BGForums

It just occured to me that the world scenarios (or any scenario containing seperate land masses) are meaningless as there can be only one winner. In all the games I have played the AI does not use sea/air transports. How then could it ever leave its own continent (country in some cases)?

I mean ,... British Isles as the AI..... it's done.. that's it, wait for the human he'll get to you in time, LOL (even if he is playing Central African Republic). Even the US... all they can ever get is South America, they can't even invade Cuba... sad.

Basically the scenario becomes a matter of when, not if you will win. You can pick the weakest of regions and you will be assured victory (it might take a decade or two, but...) Even the CIV III series AI would try to come ashore and attack you, and this game is much more sophisticated.

Don't get me wrong, its an awesome game, especially on the smaller scenarios, but the world scenarios are worthless in my opinion. Too bad, especially with the 32 player option on update 5. Could have been complete mayhem, and very challenging with a small country.

Yeah. The AI needs a complete new naval AI.

Rather than waiting for the human to take some naval territory and repond to that it should activley patrol and attack human naval territory.

AI absolutely needs to learn how to transport troops via sea or air. It's really essential to SP experience and IMHO should have highest priority.

Some thoughts about this matter that might help in coding such AI:

- Important note, part of the necessary mechanics to accomplish that already exists in game. When you select a unit, right click on destination and select air transport the game will look for available free transports, select closest one to the unit pick it up and drop it at destination. It's a process that AI could be learned to use.

- If unit is on foot and distance to travel greater then 2 hexes it should try air transport first and if none available (return 0) then use normal move. It would be great if those units would use trucks for transport too but it may be too much to ask.

- Strategic AI could use same way of selecting targets over a body of water as it does with targets on the same landmass. E.g. UK AI could select some seacoast city in France as a target and attempt to send troops there. They would try to use air transport, if not possible, naval transport and if not possible either, abort. If aborted AI should get a priority to make transport units. If none exist, research them.

- Units trying to use sea transport would select their target on the other side of the water, move to the location where they can be picked up (dock or a port), transport would arrive to that destination and pick them up. Then transport would drop them at such destination.

- These processes should be automated so that AI only needs to select its target for a unit and unit would arrive there with best means available.

I hope some of this will help.
Just don’t tell me that it’s out of limits of this engine because if the engine isn’t able to have AI that can navigate a body of water which is essential part of the game on any map with any sort of islands on it, the whole engine is fundamentally flawed.

Obviously this point has been discussed on the forums before, and here with the dev team we've had even more extensive discussions about it.

First off, I entirely disagree about how this makes the world scenarios 'useless'. What does happen is that the world level plays in a certain way which puts the human player into a role where he/she must make the moves and take the initiative.

A human can't "lose" the world scenario much like the US could never "lose" World War II (barring the possible exception of a German atomic bomb). That didn't necessarily mean that the US was guaranteed to "win" WWII, just that losing it (as in being landed on and taken over by German or Japanese forces) was pretty much not possible.

To use another example, the US didn't "win" in Vietnam either, but that doesn't mean that the US territory fell.

In the World levels of SR2010, there are a number of challenges; one is simply the challenge of taking over the world... you still need to be able to pull off the military and economic successes to do this. Then make sure you run as a democracy and have to win your elections every four years. Then maybe consider a time challenge - take over the world in 10 years? or 5?

I don't see the fact that cross-continent landings are not initiated by the AI to be the 'game-stopper' that is being suggested here. And if we did implement it in the current engine, it would in my opinion turn out pretty bad and actually make the game easier to win - the AI would lose a lot of forces in vain attempts. (Not that this is necessarily bad from a gameplay standpoint, since the occasional sea or air landing would certainly keep humans interested & amused; but that's a different point really.)

-- George.

I gotta disagree George... your analogy for WWII is somewhat bad. What if I'm playing Europe? Historically, the USA has launched an Amphibious Invasion of Europe, and won! As you said. So if I'm playing Europe, why is that impossible?

It is not a game-breaker, but it takes a lot of the charm out of a World Scenario.

its not useless and the amphibious invasion was launched from england and took nearly a year or more to get ready.

However complaints aside the AI does need to use transports more often, even if it is in a zerg like fashion. But would need to keep enough forces at home to defend itself if invaded.
Would bring more of a requirement to stock up on the old radar and scouts et al.

But Operation Torch, the American Amphibious Invasion of North Africa, was launched straight from the USA.

Let's face it, there's no reason any nation on the World Scenario should not be able to pull off an Amphibious Invasion.

This reeks off ''We couldn't code it so we'll say that the lack of gameplay is actually a feature''

The game lacks everything beyond a 2-4 year gamescope, economicly, militarily, diplomaticly. The shortterm is great facet of the game though, but I would suggest focussing on long-term gameplay, similar to CIV-style. I'd rather play the game for 5 to 10 hours where I build an empire, then 2 hours of constant warfare to emerge as a victorious continent.

It actually puts player in a role where he plays game while AI sits and does nothing - doesn't play the game. If AI declares war at you it should be able to attack you. Also, in WWII there were lots of both amphibious and air attacks done by boths sides of the war but in SR2010 you will only get those from the player. AI as it is reminds me lot of HOMM4 AI where you basically played a game agianst units placed on map that did almost nothing and waited for you to attack them. It gets boring fast.



Same could be said about land attacks, AI COULD lose lots of forces in vain attacks. That doesn't mean he shouldn't attack. Only made smart enough not to attack in vain. Keeping humans interested and amused is actually whole point of the game as they are the ones buying a game for that purpose alone.

You say a human could never lose the world scenario due to the challenges posed by the ocean, and also say that the human can only win the world scenario by beating the challenges posed by the ocean. Now that's nothing but saying "the AI isn't programmed to do it". Clearly it's broken in that one side is utterly helpless and unable to make the moves necessary to win.

Good points....

_________________
NO MORE NOISY FISH [unless they are green & furiously]
I HAVE STILL A FISH IN MY EAR

Im all for air and sea landing assualts by the AI.

Id even go as far as to say, "put it in , even if the Ai is horrible at it"

BUT, im also not one who will complain loudly that the AI stinks at doing it.

AND i do know that if it was put in, many people would do exactly that on these forums.

Personally, i'd be happier if the AI was made better at land battles AND we the player had more ways to lose that simply by being beat outright militarily by the AI.

If for instance the AI could soundly defeat a human landing attempt,and our population took the failed landing very seriously and could rise up against us,i wouldnt care if the AI was not capable of such landings.

Or if resources were expendable and our region could collapse due to them running out.

Or if morale automatically had a growing negative modifier the longer a war lasted.

So, summing up.
Id rather the AI was made more competant at what it can do now ,first, before it learns to do things it currently cant.
AND
Id rather we could lose in more ways, so that such weaknesses on the AIs side ,could be made less telling by allowing them to beat us in other ways.

_________________
This post approved by Tkobo:Official Rabble Rouser of the United Yahoos
Chuckle TM

I have to say that I agree that it would be nice if the AI was capable of launching amphibious attacks. I think if the world scenarion seems that "useless" to you, you might try and set up a MP game with it. I guarantee Berg will launch an amphibious assault on you as would any other human worth his (or her)salt. I'm pretty sure George could stand a break from being told what he "should" have fixed in the free update he gave us less than a week ago.

_________________
"Socialism is a philosophy of failure, the creed of ignorance, and the gospel of envy, its inherent virtue is the equal sharing of misery.” -Winston Churchill

((and I second what Tkobo just posted))

_________________
"Socialism is a philosophy of failure, the creed of ignorance, and the gospel of envy, its inherent virtue is the equal sharing of misery.” -Winston Churchill

I really don't have a disagreement with a lot of what's posted here - in particular, I've already pointed out it would be interesting (from a gameplay standpoint) if the AI did assault attempts, even the "Dieppe" style attacks that are doomed to failure. At least it would force us to garrison our coasts and backcountries.

However, as I've also said, the current engine is simply not designed for this on a couple of levels. This is something for the wishlists of future projects, not something that can be shoe-horned into SR2010.

The lack of amphib landings by AI doesn't make the world level "useless", it just removes some (albiet interesting) possibilities. SR2010 is what it is, and we work to make it better, but this is not an area that is within the current codebase to add in. I wouldn't go as far as saying ''We couldn't code it so we'll say that the lack of gameplay is actually a feature'', since I never really said it's a "feature", I just pointed out that "that's the way it is", and that there are still lots of ways to play at the world level.

-- George.