Beta news

[AlvaroSousa]

Currently I am working on the Pacific A.I. Because the pacific is mostly an island hoping, naval battling, odd land area type of map it has revealed some improvements that can be made to the A.I. in general. The Pacific A.I. improvements are being passed to the Europe WarPlan. So I have to test and test to make sure it doesn't break and things flow well.

Thus why you haven't seen a beta yet. I plan to release one end of this week. I was going to last week but I found too many inefficiencies I wanted to fix the first time around.

Now this doesn't mean WarPlan will now be at the level of Google's Deepmind A.I. (wouldn't that be something). In fact most of these changes are barely noticeable to a human player. But they do improve how the A.I. plays.

[ncc1701e]

I think the game is pretty stable now. Yes, there are still some bugs but with a minor criticity. Take a breath until players gave you their feedback about the game balance (especially the Barbarossa one).

[Franciscus]

Yes, take your time
But please release here the 07.1 patch...

Regards

[AlvaroSousa]

Well good news. I was able to resolve the issue between converting from Unity 2017 to Unity 2019 engine.

While on the surface this will not be noticed but it will have the latest internals which hopefully will resolve some of the DX crashes.

[AlvaroSousa]

Ugh found something else to fix between versions.

[kennonlightfoot]

If you can pull off making an AI that is good enough to even be passable as an opponent without giving it massive benefits, you will have done something no other developer of these type game has been able to do. Generally, these type of games are to complex for scripted AI's to even be worth using for testing. Only games like Total War where the AI plays both sides do they have a chance but it is not because they are that good, they are just that bad for both sides.

Maybe someday someone will be able to set up Google Deepmind to play like a PBEM opponent.[:)]
But then if the AI opponent becomes like the ones for Chess it may be the end of games.[:@]

[AlvaroSousa]

I wish.

[baloo7777]

I believe many more gamers purchase these games looking for an AI that is as strong as possible. Customers who pbem or other PvP methods are less numerous than those who prefer to play vs the computer. I play both ways at different times. It is often difficult to find a PvP opponent that is evenly matched with you as there often isn't a ranking system for players to use, and many do not want to take the time to play against a beginner (even if only a beginner at that particular game). The hardest decisions a developer has to make is how will this affect play-balance for vs the AI and vs a human player.

[AlvaroSousa]

It is incredibly difficult to write an A.I. for a game like WarPlan. SC3 is slightly better than mine I would say. Hubert has been doing it for a decade.

Now put my A.I. in Unity of Command it will kick your ass.... well best it can. Limited front with fixed objectives and one direction to go.

[baloo7777]

ORIGINAL: Alvaro Sousa

It is incredibly difficult to write an A.I. for a game like WarPlan. SC3 is slightly better than mine I would say. Hubert has been doing it for a decade.

Now put my A.I. in Unity of Command it will kick your ass.... well best it can. Limited front with fixed objectives and one direction to go.

You are absolutely right. I think you've done a bang-up job with this game. I do think that this (or any really) game as played against the AI is much different than if played against a human opponent. Making any changes to the game affects both ways of playing differently.

[AlvaroSousa]

The A.I. is less efficient and certainly less creative.