall in other games the computer is either constantly dropping these units and making a boring nuisance of itself, or the computer doesn't use them as airborne units at all, which is disappointing. SC3 falls somewhere in between - not often enough to become a nuisance, seldom enough to keep us on our toes.

Good job with the Airborne scripts, if such a thing exists [&o]


Thank you and this part is actually unscripted and just the AI doing its own thing, i.e. assessing the OFFENSIVE combat situation and assigning Paratroops and subsequent targets all on its own. This is what will make it hopefully interestingly dynamic from game to game.

Glad to hear it is working as expected [:)]



Nubis -> RE: How is the AI? (11/26/2016 1:05:40 PM)

The AI is pretty vicious as the Axis and I haven't beaten it yet.. but I will [;)]



The Land -> RE: How is the AI? (11/26/2016 7:26:42 PM)

On the subject of France, I found that once I had concentrated my panzers in south-east Belgium (not looking at the game map as I type but in the Charleroi/Namur area), the AI moved the French army West to protect places like Lille and Amiens. Leaving nothing between my tanks and Paris.

Really enjoying the game but I thought that was foolish!



Hartmann -> RE: How is the AI? (11/26/2016 7:34:33 PM)

That's strange as I always see them trying to maintain an unbroken line from Alsace to Calais. But of course - as they don't have precognition abilities - they don't (and shouldn't) prepare for a mighty panzer strike through the Ardennes either.



The Land -> RE: How is the AI? (11/26/2016 7:37:58 PM)

They did know the tanks were there - they'd had to go through some Belgian troops to get there.



Page: <<   < prev  1 [2] 3   next >   >>

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
1.453125