hoolaman Posted January 19, 2009 Share Posted January 19, 2009 I've been thinking about this and I think there is an unfortunate neccesity for scenario authors to script their forces based on exactly where you come from and where you very likely will go to. In several scenarios I've seen them run a bit like a movie script with an enemy unit showing up down a side alley and popping off a shot. My hypothesis is that a good scripted AI is cheating based on the scenario deigner's foreknowledge of where the player's forces start, where they are going and what they are expected to achieve. I know the answer to the question is triggers which BFC are keen to introduce in some form but I wanted to see whether anyone has experinced the things I mean and whether there is anyway around it with the tools available. 0 Quote Link to comment Share on other sites More sharing options...
Sergei Posted January 19, 2009 Share Posted January 19, 2009 "Is a pure scripted AI a break of FOW?" Not any more than giving the player information in a briefing is a break of FOW. Obviously the AI cannot read briefings, but a good scripting overcomes this limitation. 0 Quote Link to comment Share on other sites More sharing options...
Other Means Posted January 19, 2009 Share Posted January 19, 2009 Yes as it's taking advantage of knowledge that a human player would not necessarily have. But what ya gonna do? It can't be as clever as people so you've got to give it something. Now if you need to advance your M1s down a street then 20 T90's appear directly behind that's bad design - but such a scenario wouldn't get played. 0 Quote Link to comment Share on other sites More sharing options...
Webwing Posted January 19, 2009 Share Posted January 19, 2009 I think this can be *solved* by the designer if he sees it as a problem. But this is actually too generic. It has to be judged case by case IMO. Triggers are a big plus when designing missions but will bring with them even more "cheats". For instance you can make the AI react when Blue gets to X position even if the AI does not have LOS to the Blue force. Just to give you one example. - 0 Quote Link to comment Share on other sites More sharing options...
MikeyD Posted January 19, 2009 Share Posted January 19, 2009 There's good scenarios and there's bad scenarios. Bad scenarios are no more the fault of the game engine than bad novels are the fault of the printing press. I've been participating in a debate elsewhere about the utility of taking the time to do multpile AI orders sets when building a scenario. My stand is if the flow of action could go in any one of several directions the scenario designer should be obliged to explore those multiple paths. The opposing view appears to be of the opinion that the fight should be shaped in its optimal path, that packaging one 'good' AI set with one 'bad' is a waste of effort and a possible cheat to the customer if he gets the 'bad' AI set to fight. You could say these are the 'movie script' types, the puzzle makers who want there to be just one key to successfully unlock the scenario. 0 Quote Link to comment Share on other sites More sharing options...
Dietrich Posted January 19, 2009 Share Posted January 19, 2009 I agree with you, MikeyD, that if the action could flow in multiple directions, the AI plans should be multiplied accordingly. In historical scenarios, though, it makes sense to have a single AI plan based on the actual event(s). However, multiple AI plans in a historical scenario could yield rather interesting play. 0 Quote Link to comment Share on other sites More sharing options...
MikeyD Posted January 19, 2009 Share Posted January 19, 2009 I've rarely seen an account of a historical tactical situation described in sufficient moment-by-moment detail for it to be faithfully translated into a CM battle. Assumptions get made, maps get rationalized, AI-driven movements have to be accounted for. There are definite limits to one's ability to reproduce history. 'Reenactment' is a very different thing than 'simulation'. Reenactors appear to start to fetishize the event, the military historian's version of knowing all the dialog to the Rocky Horror Picture Show. 0 Quote Link to comment Share on other sites More sharing options...
Secondbrooks Posted January 19, 2009 Share Posted January 19, 2009 Triggers are a big plus when designing missions but will bring with them even more "cheats". For instance you can make the AI react when Blue gets to X position even if the AI does not have LOS to the Blue force. Just to give you one example. - It would seem to be rather simple thing to include condition that Blue in position X needs to be spotted. No? Overall i support triggers. AI in defense sucks a lot as it don't have any kind reactivity to situation. 0 Quote Link to comment Share on other sites More sharing options...
Webwing Posted January 19, 2009 Share Posted January 19, 2009 Don´t get me wrong. I am an editor freak! I prefer fooling around with the editor more than actually playing. All I was trying to say is that any feature you add can be used for "cheating". And the more complicated the editor gets the more it is prone to designers mistakes. Still, I´m all for triggers and whatever else can be made available in the editor! 0 Quote Link to comment Share on other sites More sharing options...
hoolaman Posted January 20, 2009 Author Share Posted January 20, 2009 There's good scenarios and there's bad scenarios. Bad scenarios are no more the fault of the game engine than bad novels are the fault of the printing press. Actually I think the best scenario designers are often themost guilty of this. Well tested and thought through AI will meet the player at every turn with a nasty surprise and seem to pop up in just the right place. I guess as Webwing says triggers aren't always the answer. I know ToW has them and I believe the guys that have designed missions for that game would know. Sadly I do pine for the bumling old CM AI. You knew what they would do much of the time, but they were never capable of reacting to something they couldn't know. Of course a good human opponent will also make an assessment of a CM map that will often tell them things about setup zones and avenues of advance that a real commander couldn't know either. So I guess it comes back to where to draw the line in a game sense to create a decent simulation. 0 Quote Link to comment Share on other sites More sharing options...
Apocal Posted January 20, 2009 Share Posted January 20, 2009 Still, I´m all for triggers and whatever else can be made available in the editor! I find myself wishing I could have the map editor of CMx2 and the open scripting of VBS2 put together in one game. 0 Quote Link to comment Share on other sites More sharing options...
Sergei Posted January 20, 2009 Share Posted January 20, 2009 I've rarely seen an account of a historical tactical situation described in sufficient moment-by-moment detail for it to be faithfully translated into a CM battle. Assumptions get made, maps get rationalized, AI-driven movements have to be accounted for. There are definite limits to one's ability to reproduce history. Besides, there's like a million ways that a historical battle could have gone, and in a tactical wargame all of those possibilities are just as important as the real direction the battle took. Players aren't going to stick strictly to script (especially if they're the historical losing side), and if the situation warrants choosing from more than one plan of action, it makes sense to give the AI more than one plan. And of course the game's 'die rolls' can't be fixed to match the historical 'die rolls', so you get lots of variation. Speaking of historical scenarios, they are the ultimate break of FOW. If you know what happened at Villers-Bocage and you play the CMBO classic "Villers-Bocage Tiger!" as the Brits, are you going to be surprised that there's a lone elite Tiger on rampage? 0 Quote Link to comment Share on other sites More sharing options...
cabal23 Posted January 20, 2009 Share Posted January 20, 2009 I find myself wishing I could have the map editor of CMx2 and the open scripting of VBS2 put together in one game. Wow VBS2...haven't heard that around here before. Before this game I was so into that. Just waiting on the new public version which will not be called VBS but ARMA 2. I find the current version lacking in many way, but it was never intended as a game, just a training sim. But to have access to all of the units and environments of VBS2 would make CMSF a dream. Think of CMSF as mini me and VBS2 as Dr. Evil. I think I just wet myself thinking about it. If you have never witnessed what VBS2 can do check this link...Wow is all I can say. Wow is it expensive too, but like I said it was never intended as a game and ARMA 2 will have a normal price point and still offer most of the functionality of VBS2 http://virtualbattlespace.vbs2.com/ 0 Quote Link to comment Share on other sites More sharing options...
Dietrich Posted January 20, 2009 Share Posted January 20, 2009 Besides, there's like a million ways that a historical battle could have gone, and in a tactical wargame all of those possibilities are just as important as the real direction the battle took. Players aren't going to stick strictly to script (especially if they're the historical losing side), and if the situation warrants choosing from more than one plan of action, it makes sense to give the AI more than one plan. And of course the game's 'die rolls' can't be fixed to match the historical 'die rolls', so you get lots of variation. Good points, Sergei. =) Speaking of historical scenarios, they are the ultimate break of FOW. If you know what happened at Villers-Bocage and you play the CMBO classic "Villers-Bocage Tiger!" as the Brits, are you going to be surprised that there's a lone elite Tiger on rampage? True, but that doesn't mean the outcome of the battle is a given. Oh how many times I've played CMAK Villers-Bocage scenarios as the Germans and had my elite Tiger ("LT MICHAEL WITTMANN*") knocked out in the first few turns! 0 Quote Link to comment Share on other sites More sharing options...
Secondbrooks Posted January 20, 2009 Share Posted January 20, 2009 Edit: Blah... 0 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.