Maximum AI Competitiveness
 

I play almost entirely random battles vs. the AI. Any pointers on how to make the AI more competitive?

I'm thinking of using the classic "3 multiple hex objective areas" instead of the "shotgun" cluster map configuration. I'm thinking it will keep the AI from automatically spreading his forces out to thin, so he maintains a decent firepower level. Any thoughts? Recommendations?

Re: Maximum AI Competitiveness
 

Give the AI a points advantage, try a 100% one and see if you prefer that. It's on page 2(?) of the preferences in the game launcher programme.

Additionally you can try making it harder for yourself by deliberately handicapping yourself, don't use the latest and greatest MBT, use say leopoard 1 or M48 if West Germans trying to stave off the Soviet steamroller in say the 70s/80s, not the Leo 2s. Dont use any air support, etc.

If you prefer Eastern Bloc, use the East Germans or maybe Romanians (other Warpact forces are available) with a T-55 tank basis in that era and see if that makes things moderately harder for you in installing proper scientific socialism to the Western capitalist running dogs.

Re: Maximum AI Competitiveness
 

Andy definitely gave some good advice in his last like you I've been using/playing against the AI from the very beginning and watching it get better over the years as it gets "tweaked" every so often.

I ensure the AI is on par with my player set-up throughout as in some cases it broadens out the options in equipment etc. for it.

Timed objectives were a great addition to the game that the AI can take advantage of as well.

I agree with Andy in also "shaking things up a bit" they've offered so many countries you can play against, and equipment sets you could never play them all.

East Germany as pointed out did you know the East German Army was considered by NATO during and after the Cold War the best trained, disciplined and lead of all the Warsaw Pact countries?
Also, it was one of three Warsaw Pact countries outside of Russia the Soviets operated the T-80 tanks?

The Soviets thought the East Germans to be the most motivated to fight as they felt unification would come from them in a war against the West.

Bottomline though is to maintain parity between you and the AI, to have anything less would create a "false narrative" to the facts especially during the Cold War Era when the Soviets had such a huge advantage in equipment.
https://www.cia.gov/readingroom/docs...20and%20Poland.

You'll find several posts in this thread dealing with the AI starting with the next two pages. There was a long one in here I believe from about three to five years ago that had a lot of good advice in it.

Regards,
Pat
:capt:

Re: Maximum AI Competitiveness
 

Oh - another thing you can do is respin the objectives at game setup, in order to place them where its better for the AI and make your getting to them more difficult. Or manually move them using the edit map functions.

Especially with timed objectives, putting them nearer to the AI side will make it get them and the timed points early. Thus making you have to be more agressive perhaps than you wanted to be and have to take the battle to him rather than sit back and snipe at him?.

Re: Maximum AI Competitiveness
 

Interesting comments and I thank you all. From my experience, the AI plays good, in fact quite good at times, but it's so easy to predict, it's just a machine. Maybe I should look into playing by email instead, then I'll get some real competition, against real people! I've played so many games against the AI, I probably picked up all sorts of bad habits. A human player will probably kill me.

I never played with timed objectives. That sounds interesting.

Re: Maximum AI Competitiveness
 

I have a few questions concerning about what I can do to make the AI more competitive. I have absolutely no experience using timed objectives, I understand there's a few different types, which ones should I use and how should I use them? What can I do to make the AI not so aggressive like not moving in open ground under fire? Or the opposite, like sitting in an exposed position in a tank and thinking it's impervious to attack? These last two points are particularly important because the AI will do those things over and over again, never learning not to do those things. I bet if it wasn't so numbskulled, it would win practically every game, it's so easy to predict what it's going to do in almost every situation.

Re: Maximum AI Competitiveness
 

Why? I don't know.

With "Timed Objectives" all you're really doing is deciding how many turns after you capture an objective/flag etc. you have to hold before you start getting points.

To the end of your post. seems to me what your describing is the AI having units acting as an "overwatch" to engage any enemy units trying to interfere with its units advancing on an objective.

That is something many of us "lowly" humans wished we had not forgotten to do ourselves a "time or two".

Just me being well, me :p this would probably be a good time to recommend to you that the game manual might be useful to clear up some of these questions and I would "blend" that in with a good measure of just playing the game at your local "school of hard knocks" best part there is no tuition involved since admission is free!

Regards,
Pat
:capt:

Re: Maximum AI Competitiveness
 

The main differences I've seen between AI and human players are the following:

1. AI does not use smoke skillfully, and Z fire is unheard of. Because of that it is relatively easy to creep forward to a defensive position held by AI, to disengage from a bad firefight and to make an elastic defense. The main soluition to that is IMHO to pick factions that have little to no TI for yourself and give a lot of TI to AI. This will nullify some of the AI smoke deficiency. Of course, that means you have to pick sides from late 70ies onwards.

2. AI does not use artillery skillfully. No creeping barrages, no blocking barrages etc. Solution: give AI MASSIVE artillery advantage. Quantidy vs. quality.

3. AI is too aggressive in times when it shouldn't. Most players know when the game seems like it would be a draw and pushing too hard might result in defeat and tend to be more cautious then. Ai when it loses a couple of flags does a major counterattack that results in massive losses to it when the player expects it. Solution: tough one, maybe play scenarios where a good portion of AI units are static (ie give longer than scenario reaction time), although a good players should be able to overcome them with suppression and small assaults.

4. AI cannot do the game's version of fire and maneuver ie. have a number of units fire at the enemy, draw OP fire and allow the rest to move closer and blast the enemy at close range. Solution: more points to the AI so it has the numbers that this trick does not apply easily.

5. AI suffers from helicopters (I mean, humans do too). Solution: dont buy helos, period.

Re: Maximum AI Competitiveness
 

Diversify your game alternating human opponents and AI :).
The AI ​​probably has an advantage in fights with many units because you may manage them better. The human can sometimes forget to move something in time.

Re: Maximum AI Competitiveness
 

What's TI?

Re: Maximum AI Competitiveness
 

@Dion,
T.I. Stands for Thermal Imaging. It's an optical technology that allows you to see in adverse weather conditions & see thru smoke.

Re: Maximum AI Competitiveness
 

Thermal Imaging (night vision)


OOPS - Sorry didnt see the previous reply... :doh:

Re: Maximum AI Competitiveness
 

A lot of good suggestions to combine with well documented use of player preferences. Looking forward to playing again.

Re: Maximum AI Competitiveness
 

PBEM is great. Been a while since I've played, not sure how easy it is to find opponents though...

But,

If you are willing to open up the Editor/Campaign Edit functions you can have a buddy deploy the AI forces for you. The AI will still be the AI but it'll be a lot less predictable than a generated game. This might sound like a chore but really it isn't*... :)





*Unless you catch a bad case of map-making.

Re: Maximum AI Competitiveness
 

It is not easy to get players and sometimes it is difficult to finish games when there are life problems to solve. But it's good to try. That the PBEM does not die !!!!

Re: Maximum AI Competitiveness
 

Hmmm...., I might have to get back on the PBEM horse then... :up:

Any activity at the Blitz nowadays?

Re: Maximum AI Competitiveness
 

Don't know. But here you have a victim to train. Maybe then go around Blitz again.
I went through Blitz and there is some movement.
If you want to play we do it calm or in The Blitz.

Re: Maximum AI Competitiveness
 

Sounds good! :cool:

I can set something up.

Let me know what nations and roughly what size (map and build points) you'd prefer, nations and time period. All slots open?

It does not have to be a "fair and balanced" meeting engagement game with equal points.

We can log it as a Blitz game, no problemo. Use Blitz rules too if you like, or not.

This goes for you too, Dion - if you're interested...

Re: Maximum AI Competitiveness
 

I adapt to what you have in mind. No Blitz. If it were for Blitz, I would go to me more secure for me but I like to face challenges. That is why I join the unknown Lol.
I like more current times but you choose year and sides.
Also choose map size and quantity of points, but at larger size, greater delay.
I am with a expectant anxiety because I know that you are a scenarios editor so you can have brilliant ideas.

Re: Maximum AI Competitiveness
 

Location:
Schleswig-Holstein, WEST GERMANY

Date:
November 1985

Type:
NVA/Warsaw Pact delay (XXX Battle Points) vs Spain/NATO advance (30 000 Battle Points)

Air Sorties:
Warsaw Pact XXX
Spain/NATO 10

Map:
Grossenbrode, Germany 200x160

Visability:
53

Turns:
65




Background:

It is November 1985.

WWIII began in early August the same year with a Warsaw Pact attack on western Europe. NBC weapons have not yet been deployed by either side.

While initially enjoying some success on the ground Pact forces have failed to knock NATO out of the fight and in the high intensity warfare of this sort munitions, fuel and troops are consumed at alarming rates.

Warsaw Pact competence have also been found wanting. Especially the logistics and air warfare domains suffer from the extreme strains of large scale prolonged combat. By October the weight of NATO air strikes have had a crushing effect on Pact ability to bring munitions and troops forward leading to a culmination of the Pact offensive. NATO counterattacks have begun to roll back earlier Pact gains.

Pact morale has not yet fallen apart. But things are looking very grim. Ground is being lost and there seems to be no reprieve for the depleted Communist formations at the front.

Things are different on the NATO side. Some of the forces reinforcing the NATO forces in Germany come from Portugal and Spain. They are received with some suspicion, but are nonetheless a welcome sight.

The Spaniards, having recently joined NATO are initially given tasks believed to be less risky. One Spanish contingent ends up on the German Baltic coast and is tasked with reducing a Warsaw Pact force believed to be of very limited combat value…

The Pact troops have been bypassed by the main NATO offensive and have their backs against the Baltic Sea on a small piece of land just in front of FEHMARN Island. The bulk is believed to be made up of battered remnants of the East German Nationale Volksarmee’s 19th Motorised Rifle Division, a reserve unit that by circumstance ended up on the Baltic coast.

There is a slight chance that the Pact forces will be supplied over, or possibly withdrawn by sea and could thus pose a problem. The mission of the Spanish forces is to defeat the Pact forces and reach the bridges to FEHMARN Island.

Re: Maximum AI Competitiveness
 

You have a decent amount of battle points, but you don't know what you are facing...

I couldn't upgrade to v17,5 - I am still using v17 - so I set it to basic security. I won't have time to troubleshoot it tonight.

Take your time with the setup, I'll be out of town for a few days.

If there's a game version conflict I can set up this as a scenario to be played against Comrade AI.

Re: Maximum AI Competitiveness
 

Thank you. Unfortunately there are version problems. I wait for you to return and update the game.
https://i.imgur.com/K6qYdeh.png

Re: Maximum AI Competitiveness
 

Yeah..., I sent you the wrong files. Try this. Might still get an error message I guess...

Re: Maximum AI Competitiveness
 

Works. Do you use discord? We can exchange files there. And I also made a server for players. File exchange is easier. It seems to me.
Any rules?
Artillery limit?
Is extra ammunition allowed?

Re: Maximum AI Competitiveness
 

No rules, unless there is something you'd like to play with.

Your opposition will not be too outlandish. I'll be the AI for this game.


Everything is allowed.

I set you up as Spain but feel free to use other nations too if you wish - I think other NATO nations would be more fitting than say Afghan Mujahedin..., whatever floats your boat. This game allows for a lot of freedom.

There is no 'play balance'. The Spanish forces are previously uncommitted and largely intact. The NVA forces are depleted, abgekämpft, with failing logistics and their backs against the water.

How many there are and with what kit exactly is unclear, if they have air support is also not known, or the amount of artillery they have, but they are not dug in and there are no mines deployed..., it is believed they have, all in all, a low combat value.

I'll set up Discord within a few days (when I get back)...

Re: Maximum AI Competitiveness
 

Ready. I did not use 30 thousand points. Where did the file pass?

Re: Maximum AI Competitiveness
 

Okay. The Spanish army will only attack because the communication of the allies did not arrive on time.
I await your discord configuration and the exchange of missile files.

Re: Maximum AI Competitiveness
 

This thread has seriously drifted from the initial topic!.

You guys should take this PBEM game setup stuff to private messages, methinks?

Re: Maximum AI Competitiveness
 

It's true. Private or perhaps scenarios and maps as Wulfir's proposal seems interesting. Could you wait for him to think?

Re: Maximum AI Competitiveness
 

Sure thing 👍

Roman: I'll download Discord when I get back home, probably late Friday.

Re: Maximum AI Competitiveness
 

Anybody else wanting to play the scenario, either as a PBEM or against the AI let me know. I can set it up.

The map is max size but limited by water on two sides. Made with Venhola.

Re: Maximum AI Competitiveness
 

After we play I will want the scenario version vs ia

Re: Maximum AI Competitiveness
 

In some campaigns I've set up for myself I've made use of auto generated single battles renamed from a save file to a scenario file - then strung together as campaign nodes in a user made campaign.

It's a somewhat busy way of setting up a campaign, but I do it for two reasons - I don't want to know the composition and deployment of the AI and I wanted the AI to have a massive build points advantage beyond the first battle.

Re: Maximum AI Competitiveness
 

You can also "save as scenario" a generated battle on the first turn as an alternate way of doing that

Re: Maximum AI Competitiveness
 

Very good. Taking note

Re: Maximum AI Competitiveness
 

It can be saved as a scenario on any turn but in the context of what Ulf was discussing it can be a scenario save on the first time which gives you the same as renaming the save game to a scenario file

Re: Maximum AI Competitiveness
 

I did for real not know this! :up:

I guess in theory one could save it as a scenario on the last turn?

This would not make the AI more difficult, likely the opposite in fact, but what a time saver in any set-up where you'd want the next battle to take place without any sort of refitting period...

I guess damage to the map like destroyed bridges etc would be carried over also...?

Re: Maximum AI Competitiveness
 

You could save it wherever you like and everything would be saved as it was when you saved it......... this has been in the game since the 2018 release