kevinkin
-
Posts
3,208 -
Joined
-
Last visited
-
Days Won
2
Posts posted by kevinkin
-
-
I guess my question is how much ground needs to be covered and in what time frame?
Kevin
-
element of chance or unpredictability
The editor allows for 5 separate AI plans per battle These are randomly executed depending on the designers settings. The TACAI delivers somewhat
unpredictable fire results and movement. Indirect artillery and air support are not 100% predictable as well. LOS/LOF are difficult to master and while predictable offer a lot of surprises. One thing to try is to play a battle where you only give orders every 3 or so turns and let the AI try to carry out your
intentions. Another thing to try - which can be wild - is to play a battle only from ground level. Take a screenshot of the map at the beginning and try to keep the angle of view down low. (No level 2 buildings). We used to game this way back in CM1x now and then.
Not sure where you would like to see more chance/unpredictability but these are my thoughts.
Kevin
-
When playing vs the AI, wouldn't that screw things up since the AI plans anticipate an initial basic enemy set-up. See Mark's suggestion above about making a new map to cover these situations. I don not think this is a big deal for the player especially if it gives a good fight on a quality map.
BTW,I am making progress today so Thanks for the help guys.
Kevin
-
Thanks ... each posting is sheading light on the system. Looks like the TAC is very active to a point where it may trump the AI plans.
I still have no clue why the AI waits till the 7th minute every other attempt. It must be related to (like you mention) map size and tactical
situation. But in any event the troops are moving well and I have something to work with over the next week or so. I am noticing a lot - almost all - assaults and attacks are for the Soviets - this make sense given the offensive being simulated. I may add a few German attacks for city, village, forest/hills and open battles to the list of things to work on.
Kevin
-
The nature of the project is described in the post "Quick Battle replay" and has received a small but positive response.
It involves adding new AI plans to the existing QB maps for replay value. Since the maps are already very nice I thought to spend most time
on the AI and crank out a dozen maps with 3 new AI plans each at a time for posting. But this has slowed down due to my rookie understanding of the on the inner mechanics of the QB system.
Thanks for the help. My next question is I have a set-up zone set for a move between 0 and 3:00. Half the time they wait till the 7th minute to begin.
The otherwise they seem OK. Its as if the TACAi is taking over. I never saw this in regular scenario design so if you or someone can help it would be great. BTW, order 2 is right under the nose of the start up zone. All plans are turned off and other than number 1.
Kevin
-
Mark
Why are the set up zones ignored? What tells the AI always attack from the east regardless of nationality? If I select an Axis attack then they set up in Soviet zones. If the user goes this route can the orders and objectives be ready to provide the right battle even if the battle is not set is as specified? On this type of map should German attackers and Russian defenders not be recommended?
Sorry for the questions - a have a nice project that depends of further understanding of the system.
Kevin
-
Like express above, I have no interest in 1945. Doctrine was established and the war over. If we could go hypothetical with a often mentioned scenario of the Germans surrendering to the west and attacking with the allies to the east as part of the 45 product I would be happier. Non-nuke skirmishes to decide the border between east and west. I think starting in 44 was a good idea in that the new engine is getting a great test drive and the 41-43 add-ons better for it.
Kevin
-
I noticed that the QB Light Forest-Farmland Probe 001.btt describes all attacker set-ups are East and Defender are West.
As I try to modify the map AI I noticed this to be true. The AI ignores the owner of the zone and will place (in my case) the German attackers in the eastern Soviet zone and the Soviets defenders in the west German Zone.
How can this behavior be turned off? Right now the Germans attack from the East as if cut off. I guess the bigger question is how to design QB AI to accommodate the four possible user selections:
German attacker (Human or AI) vs Soviet defender
Soviet attacker (Human or AI) vs German defender
Trying to put together an AI plan for both defense and offense is confusing. If you know that a side will always set-up using your programed zones it would be a little easier. It would be nice to be able to design offensive and defensive operations into separate AI plans so the appropriate plan is called up depending on user selection of sides at the beginning.
If I select a QB and ask for Human Soviet Defender a German Offense AI plan is loaded into the game.
Kevin
PS Once placed the Germans follow the Soviet plan from the Soviet zone and the Soviets the German plan from the German zone if the orders are spaced just right. This is really crazy.
-
Thanks for the thumbs up Mark. I have considered both points and finished one this morning. Here is the new syntax for the file names:
AI Mod Small City Assault 045.btt
Terrain - small changes to south map edge
AI: 3 new plans
Kevin
PS ... Just need a few of Santa's helpers.
-
I would like to see 41-43 since I have little interest in the final battles other than reading about them.
But the list correctly identifies
1 Waffen SS
2 Romanians and Hungarians
3 Partisans
7 Winter weather
BTW some of the coolest small unit battles can be designed from fall/winter 41 aka gates of Moscow.
Kevin
-
Just send a message to kevinkin@comcast.net.
Have to finish the briefings but they are started.
Thanks.
Kevin
-
Thanks for the feedback guys. I will look for more folks to comment. BTW, I mistyped and mean that I find some QBs could use an AI tweak anyway.
Kevin
-
Has anyone edited the games QB maps for the community with respect to the AI plans to extend the replay and surprised given to the player?
I was thinking of teaming up with someone and start with the 30 min smaller maps and post the files renamed for transfer to the QB folder. Perhaps small terrain changes based on the new AI plans would make for better battles. Any thoughts are welcome.
Kevin
PS: I have found some QBs some tweaks in the AI anyway.
-
hm_stanley
These types of scenarios are a bear to test due to their size and length. You might want to consider a practice scenario (1/3 final size) were you can watch the AI attack and estimate the force ratio and length of time for the battle. This information may not hold completely for the full size battle but it will help. You might be able to use the final map but limit the action to 1/3 of the area. You can even see how long it takes for the AI attacker to move through all the defenses with no enemy intervention which can serve as a baseline.
Kevin
-
I am using Panzer Blitz Vassal for several reasons knowing I have to modify the rules for more up to date game play: Its freeware with a lot of free add-ons and rules to be found on the internet. (Panzer Campaigns Minsk is 40 USD). I added simple moral and activation rules. The units fit into the CMRT timeframe and there are add-ons available for different years/fronts. I agree about simplicity knowing layers of complexity (some call it realism) can be added. A rules system that provides a framework for deciding when and where to fight while protecting rear area installations is what I am trying to develop. Results of tactical battles in CM would add unpredictability to the affair and utilize the players CM skills. Adding the right reinforcement schedules can produce operations that last a long time and would make rear area protection all the more important - food, ammo, communications etc, would become targets.
Kevin
-
I think single player against an AI is out. How do you stop the AI from having combat where a CM QB needs to be fought? You would generate double results. Head to Head and Solitaire have a chance. But I need to know better the mechanics of using the OP layer software ie when would a CM QB take place and how would the results be applied at the OP level. Getting down to the nitty gritty and applying CM casualties back into the OP level may be overkill. Rules that use the CM victory levels and perhaps overall CM losses may be smoother to implement.
Kevin
-
It would be easier to envision this knowing what commercial game would provide the operational layer. I think I read PzC for example. Two things: combat has to be resolved in the OP layer and TAC layer each turn. I used to have a copy of Smolensk but forget how this would be achieved. Can you turn combat off and on and decide when the OP layer resolves or when to fire up CM for resolution? All combat cannot be resolved at the TAC level within CM - the game would last forever.
Those types of files mentioned would be interesting to play with and I would like to hear what folks think a OP layer really means.
Kevin
-
76 mm
I agree a computerized OP layer would be ideal based on company scale combat. The program would handle FOW and the book keeping needed to keep the game going.
The RT files mentioned ....
BF would not make these files available to the public I believe. Perhaps to a small design team under NDA. But they been down that route. I think the situation is different than with graphical MODs in that there is a long history with these being released as freeware successfully across differing games. Files that will manipulate gameplay are a different matter.
Again there are two fundamental questions:
- What do we mean by operational?
- Will the efforts result in freeware or payware?
I see a few benefits using Cyberboard plus "old fashion" rules sets as the OP layer
- The software is free. The gameplay is well known at least to the older of us.
- It fits the KISS concept to a tee
The downsides are lack of FOW and keeping track of the counters as they move and engage the enemy. "Darn I forget if that unit has enough action points left to fire?"
After two single board trials switching from PB to RT and back I have to say the system will work - but the devil is in the details. I have been writing the rule changes down as I test. The basic concept is PB platoons represent companies where MPs are 1/2 and the hexes are 500 meters. Turns are 15 mins real time - about 2x of PB. Combat is from adjacent hexes. Several long range units can fire beyond adjacent hexes but as of now are not included in the RT QB. Everything will not be standardized and player judgment will be needed e.g. setting up the QBs One to one combat odds came up a lot in the OP layer so that was good - perhaps too many times. The best thing I can say is I had fun playing this way.
Kevin
-
Here is what I posted 2 months back:
=================================================
Hello all:
I have been away from wargaming for a long while and finally back at it. Back in the day I was on the test team for CMC. I am still under NDA, but wanted to add a few observations to this thread.
First, programing another layer to CM2x is a very difficult task. I am not sure but getting the right data of the right format from CM2x, using it and passing it back to CM2x seems mind numbing especially for a few good programmers working part time. One thing that these brave wargamers have to decide is if the result of the hard work will be commercial or freeware which will decide how polished the product becomes.
Second and perhaps most important, what do we mean by an operational layer to a tactical wargame. I have played the total war products and like many always wonder how this could be executed with CM. I am not sure who defined operations as warfare activities beyond the reach of direct fire and only within reach of long range arty and air power.
This definition creates a problem. Over the weekend I took a look at many wargames - even older ones made with Cyberboard. Their are really two types: those where combat takes place in adjacent hexes and those (fewer) that use ranged fire like Squad leader and Panzer blitz. Using the adjacent hex model would mean large scale hexes and small/narrow operation maps with large CM maps and OOBs (depending on the hex scale) that would be very cumbersome to play. A ranged model would mean gathering smaller units from multiple nonadjacent hexes into a tactical CM battle. This might be OK if say the game is at the division/corps level having battalions on the operational map and combat companies and platoons are grouped into a force and passed to CM for tactical resolution on an appropriate map for the terrain. Here the campaign map would look like a EF map. So choosing the size of the operational units might be key and perhaps company sized units fighting from adjacent hexes would strike a good compromise. I am trying to find a WW2 company scale game to look at the mechanics.
Great to be back...Kevin
====================================================
Using Cyberboard PB I modified the longstanding rules to provide company level combat from adjacent hexes. The rules are abstraction and don't provide FOW. One to one odd combat is resolved in RT using a quick battle. The terrain is town, woods, hills and open. I have had fun with this and the rules shows promise for making quick campaigns. I have to add supply to the rules and other ideas as they come up. Its mostly a solitaire system with some player judgment needed to create the QBs.
Kevin
ps rules are available for comment.
-
Phil
"There have been a couple of reports - specific to particular game modes"
The files I sent were SP. Can you speak to those modes? Or is more time needed.
Plan to test some recent scenario designs this week and will save as I go SP.
Kevin
-
From the standpoint of setting fires ...
I believe this was in Squad Leader (COI?) under the right environmental conditions.
As for CM, I can a imagine a burning wheat field on a hot/dry day moving toward an all wooden village during high winds. The situation was probably too rare to simulate but it would be cool to watch a few times. Fires from left over battles might be more practical. But how long did they burn for? May not fit into even the CM time scale.
Kevin
Kevin
-
Hi there .. are the German units enough to cover the map in line north to south? To effectively block them you will need to know how they will arrive near the exit - spread out or bunched up. I would sub-divide the Germans into a few groups to help you predict somewhat how they will arrange themselves as the move to the exit. Not sure if it matters, but its a good idea to check the map edges - in this case North to the Reds and west to the Germans. Setting the scenario length and the speed of the German retreat will take a bit of testing to get right. An idea is to start with a smaller size battle and apply what you learn to the larger one. Another is to program the German AI and watch them retreat across the map to the exit without Red intervention. This will sort of give you the minimum length of the scenario. Adding in combat will delay the Germans so you will probably have to add significant length once you add in Red combat.
Kevin
-
"I also have not since a building as a destroy objective, AFAIK. Occupy a building, sure."
If you have shock force, try the battle "Time to go" from the US side. Quick to play. One of my first after a long while away so I started out designing small battles.
Kevin
-
I was thinking more on the lines of wooden structures that could collapse or be partially destroyed. BTW, I have not seen any fire even from a flamethrower into a wooden building.
Kevin
Anyone use a system for the element of chance?
in Combat Mission Red Thunder
Posted
Zenomorph
Yesterday I set up a very simple chit draw system with random unit activation. Chits were made for each HQ in the game for the Germans (me). I played the AI in a small QB. In this case the company (Tank) HQ once activated automatically activates all units under its command. If the company fails to activate (one attempt per turn if chit is drawn) then the platoons can try to activate their AFVs if their chit is drawn. In my case there were one company and two platoon chits with one "wild card" chit that can be used to activate any single AFV (or infantry squad etc) that remains un-activated at the end on the chit draw process. So there were 4 chits but only three were drawn per turn. The random activation was >= 50% (0-99.99% using a random number generator) with the -2 to 2 leadership values serving as modifiers ie -10 subtracts 10% while 2 add 20%. So a HQ lead by a -2 leader would fail activation on a random number < 70%. Ideally for the player the highest level HQ would activate and the game plays as normal. Un-activated units can not target of plot moves but can finish moves and fire as instructed by the games AI.
The system works but really only slows down the forces being regulated by it. I imagine rules could be established to activate on map reserves hiding behind terrain sort of like AI triggers. Off board reinforcements will not work - none allowed in QBs and they are pre-set in scenarios by the designer.
Kevin