Falklands War (1982) – Beta Playtest Update

Above are some of the highlights from today’s game. I’ve been play testing the new Falklands War campaign and made some modifications which will be uploaded shortly.  One key change is I’ve added the visibility to the exploration mode of the game.  As much as this decision pains me, the AI is just hopeless without it.  Enabling the entire map made the game no fun since you knew where everything is, and turning it into exploration mode saw the AI retreat to the corners of the map for no apparent reason.

So, while it makes the map awkward to look at, it’s the only way I could find to make the AI work as intended.  And by “intended” I mean it will function as a reasonable opponent.  It still constantly trips over itself, which creates an amusing display of incompetence if you want to watch an AI vs AI match.

I’ll finish off this update with a quick summary of what I’ve noticed from my breakthroughs.  This is all reflective of the map I’ve just uploaded, and all screenshots I took above came from the playthoughs I’ve done.

British AI

I simply cannot get the British AI to work this map properly.  In a cruel irony, the Classic AI handles them best but won’t offload troops so it never goes anywhere.  Needless to say, I spent the majority of my play testing tweaking this part of the game.  It’s the reason for so many awkward paths, and the reason all submarines start surfaced.

Notes on AI Behaviour

• Opening moves:
  • Assault on San Carlos: 75% of games played
  • Assault on Salvador Water: 75% of games played
  • Suicide run directly into Stanley: 95% of games played
  • Assault on Port Howard: 20% of games played
• Games where carriers played a key role: 0%
• Games where cruisers played a key role: 95%
• Games lost through capitol capture: 15%
• Games Won in less than 400 turns: 0%
• Games Won in less than 1000 turns: 2%

Additional Thoughts

The AI cannot win as Britain.  I had to do some heavy manipulation of the visibility options to get it to a point where it even stands a chance at all.  It plays the most competitive game when you reveal the entire map, but this is a walk in the park if you do since the AI cannot suprise you.  With the options set the way they are now, exploration mode does a decent job of mirroring the options the real British fleet considered during their assault.  Prior to this, it would only do suicide runs to Stanley or get lost near Rincon Grande.  Every once in a while, it will be able to capture everything north of the Top Malo House, and the AI typically fairs better on the Western Continent when it tries.  But it’s inability to use carriers properly is a serious hindrance to its ability as an opponent, and beyond my ability to influence I’m afraid.

Do not be afraid to give this AI substantial combat and production advantages.  It needs it to succeed.  It was a lot more competitive when you removed the drain restrictions as well, so consider doing that if you want a challenge.  Changing all of this isn’t all that unrealistic – the population of the Falklands was really British citizens, their forces outperformed the defenders, and their economy is larger in real life.

Recommended Settings for this AI

• Suggested Handicaps for Success: Production +3, Combat +1, Civilian +3
• Enhanced AI has the best chance for success
• Consider removing resource drain when playing against this AI

Argentinian AI

This AI plays the map decently well.  It’s strategy has varied on occasion, and starts off with a powerful assault to capture as many cities as possible.  It even makes a reasonable effort to win capitol victories.  Still it trips over itself far too often, mostly the navy, and once the initial assaults subside the AI just becomes lazy.

Notes on AI Behaviour

• Opening moves:
  • Defense at San Carlos: 95% of games played
  • Defense at Teal Inlet: 85% of games played
  • Capture of all East Falkland cities: 50% of games played
  • Defense of Important Cities: 90% of games played
• Games where cruisers played a key role: 25%
• Games where battleships played a key role: 25%
• Games lost through capitol capture: 2%
• Games Won in less than 400 turns: 10%
• Games Won in less than 1000 turns: 40%

Additional Thoughts

The Argentinian AI plays the defensive war very well.  Some of it I think is a fluke, but nonetheless it works.  It’s common to see it set up a strong artillery battery near Teal Inlet, especially if it loses at San Carlos, which usually stops any offensive.  It also keeps most of its defensive forces locked up near Stanley.  But most impressive is its tendency to make a suicide run at Ascension and surprisingly it worked a few times.

There are three frequent occurences which I think are completely random.  First is the tendency of their capital ships to get locked up in Berkeley Sound.  This turns out to be the ideal position to spot and destroy those pesky British suicide assaults, but I actually think the AI pathing is getting stuck.  Second is the occasional use of the General Belgrano to absolutely wreck the British shipping channels on the northern end of the Falklands, especially near Salvador, but I suspect (and I’ve actually seen) it trying to reach Stanley through the shortest path possible.  And lastly is the amount of helicopters used to attack any attacking ships near Stanley.  It’s fun to watch the Endurance fall for this every time.

The one time I saw the Argentinian AI lose was when it unsuccessfully assaulted the British occupied Teal Inlet where the British cruiser force was able to destroy all attackers, then shortly after it make a run through to Stanley.  I have a feeling this was more luck for the British, as most of times the number of defenders in this area simply obliterate any attacker.

Nonetheless, the Argentinian AI offers a decent challenge.  But once you figure out how to beat it with overwhelming force, it’s a walk in the park.  What I found to help make it more challenging is to keep it under the 400 turn limit.

Recommended Settings for this AI

• Suggested Handicaps for Success: Default Settings OR slight combat disadvantage
• Enhanced AI has the best chance for success
• Keep resource drain off or else the island will be swamped with infantry.