Is CMSF Fundamentally flawed beyond patching???

[MarkEzra]

Originally posted by Rollstoy:

Without caring too much about QBs, I cannot help but make the following observation:

The reason, why QB troop selection is so restricted is, that the OOB is very important in CM:SF, right?

Now, especially playing Veteran, because Elite is too much clicking for me, I fail to see any significant influence of the OOB on gameplay at all! Heck, I do not think I ever paid any attention on how my companies are organized, I just spread them all over the place, without any penalty that I am aware of.

I wonder, if others perceive this differently?

Best regards,

Thomm

Point well taken. If there is a difference in C&C I don't get a sense of it. Build, play and test daily....

[MarkEzra]

Sirroco said "The random maps I think are gone because of the greater complexity of the AI plans. But if enough are generated I don't see a problem with it. You don't like the map, close it and start again, as with QB's in CMx1 now. I think it has to ship with enough maps for it to work out of the box, though, and not rely specifically on the community to make it work."

You have it exactly right....When people say "Yuck" to user made maps...I no longer start to sniffle and sob...It's likely they have never seen one...good bad or indifferent. It's likely they are comparing to CMx1 or some other game they have played. This is an outstanding editor...best I know of. There are now over 200 maps available ( www.CMMODS.com ) for DL ready and designed specifically for QB play. Also ANY scen map can be used as a QB....before you try this at home, kids, be sure those setup zones are big enough to handle the Battalion you plan to use!

[acrashb]

Originally posted by Paper Tiger:

Have you any idea how CRAP a random map would be?

[...]

What about it's AI plan?

[...]

your remarks about the Strategic AI are completely wrong. Since you don't have the game

1) only as crap as the programming; there's no evidence that it need be crap.

2) that's part of the point - having to hand-develop an AI plan in some detail is undesirable - but certainly a map generator could also develop an AI plan if that's what it takes.

3) true, I only have the demo. Having said that, between studying the demo and reading here, it's easy to have a lousy AI plan and game it. It looks like some very well-made scenarios act OK from the strategic perspective, but a lot of them don't. The fact that the game requires an AI plan, rather than being enhanced by them, strikes me as "fundamentally flawed" because of the ripple effect on QB's.

Those who play exclusively HTH won't be affected.

[Sirocco]

Random maps for QB solo play in CMx1 were useable, as long as you attacked. You just didn't get a decent game out of defence, unless you enjoyed mowing down mortar crews leading human waves.

If a random AI plan was created along with random maps either BFC would have to spend an age coding it, at the expense of other features, or you'd end up with the situation above.

[Hev]

What we needed (in my humble opinion) was something in the middle of old and new. The exact system that was used for selecting QB's in CMx1 with no random maps providing you could pick a named map to play on.

[Paper Tiger]

acrashb: I'm not going to get drawn into a negative debate with you about this. If you want to riposte what I'm about to post, fine but I'm done.

quote:

"1) only as crap as the programming; there's no evidence that it need be crap."

Have you any idea how long that would take to programme? Let's be realistic here: Battlefront need to have some income to survive. To hold back publication of a title for 6 months+ to develop a frankly unnecessary and massively complex piece of programming is unrealistic. You REALLY need to look at the scenario editor to see how complex it is to qualify your opinion here. Lets see if anybody else jumps in here to back you up on this one.

quote:

"2) that's part of the point - having to hand-develop an AI plan in some detail is undesirable - but certainly a map generator could also develop an AI plan if that's what it takes."

undesirable for you maybe but not for a lot of us. I actually find it an exciting step forward with plenty of potential for more sophistication. And you want to add to this random map generator a planning generator too? No doubt, it would have to be a good plan to to keep you happy. Whew, nobody's going to accuse you of having low expectations. Good luck getting support for that too.

quote

"3) true, I only have the demo. Having said that, between studying the demo and reading here, it's easy to have a lousy AI plan and game it. It looks like some very well-made scenarios act OK from the strategic perspective, but a lot of them don't. The fact that the game requires an AI plan, rather than being enhanced by them, strikes me as "fundamentally flawed" because of the ripple effect on QB's."

Would you care to qualify your statement that a lot of them don't? Which ones? And if some well designed scenarios do work as you admit, then your point is invalid. You're just floundering here. You said at the end of your first post that you would welcome factual correction because you probably needed it. I don't think you want to be corrected at all.

[Guest Guest]

deleted per user request

[Sirocco]

I'm as big a fan of QB's in CMx1 as anyone here, but I can see how dropping random maps makes sense, from a technical standpoint. Again I'd qualify it with the caveat that the game should have shipped with a lot of pre-generated maps.

[MarkEzra]

AB: You said "quote

"3) true, I only have the demo. Having said that, between studying the demo and reading here, it's easy to have a lousy AI plan and game it. It looks like some very well-made scenarios act OK from the strategic perspective, but a lot of them don't. The fact that the game requires an AI plan, rather than being enhanced by them, strikes me as "fundamentally flawed" because of the ripple effect on QB's."

That you don't like player directed AI pathing is a personal preference. I don't pretend to understand your view. You need to be aware that the AI for CM:SF QB is the same found throughout any CMx1 game: QB or Scen. I play CM:SF QB very often and see the same wonderfully exciting and plain dirt stupid AI that is the center piece of the CMx1 franchise. Doesn't mean I don't love it old or new. It is what it is. I love the new AI for scen. and consider it a huge step forward for CM.

[Guest Guest]

deleted per user request

[MarkEzra]

Originally posted by Adam1:

A huge step forward from CM's original AI it is. But it needs a lot more player control to work. Right now it's in this middleground where it's not intelligent enough to work without player scripting (which is ok) but there aren't the tools to script. I'm curious what Aaron thinks about the "AI" in Steel Beasts, which has a lot of scripting tools but won't do anything unless you write the doctrine for each scenario. And what is an example of a good wargame AI? I respect AB and his opinions so I'm not asking this rhetorically.

Couldn't agree more. I will definitley want to put my views forward. But for now I'm very pleased

[CaptainBly]

Originally posted by MarkEzra:

Hev's got it right...To hear some folks talk you'd think that BFC's looking for a fast buck and faster exit. To hear some folks talk you'd think that good wargaming companies are on every street corner. To hear some folks talk you'd think that the world is just full of people begging to play pixel war. To hear some folks...why it's enough to make you go deaf.

You sound like a lot of some folks. hahahaha

And what is an example of a good wargame AI?

A good wargame AI is one that plays to WIN not to just delay how long it takes the human player to win.

What we needed (in my humble opinion) was something in the middle of old and new. The exact system that was used for selecting QB's in CMx1 with no random maps providing you could pick a named map to play on.

Sorry, but, I totally disagree with this. No random map generator = no buyee for me an my wargaming group. We'll just play CMAK or CM-ETO instead of shell out good money for less than we've gotten before. This module business is just money grubbing anyways. Maybe they'll get it right by the time they release the WWII stuff.

[ November 29, 2007, 09:10 AM: Message edited by: CaptainBly ]

[Chelco]

The AI editing tools in SBProPE are far superior than the ones in CMSF.

The implementation of triggers in CMSF's AI tools is sorely absent. For the scenario designer, there is no option to change the AI plans in the middle of the game because of an specific player’s action. I am growing more and more distant to this game due to the non-reactive, follow-waypoint-to-death AI opponent.

There is no doubt that making a good wargame AI is a great endeavor. It can be done, though (have you guys tried any of Panther Games?). Lately, I don’t wish developers would make a super-smart computer opponent, I just wish they would make a “credible” one. One computer opponent that fears the bullets of my grunts (don’t enter a kill zone if he sees his mates being ripped apart), one that fears his retreat routes being cut (reaction to turning movements), etc. I would be more than more than happy if CMSF’s AI editor would allow me to (even painstakingly) script that type of behavior.

The best tactical (company sized engaments) AI in a wargame is in the HPS' Squad Battles series. Pity those games are IGOUGO.

[helm123]

Originally posted by Chelco:

The AI editing tools in SBProPE are far superior than the ones in CMSF.

The implementation of triggers in CMSF's AI tools is sorely absent. For the scenario designer, there is no option to change the AI plans in the middle of the game because of an specific player’s action. I am growing more and more distant to this game due to the non-reactive, follow-waypoint-to-death AI opponent.

There is no doubt that making a good wargame AI is a great endeavor. It can be done, though (have you guys tried any of Panther Games?). Lately, I don’t wish developers would make a super-smart computer opponent, I just wish they would make a “credible” one. One computer opponent that fears the bullets of my grunts (don’t enter a kill zone if he sees his mates being ripped apart), one that fears his retreat routes being cut (reaction to turning movements), etc. I would be more than more than happy if CMSF’s AI editor would allow me to (even painstakingly) script that type of behavior.

The best tactical (company sized engaments) AI in a wargame is in the HPS' Squad Battles series. Pity those games are IGOUGO.

Follow waypoints to your death is the best description anyone has put out. This is the exact experience I had with the demo and is exactly why am not a owner of the game.

The second scenario in the demo gives you a great feeling for the scripted waypoints that lead to units following them to the death.

[thewood]

I think the scenario editor is a great step forward, but have to agree that lack of triggers of some kind really limits it.

I also don't think squad battles is a great example of AI. None of the HPS SQB series have good AI.

[Webwing]

Chelco,

Originally posted by Chelco:

The AI editing tools in SBProPE are far superior than the ones in CMSF.

I don't know about the ProPE version but I also really like the SB(1) editor. Long time I don't use it but I used to like it a lot.

I wouldn't say that it is far superior to the CMSF editor though.

The idea behind the CMSF editor is to be user friendly and to leave more decisions to the TacAI and less burden on the designer. SB for instance doesn't hava a TacAI, therefore the need for triggers, etc.

I think they achieve a pretty decent compromise between a user friendly interface and some powerful features. But again players only, think it is too complex. Designers that it lacks features.

The implementation of triggers in CMSF's AI tools is sorely absent.

I agree.

But then again you fall in the issue I mentioned early. Not many people want to spend their time designing missions.

Although they actually have the time trigger. They don't have conditionals.

(have you guys tried any of Panther Games?

Airborne Assault is one of my favorite war games ever!

Can't compare though. Apples and oranges.

-

[ November 29, 2007, 10:02 AM: Message edited by: Webwing ]

[Webwing]

Originally posted by helm123:

The second scenario in the demo gives you a great feeling for the scripted waypoints that lead to units following them to the death.

I think sometimes there is a misunderstanding about this whole issue.

"This is too scripted" or the other side "There is no scripting/triggers etc. in the editor"

A heavily scripted scenario can be a good thing. If done properly. You can ensure a lot of replayability value by including random events and triggers. BUT, and it's a big BUT, It takes ages just to do one mission but it sure can seem like you are against a very clever opponent. If its not done properly it will feel robotic and... 'scripted'. Also you rely on the designer alone. A game engine that does not have a TacAI can't live without scripts, triggers and conditionals.

With a TacAI in place one can only give general directions to the computer opponent and leave it to take the rest of the decisions.

If one starts adding specific waypoints and triggers and conditionals, the TacAI becomes redundant.

Just an example:

Without TacAI

Got to point A. If you see no enemy go to B. If you see the enemy start fighting.

With Tac AI

Got to area B. Here the TacAI should know that if on the way to B they find an enemy they should start to shoot. Or retreat, or panic... depending on the moral, training, tiredness of the troops.

I'm not saying one is necessarily better than the other.

It would be possible to more features and options to be used if the designer decided it was needed in specific situations. That would be the best of both worlds.

--

[ November 29, 2007, 11:13 AM: Message edited by: Webwing ]

[Chelco]

Hi Webwing,

Originally posted by Webwing:

[QB] The idea behind the CMSF editor is to be user friendly and to leave more decisions to the TacAI and less burden on the designer. SB for instance doesn't hava a TacAI, therefore the need for triggers, etc.

You make a crucial point.

What's your take of CMSF's TacAI?

What it can do and what it can't do?

I could be wrong, but it seems to me like once the TacAI orders a group to move down to a waypoint, it will assign a path to it and that's it. The poor AI soldiers assigned to this path and waypoint wilco no matter what. There is no re-assesment or re-planning of the path assigned.

You know what I mean because I am sure you have witnessed Panther's AI (apples and oranges acknowledged, but still) to attack you in one direction and after suffering losses it will pull out and attack you from another direction. Even when the AI is trying to move into the same piece of terrain during the whole scenario.

I could live with a waypoint that lasts the whole game because the lack of triggers, what I cannot live is with an apparent AI path created at the beginning of an scenario that lasts until the very end. I don't know how a scenario creator is supposed to work around this.

[Webwing]

Chelco,

I'm doing a lot of testing to find out exactly this. I need to know what the TacAI can do to be able do take advantage of it while designing the missions. It might be a waste of time since 1.05 is around the corner and might change a lot of things.

What I've seen so far is a mixed bag. Some very clever things and some pretty silly ones. Some silly ones are bugs that must be fixed. Once they are fixed the TacAI is pretty decent.

One thing I found out is that if you treat the Orders as waypoints you are putting the AI in a straitjacket. For instance you assign a squad to a single square in the grid. Next order to another one. Ok, you have control, but you leave no option of manouver to the AI. It would be best to paint a general area and let the AI decide the best spot to be, where and when to move. That's the whole idea behind painting areas as opposed to points. This is new for me. I'm trying to learn to take advantage of this system.

Right now I'm finding hard to create an effective attack. A lot easier to prepare a good defence. This is in part due to my inexperience with the system but also to some bugs like for instance the soldiers won't go to the roof although there is an especific order for that.

Once we pass the stage of fixing the major bugs then I'll be able to tell you for sure how good the TacAI is.

--

[dalem]

Originally posted by Paper Tiger:

Sirocco is right. Steve has already stated that players will be able to cherry pick forces for WW2 just like in CM1.

Link please.

I can understand Steve's reasoning for not having it in the modern era but nevertheless, the existing system still gives you some very bizarre forces, especially if you pick smaller battles. But there's no reason to believe that this won't be changed later.

I really don't understand why people are complaining so much about the lack of random maps though. (Yes, acrashb, I followed your link and read your web site) Have you guys played around with the scenario editor? Have you any idea how CRAP a random map would be? What about it's AI plan? Never mind the flavour objects. Just try randomly throwing a few objects onto a board in the editor, raise and lower the elevations randomly, stick a wall here, a palm tree there, a mosque there. What an enormous pile of sh@te! If Battlefront gave us randomly generated maps , you'd be complaining about how crap they were and shouting angrily in this forum for them to fix this 'broken' feature. It's not like you have to make them yourselves, there are plenty available to be downloaded free of charge because some people actually enjoy designing them.

That's a cop-out. Maps are nothing more than a layered matrix of numbers. A few smoothing algorithms and you have a not-insane surface and collection of surface objects. Just because BFC decided they didn't have the time or resources to do it doesn't mean it can't be done.

-dale

[dalem]

Originally posted by MarkEzra:

Sirroco said "The random maps I think are gone because of the greater complexity of the AI plans. But if enough are generated I don't see a problem with it. You don't like the map, close it and start again, as with QB's in CMx1 now. I think it has to ship with enough maps for it to work out of the box, though, and not rely specifically on the community to make it work."

You have it exactly right....When people say "Yuck" to user made maps...I no longer start to sniffle and sob...It's likely they have never seen one...good bad or indifferent. It's likely they are comparing to CMx1 or some other game they have played. This is an outstanding editor...best I know of. There are now over 200 maps available ( www.CMMODS.com ) for DL ready and designed specifically for QB play. Also ANY scen map can be used as a QB....before you try this at home, kids, be sure those setup zones are big enough to handle the Battalion you plan to use!

Don't make assumptions about what some of us expect out of a wargame.

-dale

[thewood]

One thing I agree with Dale on is that most of the lacking features we expected in CMSF were more than likely left out due to time constraints. It is possible to do them, but BFC found itself in the position of having make a very short list of things to include and hope there would be time later to add in. A few may just require too much effort at this point, but to retain or recover interest in CMSF, I hope they can get a lot in. Many of those "left out" things Steve admitted to and I am hoping this time they are taking is to do it right.

Our enemy in CMSF is not Syria, but lack of time due to the Paradox committment.

[StellarRat]

No software is flawed beyond patching. The question is only how much time and money you have to fix it and how big and numerous the patches will be.

[Guest]

I dunno, Mark. It's not so much that I say "yuck" to user-made maps (except my own, which are invariably workmanlike at best), but that I would say "whee!" to having random maps. I loves me some random maps.

In fact, part of my problem with creating computer wargames as a hobby is that I just can't NOT write a map generator. It is a pain, and time-consuming, and sometimes it's where I just throw up my hands and say "non!" and scrap the whole damn project (in fact I tend to evaluate the doability of projects by how easy it would be to write a map gen for them), but always worth it when I manage to finish them.

Seriously, if I were to write a Napoleonic wargame (something I was writing up the specs for the other night... we'll see)... I'd create a random map generator for it.

It's a sickness. Possibly an addiction?

I just like the idea that, with one push of a button, I can be playing on terrain I've never seen before, with new challenges laid before me. Mmmm. Warm and fuzzy.

Edit: Hey thewood. Honestly, I don't think BFC has any intention of spending money to "recover" CM:SF this late in the game. They're trying to fix the engine (and perhaps their reputation?) in time for their grogbase to accept CM:Marines, which I'll note maintains the same modern bent that not everyone is looking for. CM:SF improvements are a side benefit I think.

[Webwing]

Originally posted by Phillip Culliton:

I dunno, Mark. It's not so much that I say "yuck" to user-made maps (except my own, which are invariably workmanlike at best), but that I would say "whee!" to having random maps. I loves me some random maps.

I can understand that and there are many reasons to like QB.

But take a look at a user made mission like Babado. You open the mission and while it's pause look at the map. Not much to it. How long did it take to put a few walls, a road a few one storie houses? But man, what a battle!!! Why? There is nothing random there. What took ThePhantom A LOT of time, and thinking was WHERE to put the walls, etc. Where and what forces to have. At what time should they be in place. The tools are simple to use. The hard thing is how to use it to have a great effect. He is still tweaking the mission! No way you can get that with a random generator.

Like the quote: "Writing is easy. You start with a capital letter and finish with a period. In the middle you put ideas."

--

Seriously, if I were to write a Napoleonic wargame...

Have you checked Napoleon's Campaigns??? IMO best Napoleonic era wargame ever!!!! Quite a few Phillips involved in that game.

--

[ November 29, 2007, 12:30 PM: Message edited by: Webwing ]