Movement Scripts & AI Trigger

[Edwin P.]

Will there be movement scripts in the system that allow you to link unit movements to specific events.

For example, in SC1 it always annoyed me that the French Corps in Beruit would not move to UK territory after Italy entered the war, a standard human strategy.

I would like to write a script that would Trigger after Italy entered the war to move this unit to Tile xx,yy so that it would become Free French after France surrenders.

In addition, I would like to see an AI trigger added to the scripts that would influence the chance of a script being activated depending on the AI level.

For example: AI(0,0,0,80,100). Thus at Beginner AI the movement script would have a 0% to activate while at Expert level a 100%. Once activated it would still require a second trigger check to determine if the scripting event occurs.

[ August 13, 2004, 02:21 PM: Message edited by: Edwin P. ]

[pzgndr]

No AI movement scripts are planned to do this. An alternative would be to assume the unit moved when it had the chance and then bring in a Free French unit as a unit event after some date. Poof! Instant FF. And with the chance defined, this could either be a certainty or random.

No additional AI triggers are planned for the scripts. But there should be some AI % adjustment to most all of these event scripts and diplomacy based on difficulty settings. Not sure what exactly. Intelligence tech levels may also affect various %'s, for either AI or human games. Rest assured, game difficulty settings will have an effect.

[Edwin P.]

Interesting, could I write a script to say disband/remove 4 French fleets and their air unit to give the French a varied starting OB - say 1 HQ, OR 1 Armor and 1 additional Corps OR 3 Corps? This would reflect the varied strategies that players use to defend France.

[pzgndr]

This would reflect the varied strategies that players use

Now there's a never-ending task for the intrepid AI programmer, eh? Try to predict what Terif or Rambo or any of you guys will do next!

No disband unit event is planned. For solo games like this, why not start France or whoever with less units and more MPPs and see what the AI does? Or with the new unit event and some low %, just make a random event that France gets a HQ, or tank group, or other extra units? You can add dozens of these event scripts. This is about as close to a "variant" that I can think of happening at this point. You'd have a rare chance that France might get ALL these units, but not likely. A little surprise here or there is what we're interested in, yes?

[Liam]

Assigning units to make an order you give is intrigueing. If you recall in non turn games you can assign a unit to move, some will move on time increments. in SC2 if say you had a sub you wanted to take from Hamburg without micromanaging it's path all the way to say Alexandria and it moved in a fairly linear path that would be like non-turn based game. Almost like dragging and dropping and it executes turn by turn what it would do, problem is here. this is turned based, making an RTS or any kind of extreme programming like that would sapp the rest of the game. If others could come in and make add-ons it would be kewl. In other games I Play there are plenty of programmers but the SC community is relatively small. In Ages of Empires, dozens of Mods, complete alterations were made. In 1 they turned they changed every unit, terrain, building, etc... That's extensive stuff. But there are tens of thousands of Ages players, While there are hundreds if just but max a few thousand Steady SCers. Same with Hex Wargames, a lack of #s for such indepth things and a lack of 'programmers'

I'd personally love to see it all

[pzgndr]

dozens of Mods, complete alterations were made.

Some of the limitations in the SC1 campaign editor prevented the game from attracting modders. And with some of the other gameplay issues and inability to customize things like combat factors or political triggers, there was only so much you could do. Whereas HOI offered text scripts to allow modders to really get into the game. Thus, CORE really took off and did some good stuff.

You'll find with the SC2 Editor that much more can be done to customize the game or even create something new. Modders should look forward to this and we should all expect to see more Mods. :cool:

[Edwin P.]

Since there will be no AI triggers, I guess that this also applies to AI flags :

#AI_Flag=4 (Only occurs at AI Level 4 or higher)

With the proposed AI flag you can limit the script to a particular AI Level or higher.

And will there be a way to restrict the event to only benefiting the AI player? Example; a unit event that only occurs when the country is controlled by the AI, so I don't have a human player benefiting from specific unit scripts.

Example: USA AI, and only the USA AI, has a chance to receive 2 Carriers, a cruiser and a sub from the US Pacific Fleet if the UK surrenders.

Example: German AI, and only the German AI, has a chance to receive 1 to 2 extra submarine units. (ideally only at AI level 4 or higher )

Example: At Axis AI Expert Level Turkey has a 10% to join the Axis after France and Denmark Surrender to the Axis.

Example: At Allied AI Expert Level Turkey has a 10% to join the Allies after the US enters the war.

[ August 14, 2004, 02:37 PM: Message edited by: Edwin P. ]

[Edwin P.]

The case for an AI_Flag in Moddable Scripts

"Where the AI flag allows the script to run if the AI level is greater or equal to the AI_Flag value."

1. Allows the designer & modder to differentiate scripts by AI level. At higher AI levels more scripts would be activated.

2. Relatively easy to implement.

3. Easier to design AI level specific scripts and events.

4. Game difficulty gradually increases as players are exposed to a wider range of scripts and possible events at higher AI levels.

Example:

AI_Flag = 3, Script only runs at Beginner and higher AI levels.

AI_Flag = 4, Script only runs at Intermediate and higher AI levels.

AI_Flag = 5, Script only runs at Expert AI level.

[ August 15, 2004, 09:35 AM: Message edited by: Edwin P. ]

[Desert Dave]

Edwin P,

Seems as though... one time or another,

And probably only lately,

You have lived, yep,

You have dauntlessly explored! those crags

And crevices and peaks & valleys... INSIDE

Of... a computer... LOL! JK, JK!

Deconstructing anything, and especially

Logical-Positive arrangements,

CAN BE problematical.

As example, the Greek Philosopher,

Zeno I believe it to be,

Was fair certain!

(... as sly teaching device)

That you might arrive at destination "B,"

From your starting point "A,"

By... traveling, first, half-way there.

Then, from that half-way point,

Half-way again.

And again, and again.

Well, as you might soon suspect,

YOU WOULD NEVER EVER GET TO POINT B! :eek:

Opposite way around is also true,

If you see what I mean.

_______________

Anyhow, it seems to me that it would be

Most adroit and in accord with KISS principal

(... otherwise known as Occam's Razor )

To just make the AI

Absolutely the very best it could be

Under ALL circumstances,

And at ALL difficulty levels.

That way EVERYONE would benefit from

Identical AI choices and S&T excellence

No matter what game they might play. :cool:

Given our new Editor,

It is now possible to "fine-tune" and

Thoroughly "tweak" the Economic and Military

Components, so that ANYONE,

And I do MEAN - anyone,

CAN achieve a "balanced game"... for themselves.

It is very easy to do.

No major overhauls or too-exacting demands

On time or expertise.

For example, IF I discover that I can... defeat

The Allies just a bit too easily,

Well,

I simply adjust the "Industrial Modifier"

(... for either side, + or - )

By, say - 5 %

Or, change the unit costs, or the build penalty

Or the Combat target values,

Or the research limits or the diplomatic

Penalties and leanings, and,

Or or or or.

OK.

Back to the original suggestion (... and,

nearly ALL of these are quite elaborate

and X-tremely

well thought out, BTW)... perhaps it would

Be preferable to have the SAME and consistent

AI plans & mechanics no matter what difficulty

Level... IF that would detract from

The other more valuable features

That Hubert is NOW implementing.

So long as there are not any OBVIOUS "cheats"

By the AI, I am supposing

That most Players will be satisfied.

Naturally, Hubert can speak for himself,

And may well conclude that your methodology

Is something to consider.

Or, he may have ALREADY done something similar,

No way of knowing as of yet.

Meanwhile, I will go out to the park

For my daily 5-mile run.

First, though, I will have to go

Half-way there - somewhere in the neighborhood

Of that new Minor League baseball stadium,

And then... half-way again... and

Eventually... I'll be so close - a Quark's width

Distant from starting point "A," that I can

Begin planning for water at... point "B?"

[Edwin P.]

Many thanks for your most erudite feedback on an issue I find of most ponderable thought.

There are many ways to construct an AI of varing difficulty. Some give the AI advantages at higher levels - such as more plunder and more experience. While others provide the AI with more time for thought. Others give the AI access to more statigic and tactical options at higher levels. Some programmers make the AI more aggressive at higher level. As you said, there are many roads to success, and a thought provoking AI.

[Liam]

Dave has an interested way putting his point across as usual .. Civ3 is a perfect example .. You set the Difficulty level higher the AI gets Massive Bonuses akin to Cheating .. Europa Unversalis, same thing, the AI does things no Human can do cause they essentially cheat .. Both AIs can be difficult and in CIV3's case almost impossible to tackle with out more knowledge of the game engine .. EU2 wasn't as complex as CIVs and I Beat it's AI Engine but it's effective enough to create a Challenge that can destroy you if you're not patient .. In this game SC and the new Generation of SC2 I think that you're talkin about letting the AI Cheat, if you really want a good AI..

It is "Bad" We'll be cheating the AI most definitely, though I get peed at the AI in the other games. "IT'S NOT FAIR, you can build 5 Xs the units I can!" "But, it is fair that you're soooo stupid you send them all into attacking a Fortified Unit on a Mountain Top and they die !!"

[Edwin P.]

Here are some notes about the AI in the game GALCIV:

The other way intelligence matters is that it determines whether the AI is going to a) Do a thorough analysis on what players are doing Which AI scripts are available to use, and c) Determines whether the AI is going to do anything about it.

Example:

An AI with intelligence 60 will be able to check out its sensor areas and see if you are building up for an attack, fortifying starbases, trying to culturally dominate it, etc. It will then have to roll to see if it is going to do anything about it. I.e. "Do I recognize this as a threat or am I too dumb to see this?". If it rolls that it is going to recognize the threat then it has to roll whether it's goign to actually do anything about it. And if it is going to do something about it then it rolls to see if it will tell the human player about it (if it's a human player that is the threat).

IF the AI FAILS the roll, the human player gets a message from the AI letting it know it's on to what it's doing. Otherwise it quietly lowers relations and prepares for a war.

If you play on say Genius, and build up star bases and ships in a sector you're planning to attack you'll notice ships from other sectors quietly coming in to fortify. Fighters shadowing your transports. Etc.

But on lower intelligence levels, none of this happens.

If you play against "Intelligent" AI, the computer is playing at its most intelligent level (it gets to use all the AI scripts and such) and it's not getting any kind of production bonuses or penalties.

Higher than "Intelligent," if I understand correctly, it's still playing at its most intelligent level *and* it's now getting some production bonuses on top of that.

Most of all, I guess that I want to see the strategies of the best players incorporated into the game.

[ August 15, 2004, 05:37 PM: Message edited by: Edwin P. ]

[Liam]

Edwin in Civ3, I Know from reading the Forums and the proffessional Civ Players that the Advantages of the AI are basic. HyperTech, Hyper UnitBuilding, More Aggressive, More Productive... I believe also though I don't recall reading that your Dice are more unfavorable in the randomizer. GalaxyCiv sounds like a better form... Civ3 you end up facing Middle Ages units with Ancient Times Units before even the best Civ player who doesn't know how to tweak every game parimeter to get to that level...

SC2 would be annoying like this but it's effective. I've only managed to compete with the computer twice on next to Godly Level, I'd never try Godly or Sid level.. Not unless I was myself aware of strategies that can trick the Computer. SCers become so good, they can do this. Incorportating new strategies into the AI and to make it a learning Machine would be hard, some games claim this but I doubt it can be fully implemented. I think your idea about less Linear AI is good.. More Humanlike but still maybe not practical. Hyper AI is likely more effective.

[Hubert Cater]

Interesting posts all around guys... definitly a lot to consider, thanks!

Hubert

[Edwin P.]

Liam, many thanks for your comments. I find Civ3 especially good at the higher where the AI becomes much more agressive, in addition to receiving the Hyper Bonuses.

Although, a "hyperAI" is easier to program I believe that as you say a less linear AI makes for greater replayability and a much more interesting game.

The question is what is the best and easiest way to do this? Perhaps for the Allied AI's opening moves this means giving it a selection of core strategies:

1. Standard Defense (50%)

2. Corps Defense (25%)

3. Sink the Italian Navy (20%)

4. Invade Germany (2%)

5. Belgium Gambit (2%)

6. Italian Gambit (1%)

With FOW on, the Axis player is sure to be tested with this wide and unpredictable variety of operational strategies.

HC, thanks for the most informative comments.

[ August 15, 2004, 08:43 PM: Message edited by: Edwin P. ]

[Liam]

Edwin, percentages to a strategy would be nice. Especially if one Strategy is unstoppable. The AI doesn't take any chances in SC now. Maping out the various Strategies and creating a AI Strategic Move base as you have would be good. Giving them percentages and difficulty levels to perform each action. That way no matter what bonuses the AI has otherwise, it still is smart enough to make an offensive. I will not use AI much in SC2 though, though I know a lot love to.. I prefer Human oponents for a game like SC. It's a simple game, it's Highlight and Joy is real life competition.. The AI for me just a crash test dummy at best. Though I will admit Hubert's AI blocked my few tries. Heh, even the AI whooped my butt

[Edwin P.]

Liam, nice refinement, giving them percentages and difficulty levels to perfrom each action.

I agree with your comments on the AI and playing against real unpredictable human opponents, but I travel a lot and then play SC or CivIII vs the AI when I have time, especially as SC does not require a CD!

One of the changes I really like about SC2 is the new event triggers, which means that I can play a more unpredictable AI, instead of playing a MOD where I know with 100% certainty what benefits the AI has received.

In looking at the major powers, Russia probably has the fewest number of strategic options to consider. Why? As a land based power it has the fewer strategic options when compared to other Allied or Axis powers. A more powerful Russian AI will also make for a more interesting game for the Axis player.

With the new system players can easily prevent the Russian AI from investing MPPs in none critical areas such as Naval Research or purchasing Naval ships. The key though for Russia is to duplicate the key strategies of the top players - where they aim to preserve their forces until the Allies invade in the west and often times attempt to quickly take Finland.

[ August 16, 2004, 12:06 PM: Message edited by: Edwin P. ]