Glubokii Boy

Cheers ta.
Aye creating a plausible AI attack is a challenge. I'm not sure I have all the answers and at the end of the day have to work within the confines of the game engine editor and associated tools.
I accept the AI in the current CM with the AI plan tools available will never react to changes in circumstances like a human. Though in saying that in PBEMs I have seen and done some daft crap that would shame the AI if it did the same thing so...
Specifically with this scenario. I know how the Soviet attack went in as have the Soviet primary source and German AARs that corroborate it. So the intent was to create a plausible Soviet attack by the AI. Not fancy or clever or be able to react to the players every move but to make the player think - "Yikes! this is what it was like?!" Then attempt to do their best.
In this case if the German player screws up the timings by timely action then its a good thing. It buys em time. Thats what the Germans attempted to do for real. What I test the plan for initially is basic timings. i.e. ensure without any hindrance from the other side the AI attack will progress. I'm especially looking that if the player does nothing they will lose. So the basic AI should work. This involves testing in phases to ensure each part of the plan works as intended. Its time consuming and real detail work but pays off. Note the most AI units don't have loads of orders. most units in this only have 16 orders, many less. Micromanaging stuff is where the issues arise. i trsut the AI generally to do the best thing. testing seeks out the times it does weird stuff. Oh aye bookkeeping. I sue Ecel spreadsheet paper printouts (update with pencil) as I find that sueful for quickly noting stuff down and rubbing out. Overall plan though i keep track off by using windows whiteboard:

Once I have the bare bones, co-ordinated attack running well i.e.evrything works like clockwork I then play against it. here I'm seeing how the human player might check the plan. I smooth out obvious AI issues like units stopping in full view etc, co-ordinate fires etc. 
Once that is done I'll start to add triggers and now use some units as reaction forces. Again i can't account for every player action but I can attempt broad brush strokes to counter most approaches (terrain plays a large part here - once you check out the terrain the player might have limited options to 'do stuff'). This is all in the realms of art really and lots of replaying over and over. No short cuts.
Even then some player or other will come up with a novel approach. Playtesters come in at this stage. From playtests I can find what novel ways player tackle the scenario and if they come up with effective novel ways I might see how I can tweak the AI plan to make that novel approach more problematic. At the end of the day players always find appraoches i'd never envisage or the AI plan can counter. good on them.
As an aside I did the AI plans for the NTC campaign and its being insightful seeing how players tackle these. Many 9the successful ones0 used novel approaches I'd never consider ( in some early playetsing we'd to make some mountain terrain more impassable as someone managed tod rive AFvs pretty high up the mountains...). So aye good playtesters are worth their weight ten fold.
So aye the challenges you outline I have to work my way around also and I use some or all of the AI editor options available. But in large scenarios (this one uses two attacking Soviet tank brigades and two rifle regiments (all under strength so around a battalion plus tanks and battalion of infantry spread across a large map) so to make it work the plan and timings had to be kept simple as). In smaller scenarios I'll use more of the available tools. The irony is small scenarios are harder to do AI plans for than large scenarios!
Hope this helps?
Cheery!
 

Looks sweet ! 😊
Getting an attacking AI to maintain its cohesion when it comes under player fire is a tricky thing imo...
If the player plays the way you 'want' him to it is quite uncomplicated but if he throws in the odd suprise things become somewhat more difficult for the AI to cope with the situation.
Particulary the timings have a tendance to get screewed up by player fire i have found...
What options do you prefer to try and mitigate this 'complication' ?
Do you use all tools avaliable (game-clock, triggerzones, AI-order triggers) or have you found that you can solve the problems with perhaps one or two of the avaliable options ?
I tend to use all three but doing so requires A LOT ! of bookkeeping in order to keep track of all the AI-order - Exit before Aftter combinations and Triggerzones - Exit before after combinations...
Its ,tricky imo...but kind of satisfying when you get it to work fairly well 😎
Imo that is what you usually are able to do...Looking forward to this one.
 

Interesting when running these historical actions via CM. I've worked out a good position to KO advancing Soviet armour without taking much return fire. Possible this was same position taken up by 7 Kompanie when they KOd 12 enemy tanks? One Panther not so lucky but boit further forward (see right background)...

One of the things in this action was the Soviets used their aircraft in CAS - this was noted by the Germans as being a new thing. Top tip for CM - keep your AA assets close!


AI Plan after this playthrough required tweaking. Mainly because of what you can see in the above screenshot. I've also added some triggers for the Soviet AI to counter most likely German COA. Now to see if they work!

Unfortunatelly...the release of a battle-pack or two will do...very little... in the way of bringing some life back into the CM community. Not even the release of full modules (CMFI, CMRT) has done much to raise the activity around here...Possibly with the exception of CMCW...
Nothing short of the release-...or atleast... the announcement of CM3 will get things rolling again it seems...
 
 

Biggest thing CM needs is a major update to...the AI !
 

Biggest thing CM needs is a major update to...the AI !
 

Considdering that these 'packs' takes quite a long time to design...I'm pretty sure that most of us would accept loosing a few hours of developing time in orders to get a nice bone or two every now and then...😋

No...BFC making fictional games seems like a bad idea...
Let's go with...
CM Barbarossa for their next project 😎

Beautiful pics ! What can i say....👍👍
Oohh, goitch !
Something like a 100 vehicles and i'm guessing a sizable infantry force...The ai programing will be quite challenging i'm sure 🙃...This being a russian attack...AI human wave tactics might work quite well though...
Will be intresting to see the final result...Looking forward to playing this, thanks !
 
 
 

Wee update. Testing AI Plan (timings with attacking units and arty support - taking some fettling!) and the context.
"At 0900Hrs the Soviets opened a formidable barrage on all the lines of the IV.SS-Pz.Korps, in particularly against the left flank of the Totenkopf and against the entire area of responsibility of the Wiking This shelling, so intense that it was impossible to distinguish the explosions from one another, like a continuous roar of thunder, was followed by ceaseless attacks on the part of the Russian air force on the positions of the artillery. The Totenkopf and the Wiking had to face the offensive of two Soviet rifle corps, supported by a whole mechanized body, between Grabie Stare and Sulejow. Out of the six divisions of riflemen who attacked, four pressed in front of the Wiking!" Ref Trang
Images show elements from the assault group from 59 Guards Tank Brigade - 289th Tank Battalion with tankodesantniki from their SMG battalion plus elements from 301 SU Rgt, No4 battery attack into elements from the 'Wiking'  II/"Germania" at Hill 107 behind advancing infantry from 76th Guards Rifle Division..




'Germania's the 5. and 6 companies, were nearly overrun southwest of Tluszcz near Jasienica by Soviet tanks and infantry that had broken through Jadwinin. Ref nash.


 

Yepp...That will work for HE but being able to get the AI to drop a smokescreen at the right location a few minutes into the scenario seems to be...not doable unfortunatelly. 
Thanks for the reply and the upcomming scenario...Will be a BLAST i'm sure 😊

I’ve a historical OOB but given the weight of stuff that was rammed into this area by the Soviets the AI will have all the help it needs!!
re smoke screens. The Soviets have a TONNE of artillery modules. I’ve 8 AI fire support targets including three smoke missions. All come on at around 30s+
 
The AI will tend to drop stuff in a smaller impact zone than you paint. What I’m fiddling with is getting the AI smoke rounds to land on three distinct impact zones but link up the drifting smoke into one continuous screen. 
 
I’ve no real tricks I use re AI turn 1 arty apart from above. The AI will ignore plotted AI plan fire support targets and not commit reinforcement arty modules to these fire plans. In this case best bet is copious TRPs, high quality FOs and equivalent quality arty etc batteries. I’ve found can get some pretty effective AI controlled hip pocket fire support then. 
 
Cheery!

I do the same....using some simple 'AI-kind of' orders for the intended AI side. Saving the game frequently will allow for testing different approaches at the various stages in the battle quite easily. When doing initial testing I usually also keep some 'extra' units on the map (hidden out of the way) that i can quite quickly bring into the fight if i see that one side or the other needs them. If i do use these extras in my hotseat test run i will add them to the OOB before starting work on my AI plans. 
Will this be a turn one AI smokescreen ? or do you have some clever way of getting the AI to deliver a sufficient (on time, on target) smokescreen after turn 1.
Will perhaps a delayed reinforcement artillery work for a later smokescreen set for turn 1 ?

So I'm starting work on the AI Plan for the first phase of the Soviet attack into Wiking's right hand flank, SW of Tluszcz. 
As part of creating AI Plans I tend to do a quick and dirty play in hot seat mode. In this case I selected to the two assault groups from 76th Guards Rifle division (Soviet LHS - 234 Rifle Regiment CO - Guards Lieutenant Colonel Shkurko Vasily Stepanovich; and Soviet RHS 239 Rifle Regiment CO Guards Lieutenant Colonel Maxim Gavrilovich Gorb both units at around 60-70%) and just sent em forward... Arrayed against em two companies from II/Germania at around 50% strength. In RL the initial assault started a kilometer south east, with the guardsmen breaking through with no difficulty (not sure they even met any opposition apart from some German outposts). When they came into this sector the Germans were under a ferocious artillery bombardment.
It's useful to see which German units spot first and do the main killing. Then I can start to plan the arty support (in RL they used smokescreens but creating this in the AI plan and AI limitations in smoke use mean I have to ID the long range firing positions arcs of fire in game and then play the AI fire support smoke accordingly.



They did have some armour support (from what I can work out this seems to have been an overlap of the 8 GTC assault groups who the CO thinking they'd breached the main German lines sent em in) but I've left that out just now. I'll focus on creating the initial AI Plan for the infantry then 'layer' the armour onto that, using triggers etc to provide some co-ordnination but also help stall the attack if the German player can do enough to slow the initial attack.

Hello, Artkin
I'm glad you like the map and have found some use for it 😎...Have you been able to port it over to CMCW ? If so thats cool...
I have been working on a scenario using a major part of the map. Hopefully i will have that finished in a not to distant future. This time it will not be set in Odessa but  rahter a fictional battle covering the russian attack on Königsberg.
I'm trying something a little bit different (atleast for me) with regrards to the AI troops 😊...Initial testing seems quite promising. I belive this will be my last 'major' designing effort until next winter...
Many other RL things takes priority over computer gaming for me during the spring, summer so i will not really have time for any major design work during this period. I may possibly do some more work on this map or a new simular one during the summer...but nothing major and i do not want to commit to a joint effort right now because of my limited computer time.
Many thanks for your offer to help out though
Best regards/ Glubokii Boy
 
 

Hi do you plan on expanding this further? Please do, and  continue until you cannot any more. If you want a hand I would definitely help. I am having a lot of fun in CMCW. I tried to make it work with CMBS but no luck. It doesn't like the 3 story houses

Hello...I have finished the map...I will provide a link down below if anyone would like to download it and use it in their own scenarios or in H2H games...
The map in the editor

 
As mentioned at the start of this thread one of the main features of this map will be a number of 'blocks' that consist of multiple single tile
buildings combined together to build up the towns main housing complexes...These combinations of buildings will form nice fighting possitions that will allow
for full freedom of movement inside the buildings as all of them have fully funtioning internal door on every floor and in every direction. My goal here is to
try and make these buildings somewhat harder to suppres from the outside as there are so many different locations in them...with the ability to repossition within them quickly..
Some of these blocks seen from the editor...

 
And in game...

 
appart from these 'blocks' the map contains a large number of other rubbled buildings as well as a complete factory district with two decent sized factories
that also will allow for some form of movement inside them.
The map is by no means huge but it is very detailed and i contains a lot of 'stuff'...
some more pictures...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Given the high level of detail and the many rubbled buildings and flavoured objects etc...The map will be somewhat demanding on your computer i belive.
If you do decide to download this map and give it a try...Please make a short comment in this thread and tell me if your FPS performance is acceptable of if it is to demanding for your ,machine.
If the majority of people are seeing acceptable FPS i may very well decide to expand the map a fair bit...perhaps makining it usable for a static campaign or somewhat larger scenarios that would provide more room for manouver.
Also...What do you think about these kinds of buildings ? Do they provide some nice oppertunities and tactical challanges or do you considder them to be more of an waisted effort and not something that i should include in future maps...
A link to the map:
https://www.dropbox.com/s/55n385af4jloslg/Rebels in Odessa MAP.btt?dl=0

Any feedback will be apprisiated (good or bad), thanks...
Hopefully the map will work fine for you and be cool to play on...If FPS are low try to toggle shadows off and see if that helps...

No...BFC making fictional games seems like a bad idea...
Let's go with...
CM Barbarossa for their next project 😎

I don't see any usable windows or doors on those buildings (just images of them - unusable in game terms). More akin to giant boxes that will simply block movement. These are more like dioramas.

The way i did it in this test was to have the vehicle start a short distant away...I estimated that it would take the vehicle about 5+ seconds to get to the pick-up location.
At the pickup location i added a 15 second pause to the vehicle before moving on to its next waypoint.
During the same turn I gave the troops an embark order on the vehicle with a 10 second pause....
The vehicle arrived...halted...the troops boarded and the vehicle moved away...tried several times....It works !
 
But timing will be tricky in a game situation i guess.
 
 
 
 

Combat mission has its flaws but what other tactical games handle these things better in your oppinion ?
I can think of NONE 😎...

I love the east front, but one of the reasons RT gets somewhat of a bad rap is because of the timeframe BF choose to use in this game. I think most people would have preferred a 41-43 timeframe for the game, including myself. That being said, RT is a great game, there are a ton of good scenarios and campaigns. 

No...BFC making fictional games seems like a bad idea...
Let's go with...
CM Barbarossa for their next project 😎

No...BFC making fictional games seems like a bad idea...
Let's go with...
CM Barbarossa for their next project 😎

No...BFC making fictional games seems like a bad idea...
Let's go with...
CM Barbarossa for their next project 😎