How units are portrayed on the game map is really determined by the game system. You need to see whether a hex contains a stack or just a single unit, so therefore you at least need two types of graphics. For the top unit in a stack you need to give at least minimal information (NATO symbol/ship or air silhouette, country colour code, unit ID and probably but not absolutely necessary combat strength and speed). In the end, most of the information will have to exist in a purely text form anyhow, so it might be a good idea to remove actual combat values from the counter graphics (make them background, symbol and ID only) thereby removing some cluttering and making the design itself simpler. The top unit of a stack would be the only one visible on the map, once you click on the unit or stack you get additional information (all the other units graphically portrayed, and text information concerning the top unit, the others would be either accesible by clicking on their icon, or simply toggling through the stack). This might even allow for some fog of war, by providing the graphical information about an enemy stack, but not the text info.
The more I think about this the more variations I come up with, so I'll leave it at this, there are quite a few possibilities how to portray units within CWiF (without major deviation from WiF). But in the end I expect the graphic design has to happen after the major principles for the computer game are set (should FoW be included as a new optional rule? etc.). I tend to go the other way, first design my units then try to bend the game around counters, in the end I always end up either dumping the game or redesigning the units as I discovered my error (am just working on a Classic Age tactical wargame, already have 2/3rds of Alexander the Great's army on a countersheet and just decided I made a major mistake, had I started by spending an additional week on the rules frame I would not have wasted my time on beautiful graphics which I might never use now (I probably have to add more counter information which for now would clutter the counters too much).
So the first steps of the game design have to be determined before any work is started on the graphical aspects (the graphics are there to enhance game play (or even make it possible in the first place), they do not have any use on their own).
Marc aka Caran... rambling while waiting for some more books on ancient warfare to arrive in the mail
Matrix Games and Australian Design Group Seal Deal for World In Flames!
Moderator: maddog986
-
Les_the_Sarge_9_1
- Posts: 3943
- Joined: Fri Dec 29, 2000 10:00 am
Some people will call information "difficult" some people won't.
Take for instance the moron reviewer that recently totally slagged Korsun Pocket simply because it used Nato counter images. The man shouldn't have even been asked to review a wargame in the first place, he doesn't know the first thing about one.
Is there anyone here that actually doesn't now what a box with an X in it represents?
ASL has a counter that has a lot of information on it for a vehicle. But then that information is as plain as day to even a newbie after he has spent even the smallest amount of time learning what information is placed where.
RTS games use images that possess zero information at all, other than a picture graphic of what the unit looks like. I personally find having to look elsewhere to see the unit status MORE difficult. The best you get is a health bar.
And that is why I like turn based games too. In ASL my Panzer III will be a Panzer III till hell freezes over, or you actually hit it and kill it. It won't die through erosion. And I can tell all I need to know about a Panzer III ASL counter in a micro blink. Gun type, range type, mg types, movement range, movement category, armour type, vehicle orientation. None of these details are difficult at all.
The only thing that separates turn using board wargames from computer wargames, is you generally speaking are out of luck if you are to lazy to read the rules.
The computer's edge, is it knows the rules already, and doesn't have to worry about forgetting any. It can also answer questions in blinding speed. Want to see how far a unit can move, click on it, and the computer can show you in a dozen ways.
But frankly, my own eyes can see how far several dozens of ASL counters should be able to move in a glance. The key here is should be able to move of course. I am not saying I can actually SEE each units potential simultaneously.
As I can see all the counters at a glance, and I can see all of their movement rating simultaneously at a glance, and I have the advantage of knowing the game, because I learned the rules first.
You can't simultaneously see all of a computer wargames pieces potential moves all at once either, you have to click on each one one after the other.
Thats why I think having counters with the information printed on them is not the end of the world.
Some games do the real time thing well enough though that certain information is just not relevant. Close Combat only requires you to tell the unit what you expect of it. After that it tries to do that. So the unit would have no use for displaying data that only the computer needs to see.
Some games just don't need to display a lot of data. The counters in Advanced Third Reich I know for instance really only have 4 forms of potential information in most cases. Ground and air units for instance have a nation colour, a unit type in a box, a combat rating and a movement rating.
Strategic Command leaves off the movement rating, as the game can show a shaded radius.
But it uses a text area at the bottom of the screen to give detailed unit particulars. Still those details are fairly basic information.
But Advanced Third Reich is still a better game, even though it uses less information simply because the player is expected to actually learn the game first.
Essentially, a game will only be as good as the player.
Take for instance the moron reviewer that recently totally slagged Korsun Pocket simply because it used Nato counter images. The man shouldn't have even been asked to review a wargame in the first place, he doesn't know the first thing about one.
Is there anyone here that actually doesn't now what a box with an X in it represents?
ASL has a counter that has a lot of information on it for a vehicle. But then that information is as plain as day to even a newbie after he has spent even the smallest amount of time learning what information is placed where.
RTS games use images that possess zero information at all, other than a picture graphic of what the unit looks like. I personally find having to look elsewhere to see the unit status MORE difficult. The best you get is a health bar.
And that is why I like turn based games too. In ASL my Panzer III will be a Panzer III till hell freezes over, or you actually hit it and kill it. It won't die through erosion. And I can tell all I need to know about a Panzer III ASL counter in a micro blink. Gun type, range type, mg types, movement range, movement category, armour type, vehicle orientation. None of these details are difficult at all.
The only thing that separates turn using board wargames from computer wargames, is you generally speaking are out of luck if you are to lazy to read the rules.
The computer's edge, is it knows the rules already, and doesn't have to worry about forgetting any. It can also answer questions in blinding speed. Want to see how far a unit can move, click on it, and the computer can show you in a dozen ways.
But frankly, my own eyes can see how far several dozens of ASL counters should be able to move in a glance. The key here is should be able to move of course. I am not saying I can actually SEE each units potential simultaneously.
As I can see all the counters at a glance, and I can see all of their movement rating simultaneously at a glance, and I have the advantage of knowing the game, because I learned the rules first.
You can't simultaneously see all of a computer wargames pieces potential moves all at once either, you have to click on each one one after the other.
Thats why I think having counters with the information printed on them is not the end of the world.
Some games do the real time thing well enough though that certain information is just not relevant. Close Combat only requires you to tell the unit what you expect of it. After that it tries to do that. So the unit would have no use for displaying data that only the computer needs to see.
Some games just don't need to display a lot of data. The counters in Advanced Third Reich I know for instance really only have 4 forms of potential information in most cases. Ground and air units for instance have a nation colour, a unit type in a box, a combat rating and a movement rating.
Strategic Command leaves off the movement rating, as the game can show a shaded radius.
But it uses a text area at the bottom of the screen to give detailed unit particulars. Still those details are fairly basic information.
But Advanced Third Reich is still a better game, even though it uses less information simply because the player is expected to actually learn the game first.
Essentially, a game will only be as good as the player.
I LIKE that my life bothers them,
Why should I be the only one bothered by it eh.
Why should I be the only one bothered by it eh.
-
wfzimmerman
- Posts: 338
- Joined: Wed Oct 22, 2003 7:01 pm
- Contact:
A few quick comments.
1) I am quite fond of CWIF1 and would like to see as much as possible carried forward into the production version. IN particular, don't mess with the map or the counters. They are beautiful!
2) I cannot emphasize enough what a debt we all owe Chris Marinacci.
3) PBEM and net are priorities. AI is secondary.
4) I realize that Matrix will not want to give schedule projections, but can experienced observers of Matrix make a guess as to when we may Matrix CWIF?
1) I am quite fond of CWIF1 and would like to see as much as possible carried forward into the production version. IN particular, don't mess with the map or the counters. They are beautiful!
2) I cannot emphasize enough what a debt we all owe Chris Marinacci.
3) PBEM and net are priorities. AI is secondary.
4) I realize that Matrix will not want to give schedule projections, but can experienced observers of Matrix make a guess as to when we may Matrix CWIF?
Contribute to the Steve H. thank you book! http://www.nimblebooks.com/wordpress/2009/04/contribute-to-the-wargamers-wwii-quiz-book/
Too much information
Les:
I disagree about your Panzer III graphics example.
* It could be abandoned by a panicking crew but a new crew (who lost their tank somewhere else) could take it over.
* It could be immobilized through losing a tread but still able to fire.
* It could be out of fuel/ammo but able to be refuelled/rearmed within the scope of the game.
* It could be hull down.
* It could have boresighted a road and, although it can move, the owner needs a reminder that if it did it would lose the boresighting.
* It could be captured by the enemy and being used against its original owner.
Each of these situations can be (or should be!) represented on the map in ASL with additional markers. Tossing all of these factors onto a screen along with the other factors you mentioned plus a counter of the tank-killing commander and his "-2" gunnery factor and suddenly there is way too much information... and that's not even considering the actual terrain under your Panzer III. Then we start stacking multiple air, naval and land units in the same hex.
In summary, I don't think you'll be seeing all of the relevent factors sitting on the counter in cWiF... but I agree that we need more than a "Health Bar" under the counter.
Zimmerman:
You're entitled to your opinion. I just happen to violently disagree with it
If AI isn't important then what was wrong with cWiF1? Why didn't that become the definitive game? ANSWER: Most computer gamers spend a lot of time playing solo and if it doesn't have an AI then they won't give it a second look. PBEM is good but the AI is the Holy Grail!
I disagree about your Panzer III graphics example.
* It could be abandoned by a panicking crew but a new crew (who lost their tank somewhere else) could take it over.
* It could be immobilized through losing a tread but still able to fire.
* It could be out of fuel/ammo but able to be refuelled/rearmed within the scope of the game.
* It could be hull down.
* It could have boresighted a road and, although it can move, the owner needs a reminder that if it did it would lose the boresighting.
* It could be captured by the enemy and being used against its original owner.
Each of these situations can be (or should be!) represented on the map in ASL with additional markers. Tossing all of these factors onto a screen along with the other factors you mentioned plus a counter of the tank-killing commander and his "-2" gunnery factor and suddenly there is way too much information... and that's not even considering the actual terrain under your Panzer III. Then we start stacking multiple air, naval and land units in the same hex.
In summary, I don't think you'll be seeing all of the relevent factors sitting on the counter in cWiF... but I agree that we need more than a "Health Bar" under the counter.
Zimmerman:
You're entitled to your opinion. I just happen to violently disagree with it
If AI isn't important then what was wrong with cWiF1? Why didn't that become the definitive game? ANSWER: Most computer gamers spend a lot of time playing solo and if it doesn't have an AI then they won't give it a second look. PBEM is good but the AI is the Holy Grail! /Greyshaft
-
Les_the_Sarge_9_1
- Posts: 3943
- Joined: Fri Dec 29, 2000 10:00 am
I WILL agree, ASL can rapidily fill a hex with secondary counters, and those secondary counters CAN get intrusive. I was only commenting on the physical design of the specific counter symbology really.
Although, a computer would still have to deal with that data, and in all likelihood, the data would have to get displayed somewhere. Odds are in a message window. That would be no better or no worse in my opinion. It would eliminate having to look through the counters though, if the data was displayed elsewhere. That much I will give as an edge to a computer wargame.
Maybe I am odd (well of course I am odd ). But I actually DO enjoy playing wargames entirely solo ie I am the only one playing both sides.
My best example is Longest Day, a massively detailed incredibly accurate simulation. Often the only thing spoiling a solo wargame experience for me, is when there is an element to the design that really needs to remain secret from the opposing side.
Some games though, really don't have a problem with this. While some are entirely dependent on having an opponent. You can't really play Columbia Games's block games solo for instance.
I can play A3R solo almost, the diplomacy rules might as well be dispensed with while playing solo though.
It's quite impossible to play ASL solo and get any use out of Hidden units .
In the absence of a GOOD AI though, a game absolutely MUST have a very well crafted PBEM set up. Because no one wants to play an idiot AI. And the cost of a computer wargame will not fly if the AI has a reputation of only being worth about a month before besting it on a regular basis.
Although, a computer would still have to deal with that data, and in all likelihood, the data would have to get displayed somewhere. Odds are in a message window. That would be no better or no worse in my opinion. It would eliminate having to look through the counters though, if the data was displayed elsewhere. That much I will give as an edge to a computer wargame.
Maybe I am odd (well of course I am odd
). But I actually DO enjoy playing wargames entirely solo ie I am the only one playing both sides.My best example is Longest Day, a massively detailed incredibly accurate simulation. Often the only thing spoiling a solo wargame experience for me, is when there is an element to the design that really needs to remain secret from the opposing side.
Some games though, really don't have a problem with this. While some are entirely dependent on having an opponent. You can't really play Columbia Games's block games solo for instance.
I can play A3R solo almost, the diplomacy rules might as well be dispensed with while playing solo though.
It's quite impossible to play ASL solo and get any use out of Hidden units
.In the absence of a GOOD AI though, a game absolutely MUST have a very well crafted PBEM set up. Because no one wants to play an idiot AI. And the cost of a computer wargame will not fly if the AI has a reputation of only being worth about a month before besting it on a regular basis.
I LIKE that my life bothers them,
Why should I be the only one bothered by it eh.
Why should I be the only one bothered by it eh.
1. If there are 300 counters on the map and I need to check one piece of information of all 300. The best format is in an information screen that can be sorted like a data base - just like in UV.Les the Sarge 9-1 wrote:
1. Some people will call information "difficult" some people won't.
2. ...a Panzer III ASL counter in a micro blink. Gun range......, mg types, movement range.........
3. But frankly, my own eyes can see how far .....counters should be able to move in a glance.
2. In a computer game, you click on the counter. The places that a unit can move to are highlighted and the rest of the map is shaded. And the same for the range of the gun. I REFUSE TO COUNT HEXES.
3. Yes? Really? But the weather might change from sunny to mud to ice in a day. Then the distance that the unit can move is altered from one turn to the next. Read point 2 again. A computer game makes this simple - a board game makes it hard work.
I am not concerned either way about an AI. I stopped playing seriously against the AI years ago. I only do so to test a feature. Serious games are played against humans.
[QUOTE=Greyshaft]
Les:
I disagree about your Panzer III graphics example.
* It could be immobilized through losing a tread but still able to fire.
* It could be out of fuel/ammo but able to be refuelled/rearmed within the scope of the game.
* It could be hull down.
* It could have boresighted a road and, although it can move, the owner needs a reminder that if it did it would lose the boresighting.
a screen along with the other factors you mentioned plus a counter of the tank-killing commander and his "-2" gunnery factor and suddenly there is way too much information...
QUOTE]
Of course KP is an operational level game. But as variables change, the defence factor of the unit changes.
I can check each end every factor and there is no need for me to add them all up - the computer does it for me.
Rather, I simply look at the info screen and it tells me the current defensive value.
In ASL you have to manually add up all the values - wasting valuable playing time.
Les:
I disagree about your Panzer III graphics example.
* It could be immobilized through losing a tread but still able to fire.
* It could be out of fuel/ammo but able to be refuelled/rearmed within the scope of the game.
* It could be hull down.
* It could have boresighted a road and, although it can move, the owner needs a reminder that if it did it would lose the boresighting.
a screen along with the other factors you mentioned plus a counter of the tank-killing commander and his "-2" gunnery factor and suddenly there is way too much information...
QUOTE]
Of course KP is an operational level game. But as variables change, the defence factor of the unit changes.
I can check each end every factor and there is no need for me to add them all up - the computer does it for me.
Rather, I simply look at the info screen and it tells me the current defensive value.
In ASL you have to manually add up all the values - wasting valuable playing time.
-
Les_the_Sarge_9_1
- Posts: 3943
- Joined: Fri Dec 29, 2000 10:00 am
I guess I will just have to risk sounding arrogant then by saying I can view hundreds of counters and in seconds know most of the potentials of each counter to a degree of clarity that I am able to see what their overall movement potentials are suifficient to know enough about them that I can manage.
Yes sounds arrogant, but after many years, I also know most wargamers that have played board wargames are the same.
I guess it is like my brother being so good with math, that most people can't keep up with him even if they DO have a calculator whilst doing basic math functions. He merely thinks that fast.
There is something that is being assumed here, that, that Panzer III counter is unable to signify details you would already know.
If it wasn't bogged down last turn, odds are you won't suddenly wonder if it is. Especially if it did nothing to allow it.
If it is standing in the middle of a grain field, odds are you won't even be wondering if it will in its next move be bogged down.
If last turn I didn't boresight the hex that is the cross roads, then hey how about that, it won't be next turn either.
Those examples are misleading. They assume details that I would already know, I would suddenly forget.
I can't speak for others, but I don't suddenly forget the previous turns actions/results constantly turn by turn through a game. Nor do I constantly need to remember each and every non relevant rule each and every turn, and wonder if it applies to situations I know it won't apply to.
While driving that Panzer III down a road, I won't be contemplating details that won't have any application. Do I have any movement points left?, well considering I only moved one hex, and nothing happened as a result, odds are no, I won't have run out of movement points.
Fortunately my brain doesn't require the same question/answer sequence be used during the performance of each and every action. I am not a machine.
Yes sounds arrogant, but after many years, I also know most wargamers that have played board wargames are the same.
I guess it is like my brother being so good with math, that most people can't keep up with him even if they DO have a calculator whilst doing basic math functions. He merely thinks that fast.
There is something that is being assumed here, that, that Panzer III counter is unable to signify details you would already know.
If it wasn't bogged down last turn, odds are you won't suddenly wonder if it is. Especially if it did nothing to allow it.
If it is standing in the middle of a grain field, odds are you won't even be wondering if it will in its next move be bogged down.
If last turn I didn't boresight the hex that is the cross roads, then hey how about that, it won't be next turn either.
Those examples are misleading. They assume details that I would already know, I would suddenly forget.
I can't speak for others, but I don't suddenly forget the previous turns actions/results constantly turn by turn through a game. Nor do I constantly need to remember each and every non relevant rule each and every turn, and wonder if it applies to situations I know it won't apply to.
While driving that Panzer III down a road, I won't be contemplating details that won't have any application. Do I have any movement points left?, well considering I only moved one hex, and nothing happened as a result, odds are no, I won't have run out of movement points.
Fortunately my brain doesn't require the same question/answer sequence be used during the performance of each and every action. I am not a machine.
I LIKE that my life bothers them,
Why should I be the only one bothered by it eh.
Why should I be the only one bothered by it eh.
As I recall, those maps and counters were not final. If CWiF1 had ever been published it would have had greater quality counters and maps (not that Chris' were that bad).wfzimmerman wrote:A few quick comments.
1) I am quite fond of CWIF1 and would like to see as much as possible carried forward into the production version. IN particular, don't mess with the map or the counters. They are beautiful!
2) I cannot emphasize enough what a debt we all owe Chris Marinacci.
3) PBEM and net are priorities. AI is secondary.
4) I realize that Matrix will not want to give schedule projections, but can experienced observers of Matrix make a guess as to when we may Matrix CWIF?
On the rest I agree with you. I don't think can guess at CWiF2's publication date at this time. I doubt it's even in the first stage of development.
Marc aka Caran...
Marc aka Caran... ministerialis
Days of Decision
Another consoideration is the inclusion of the political wrangling that preceeded WWII. I always find it vaguely unsatisfying that I have foreknowledge of the inevitability of Russian and American involvement in the final settlement of the 1939 Polish border dispute.
Anyway, I'd like to understand how the diplomatic model will work in the war. At this stage I presume it'll be a straight copy of WiF with the a few options for Turkey/Spain/etc to be involved.
Anyway, I'd like to understand how the diplomatic model will work in the war. At this stage I presume it'll be a straight copy of WiF with the a few options for Turkey/Spain/etc to be involved.
/Greyshaft
CWiF thoughts and AI ideas
Hello All. First post. I'm glad to see that CWiF may finally happen. I'd like to be more confident but long suffering followers of this saga will probably forgive my caution. I bought a computer in 1998 in anticipation of its imminent release. Congratulations to Matrix for taking it on. Just don't screw it up I enjoyed playing Chris' beta and would like to thank him for his efforts. I didn't mind the interface and liked the maps and counters.
While playing CWiF1, I thought about how the game would be best adapted to incorporporate the advantages that computers bring. Ultimately I decided that as little change as possible would be best. As has been pointed out here, this system has been honed over many years by a huge and knowledgeable pool of gamers. We have been waiting forever for this sucker and I think it would be best to implement something that works. Only change stuff that prevents computerization or to facilitate a radical improvement in gameplay. Its corps-level chunkiness makes sense over the long haul and global scale.
I'm one of the people who wants an AI and it seems to me that there may be two approaches. Both would combine probability calculations for the unknowable but estimateable bits.
The top-down method....
1. Analyse possible strategic goals given the current situation and choose/modify the best one/s
2. Work out the most efficient operational/political/production sub goals to accomplish 1
3. Work out the best tactical moves to accomplish 2
4. Go to 1
The current situation could involve initial setup.
The bottom-up method...
1. From the current situation, choose moves/attacks that are likely to maximize tactical success.
2. Make operational decisions based on the results of 1 (or perhaps the probabilities of results). Produce units/choose political options that are likely to be useful for these plans.
3. Modify strategic goals according to 2.
4. goto 1
I think humans do a bit of both when they play WiF. Perhaps combining them is possible. The bottom-up approach is more like how a chess computer plays and given the chess-like flow of WiF, that may be a salient. In fact, chess computers use a variant of the minimax algorithm. They find their best move, look at your best countermove, find their best counter-countermove etc. I think something similar may be beneficial in the algorithms above.
I'm certainly no expert on this stuff and I know CWiF is a lot more complex than chess. I just have a sneaking suspicion that the impulse/turn mechanics of WiF lend themselves to an AI implementation. I might also be completely wrong and would welcome any feedback. How about a neural network based AI that learns by playing dozens of WiF experts? Just a thought
I enjoyed playing Chris' beta and would like to thank him for his efforts. I didn't mind the interface and liked the maps and counters. While playing CWiF1, I thought about how the game would be best adapted to incorporporate the advantages that computers bring. Ultimately I decided that as little change as possible would be best. As has been pointed out here, this system has been honed over many years by a huge and knowledgeable pool of gamers. We have been waiting forever for this sucker and I think it would be best to implement something that works. Only change stuff that prevents computerization or to facilitate a radical improvement in gameplay. Its corps-level chunkiness makes sense over the long haul and global scale.
I'm one of the people who wants an AI and it seems to me that there may be two approaches. Both would combine probability calculations for the unknowable but estimateable bits.
The top-down method....
1. Analyse possible strategic goals given the current situation and choose/modify the best one/s
2. Work out the most efficient operational/political/production sub goals to accomplish 1
3. Work out the best tactical moves to accomplish 2
4. Go to 1
The current situation could involve initial setup.
The bottom-up method...
1. From the current situation, choose moves/attacks that are likely to maximize tactical success.
2. Make operational decisions based on the results of 1 (or perhaps the probabilities of results). Produce units/choose political options that are likely to be useful for these plans.
3. Modify strategic goals according to 2.
4. goto 1
I think humans do a bit of both when they play WiF. Perhaps combining them is possible. The bottom-up approach is more like how a chess computer plays and given the chess-like flow of WiF, that may be a salient. In fact, chess computers use a variant of the minimax algorithm. They find their best move, look at your best countermove, find their best counter-countermove etc. I think something similar may be beneficial in the algorithms above.
I'm certainly no expert on this stuff and I know CWiF is a lot more complex than chess. I just have a sneaking suspicion that the impulse/turn mechanics of WiF lend themselves to an AI implementation. I might also be completely wrong and would welcome any feedback. How about a neural network based AI that learns by playing dozens of WiF experts? Just a thought
Cheers, Neilster
CWiF AI
This was from: "incywif" at the Yahoo wifdiscussion
>If it was up to me, I'd go with a limited (15-30) number of agents
>handling different types of decision processes.
>The agents would interact in no particular order, but more downwards
>than upwards (i.e delegation of information-gathering, analysis, sub-
>tasks, etc would be common). Different agents would operate over
>different timeframes, so that longer-term goals and strategies are
>more achievable. A number of state machines should also be
>maintained.
>Extensive use of fuzzy logic provides an easy and reliable way of
>modeling logic that handles many interacting events with known
>probabilities. It also does away witn many predictability issues.
>That said, each agent should be modeled using whatever technique or
>strategy that is best suited for implementing the given agents
>responsibilities.
>Ingebrigt
Are you looking for staff Matrix? This person sounds like they know what they're on about. If not, feel free to shamelessly rip off this idea. After all, it wasn't mine and I want be beaten up by your AI (at least at first).
Also, if you do a computer opponent, could you please implement a highly watchable computer vs computer mode. With a game like CWiF this would be most entertaining and probably educational. Make it so one can jump into the game and take control of a side at any time. I'd also love a flexible replay feature. Cheers...Joey
>If it was up to me, I'd go with a limited (15-30) number of agents
>handling different types of decision processes.
>The agents would interact in no particular order, but more downwards
>than upwards (i.e delegation of information-gathering, analysis, sub-
>tasks, etc would be common). Different agents would operate over
>different timeframes, so that longer-term goals and strategies are
>more achievable. A number of state machines should also be
>maintained.
>Extensive use of fuzzy logic provides an easy and reliable way of
>modeling logic that handles many interacting events with known
>probabilities. It also does away witn many predictability issues.
>That said, each agent should be modeled using whatever technique or
>strategy that is best suited for implementing the given agents
>responsibilities.
>Ingebrigt
Are you looking for staff Matrix? This person sounds like they know what they're on about. If not, feel free to shamelessly rip off this idea. After all, it wasn't mine and I want be beaten up by your AI (at least at first).
Also, if you do a computer opponent, could you please implement a highly watchable computer vs computer mode. With a game like CWiF this would be most entertaining and probably educational. Make it so one can jump into the game and take control of a side at any time. I'd also love a flexible replay feature. Cheers...Joey
Cheers, Neilster
-
Les_the_Sarge_9_1
- Posts: 3943
- Joined: Fri Dec 29, 2000 10:00 am
I have been watching the progress here, and sometimes the comments lose me.
Admittedly I am not nearly as versed with AI design as I would like to be. That, and I have 10 times as much time with A3R as I do with WiF.
But one thing I have learned with wargames. The further you go up the chain, the further you go from tactical, the more important it becomes "just where is the unit when the game starts".
I have been seeing this a lot in Strategic Command. The game always sets up units in EXACTLY the same location. And as a result, players play a campaign a few times, always know that such and such a unit will be exactly here, and thus that combat becomes slide rule perfect.
Picture if you can, if the 2 sub counters in the atlantic could be placed where ever the German player wished them to go. Currently, the allied player simply sails sufficient shipping to the spots the subs either were, or could get to with easy prior knowledge. And 9.9 times out of 10, sinks them all by turn 2.
If those subs are around by turn 3, either the allied player is using a bold strategy with his own ships, and the subs are just not interesting enough, or your are playing a clumsy allied player.
But if those subs were located somewhere elese, yes it might take a couple of turns more to find them. And in grand strategy, those turns means something.
And this applies to every unit.
If the game allows for random set up option, then it will be a great deal harder to have the game get mired in must do turn one moves that readily and easily make a game dull.
Admittedly I am not nearly as versed with AI design as I would like to be. That, and I have 10 times as much time with A3R as I do with WiF.
But one thing I have learned with wargames. The further you go up the chain, the further you go from tactical, the more important it becomes "just where is the unit when the game starts".
I have been seeing this a lot in Strategic Command. The game always sets up units in EXACTLY the same location. And as a result, players play a campaign a few times, always know that such and such a unit will be exactly here, and thus that combat becomes slide rule perfect.
Picture if you can, if the 2 sub counters in the atlantic could be placed where ever the German player wished them to go. Currently, the allied player simply sails sufficient shipping to the spots the subs either were, or could get to with easy prior knowledge. And 9.9 times out of 10, sinks them all by turn 2.
If those subs are around by turn 3, either the allied player is using a bold strategy with his own ships, and the subs are just not interesting enough, or your are playing a clumsy allied player.
But if those subs were located somewhere elese, yes it might take a couple of turns more to find them. And in grand strategy, those turns means something.
And this applies to every unit.
If the game allows for random set up option, then it will be a great deal harder to have the game get mired in must do turn one moves that readily and easily make a game dull.
I LIKE that my life bothers them,
Why should I be the only one bothered by it eh.
Why should I be the only one bothered by it eh.
Random Nirvana
Les:
Ahhh... the Holy Grail... a truly unpredictable wargame. I dislike the hindsight we are blessed with in WWII strategic games.
* Stalin thought Hitler wanted Peace with Russia. We all know he didn't.
* Hitler thought his Panzers were unbeatable. We all know they weren't.
* Some scientists didn't think an atom bomb was possible. We all know it was.
Imagine a WWII game where airpower revealed itself to absolutely worthless and the Italian frogmen were the ultimate war winning weapon. We'll never see it because all players want a "fair" chance of winning which implies that the game must model history. However we can be random as Les has suggested. Random deployments. Random Resource Points... up or down by (say) 10% per turn.
I dislike games which are suceptible to forming perfect plans. They ruin the long term playability of a game.
Ahhh... the Holy Grail... a truly unpredictable wargame. I dislike the hindsight we are blessed with in WWII strategic games.
* Stalin thought Hitler wanted Peace with Russia. We all know he didn't.
* Hitler thought his Panzers were unbeatable. We all know they weren't.
* Some scientists didn't think an atom bomb was possible. We all know it was.
Imagine a WWII game where airpower revealed itself to absolutely worthless and the Italian frogmen were the ultimate war winning weapon. We'll never see it because all players want a "fair" chance of winning which implies that the game must model history. However we can be random as Les has suggested. Random deployments. Random Resource Points... up or down by (say) 10% per turn.
I dislike games which are suceptible to forming perfect plans. They ruin the long term playability of a game.
/Greyshaft
-
Les_the_Sarge_9_1
- Posts: 3943
- Joined: Fri Dec 29, 2000 10:00 am
One aspect I liked about WiF was how the military units were even not entirely set in stone as well.
You knew what the unit was, "sort of", but it was an unknown till it was employed.
What I want in a grand strategy game, is the same conditions, but not an identical starting point set in stone.
I don't wish to meddle to deeply though, I expect the Maginot line to be there for instance. But I want to know, if I attack Russia straight off in say 1940 after a stunning win against the French, will it make a difference in the course of the war.
Can I ignore the delay Yugoslavia would create and not attack them.
A perfect game makes me feel like I am fighting WW2, but at the same time not reading a text book.
Some historical events are important, while others while actually significant, occurred only thrtough random chance.
Much is said about the intrusive nature of the "Russian first winter".
But when was the last time you played as Allies and were told that a few weeks after you do your "Overlord" your supply will be drastically punished thanks to the Channel storm the destroys one Mulberry and bashes the heck out of the other.
Any that have played Overlord based wargames knows that those Mulberrys are at risk eventually.
The storm really happened, and the Mulberry's really were severely damaged too.
The hardest aspect though of grand strategy games, is the politics involved.
Countries are allowed to make war, but I have not seen much in the way of games that allow peace treaties to be signed.
A3R has provisions by which the British and the Russians both have conditions that if met will force them out of the game. They also have provisions that will allow them to re enter the fight.
You knew what the unit was, "sort of", but it was an unknown till it was employed.
What I want in a grand strategy game, is the same conditions, but not an identical starting point set in stone.
I don't wish to meddle to deeply though, I expect the Maginot line to be there for instance. But I want to know, if I attack Russia straight off in say 1940 after a stunning win against the French, will it make a difference in the course of the war.
Can I ignore the delay Yugoslavia would create and not attack them.
A perfect game makes me feel like I am fighting WW2, but at the same time not reading a text book.
Some historical events are important, while others while actually significant, occurred only thrtough random chance.
Much is said about the intrusive nature of the "Russian first winter".
But when was the last time you played as Allies and were told that a few weeks after you do your "Overlord" your supply will be drastically punished thanks to the Channel storm the destroys one Mulberry and bashes the heck out of the other.
Any that have played Overlord based wargames knows that those Mulberrys are at risk eventually.
The storm really happened, and the Mulberry's really were severely damaged too.
The hardest aspect though of grand strategy games, is the politics involved.
Countries are allowed to make war, but I have not seen much in the way of games that allow peace treaties to be signed.
A3R has provisions by which the British and the Russians both have conditions that if met will force them out of the game. They also have provisions that will allow them to re enter the fight.
I LIKE that my life bothers them,
Why should I be the only one bothered by it eh.
Why should I be the only one bothered by it eh.
Les the Sarge 9-1 wrote: If the game allows for random set up option, then it will be a great deal harder to have the game get mired in must do turn one moves that readily and easily make a game dull.
Les,
This is an example of taking advantage of computers to make a better wargame.
Glad you have seen the light
-
Les_the_Sarge_9_1
- Posts: 3943
- Joined: Fri Dec 29, 2000 10:00 am
Well remember Joe, computers don't always employ their capabilties hehe. That sub example in a slightly earlier post, was from Strategic Command, an admittedly great game, and a computer game too, but it highlights, that computers can sometimes have weaknesses board games don't have.
The starting set up for the Polish in A3R is to some extent obvious if you wish to max out known potentials, but the player still gets to do set up as they see fit each and every time.
Whereas, in SC, the Poles are initially set up more or less as they were historically, but it is the exact same set up each time, and is thus well established, and entirely predictable.
Employing a random element would make that set up not so predictable.
For all its known limitations shortcomings, Third Reich PC still makes the player set up their initial forces just as it is done in the board game.
That said, cWiF will want to preserve that element. As it is obviously possible to make it an option in a computer program.
The starting set up for the Polish in A3R is to some extent obvious if you wish to max out known potentials, but the player still gets to do set up as they see fit each and every time.
Whereas, in SC, the Poles are initially set up more or less as they were historically, but it is the exact same set up each time, and is thus well established, and entirely predictable.
Employing a random element would make that set up not so predictable.
For all its known limitations shortcomings, Third Reich PC still makes the player set up their initial forces just as it is done in the board game.
That said, cWiF will want to preserve that element. As it is obviously possible to make it an option in a computer program.
I LIKE that my life bothers them,
Why should I be the only one bothered by it eh.
Why should I be the only one bothered by it eh.
LoL, the author of the above would be me. Matrix can of course shamelessly rip off anything of the above, nothing there a decent AI programmer wouldn't know about! My reply was more aimed at the "decent-AI-is-impossible"-crowd over at the WiF-list, and is not a serious try at a masterplan for a new, brilliant CWiF AINeilster wrote:This was from: "incywif" at the Yahoo wifdiscussion
>If it was up to me, I'd go with a limited (15-30) number of agents
>handling different types of decision processes.
>The agents would interact in no particular order, but more downwards
>than upwards (i.e delegation of information-gathering, analysis, sub-
>tasks, etc would be common). Different agents would operate over
>different timeframes, so that longer-term goals and strategies are
>more achievable. A number of state machines should also be
>maintained.
>Extensive use of fuzzy logic provides an easy and reliable way of
>modeling logic that handles many interacting events with known
>probabilities. It also does away witn many predictability issues.
>That said, each agent should be modeled using whatever technique or
>strategy that is best suited for implementing the given agents
>responsibilities.
>Ingebrigt
Are you looking for staff Matrix? This person sounds like they know what they're on about. If not, feel free to shamelessly rip off this idea. After all, it wasn't mine and I want be beaten up by your AI (at least at first).
That said, I do have a background as an AI developer, was a CWiF1 beta playtester, and have played WiF for about 15 yrs.
I certainly wouldn't mind getting involved in the CWiF AI, even (or especially) if it just amounted to an occational chat with the matrix AI developers and a look at their ideas.
cheers,
Ingebrigt