SCW and the limits of AGEOD engine

[Blancador]

Yes, increase patrol value and also is needed high loyalty level in the regions, but in España:1936 great Campaign we start with a 50% loyalty in most of the regions. In a mid war scenario this is much easy to do. What will help a lot is to make slower the movement in enemy controlled regions, specially in 100% enemy MC, but i think this needs changes in the Exe file.

That sounds like a very interesting solution. I am looking forward to test it :-D

[Mikkel]

One thing I noticed in my first extensive game is that the Republican AI is not all that committed to defending Madrid. I was able to capture it as the Nationalists by December 1936. He has a massive stack that could steamroll through my lines (or hold Madrid for a long while), but he sent it north to the Basque country. That definitely caused a lot of problems on my end, but given the immense value of Madrid, it still doesn't make too much sense.

Is there any way to have the AI devote more attention to Madrid?

[Leibst]

I'm thinking about making a AI Mod, non official. And if it works make it official. I can make a mod of my own game

[Mikkel]

That could work!

Otherwise I have to say I'm quite pleased with the AI overall. I ramped up the difficulty setting and although I'm not fighting for my life and although I have Madrid, the war is still far from over and I had a few nasty surprises. Then again, I am playing as the nationalists, and I think taking the republican side has to be a considerably bigger challenge.

[TheDoctorKing]

Another thing that would reduce deep penetration raids is to lower the amount of supply that units can carry. Make supply units more expensive too. This way, people could only afford one or two for each stack, and then they would only have a couple of turns of supply at the most. Deep raids would then be too dangerous.

The deployment of the Republican units in the north tends to set up a line.

Another idea is to make entrenching easier and quicker. Then, a Republican brigade or two with supply unit and maybe an artillery could reasonably expect to hold a region against anything but the heaviest Nationalist assault. That would bog things down right quick once the Republicans get their brigades and the Nationalists get some numbers, enough to put a regiment or so in each front-line region. Then, try to convince the AI to divide her forces like that.

I answered that poll "mostly plays multi, but sometimes against the AI". I don't have any objection to playing two turns a day (the way I set up PBEM games, I get a file with the opponent's orders, give my own orders, execute the turn, then do my orders for the next turn and send the files back to my opponent, so we alternate "hosting" - goes twice as fast that way). I rarely have more time than that anyway, particularly if I have two or more games going on at the same time. And, while I can beat the AI in this game or any other AGEOD title with relative ease after a couple of warmup games, I have lost three or four straight PBEM games. Nothing like a real human opponent.

[Taillebois]

I only ever play against the AI.

For me the AI is easily hard enough - I wish there were an easier level or the ability to mod it to make it easier.

[Mikkel]

I only ever play against the AI.

For me the AI is easily hard enough - I wish there were an easier level or the ability to mod it to make it easier.

I have to admit, even though my Madrid observation still stands, the AI can put up a fight if you crank up the difficulty level. It pulled a couple of punches recently that I was not expecting at all.

[Durk]

Is interesting to see those who love solo play.
I do not exactly understand why anyone would think a computer program is a better challenge than a pbem opponent, but we all have our preferences.
As I have already said, I find playing against myself in all games, not just Ageod, is better than playing versus AI. I wonder why players put such demands upon a theoretical AI which even top AI researchers could not meet.
AI is introductory, not challenging.
Well, enough said.

[PJJ]

Is interesting to see those who love solo play.
I do not exactly understand why anyone would think a computer program is a better challenge than a pbem opponent, but we all have our preferences.
As I have already said, I find playing against myself in all games, not just Ageod, is better than playing versus AI. I wonder why players put such demands upon a theoretical AI which even top AI researchers could not meet.
AI is introductory, not challenging.
Well, enough said.

We're not talking about a true artificial intelligence here. That HAL 9000 stuff is still science fiction and won't be reality in many decades, if ever. It's another thing entirely to have a well-programmed computer opponent in a wargame. That's definitely doable, and there are already some wargames available with a computer opponent that give a good challenge even to the best of players. Is it the same as playing against an experienced human opponent? Of course not. But that doesn't mean all work on programmed opponents is pointless and that people who for some reason can't or won't play against other humans (or against themselves!) shouldn't be able to play at all in an enjoyable way.

I feel that this is the whole point of this AI discussion - to give also non-PBEM players a satisfying experience with a wargame. It doesn't take anything away from the PBEM-only guys.

[TheDoctorKing]

And another way to create a solid line is to give both players enough 2-star and 3-star generals that they can have a "corps" stack in each region. Then, any attack on any region will gather support through MTSG from neighboring regions, and from "front" units one region to the rear. That will bog things down right quick. Play the RUS Drang Nach Osten scenario for an example of how this works.

[Leibst]

That's very similar to the historical deployment in 1938.

[veji1]

Is interesting to see those who love solo play.
I do not exactly understand why anyone would think a computer program is a better challenge than a pbem opponent, but we all have our preferences.
As I have already said, I find playing against myself in all games, not just Ageod, is better than playing versus AI. I wonder why players put such demands upon a theoretical AI which even top AI researchers could not meet.
AI is introductory, not challenging.
Well, enough said.

I am sorry but you fail to grasp the essence of Computer gaming. The whole point was that suddenly one could play alone. Sure it will always be better to play against a human, but the freedom of computer gaming is to play alone against the computer, wherever and whenever, might be 2 am and you have insomnia, or between 2 and 3pm while the baby has a nap, or on the plane during that business trip, the one and only moment you have time to play in your hectic life, etc... Preference isn't the right word here, it's possibility and freedom. Committed PBEM requires organisation, timekeeping, commitment, etc.. Rewards are great, but it is often hard to put everything together in real life. And again with computer games lots of people are just looking for the freedom of the quick break from their daily lives, not the commitment of a PBEM.

[Taillebois]

It's always better playing against AI than humans. Humans are horrible. They are either useless, if you beat them easily; or gloating bully boys who just know the rules better than you. And they actually get upset if you blow up their favourite boat/tank/plane/horse in what they say is an unrealistic way and would never have happened if it wasn't only a game. I'm sure the AI sheds tears over the loss of its pixel troops just like any mother would.

I bet few people here who moan about the AI can beat a Chess or Go AI if it is not handicapped. Find a game or an AI level you like. Most games don't have enough variation in the levels available be they from Arcade/Newcomer to IronMan/Grognard.

[Erik Springelkamp]

It's always better playing against AI than humans. Humans are horrible. They are either useless, if you beat them easily; or gloating bully boys who just know the rules better than you. And they actually get upset if you blow up their favourite boat/tank/plane/horse in what they say is an unrealistic way and would never have happened if it wasn't only a game. I'm sure the AI sheds tears over the loss of its pixel troops just like any mother would.

I bet few people here who moan about the AI can beat a Chess or Go AI if it is not handicapped. Find a game or an AI level you like. Most games don't have enough variation in the levels available be they from Arcade/Newcomer to IronMan/Grognard.

The worst thing about playing against a human is losing

[Taillebois]

Correct. At least with wargames you can take some solace from probabilities, but chess, draughts, Go - it is unremittingly you being beaten.

[Blancador]

That's very similar to the historical deployment in 1938.

Therefore, the interest of historical scenario starting in late period :-)

[Blancador]

I am sorry but you fail to grasp the essence of Computer gaming. The whole point was that suddenly one could play alone. Sure it will always be better to play against a human, but the freedom of computer gaming is to play alone against the computer, wherever and whenever, might be 2 am and you have insomnia, or between 2 and 3pm while the baby has a nap, or on the plane during that business trip, the one and only moment you have time to play in your hectic life, etc... Preference isn't the right word here, it's possibility and freedom. Committed PBEM requires organisation, timekeeping, commitment, etc.. Rewards are great, but it is often hard to put everything together in real life. And again with computer games lots of people are just looking for the freedom of the quick break from their daily lives, not the commitment of a PBEM.

I follow completely on this point. When I have time for 2 players wargaming, I prefer to come back to the old method: A good friend (and fair player) a tabletop, a hundred of miniatures, good alcool and a cigar ;-)

[andatiep]

----
About the simulation of front lines in AGE games :

I think the corps' reserve movement system ("March to the Sound of the Guns") should be use in a more and more automatical way, the more the warfare use modern transports and, after all, transmissions.

RUS is currently beta tested with almost 100% chance for a corps to participate in a fight in a neighboring region if the stack can reach it in one turn. I'm currently satisfied with the results : It allow (and push the player to do so) the constitution of small WWI fronts of 3-4 regions (which was the WWI Eastern front standards). This fronts can't be cross without fights, but you still can turn them if you have time, supply and fast units.
I suppose it would be easy to simulate the WWI Western fronts (and so more or less the SCW fronts) this way : like said TheDoctorKing, just give to both sides much more 2 stars Corps leaders in the setup of the scenarios, and you will get many "uncrossable" front lines.

The same way, the AI could be helped to use/simulate this strategy if you give it a lot of 2 stars leaders commanding the forces at the start of the game, which come already as corps leaders (i suppose the AI will keep them as corps leaders if they are in an Army range).
To be sure they are in an army range, there is also the idea to change the 3 stars army leader system (we will test locked 3 stars Army leaders (as "Big Theater Army leaders") with a huge range of action, so that 2 stars leaders can form corps which much less limitations and so that players (and IA) don't meet anymore the "bug" which don't allow a 3 stars leader Army stack to attack if there is another friendly unit in the region).

If most of the neighboring stacks of the IA help each other, you should get more challenge for small time design. Doesn't mather actually if your human's front line is facing properly the enemy like IRL or if the AI front lines looks like a worm-column.



----
About PBEM vs play-AI debate

True that there is a majority of only-AI players, but there is obviously a minority of it to show up as volunters and MODers to improve the play of the AI in the games. And this is not a critic, this is just a fact, probably because, like it was said above, AI players are less involved in one single game but like to try many of it. And of course because AI design need a lots of time for each possible situations.

I understand all the reasons why to play against the AI.
And i (always) like the one who called for freedom and choice .

But i don't understand some reasons why not playing PBEM. Especially the ones who says that it takes not enough time to play its 2 turns before sending the files to the opponent. I'm so perfectionist that i can't spend less than 4 hours to play my turns, so i'm never in a situation where i need the AI to play more turns in the week or in the month.... .

The other reason i can't understand is the one which say that PBEM humans are always unfair bad guys making troubles.
Boys apy:, how are we supposed to build and improve good games if we are not able to play it together ?

I'm born boardgamer long time before the PC and an AI could propose any good historical simulation games on earth. And i will not change that.
My free contributions are following the hope that the boring time i have to spend on the play of the AI in the scenarios should provide a good introduction and historical feellings for beginners, as well as a good training for the Big Game : the games with and against gamers.
That's my two cents.
I'm ready to follow anybody who wants more from the AI, i'm just still waiting for the AI fans to join the coding party .