Working and winnable campaign

Campaign editing

Moderators: Lone Wolf, Snake Man

Snake Man
Commander-In-Chief
Posts: 9338
Joined: 2000-07-31 22:01:01
Gaming Interests: ArmA, ArmA 2, Falcon 4.0 and OFP.
Editing Interests: All, I (try) to edit everything.
Location: PMC
Contact:

Re: Working and winnable campaign

Post by Snake Man » 2009-01-20 22:26:52

Yes absolutely keep your .cam file references on SP4.2 data.
PMC Tactical Forum New User Registration please read new info here.

PMC since 1984

Editing knowledge, visit PMC Editing Wiki
The leading, most detailed and comprehensive modification made for the Vietnam War - Vietnam: The Experience homepage
View our videos in PMC Youtube channel

PMC Tactical forum Advanced Search is power.

"ALPHA BLACK TO PAPA BEAR. ALL RUSSIANS ARE TOAST. OVER."

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 05:27:10

Note, the following edits took place in FreeFalcon 4, Nevada/Redflag1/Save 2:

Alright gents, I think we're getting somewhere here.

I made too many edits to say, specifically, what is working and what is not. BUT, I think there's some great info to be gleaned nevertheless.

First: ccc is right!. I think (in fact, everything from here on out is prefaced with "I think") that a working road and link network is simply fundamental to getting this all to work. The first picture illustrates this (SM, I'm going to great lengths to re-format these pics to fit...I think this is important, please grant some latitude if you can...).

I was bothered by the fact that my ground units last night were routing around BAKER airfield. Why did they always insist on avoiding this field? Look at this pic:

Image

Despite the link, there is an obvious break in the road between Ft. Irwin Junction and Baker; thus the units can't go there!

****************

I opened up the file in Tacedit, with mapview on. I can't believe how much easier editing is in this view.

Like I said, I made alot of edits, so I can't isolate one in particular that made the differnce. But, that's ok. This series of edits encompasses alot of the ideas I put forth earlier today.

First, I edited the Baker objective (obj) and the two obj nearby to be under bluefor control. I wanted to eliminate this broken link. Then I added alot of T-80 armor battalions:
I added batt. to support and defend (two batt for each obj) for the choke points that I kept getting: China Lake Br, China Lake, Ft. Irwin, Ft. Irwin Jct., Palm Springs, and so on.
Added batt. with secure near many of the same places.
Added about 4-6 batt. with variations of secure and capture, some with 1189 (LV) some blank.

Altered objectives: I went along a line from Yucca Valley all the way up to around Bishop, and added 'Border', 'Beach', and in some 'manual' to the obj dump.

************************************

Started new campaign. No input from here on out.

Notice at 11:50 or so. Less than 3 hours in. Some units are going into defensive positions. But this one single armor batt. goes "reserve Las Vegas". Yes, I said "reserve".
Image
Last edited by toonces on 2009-01-21 06:18:35, edited 1 time in total.

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 05:41:09

Now, note that this batt (psycho batt) is just cruising on to LV. He's captured the junciton in the way and is headed to capture the outskirt obj's on the way to LV:

Image

As the psycho batt makes its solo run to LV, another armor batt suddenly gets inspired. Check this guy by Tecopa. He's changed his tasking from some sort of reserve task to capture Boulder City. What caused this inspiration? I think it has to do with psycho batt capturing that bridge; somehow that opened up the way. OR, perhaps the batts that reached their defensive positions allowed the units to start their offensive operations. Like I said, the perimeter was established...

Image

Note this guy in the next picture. He's taking this zigzag route (which shows the linkages very well) to reserve this bridge. It turns out the Yucca airfield, that southernmost airfield, is the last border obj on the map. Why does the engine skip that first (east) bridge to reserve the west bridge? Maybe it finds the obj behind the border obj, but close to a town/city/village?

Image
Last edited by toonces on 2009-01-21 06:19:59, edited 1 time in total.

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 05:54:28

The following pic simply reinforces a point that ccc has made at least a million times since I've been on this forum. Check the convoluted route this unit is taking...this is because of that broken road in the first pic I posted. It can't go south, so it detours all the way around the lake to get to the obj.

Image

The following pic is very interesting. A third armor batt has become inspired. Psycho batt is virtually on the outskirts of LV, but then bores of the sport apparently. He decides to retire south into a secure position to the south. Batt-2 is still secure Boulder City. Batt-3 (new guy) is capture Riviera (through about 3 other bluefor obj's). So, 3 batts have gone on offensive now.


Image

This is the psycho batt that decided that instead of capturing LV, he wanted to defend Searchlight all of a sudden. Perimeter?

Image

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 06:04:55

Just for kicks and giggles, I decided to try and task the psycho batt to capture LV. As you can see, it accepts the tasking (for about 10 minutes) but then rejects it for engine tasking; it reverted to Searchlight shortly after this pic.

Image

By now you're probably thinking you've seen everything. We're 18 hours into a campaign in which the player (me) hasn't flown a mission yet. And look what happens. Batt-4 decides that enough perimeters, secures, and who-knows-what-else has been established, and suddenly gets inspired to go for LV itself.

Image

This, finally, is where I must leave you because at this point the campaign engine decided it needed some input from me. The campaign stalled (didn't stop), but units sort of stopped moving. That's understandable. I haven't flown a single sortie for this campaign yet.

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 06:16:06

Lessons learned:

It's hard to determine what worked, and what didn't. There's no control; I didn't test things methodically. I did a full-on edit with new objs, obj edits, new units, unit edits, etc. What worked? What was coincidence?

Here is what I think:

1. CCC is spot on. Fundamental to everything is a working road network, terrain network, link network. Everything has to be connected logically and in a non-broken manner. Paths have to exist for sensible movement.

2. Borders are important. I don't know how or why, but I think borders have to be established. When the border ended, units picked up a defensive position.
And I didn't even experiment with borders on both sides of the (Nevada) dmz.

3. Units have to be sufficient for offense and defense. Enough units have to be defending junctions/nodes/objs that the engine determines are important such that the engine can task offensive movement. There are in-game triggers (independent of the .tri file) that let units go offensive. The way psycho batt goes from offensive on LV to secure Searchlight is telling. It says to me that the engine calculated what it needed to capture and hold LV, decided it was overextended, and then decided to secure its line of advance. Why it sends one unit back and one unit from the rear forward, I can't explain. Could be an engine hiccup, could be something I did with tasking. But you can see the units use a very standard flank defense on their offensive. If I put 15 more battalions in there, I'll bet it would be alot more interesting.

4. Unit tasking is important. The engine is pooling resources for reserves and defense. Putting units in defense at China Lake Br. and Ft. Irwin Jct. allowed offensive units to go forward.

I saved everything from this edit. This is all backed up so that if I foul things up from here on out I have this mod to fall back on.

Next: change all secure/capture tasking to 'reserve' in tacedit and see what the engine does.

Next 2: Incorporate Brigade structure.

Next 3: Actually come up with a sensible OOB for defense.

Next 4: Put some bluefor in there and see how the engine handles some opposition.

I welcome your comments and criticisms!

ccc
Chief of Staff
Posts: 4857
Joined: 2000-08-06 22:01:01

Re: Working and winnable campaign

Post by ccc » 2009-01-21 07:56:37

Ah - thanx for checking and understanding the key role of road network. :mrgreen:

Again, my conclusion and solution after testing several PMC mods - remove bridges( just leave some for choking point in 2d work), re-link objectives, and erase river-bridge definition on terrain tiles. don't bash me for not doing the tweak.. the modding work is boring and exhausting, i need a long break after the first try- failure- few months ago.
2. Borders are important.
it's news to me.
3. Units have to be sufficient for offense and defense.
agree. even in default korea campaign variants, if your ground units are low- most used on secure/defence, then you won't have spare units for another offensive. in this case, adding reinforcement units periodically may help the game rolling.. you don't have to add all units at the begining of war.
4. Unit tasking is important.
hmm.. i doubt it. last time i checked default korea campaigns, most units have no task assgined.. campaign commander AI will assign them automatically. If you pre-assign a task, that's make it like a TE, not a fluid campaign. However, like i said in Taiwan mod thread, some tasked units do help the game rolling( if the unit follow your order loyally).. as you noticed, one tasked leading unit's movement may inspire the maneuvoring of rest, or following units. i also noted this behavior in recent test on DoctorX's three korea campaigns..[order a unit to move ahead-capture some obj, then more units follow] > this human intervention shortens the time to win the war. if you let AI do the order, the war may extend few more days!!!

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 17:44:56

Points taken ccc.

WRT point 2: All I can say for certain is that there are alot of parameters assigned to objs in Korea that aren't in Nevada objs. The border and beach parameters struck me as interesting and purposeful. Without knowing how the campaign engine is programmed, I can only guess at their true purpose. The movement of that armor unit to the western-most bridge south of the southern-most border obj (Yucca Valley) suggests something. More testing would be needed of course. But something tells me that there is some fundamental reason for the border, beach, frontline, secondline, etc parameters for objs.

WRT point 4: I have two thoughts on this. First, even with user-supplied tasking at campaign start, the engine takes over the tasking for the units once the sim starts. So, I disagree that it is necessarily TE like. I see what you're getting at, but I think that once the sim gets moving, the engine by its nature changes tasking to suit its plans and dynamic's the campaign. I need to test this more. Second, I absolutely think that tasking units to defensive and support positions is important. Everytime I ran the campaign previously, my ground units nearly all went to reserve China Lake Bridge. Once I put some units tasked with defense on those positions, all the units found new things to do. Again, I changed too many things at once to say that what happened was definately caused by X, Y, or Z, but my gut tells me there's something there worth investigating more.

I only have the FF4 campaigns, so I don't know what exists in the SP4.2 campaigns. But I can assure you that the units in Korea DO have tasking associated with them. Are you checking brigades as well as battalions? For example, I've seen frontline brigades in Korea with the brigade tasked with secure, and 3 battalions tasked with reserve, one tasked with support and one tasked with defend. Or something like that.

Question: What PMC theaters have the best road/link network? Have you seen any that are better than others?

The southern portion of Nevada south of Las Vegas seems to have a few working roads, I can keep running experiments in that region for the time being.

You've got alot more experience with this than me, so please share your experiences freely.

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-21 17:51:42

I keep reading about a "working ITO2 campaign" on these boards.

Somebody (Biker?) has cracked the code on how to generate something that "works"...

I'd love to get his inputs on what items he found to be fundamental to getting purposeful ground movement. I mean, why am I trying to reinvent the wheel here if someone has already done it?

User avatar
Phoenix711
Captain
Posts: 186
Joined: 2007-10-03 07:52:29
Location: Istanbul Turkey
Contact:

Re: Working and winnable campaign

Post by Phoenix711 » 2009-01-21 18:18:51

Gentlemen I've just runned some tests.

Aegean theater,

1. THR WITH ROAD INFO BUT NO RIVER INFO!
With this THR, an M1A1 brigade can move across the map like 500 km and capture everything on its way (including bridges). But I've spent a lot of time on the road network so the road network is complete.

2. THR WITHOUR ROAD INFO BUT NO RIVER INFO
With this THR, same placement as above, if the units start on an objective, no move at all. If they start near an objective, move to the nearest obj. on the way, capture it and stop there. This obj, is not the assigned obj.

When I save at that position and change the THR to the one with road info, voila, the units capture the nearest obj and continue moving without even stopping.

Other thing interesting is, Falcon engine draws exactly the same route with or withour road info in the THR file. When I checked, I saw that the route info for ground units is calculated based only on tile types, and not roads. It tries not to go into forrest areas when possible.

My conclusion so far is, road network is important, especially objects must be connected with roads, or your units will be stuck over the first object captured.

I'm working now on the tool to make it so that you can select

CREATE THR OPTIONS
-------------------------------
ROAD INFO - NO ROAD INFO
RIVER INFO - NO RIVER INFO

etc.

Some other functions can be added, like all farm tile info THR etc. I don't have much time to test, maybe you can use this tool to create your own THR's as you want without changing all the AREA / PATH info on the tiles themselves.

I can just add, everyone check your THR files at 2x zoom level in Tac Edit Map View (ALT+R, Relief Map) to be sure your THR files are OK.

Regards.

ccc
Chief of Staff
Posts: 4857
Joined: 2000-08-06 22:01:01

Re: Working and winnable campaign

Post by ccc » 2009-01-22 01:44:38

Question: What PMC theaters have the best road/link network? Have you seen any that are better than others?
Each time i got new pmc mods, i checked the road network roughly, to see if the main combat axis/routes is open and complete. So far Kurile and Georgia look better- not complete yet. the rest- especially 512 ones, need extensive edits. phoenix7111's post is very promising.
I keep reading about a "working ITO2 campaign" on these boards.
Somebody (Biker?) has cracked the code on how to generate something that "works"...
Biker can answer it himself. AFAIK, no code edit.
1. THR WITH ROAD INFO BUT NO RIVER INFO!
With this THR, an M1A1 brigade can move across the map like 500 km and capture everything on its way (including bridges). But I've spent a lot of time on the road network so the road network is complete.
GREAT JOB. :D I''ve asked for such tool for a while. :wink: as your ground units can move freely, your mod has entered a new level- tell them the targets and get their maneuvoring organized and orientated.
Other thing interesting is, Falcon engine draws exactly the same route with or withour road info in the THR file. When I checked, I saw that the route info for ground units is calculated based only on tile types, and not roads. It tries not to go into forrest areas when possible.
yes AI pick the easiest and lowest "cost of movement" way to their target. the value is list in TacEdit- you may change it freely.. to modify the desired combat axis/route/area.
My conclusion so far is, road network is important, especially objects must be connected with roads, or your units will be stuck over the first object captured.
yes, in 2d world/Tacedit, the campaign is designed as a wargame. in 3d world/terrainview, all units/items turn into visible stuff and act with 3d world rules. the sweetest part of falcon4 engine is, it merge the two worlds nicely.

bear with me, when looking at 2d world/TacEdit, the map is nothing but numerous obj dots connected by link lines. the bridges, are key objs that once destroyed the link broken and ground units can not move across it.(unless you use AF that can repair bridges)

when you jump into 3d world/terrainview and tile editor, the bridge obj choking point, is further defined by the tile definition(pathmaker tool). the tile has river line defined- troops can not cross it, the tile has bridge postion defined- to help placing the bridge(IIRC), so when the ground units try to cross river, they must line up, or gather closer to cross the defined-bridge-area on tile, in 3d world. i did notice ground units crossing a destroyed bridge in 3d world- yet i don't remember they showing crossing the same bridges in 2d map view..they just stopped there..probably means 2d world rules the 3d world.

User avatar
Phoenix711
Captain
Posts: 186
Joined: 2007-10-03 07:52:29
Location: Istanbul Turkey
Contact:

Re: Working and winnable campaign

Post by Phoenix711 » 2009-01-22 08:49:06

Great bunch of info CCC. I'm sure that even in 2D map, I've seen units trying to line up with the bridges.

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-22 17:10:18

I ran a few more experiments last night.

Nothing definitive. I added several brigades, all tasked with reserve, with their component battalions tasked with reserve.

None went offensive initially. Psycho battalion from the other posts still drives right up to Las Vegas. Right around Boulder City Bridge, everytime the psycho batt. gets there, the tasking for units all over the map changes to an offensive posture. A few of my new battalions-within-brigades started on offensive movement to capture objectives after psycho batt. hit that point. I replicated it twice.

I tried changing ownership of Boulder Bridge to Opfor, but the trigger for offensive movement wasn't tripped until psycho batt hit that point.

I have to figure out what is significant about that particular point.

I also want to play around with the reserve tasking for some of the brigades. Korea starts with the units sitting 5 miles from the enemy objectives, Nevada has them 100 miles from objectives. I want to see if there's something to that.

Snake Man
Commander-In-Chief
Posts: 9338
Joined: 2000-07-31 22:01:01
Gaming Interests: ArmA, ArmA 2, Falcon 4.0 and OFP.
Editing Interests: All, I (try) to edit everything.
Location: PMC
Contact:

Re: Working and winnable campaign

Post by Snake Man » 2009-01-22 18:00:01

Can you use Tactical Engagement (TE) to verify if the links are working?

So far you only checked in Tacedit the one trouble spot using relief map which is good, but you should check them ingame in TE before continuing the painstaking campaign observing / testing. TE will show results immediately if you can reach objective A from objective B.

Oh btw, just to be sure, have a quick look of the te_new.tac in Tacedit also, is it identical objective linking vise to save0.cam for example.
PMC Tactical Forum New User Registration please read new info here.

PMC since 1984

Editing knowledge, visit PMC Editing Wiki
The leading, most detailed and comprehensive modification made for the Vietnam War - Vietnam: The Experience homepage
View our videos in PMC Youtube channel

PMC Tactical forum Advanced Search is power.

"ALPHA BLACK TO PAPA BEAR. ALL RUSSIANS ARE TOAST. OVER."

Closter
Captain
Posts: 154
Joined: 2007-01-07 00:13:42

Re: Working and winnable campaign

Post by Closter » 2009-01-22 18:46:00

My conclusion so far is, road network is important, especially objects must be connected with roads, or your units will be stuck over the first object captured.
...
I can just add, everyone check your THR files at 2x zoom level in Tac Edit Map View (ALT+R, Relief Map) to be sure your THR files are OK.
This issue was slightly oversighted when we discussed Taiwan's campaign problems. I pointed to this about a bridge that units were not able to cross. My point was that 2D engine may need to see a road bit info into the THR under the bridge to allow units to cross. That way 2D engine knows how to make a way for the unit through the river.

My take on this issue with other terrains is THR usually is badly formed, and has many holes in his road network, THR level. Not necessarily a problem of tiling, .bin or bad linking.

But as Tacedit seems to use the THR to calculate the travelling costs that are stored at the time linking is made, if THR is badly done then perhaps linking is corrupted. THR and linking could need reworking, as a result, in that order.

Another source for the problems may be the slope info, that's into the L2 file. Too much slope and engine could decide units are not able to travel across this tile.

I would like to prove me wrong or right myself, but as my real life and my early steps in spain theater dev are needing me bad, I leave this to you ;-)
Image

ccc
Chief of Staff
Posts: 4857
Joined: 2000-08-06 22:01:01

Re: Working and winnable campaign

Post by ccc » 2009-01-23 01:37:43

My take on this issue with other terrains is THR usually is badly formed, and has many holes in his road network, THR level. Not necessarily a problem of tiling, .bin or bad linking.
Closter pointed out a possible bug..maybe the THR should be rebuild once you tweak the obj linking in TacEdit.

reason is, last time my tweak on Taiwan and Vietnam was, 1. remove some bridge obj and re-link nearby obj in TacEdit, 2. i erase river and bridge section defintion on Tiles and re-draw the road line on those" river-road-bridge" tiles with Pathmaker tool. The test result failed.. quite disappointing and i really hesitate to test the idea again. it makes me wonder what going wrong in first or second step. now re-building THR sounds reasonable..how?

pheonix711 said he's a tool to erase all river info in THR.. wonder if the tool also remove those river-road-bridge junction data and keep the obj linking/road network intact?

User avatar
Sherlock
Lt. General
Posts: 1167
Joined: 2006-05-24 22:01:01
Gaming Interests: Falcon 4.0
Editing Interests: All, I (try) to edit everything.
Location: Arizona, USA

Re: Working and winnable campaign

Post by Sherlock » 2009-01-23 02:53:36

ccc wrote:
My take on this issue with other terrains is THR usually is badly formed, and has many holes in his road network, THR level. Not necessarily a problem of tiling, .bin or bad linking.
Closter pointed out a possible bug..maybe the THR should be rebuild once you tweak the obj linking in TacEdit.

reason is, last time my tweak on Taiwan and Vietnam was, 1. remove some bridge obj and re-link nearby obj in TacEdit, 2. i erase river and bridge section defintion on Tiles and re-draw the road line on those" river-road-bridge" tiles with Pathmaker tool. The test result failed.. quite disappointing and i really hesitate to test the idea again. it makes me wonder what going wrong in first or second step. now re-building THR sounds reasonable..how?

pheonix711 said he's a tool to erase all river info in THR.. wonder if the tool also remove those river-road-bridge junction data and keep the obj linking/road network intact?
Based on previous discussions I have had with Dr. Fred Balding rebuilding the THR file is the answer to roads/links that work.
Sherlock
Victurous te Saluto

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-23 03:00:50

Unfortunately, you guys are rapidly getting out of my knowledge zone.

If we start talking about rebuilding thr files, I think that has to come from some sort of official PMC developer/development. I mean, it's one thing to tweak the ownership of objectives and alter the OOB, but once you start rebuilding the terrain that sort of crosses a line imo.

I'm curious what you guys think.

It seems like we have a good dialogue going and are actually making some progress; we might as well hammer out how to do this for real.

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-23 04:41:50

Just some quick results from some tests tonight.

First, I tasked 3 Brigades (about 3-4 battalions each) to "Capture 1189 (Las Vegas)". I gave no tasking to the component battalions; they were all tasked as reserves.

All 3 brigades started towards LV right at the campaign start. That's good because it tells me the campaign is accepting tasking. BUT, after a while I noticed that battalions were splitting off on various reserve and defend tasks. There were a good 4 battalions from 3 of the brigades that beelined for LV. But as they worked their way towards LV, they changed their tasking to capture/defend objs enroute: Searchlight, Boulder City, Boulder City Bridge, etc.

As I searched around, like before, as the battalions cleared the route and the objs around LV, other battalions started accepting capture LV tasking from their reserve status from the campaign engine- no input on my part.

Now, I still have a bunch of guys that are going into reserve/defend mode away from the FLOT. I manually changed the destination of two separate battalions to see what would happen. The result is below. 1) It seems that the roads/links are broken in some areas. No surprise there, we already knew that. But look at the individual battalions along the convoluted route each unit takes. 2) It appears to me that what is happening is that battalions are being individually tasked from the campaign engine to secure objs along the working links (edit: actually, the link NODES. some of those routes won't work- it's looking at the link nodes, not roads). It could be a coincidence. But it definately looks like the engine wants to set up defensive units along working routes. Some of the other unusual tasking and units gettting grouped and stuck might be a simple matter of the engine tasking them, but then getting confused because the route/link is broken...

Image

Image

ccc
Chief of Staff
Posts: 4857
Joined: 2000-08-06 22:01:01

Re: Working and winnable campaign

Post by ccc » 2009-01-23 04:48:56

Sherlock wrote: Based on previous discussions I have had with Dr. Fred Balding rebuilding the THR file is the answer to roads/links that work.
yeah..i got the same answer after tweaking obj linking and tile definition edits.. i just can't remember if i ever did the THR rebuilding.
:oops:

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-23 17:08:51

Alright guys, I have mucho ground movement.

I used the same set up as above. I added about 6 or 8 battalions of infantry at each of the kinks in those links where my units kept getting tasked with reserve. I tasked the infantry units to defend their respective obj.

I restarted the campaign and most of my brigades and battalions went on offensive operations, cleaning up alot of the map.

The next step is to try to figure out how/why the campaign picks certain objs over others for offense. For example, psycho batt when straight up to LV as usual, but after it captured the LV Factory it went to defend Boulder City. The other brigades/batt seemed to waver between LV and Needles. Why Needles?

I have to get into tacedit and look at priorities and stuff some more. One thought is that the path of the ground units to LV leaves their southern flank very exposed and the engine seeks to defend the flanks. Or, it could be that that is what I want to see and I'm projecting my own ideas onto the engine.

At any rate, I had alot of promising activity last night, without alot more tweaking.

One more thought is that the engine seeks to defend alot of objectives. But, the number of objectives is so overwhelming compared to the number of units available for tasking that the engine gets confused. I think this is what is happening. To test it, I really need to insert alot more defensive units around objectives and see if this free's up the offensive units for unrestricted operations.

I think we're getting somewhere...

Snake Man
Commander-In-Chief
Posts: 9338
Joined: 2000-07-31 22:01:01
Gaming Interests: ArmA, ArmA 2, Falcon 4.0 and OFP.
Editing Interests: All, I (try) to edit everything.
Location: PMC
Contact:

Re: Working and winnable campaign

Post by Snake Man » 2009-01-23 17:47:34

toonces wrote:One more thought is that the engine seeks to defend alot of objectives. But, the number of objectives is so overwhelming compared to the number of units available for tasking that the engine gets confused.
Well, one brigade can defend 5 objectives. I recall default Korean campaign having stupid amount of ground units far far in the south where there are no action, maybe this is the same thing you're talking about.

How high priority those defended objectives have, I mean does it need to have like 40-50 or up to be defended?
PMC Tactical Forum New User Registration please read new info here.

PMC since 1984

Editing knowledge, visit PMC Editing Wiki
The leading, most detailed and comprehensive modification made for the Vietnam War - Vietnam: The Experience homepage
View our videos in PMC Youtube channel

PMC Tactical forum Advanced Search is power.

"ALPHA BLACK TO PAPA BEAR. ALL RUSSIANS ARE TOAST. OVER."

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-01-23 18:32:07

@ SM,
That is precisely what I'm talking about. The more I dig into this, the more I find myself realizing how little I know about Korea. But, yes, I think that the engine is looking to defend as many objs as possible. I don't know what the priority is. Bridges and factories seem to be getting alot of attention in Nevada.

I think it's a combination of priority, links in roads or 'links', and flank objs.

I've been thinking some more about this. We have been pushing to see how big we can make a theater; I wonder the converse- how small can we make a theater?

It's not a trivial question. There is a certain scope the campaign has to have in order to manage the air war. If you make the fighting area too small you end up with an air war that takes place in a goldfish bowl. But, if you keep the theater small (or, more specifically, the number of objectives relatively small), then the amount of ground units that need to be placed to defend everything is much easier to manage. But how small can we go before the engine barfs in the other direction- it doesn't have enough space or units to generate the war it was designed to fight in Korea?

******************

I'm trying to think about this as wargaming 101. In order to design a truly robust dynamic campaign engine, there has to be a robust AI underneath that determines offense and defense. Rule 1 would be to protect flanks and retain a defense at your own objectives- factories, bridges, junctions, etc. Rule 2 would be to keep your units on reserve until you could amass a 3:1 force ratio on the ground. Rule 3 would (probably) keep the ground war in tempo with air superiority. I have no evidence specifically to defend this statement, but it seems that if your side gets the snot kicked out of its air force, the ground war stalls.

So, in conclusion, I think that my next step tonight will be to go through and add battalions all over to cover objectives. So far that seems to be working well. I'll keep at it.

I can't decide if this is actually starting to make sense, or if I just want it to make sense...

Snake Man
Commander-In-Chief
Posts: 9338
Joined: 2000-07-31 22:01:01
Gaming Interests: ArmA, ArmA 2, Falcon 4.0 and OFP.
Editing Interests: All, I (try) to edit everything.
Location: PMC
Contact:

Re: Working and winnable campaign

Post by Snake Man » 2009-01-23 20:14:55

Well, 8x8 segments, dunno I guess Dem2terrain is capable of doing 4x4 too.

Code: Select all

Theater name: my_asia1
Theater size: Regular
Size (L2 segments): 8 x 8
Size (L2 tiles): 128 x 128
Size (Nautical Miles): 69.0471 x 69.0471
Resolution: 0.134858 Nautical Miles
Centered on: Lat 38 Long 127
NW: Lat 38.5754 Long 126.264
NE: Lat 38.5754 Long 127.736
SW: Lat 37.4246 Long 126.275
SE: Lat 37.4246 Long 127.725
Highest point: 4175ft
Lowest point: 0ft
Not sure which one would be practical anymore, even for testing purposes.

Do you want me to create such tiny theater for campaign testing?
PMC Tactical Forum New User Registration please read new info here.

PMC since 1984

Editing knowledge, visit PMC Editing Wiki
The leading, most detailed and comprehensive modification made for the Vietnam War - Vietnam: The Experience homepage
View our videos in PMC Youtube channel

PMC Tactical forum Advanced Search is power.

"ALPHA BLACK TO PAPA BEAR. ALL RUSSIANS ARE TOAST. OVER."

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-02-22 18:34:02

Sorry it took me so long to get back to this- I had a major computer malfunction the last month or so.

It might be interesting to generate a smaller campaign terrain. Perhaps a terrain centered on Las Vegas, but about half or less the size of the current theater. I don't know how much work it will be for you, nor how much use it will be. But, if you make a smaller version based on the Nevada theater based on the setup of the Redflag1/save2 campaign, I'll try populating it and re-running these experiments.

I got pretty far with these edits. I had plenty of offensive ground movement, but I had trouble getting the engine to focus on Las Vegas instead of Searchlight. This could have been related to the .tri file.

[edit]: I don't know if 70x70 will be large enough. How about 140x140?]

User avatar
Phoenix711
Captain
Posts: 186
Joined: 2007-10-03 07:52:29
Location: Istanbul Turkey
Contact:

Re: Working and winnable campaign

Post by Phoenix711 » 2009-02-22 21:02:37

This post may be valuable for campaign creator / editors. Please check.

viewtopic.php?f=63&t=21784&p=180746#p180746

Snake Man
Commander-In-Chief
Posts: 9338
Joined: 2000-07-31 22:01:01
Gaming Interests: ArmA, ArmA 2, Falcon 4.0 and OFP.
Editing Interests: All, I (try) to edit everything.
Location: PMC
Contact:

Re: Working and winnable campaign

Post by Snake Man » 2009-02-26 15:38:17

toonces wrote:It might be interesting to generate a smaller campaign terrain. Perhaps a terrain centered on Las Vegas
No, the campaign testing theater would be just few roads and objects, no real life reference.
I don't know if 70x70 will be large enough. How about 140x140?
Nevada is 64 x 64 segments, so umm... what are you talking about?
PMC Tactical Forum New User Registration please read new info here.

PMC since 1984

Editing knowledge, visit PMC Editing Wiki
The leading, most detailed and comprehensive modification made for the Vietnam War - Vietnam: The Experience homepage
View our videos in PMC Youtube channel

PMC Tactical forum Advanced Search is power.

"ALPHA BLACK TO PAPA BEAR. ALL RUSSIANS ARE TOAST. OVER."

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-02-26 17:08:51

I was exposing my ignorance again, that's all.

In the green text you posted above you have "69.047x69.047" miles as the size. I was thinking double that might be more appropriate. I can get units to move 60 miles no problem- getting them to capture objectives along 120-140 miles I haven't managed yet.

I don't want to create more work. But if it's easy to do, I'll play with it as a control experiment. If I can garrison all of the objects in the theater and then create some offensive and defensive units, I can do the tasking like I did with Nevada above and see how it all plays out with the flanks protected.

But, I don't know if it will be useful or just waste an afternoon of your time.

Snake Man
Commander-In-Chief
Posts: 9338
Joined: 2000-07-31 22:01:01
Gaming Interests: ArmA, ArmA 2, Falcon 4.0 and OFP.
Editing Interests: All, I (try) to edit everything.
Location: PMC
Contact:

Re: Working and winnable campaign

Post by Snake Man » 2009-02-26 19:01:28

Ah I see now, I was talking about segments as we always do, not the miles.

Making a campaign ground unit movement test theater would just mean like straight roads between perhaps 4 objectives or something like this as the real problem is to concentrate your focus in certain area in real theater is difficult, if not impossible.
PMC Tactical Forum New User Registration please read new info here.

PMC since 1984

Editing knowledge, visit PMC Editing Wiki
The leading, most detailed and comprehensive modification made for the Vietnam War - Vietnam: The Experience homepage
View our videos in PMC Youtube channel

PMC Tactical forum Advanced Search is power.

"ALPHA BLACK TO PAPA BEAR. ALL RUSSIANS ARE TOAST. OVER."

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-02-27 04:52:47

I tell ya SM, this would be so much easier if I could just sit down and buy you a beer and chat.

WRT the testing I did so far with Nevada, getting ground units to move from one obj to another really isn't a problem. Like in the screenies I posted above, setting brigade tasking to capture or secure will get the ground units offensive. Provided the roads and links are unbroken the units will move (it seems) from offensively from one obj to another.

The problem I had was getting the offensive to focus on Las Vegas rather than Searchlight and the southern cities. I remain convinced that the wargame campaign engine underneath the Falcon hood is seeking to secure its flanks before it moves on Las Vegas. I can't prove it, but it makes sense to me. Furthermore, I think the offensive tasking for units comes after sufficient garrison units take up reserve/defensive positions along each kink in a road or obj linkage.

If that is the case, one of the problems I can't overcome easily is the huge number of objs and terrain that needs to be garrisoned before my front line units will either go on or stay on offensive. A smaller terrain will allow me to garrison everything easily because the number of objs is smaller. Then I can set two sets of ground units against each other and see how things play out with variations of the force levels, etc.

A couple of straight lines between a handful of obj's isn't going to show anything new.

I hope I'm explaining this well. This is where an hour over beers would really make this so much easier.

Before you go burning time on a new test theater, let me keep experimenting on Nevada. I just need time to sit down and immerse myself in it again. With FF5 coming out any day now I'll have a stable install to work with for a while.

If I can't crack the code in Nevada, we can revisit the test theater thing...all in my opinion of course.

User avatar
Sherlock
Lt. General
Posts: 1167
Joined: 2006-05-24 22:01:01
Gaming Interests: Falcon 4.0
Editing Interests: All, I (try) to edit everything.
Location: Arizona, USA

Re: Working and winnable campaign

Post by Sherlock » 2009-02-27 06:12:32

toonces wrote:The problem I had was getting the offensive to focus on Las Vegas rather than Searchlight and the southern cities. I remain convinced that the wargame campaign engine underneath the Falcon hood is seeking to secure its flanks before it moves on Las Vegas. I can't prove it, but it makes sense to me. Furthermore, I think the offensive tasking for units comes after sufficient garrison units take up reserve/defensive positions along each kink in a road or obj linkage.

If that is the case, one of the problems I can't overcome easily is the huge number of objs and terrain that needs to be garrisoned before my front line units will either go on or stay on offensive. A smaller terrain will allow me to garrison everything easily because the number of objs is smaller. Then I can set two sets of ground units against each other and see how things play out with variations of the force levels, etc.
What are the city priority numbers set to Las Vegas versus Searchlight and the southern cities? Post them (I don't have that theater installed) and let's examine them. There may be an issue there.
Sherlock
Victurous te Saluto

toonces
Brig. General
Posts: 484
Joined: 2008-07-20 19:43:12

Re: Working and winnable campaign

Post by toonces » 2009-02-27 17:30:47

Sherlock,
I don't have it installed right now- I had a major computer crash last month.

Right now, Tacedit isn't reading my backed up .cam files- it might have to do with a corrupted or missing Falcon install. I don't know how Tacedit works, so I don't know what the problem is.

I remember checking the priorities for both cities and making LV either 80 or 100, and Searchlight was much less- I want to say 50 or so.

When I get my install working again, I'll see if I can get those backed up files to open (or I can email them to you if you want to pm me your addy).

I'm PCS'ing from CA to VA this week coming up, so I'm not going to get much done until at least mid-March.

User avatar
Sherlock
Lt. General
Posts: 1167
Joined: 2006-05-24 22:01:01
Gaming Interests: Falcon 4.0
Editing Interests: All, I (try) to edit everything.
Location: Arizona, USA

Re: Working and winnable campaign

Post by Sherlock » 2009-02-27 20:11:51

toonces wrote:Sherlock,
I don't have it installed right now- I had a major computer crash last month.

Right now, Tacedit isn't reading my backed up .cam files- it might have to do with a corrupted or missing Falcon install. I don't know how Tacedit works, so I don't know what the problem is.

I remember checking the priorities for both cities and making LV either 80 or 100, and Searchlight was much less- I want to say 50 or so.

When I get my install working again, I'll see if I can get those backed up files to open (or I can email them to you if you want to pm me your addy).

I'm PCS'ing from CA to VA this week coming up, so I'm not going to get much done until at least mid-March.
toonces,
I don't have Nevada installed and don't plan on installing it anytime in the near future (I have too many other irons in the fire). So it wouldn't be of any use to email them to me. Just was trying to help track down the problem and the priority number thing occurred to me.
Sherlock
Victurous te Saluto

ccc
Chief of Staff
Posts: 4857
Joined: 2000-08-06 22:01:01

Re: Working and winnable campaign

Post by ccc » 2009-04-22 12:29:04

details of my rough tweak on Kurile mod here..
viewtopic.php?f=65&t=21867

my summaries for a "possibly" working campaign...

- check the road network first. use Terrainview, load feature tdf files, and use a road-highlighted textxure.zip to check road network in all - or major combat areas.

- if you choose to keep bridges, you should check the placement of bridges. and make sure once some bridges destroyed, your falcon version/campaigns have engineer units to repair them, or the campaign will stuck halfway.

- if you choose [ delete bridges ], you have to do obj linking again in TacEdit. Also you have to erase/edit road/river/bridge definition on road-river tiles with Pathmaker.

- test road network with TE, make sure road network in main combat area is open. you may fine-tune the [ move cost value ] with TacEdit, to force ground unit movement within a desired battle area.

- along the main battle routes, objectives for victory condition should have higher priority, to draw more ground units to capture/defend them.

- for campaign rolling, both sides should have sufficient ground units. many will be used on defend/secure high-prority obj and become immobilized for a long while.. only surplus units can take the offensive order. Once both sides lost too many units, it could happen that nobody has surplus units for offensive, and campaign hanging there infintiely. in this case, reinforcement units should help pushing the war to an end.

- the ground units should be organized in a good division-brigade-battlion system. when campaign AI gives an order, it seems brigade level will receive and excute the order, and its battlions do the real job in a coordinated way. we don't know if HQ unit is a must-have in ground operation yet.

- when placing ground units, better avoid putting them on water(river, lake, sea, near coast), or moutain( rock, thick-forrest) area. some ground units won't be able to move as they trapped in water or thick forrest, and turn into totally unresponsive to campaign AI order. in this case, move them to flat area or close to city/roads.

- To make a hot working campaign, first start with adding ground units for both sides. i suggest give one side more units first, then check if the stronger side can win the war - make sure the campaign mechanism is working right. Then you may add more ground units to the weaker side, to increase the intensity of fighting. When adding air units, keep in mind ac with CAS/interdiction capability should be added with care - one sq of attacker may neutralize many ground units and tilt the balance of war.

Post Reply

Who is online

Users browsing this forum: No registered users and 14 guests