A challenge - playing on Impossible difficulty

[JanSorensen]

I would like to challenge others to see who can oppose the AI most impressively.

The challenge is to play
- both Axis powers with the AI taking control of the Allied powers
- difficulty on impossible
- advanced supply and fuel on
- AV and end date off.
- v1.040 or later. The specific version isnt too importent.

How long can you last - or can you even go on to actually conquer the Allies fully and if so how fast?

Play as many times as you like and save and reload all you wish - for once this wont be considered cheating. Lets hear it from the world conquerors - how well can you perform against such rough odds?

There is no prize except bragging rights and the prize of having had a fun time while pushing for the limits. Feel free to post if you accept the challenge before playing as well as after you have played through the game.

[Lebatron]

If I was to try this, I would as my first mission take both Egypt and Gibraltar. The AI is very poor at saving Gibraltar so take advantage of that. Trying Sealion would probably be pointless. The combat penalties are so bad you would probably bounce. So the Med is the place I would focus until its time to attack Russia.

The best Jap strategy I think would be to attack Russia. Doing the double squeeze to limit Russian production is a must to beat the AI at this level. Also you must bomb all of China's factories to nothing so they no longer bother you the rest of the game.

Any comments?

[SGT Rice]

Wouldn't the Chinese be swimming in supplies, and therefore be able to repair any infrastructure damage easily?

[JanSorensen]

Its good to see some interest

I hope you make an actual attempt - I am definitely looking forward to hearing how it goes.

[Lebatron]

ORIGINAL: SGT Rice

Wouldn't the Chinese be swimming in supplies, and therefore be able to repair any infrastructure damage easily?

Oh yes I forgot about all the free supply they would recieve. I was just thinking about the combat bonuses and penalties that apply under the difficulty settings. Perhaps, we should change the requirements a little[:)] Turn off the extra supplies the Allies get or just in China's case turn them off. Otherwise Japan could not invade Russia as they would be so busy with the Chinese. I can't see how it could be done otherwise.

[JanSorensen]

I am already well into my current game but if you prefer then certainly lowered the extra supply for the Chinese to say just 5 to avoid supply being meaningless for them. I dont think you should completely remove the free supply for China though nor lower it for the others. That would make too large a change in the challenge.

[JanSorensen]

I was into 1944 and doing well - when I decided to close the window without saving!!! Silly me, I guess I will have to pick from somewhere in 1942 when I last saved, ugh.

I am fairly certain I can win the war eventually - but how long it will take is still an open question. I am playing with the upcoming change to tech that only lets you use 3 points when trying to go from +2 to +3 - and its definitely making it much harder but also more fun.

[JanSorensen]

Spring 1950 - by then the Allies had no production left at all. I figure thats a fair enough place to call a win rather than have to conquer every single desolate place. I wont spoil it for others by telling what strategy I used except for what can be seen on the world map at the end of the game. Now, someone else beat this!

[Drax Kramer]

ORIGINAL: JanSorensen

I would like to challenge others to see who can oppose the AI most impressively.

The challenge is to play
- both Axis powers with the AI taking control of the Allied powers
- difficulty on impossible
- advanced supply and fuel on
- AV and end date off.
- v1.040 or later. The specific version isnt too importent.

How long can you last - or can you even go on to actually conquer the Allies fully and if so how fast?

Play as many times as you like and save and reload all you wish - for once this wont be considered cheating. Lets hear it from the world conquerors - how well can you perform against such rough odds?

There is no prize except bragging rights and the prize of having had a fun time while pushing for the limits. Feel free to post if you accept the challenge before playing as well as after you have played through the game.

Winning in such scenario is not going to be an endorsment to the AI in this game.

I recognise that any major upgrade of AI is unlikely to happen, but am certainly sorry that this game has turned into a PBEM tool so quickly. I am hardly an experienced player, yet I won every single 1941 and 1942 scenario playing as one power in first attempt.

German AI is incapable of anything (even against AI) but delaying for few years before fortifying Ruhr and trying to beat the clock.

Allied AI is incapable of invading Italy, invades Finland, but not Norway even if the latter is virtually undefended, leaves transports unprotected as late as 1945, fails to stage a meaningful Pacific campaign etc.

Unless someones convinced me otherwise, my firm belief (based on reading various online and printed sources) is that significant majority of players play their wargames against AI. PBEM is played by best players who are only the most vocal minority. Yet, after few scenarios played, a player cannnot lose against AI on "normal" level.

I do hope that upcoming projects related to ACW and Eastern front will have more work dedicated to AI or that at least AI was not going to be hardcoded to do things that would have been deemed stupid by both sides in the actual conflict.


Drax

[JanSorensen]

Drax,

I never intended for this to be an endorsement to the AI. I think its well known that its very hard if at all possible to design an AI for this kind of game given the kind of time/money that can be put into it. Just consider how much time and effort it has taken to finally let a computer beat the top chess players - and chess is a couple of magnitudes simpler. As such the AI is likely to only give you a decent run for your money while you are learning the game or if you give it some pretty large advantages.

Once you know the rules you definitely should not play on "normal" anymore. I urge you to try playing the Axis on "Impossible" though. I think you will find the challenge quite large.

You may be correct that most people play against the AI rather than PBEM. Personally, I do both - and I urge everyone to atleast try out PBEM before deciding its not for them. If you can find opponents of roughly your own skill PBEM can be a nervewrecking experience.

It would obviously be nice if the AI was better - I am just not sure if its realistic to ask for an AI that can stand against any decent player without being given specific advantages.

[dobeln]

First, it is very good to hear the research system is being worked on - that will solve a lot of problems

Second, in the meantime, I have found Lebatron's Franco's Alliance mod a great PBEM play (as Axis) so far. It provides for a meaningful Med game, as well as some zone changes that really help. (Tougher going for a Japanese invasion of the Soviet Union, easier defending the UK from air attack, etc.). In addition, there is already a fix in place for super infantry. In short, if you feel the game has problems right now, try out one of the mods.

[ravinhood]

Hey Jan, lemme see you do that with the impossible difficulty of "SPARTAN" (not Gates of Troy but Spartan). There's a campaign game that you can bypass the scenarios and just play "conquer the world". Lemme see how well you do against that AI. I play it on both Hardest and Impossible, but I do use v1.013 vs 1.017 I think 1.017 is easier, but, nevertheless, I'd like to see how well you do. And you need to take a "weak" nation, one that only starts with 1 or 2....3 max cities. hehe There's one nation up in the north surrounded by barbarians, it's actually the hardest scenario victory condition in the game, has something to do with getting a level 7 wood factory. But, you start out with ONE city. heh Lemme see you conquer the world with that one on impossible settings. It is doable, just like you did with GGWAW. But, I'll cut you some slack and let you choose a 1 to 3 city nation to start out.

[JanSorensen]

Hi Ravinhood,

I actually own Spartan too and have played that a little. It never did catch my interest though so I abandoned it after playing a couple of campaigns. I am certain that can be extremely hard to play though due to the extreme difference in position the various countries have.

[Battlebyte]

This is exactly the setup I was using for my current/last game, except at standard difficulty level. Now it's 1942 and Germany controls everything from Finland to South Africa, from Portugal to India (except Russia). Japan is in excellent shape as well, except having lost Foochow to a Chinese attack. The Allies have only 16 transports left and Italian fleets are operating unhindered in the Atlantic from bases in Gibraltar and Western France while the Luftwaffe bombs London.

I'll give it a shot with the harder difficulty. I've never tried it because I don't like the idea of giving one side free supplies. In strategic games, especially WWII, I've always liked the idea of using submarines and strategic bombing to make logistics a major difficulty for my opponent, making it much easier for me to take and maintain the initiative when and where desired. Of course, in WaW, supplies don't exactly work that way...but it still seems like giving free supplies would detract from the logistical part of the game. I'll try it though.

As for making more challenging AI opponents in wargames, I have a very simple suggestion:

1) Select an existing general-purpose scripting language (javascript, VBscript, or something else). You could create a custom one, but it would be quicker, cheaper, and easier to use one of several available, well-documented, and commonly-known ones that already have interpreter libraries that are fairly easy to incorporate into your program.

2) Create an interface into the data of your program accessible by the script engine.

3) Document the interface. Now anyone who says "the AI's not challenging" can attempt to write their own. I include step 3 because most scriptable games that I've seen don't document the capability well enough for it to be useful to anyone except the original programmers, which kind of defeats the purpose.

Of course this would be some added expense and difficulty, but I think the payoff would be worth it. You could have competitions between modders' AI scripts and some really good techniques could emerge.

Failing that, make it possible to programmatically access the UI, so that "bot" AIs could be created by hobbyist programmers in whatever language they prefer to use. Same effect, but probably cheaper and easier for the game developers, although fewer modders could take advantage of it.

[vlale]

Fall 1949, beat ya

Kind regards,

Rasmus, da record keept in house country wise. No one else up for it?

[JanSorensen]

Great work Rasmus, my compliments.

[Daykeras]

Am I allowed to cheat?

I'm total rubbish at this game, I just love it!

[Lebatron]

Day,
I bet your better than you think[;)]

[Daykeras]

Well I'm better than I let on, sure... but I'm no master. I know what I _should_ do. I just never do it.

Thanks for the vote of confidence, Lebatron