Afrika Korps beta test

[Erik2]

Hey bru

Seaborne AI invasions...
I'm somewhat confused about this process.
My 'coastline' hexes are simply shallow water hexes, it could as well be deep water.
I would expect the debark hex to be a land hex (next to a shallow/deep water hex).
Apparently this does not work as you have firmly documented.
Your solution, which works at least once , is to actually set the debark hex to a water location.
Not the way I would do it in real life
It looks like there may be a game issue with this.

I'll test AI seaborne invasions a bit more and see if I can find some reliable process.

Edit:
Tried the bru-solution in Persia where each seaborne unit has its own AI-team. I also put supply on land hexes next to the debark hex since the AI is very fond of moving towards supply hexes.
Didn't work. None of the 6 units made it onto land.
I also tried my original land debark hexes, this time with supply. No cigar.
So at this point I think the AI debark process simply do not work (reliably at least).
I would love some-one proving me wrong, of course

[Erik2]

Edit2: Solved it
1) Used bru's Move to and Debark hexes.
2) Added a 'Check unit at debark hex'.
3) Added a Move to inland/supply hex.

finally.jpg (226.38 KiB) Viewed 3554 times

[Erik2]

Persia:
Fixed AI seaborne invasion.

The Road to India:
Fixed sec objs.
Added missing graphics.
Fixed Diversion stuff. And yes, I'd very much like you to another look at it.

Got to go, the wife is hollering
Link updated to 1.8

[bru888]

Edit2: Solved it
1) Used bru's Move to and Debark hexes.
2) Added a 'Check unit at debark hex'.
3) Added a Move to inland/supply hex.


finally.jpg

I am very glad about this. It does make sense when you think of it this way: You are telling a transport ship to move to a coastline sea hex that it can reach, of course (deep or shallow water regardless). Apparently, and what we did not know, is that the game is programmed so you should then be telling the AI to disembark on that same hex, not inland the way you had it. You (and probably I, if I go back and look at Oahu Invasion and The Aleutian Crisis) were anticipating telling the AI to disembark on an inland hex by mistake. The invasions never worked because the transport ships could not reach those inland disembark hexes. Instead, the ship must disembark on the coastline hex and it knows to place the transported unit on an inland hex without us telling it to do so.

[bru888]

Road to India (revisited) v. 1.8

I see you took the "make separate triggers for each objective and make three Check Hex Owner conditions for each trigger" suggestion for Ax Masqat and AFs and Ax Al Qalif adn AFs. These should work, I think.

Is there a reason why you need these capture points? Maybe you were using them as markers as you were working on the Diversion? No big deal but perhaps remove them now because they are giveaways that "something is up" (or will be up) when you look at the map:

Screenshot 1.jpg (111.94 KiB) Viewed 3503 times

I still think that the Diversion spawn trigger is probably never going to fire because you are requiring too many British units of all types to be eliminated from the scenario; over 40 of the 75 or so total including planes, ships, bunkers, and fortress. I would at least filter for land units only but you are the master chef of this casserole dish!

I see that you made individual AI teams for each transport in the seaborne invasion which I believe is a good idea. I noticed that you placed the move to destinations one hex out to sea with the disembark spots one hex forward on the coastline. As I said earlier, I would have placed both on the same hex but let's see if this works. At least we are no longer trying to get the ships to sail inland one hex in order to disembark:

Screenshot 2.jpg (98.43 KiB) Viewed 3501 times

No, my friend, I am sorry but this is not working entirely correct. I tested the Diversion by eliminating all British units at the start. Three out of four transports did disembark, proving your method is sound, but the fourth idled offshore. Moreover, despite a quick run to seize an airport by one unit, the invasion force camped on shore and did nothing for 12 turns. (See the sequence here, if you'd like: British Seaborne Invasion Test - it's a PDF file of 12 screenshots.)

Screenshot 14.jpg (428.24 KiB) Viewed 3497 times

The reason why the Diversion Group didn't do anything but sit after a while (despite the one unit grabbing low-lying fruit and returning to shore) is because your triggers have them moving around in the landing zone but you are lacking a trigger that would order them to move to Masqat. And without taking Masqat, they don't activate the trigger that you do have to move on Al Qalif.

When you fix that situation, you should also correct the error in this trigger; this effect is a duplicate of the one above it in that they both address Diversion inf-2 and you want this one to apply to Diversion inf-1:

Screenshot 9.jpg (334.99 KiB) Viewed 3488 times

At this point, here's another break in the action with this non sequitur post to follow:

[bru888]

I have to admit that I am truly stumped by AI seaborne invasions in this game. Read the end of this post for a theory of mine as to why they don't work.

So first, I tried my method of placing move to and disembark hexes together. Same result as you:

Screenshot 1.jpg (365.49 KiB) Viewed 3494 times

Screenshot 2.jpg (412.64 KiB) Viewed 3494 times

Next I thought, "Maybe it's shy about the initial lack of supply" so I placed a couple of supply ships and rerouted one of the transports. That one, and the one that's been bugging us, stayed idle:

Screenshot 3.jpg (437.62 KiB) Viewed 3494 times

Then I said, "Maybe it's the shape of the land, that promontory is not suitable?" So I terra-formed your coast and moved up the hexes. Same result once again:

Screenshot 4.jpg (407.36 KiB) Viewed 3494 times

Screenshot 5.jpg (405.04 KiB) Viewed 3494 times

Screenshot 6.jpg (404.04 KiB) Viewed 3494 times

Screenshot 8.jpg (418.57 KiB) Viewed 3494 times

So here is my theory, followed by a recommendation:

Theory: The lead programmer of this game, and the guest star designers that have produced the official DLC, have rarely been faced with automating an AI seaborne invasion. Maybe a unit or two here and there (the Japanese end run in Okinawa comes to mind) but nothing in significant numbers. And goodness knows whether the few AI seaborne invasion operations that they have created actually work or not; chances are that flaws would go unnoticed. The major seaborne invasions in OOB have been designed for the human player and thus the programmer and designer never encountered these difficulties nor did they ever perfect the AI mechanisms to overcome them.

Recommendation: As I mentioned earlier, I ran into these problems in both Oahu Invasion and The Aleutian Crisis. I recommend that you adopt the solution that I finally went with in both scenarios for your own Persia and Road to India; that is, to spawn the invasion units on shore.

NOTE: PLEASE DISREGARD THIS EXTRA SCREENSHOT. I CANNOT DELETE IT:

[bru888]

I notice that you were abruptly and definitively interrupted by spousal intervention before you could get to British India but I will call your attention to it because that post is now on a previous thread page so don't miss it. The most important things are replacing the All 4 objs trigger and linking bandits and rioters to an AI team but there are a few other minor things as well.

You know, you had me stumped there for a while. I kind of thought British India looked like the ending mega-scenario but it was numbered only #20 and there are 24 scenarios in the campaign. The next is #21, El Alamein, and we're back in 1942!

Then I looked at the campaign tree and realized (I think) what you are doing. Apparently #21-24, El Alamein, Kasserine Pass, Mareth Line, and Tunisian Bridgehead you added later to expand the North African portion of the campaign?

So after Herkules, which, by the way, you may want to force a replay instead of scenario defeat leading to campaign defeat - the same as you did elsewhere . . .

Screenshot 10.jpg (181.87 KiB) Viewed 3480 times

. . . you have this branching point for #8 Halfa in that a scenario defeat or draw diverts the player to #21Alamein:

Screenshot 11.jpg (158.37 KiB) Viewed 3480 times

Alamein leads to Kasserine Pass, Mareth Line, and Tunisian Bridgehead . . . and then, The End:

Screenshot 12.jpg (174.78 KiB) Viewed 3480 times

No Suez, no Palestine, no Syria, no Iraq, no Persia, and no India. Is that what you intend, you harsh man? If so, fine, but I would still review the campaign tree because it looks like it's got some inconsistencies and oversights.

Nevertheless, I will assume that you would like to include scenarios #21-24 in my CSI sweep, so I hereby don by pith helmet, sun goggles, shorts and knee socks, and trudge back to the desert.

[bru888]

El Alamein v. 1.8

Here's a clue, perhaps, to what I was saying in the previous post about the campaign tree diversion: "General Montgomery, who defeated you at Alam Halfa . . ." So maybe that diversion is by design, a holdover from wherever you got the prototype for this campaign? Seems a bit rough, though, to promise people that they will fight all the way to India only to have them wind up in Tunisia. Your call, maestro.

"Eight[h] Army's anger unleashed on Panzerarmee Afrika. Can Montgomery be stopped?"

"Berlin will insist you hold fast, I can assure you; but old soldiers like us know the better course.
[T]he hour is coming."

I haven't been mentioning it but when there are no secondary objectives, there is no distinction between Major/Minor Victory/Defeat. See, this is what happens; having John Connor'ed my way to the end, I expected an unadulterated Victory and was bitterly disappointed:

Screenshot 13.jpg (362.97 KiB) Viewed 3475 times

I also haven't been mentioning this but will do so now: The Victory, Draw, and Defeat messages are generally too long and they get truncated in these windows with no scroll bar to use in order to read the endings. I think you mentioned that these blurbs are another feature that came with your prototype. Again, your call but it was the same in Desert Rats and some of your other campaigns.

[bru888]

Kasserine Pass v. 1.8

"The ambitious German offensive to cut off [the] Allied Army in Tunisia."

I believe this has to be Value > 2 because Kasserine will count as the first primary VP. It stands to reason that both Djebel Chambi and Djebel Semmama would need to be taken in order to "Secure the Pass" so that would be a total of 3 primary VP's:

Screenshot 14.jpg (451.3 KiB) Viewed 3473 times

By the same token, you are requiring the Allies to hold both Djebel Chambi and Djebel Semmama at the end of the scenario for them to win. I think you need two triggers, one for each point, and if the Allies are in control of either at the end, they win and the human loses.

Not the same with Thala and Tebessa, however; here the Allies need to be in control of both because the Germans need to take only one of them. There is a slight error in this trigger, however:

Screenshot 16.jpg (210.2 KiB) Viewed 3473 times

Your "Hold Kasserine at all times" objective and Ax Kasserine trigger should be redone in my opinion. I would mark the objective as completed at the start and have the trigger fail the objective only if Kasserine falls to the Allies. Maybe use the same trigger to award the AI objective to the Allies as well. (Actually, the All Kasserine trigger does this and maybe that's all you need as long as the objective is marked completed at start.)

[bru888]

Mareth Line v. 1.8

"Rommel ob[v]iously expects our troops to[ ]exhaust themselves on the defenses," plus the briefing is cut short. Rather than leaving the empty caption, I have injected something that I think is appropriate:

Screenshot 17.jpg (244.26 KiB) Viewed 3467 times

Which bit of levity ignores the fact that somehow we are listening in on the Allies briefing instead of our usual Generalfeldmarschall! You apparently adapted this scenario from Desert Rats and forgot to change the briefing, as evidenced by the outcome descriptions (which I feel the need to point out include misspellings "letal" and "obstackle" ).

These are both wrong in that, since you have those objectives marked as competed at the start, they should only fail those objectives if a tank or a plane is destroyed:

Screenshot 18.jpg (218.85 KiB) Viewed 3463 times

Screenshot 19.jpg (208.07 KiB) Viewed 3463 times

This one should be VP Type: Secondary and Value > 2 and it should evaluate on Turn 11, not before Turn 11:

Screenshot 21.jpg (234.71 KiB) Viewed 3460 times

Screenshot 22.jpg (217.15 KiB) Viewed 3460 times

All of your RAF triggers are set for Any Event; usually you use Turn Start.

[bru888]

Tunisian Bridgehead v. 1.8

Orphan:

Screenshot 23.jpg (98.04 KiB) Viewed 3457 times

Not that it matters much, but on US T4, you award the British with an income point for a US unit.

On US air T2, you have a P40 joining the [12] Bombers AI team.

[GabeKnight]

Guys, maybe I don't see some specific problem with Erik's scens, but I don't understand the whole "seaborne landing" hassle. I've made a simple test scen with two invasions on different turns with two landing sites and two targets to attack. Everything went smoothly. I'll attach the scen (just hit the "next turn" button and let the AI play).

Please don't forget, that the AI will guard every supply hex disregarding other AI orders! If you have four units and four supply hexes, I guess only one (if there's an opportunity) will ever abandon guarding them.

Screenshot 709.jpg (491.85 KiB) Viewed 3433 times

[bru888]

Gabe, you may be onto something. Erik, it may be worth another try if you are willing to take a look and see how Gabe accomplished it.

[Erik2]

Gabe

Just a quick note.
In one of the AK scenarios I have 4 units landing and there are about 3 supply locations added to/around each landing site.
All 4 units move to their next objective, the AI does not hold any unit back to guard any of the supply sites.
There is of course supply available at the new objective.

[Erik2]

Road to India:
Removed the temp sea invasion flags. Used during construction.
Changed Diversion spawn trigger.
Added/fixed missing Diversion move triggers.
Noted that the AI will not land on the tip of a peninsula.
Ran the scenario in editor mode, the Kiwis now follow orders.

British India:
Fixed orphan.
Fixed Typhoooooon.
Cleaned up briefing.
Added primary obj map marker.
Added Axis 10 obj fail trigger.
Added missing graphics.

Campaign changes:
If the player only gets a defeat at Alam Halfa the Afrika Korps offensive towards India is over. The player then needs to go on the defense from El Alamein to Tunisia.
A draw at Halfa and a major victory at El Alamein puts the player back on track to Suez.

Herkules:
Added replayability.

El Alamein:
Fixed scenario description and briefing.
Fixed victory/defeat labels.
I'm still waiting for the devs to add a scroll bar to the victory/defeat messages. Should be a simple fix to allow for proper battle reports

Kasserine Pass:
Fixed scenario description.
Fixed obj value >2.
Fixed victory/defeat labels.
Fixed Allied pass objectives.
Fixed Allied HQ obj.
Fixed Kasserine obj.

Mareth Line:
Replaced the old Allied briefing with a proper German one.
Moved Panzer/Luftwaffe to sec objs (and fixed them).
Fixed 'sec obj'.
Fixed RAF triggers.

Tunisian Bridgehead:
Fixed orphan.
Fixed US air.

Link updated to 1.9.
The usual big thank you to Bru for the forensic work.
I'd give you my first-born, but she may have reservations

[WarHomer]

I´ve started this DLC and 4-5 missions in I´m enjoying it immensely. Playing highest diff as usual.

Notes to mission 1 I think it was. After getting a few secondary VP´s I get a new primary objective to capture Tobruk. This was very hard to do, since I got it quite late as I was positioning for all the 5 secondary objectives (and not for Tobruk) and maybe this new objective should only be a secondary or at least be advertised in the mission description or secondary objective description. Also I couldnt quite figure out if it was worth the hurt.

Notes in generel.

You get almost all of your units handed to you. Maybe it was better for replayability and diversity if you got more prestige instead. I also found myself disbanding all my given air units to be able to take advantage of the air specialisation (1 star xp).

Maybe its more historical correct I dont know, but in regards to land CPs you get a lot more German than Italian. I´m only 4-5 missions in, so this could change later on, but I´m already finding myself a bit limited in regards to my Italian OOB and not quite sure how diverse I can let myself be in regards to Italian units.

[Horst]

I wasn’t fond of the constant gifted core-units in Desert Rats either. Players can easily waste RP on early purchases which are simply gifted one mission later. At some point during Rats, I had disbanded more than half of the core units, in particular the many useless infantry units.
I’d appreciate it too that developing a core should be in the hand and motivation of the player. Getting a secret unit gift once in a while is always fine. If naval units don’t play such an important and lasting role during a campaign, like Desert Rats, then I’d like to have them as aux units either.

[WarHomer]

I´m okay with core naval units, but otherwise agree with Horst.

[WarHomer]

mission 5.jpg (368.21 KiB) Viewed 3286 times

Trigger error. Mission 4.

[bru888]

Trigger error. Mission 4.

There's nothing wrong with the trigger per se. I tested the trigger using #johnconnor and it worked. However, when I delayed capturing the eighth VP until turn 18 in a slightly revised version of this scenario, it failed.

My conclusion is this: The Turn Start / Check Turn / Scenario Turn Limit condition is flawed in that it is adjudging the situation at the beginning of the final turn rather than at the end. I think we have all seen this before and I am guessing that is what happened here.

When I tested it again using Turn Start / Check Turn / Turn < 19 (even Turn = 19), I got the same result. Turn Start means "start of turn" and if you fulfill an objective during that last turn (which is 18, not 19 despite what it says in the display), apparently it does not count.

This makes it tough for those "Hold objective until scenario's end" situations but I don't have any further advice.

Erik, here is a file containing screenshots of how I tested this in case you wish to contact Tech Support about it. If you agree with my analysis, of course.