Over the Reich - Creation Thread

I have the events.

 

The events are available.

Changelog 20 May 2019

Commented out the old freighter generating system. Will wait a short time before deleting it.

Deleted some commented out code in civ.scen.onTurn I doubt we'll need it, and if we do, there are tons of older event files around to get it.

local improvementAliases = {}
improvementAliases.militaryPort = civ.getImprovement(34)
improvementAliases.civilianI = civ.getImprovement(4)
improvementAliases.civilianII = civ.getImprovement(11)
improvementAliases.civilianIII = civ.getImprovement(14)
improvementAliases.refineryI = civ.getImprovement(5)
improvementAliases.refineryII = civ.getImprovement(10)
improvementAliases.refineryIII = civ.getImprovement(22)


specialNumbers.convoyArrivalSquare1 = {5,15,0}
specialNumbers.convoyArrivalSquare2 = {2,66,0}
specialNumbers.convoyArrivalSquare3 = {4,126,0}
specialNumbers.trainsPerConvoy = 4
specialNumbers.baseFuelPerConvoy = 300 -- basic fuel per convoy unloaded (before military port reduction)
specialNumbers.convoyRefineryFuelBonus = 300 -- bonus fuel per refinery when unloading a convoy in a city
specialNumbers.unloadTrainsPerCivilImprovement = 2
specialNumbers.unloadTrainsForPort = 1

also moved

specialNumbers.maxGermanPorts = 19
specialNumbers.maxAlliedPorts = 13

further down the "list" of special numbers

Battle of the Atlantic Mechanics:
Turn/Location 1 2 3
1___________ 1 4 7
2___________ 2 5 8
3___________ 3 6 9

If the number of German Ports divided by 3 is less than the corresponding value for the turn and location, then a convoy is generated. So, since 7 is always larger than 19/3, all the convoys are generated every third turn, unless the Germans capture some English Ports and build military ports there.

Each city has a number of trains it can unload each turn. 2 trains per civilian improvement, and one more for the military port itself. If the port has unloaded fewer trains than the limit, the convoy can dock, otherwise it can't. If the convoy will unload more trains than the city can accommodate, the extra trains will be unloaded, but not have any movement points available.

Unloading convoys will also generate a fuel bonus worth 300 plus 300 more for each refinery in the city, and worth 100 less for each military port the Allies are "missing".

The following units will be destroyed if they are in a city when its military port is destroyed. This should act as a counter to "hiding" units in cities, especially convoys until they can be unloaded.
local killWithMilitaryPort = {
[unitAliases.Convoy.id] = true,
[unitAliases.FreightTrain.id] = true,
[unitAliases.GermanTaskForce.id]=true,
[unitAliases.AlliedTaskForce.id]=true,
[unitAliases.UBoat.id]=true,
}

Killing a convoy in this way still increments the German score.

Eliminated variable production cost for submarines and destroyers.

When a Wolf Pack is destroyed at sea, a city among all the cities with military ports is chosen. If the city is owned the Allies, the Wolf Pack is killed. If the city is owned by Germany, the unit is re-created in that city. If we keep this mechanic, we probably want to have wolf packs be very expensive (I'm thinking ~900).

specialNumbers.extraUBoatLoss = 0 -- Counts as extra enemy cities when determining chance of u-boat returning to port.

 

I made a few small tweaks but the events are available.

I very much enjoy these changes. I think they have a lot of application for other scenarios, perhaps even in your Imperialism scenario you've been considering. One could almost use this to duplicate the "districts" mechanism in Civ6 and have cities sprawl beyond simply their tile by connecting certain outside units to inside improvements (consider if killing a marketplace unit killed all traders for the city).

I do have one question:

specialNumbers.convoyArrivalSquare1 = {5,15,0}
specialNumbers.convoyArrivalSquare2 = {2,66,0}
specialNumbers.convoyArrivalSquare3 = {4,126,0}

Would {{5,15,0},{5,16,0},{8,18,0}} etc. also work here or does it just have to be one tile? I see that you've used the "Move to Adjacent" which will prevent the Germans from stopping on just that tile but I'm curious if I'll be able to randomize it any. I'm envisioning drawing some rectangles where the convoys will generally appear somewhere within the rectangle (although, perhaps that isn't even necessary to draw it).

JPP 5/21 AM CHANGE LOG
-Added Carrier to list of units that can be killed in city if military port also destroyed
-Changed cost of Wolfpacks to 900 shields
-Decreased cost of gun battery slightly as it'll be less impervious to attack in future versions - down to 600 shields from 750
-Made Army Groups and Task Forces non-disbandable

 

I think we're making some progress towards fixing some issues the playtest unearthed/confirmed:

-late game crawl should be reduced with very few army and sea units walking around on the map and a viable aerial option (fighter bombers) to take out flak.
-Predestination of Allied victory (or at least impossibility of German victory) should be fixed...
-Germany now has some very, very good reasons to attack Allied military ports and fuel refineries.
-Germany should now present a much more dangerous challenge in the Battle of the Atlantic. I could easily see the Allies having to spend a considerable effort early to mid game on trying to keep U-Boats at bay.
-Payload should fundamentally change the game - one bomber, one mission.

I am wondering if we've done enough to keep the game fun for everyone because we are trying to do something a little unique here - the German player has to basically stomach being ground into the dirt which is a bit unusual. I think reducing the number of turns from 175 to something between 90 - 120 will probably help this too since they have a good chance of holding out. I think this turn reduction will be balanced by reintroducing fuel bonuses from freighters so that there are less turns of "down time" while the Allies are building up a massive force.

 

I can make the choice of tile random from a set of tiles in a table, or random within a box that you define. Just tell me what you would prefer.

I agree that we should give some more thought to the German player's experience. Maybe we can use some of the extra unit slots to give the Germans some one-off units, either "prototype" planes that were never actually produced, or a few "experten" units like you suggested earlier. What might be interesting is a sort of "solitaire" mission for the German player that occurs within the context of the war, and is affected by the war, but doesn't itself impact the war. The Allied player wouldn't care, but the German player would have some extra reason to want to play and prolong the game. I don't have any idea what this solitaire game would be, unfortunately.

 

You had mentioned quite a while ago an intention to make this scenario playable single-player. Most of what you're discussing sounds like game strategies best utilized with two human hands behind the players. How is the progress (if still being pursued) of having the game be single-player-friendly without an incompetent or inexplicable AI opponent?

 

It's a tough thing to figure out, for sure. There aren't that many scenarios or games out there in the world where you're basically trying to stave off defeat and call that a victory, and there are even fewer people who would enjoy it, but there can be something kind of fun about it. It's definitely a different experience trying to scrape together what you can and fight something much more powerful.

I do think that "nerfing" the tactical bombers was important. My morale more or less broke in our game when you had your large build-up of tactical bombers and wiped out my main jet base. @McMonkey may or may not feel the same way after my last turn's attack on Stolberg AF (although I do expect him to in turn wipe out about 100 of my aircraft that are all stuck next to it). I suspect that making short-range fighter bombers the only aircraft that can really expose those bases will help a lot in that regard. It shouldn't be possible to do that to "hub" bases now - only spokes.

I also think we might want to implement the chance of special targets being discovered just because he voiced so many concerns about those, and get rid of Operation Jericho completely as there's no chance for it to be fair.

We might consider using the "prototype" trade units in a different way as well. Perhaps instead of being true "trade" units, we make them 'k' units of a sort and change the mechanism a bit... Perhaps there is a point counter of some sort with them, and what you would do is build them and then bring them to Peenemunde and press 'k'. If that key is pressed on the same square as that city, the counter ticks, and there is a science beaker boost. Perhaps this could function also as a "parallel" tech tree. It would be linear, but bringing enough of these to Peenemunde over time would unlock some prototype, "secret weapons of the Luftwaffe" type units... Culminating, even, in the atomic bomb (only the Germans can get this for the sake of this scenario).

This would serve a few functions:

1. The German player would truly have a science advantage because they'd have their regular tech tree plus this "parallel" tech tree.
2. The Germans would get some nifty "what if" units out of it.
3. The Germans might have a chance to turn the tide at the last moment with these units.

Perhaps to balance it, the prototypes now have to move on railroads, so the Allies would have a good reason to target railroads in Germany leading to Peenemunde.

Edit-I just checked what art we already have. @Fairline has created:

Me262 HG III
Me P.1101
Ta-183
Ta-283
Arado E.555-1

Plenty of very interesting, very advanced fighters if we go this route...

I do think the Experten is also a good idea. Even after you consider that all air forces clearly over-estimated their kills, the Luftwaffe's kill counts still dwarfed USAAF and RAF aces. Maybe we could tweak the "veteran from offense" mechanism for the Luftwaffe to give a 1/10 chance of the unit becoming an "Experten" unit (that would fire a more powerful munition than the Spitfire). 50/50 chance of Ace, 1/10 chance of Experten?

Everything being discussed at this point is 100% for multiplayer and I haven't put in much thought into single player. I was a playtester for Nemo's Blitz and am aware of what went wrong with it (chiefly, the AI stacked all its units on one tile and they would then be wiped, 100+ at a time). I think Lua could, to an extent, be used to overcome that issue. I don't know how the AI would manage with multiple maps but I doubt they would do well with it (at least the high-alt/low-alt concept). It would probably require a complete redo of the scenario from the ground up (quickened of course given how much ground work has been laid), and require several other tweaks.

I'm not really sure what the AI would do better at - defending or attacking. I do think the strategic bombing mechanism would be easy enough for the AI to use (from either side), because it's simply "attack unit" and "build improvement," both of which they can do. But getting them to attack with enough force to do something would be tough without heavily scripting the scenario. Also, the entire 'k' unit and payload mechanisms wouldn't likely work with the AI.

I would love to make this single player at some point but to be honest if that happens I see the path not as converting OTR immediately, but building a different scenario - The Battle of Britain - as a single player scenario first and foremost (and eventually converting to MP), and then taking the lessons learned from that and reapplying to OTR. BoB would be much easier to playtest given the condensed time frame, significantly fewer units, smaller tech tree, etc. OTR is truly a monster of a scenario in scope. There are 12 lua modules working together to make it run, and its main event is nearly double the size of Napoleon's, and about four times the size of Caesar's for example (and both of those are pretty darn large scenarios themselves)! Trying to convert something like that while simultaneously testing how the AI handles even a tiny percentage of what we've incorporated just seems monumental, hence my thought of trying a small-scale test with BoB (that would also give more of a reason for BoB to exist beyond just being a "reskin" of OTR).

 

Let's do three small boxes:

1,1,0 7,1,0 7,13,0 1,13,0

1, 59, 0 7,59,0 7,71,0 1,71,0

1,105,0 7,105,0 7,117,0 1,117,0

 

I don't remember discussing a single player implementation of this scenario, and certainly haven't included that possibility it in my ideas for mechanics or in my code. That said, I have thought a little bit about it lately, and here are my thoughts so far (many of which occurred to me as I wrote others down).

1. I'm pretty sure that we can make the AI "move" units by using the onActivate trigger to control the unit before the AI has a chance (I could be wrong; the code might start to run while the AI does something else with the unit). We might be able to set a move-to destination, but in the worst case we can teleport the unit to where it is supposed to go and expend its movement points. We can also make the computer "press a key" by event by simply running the onKeyPress code with the keyID of the button it is supposed to press.

2. If we can make AI units move, then we should be able to have the AI conduct bombing runs. If a bomber activates "behind the lines," have it rendezvous to a "staging area" airfield. If a bomber activates in a staging area airfield and there are "enough" bombers in the airfield, the bombers depart for a target. If a bomber activates with a go-to command to a target, it proceeds towards that target. If it activates close enough to a target to attack that turn, move to an adjacent square and run the code to "press" 'k' or 'backspace'. We can make munitions attack upon activation if that doesn't happen automatically. NONE bombers return to a staging airfield, NONE bombers that activate in an airfield are assigned a home city from an available airfield (not necessarily the one it is in).

3. We can have the AI cheat with fighters assigned to "guard" targets. As long as the fighter is at full health, just let it keep hovering over its assigned location. If enemies stray nearby, the "guarding" fighters attack with similar mechanics to bomber attacks. A problem is how to choose where to place the fighters.

4. "Realistic" escorting of bombers by fighters is difficult. Perhaps bombers "in range of fighter cover" are considered high value targets to defend, using the same mechanics as guarding strategic targets.

5. Trains travel to airfields and are "disbanded" automatically by code.

6. Getting the AI to "legitimately" produce what it "should" is difficult. We might be able to direct somewhat by using canBuild to limit its options, but we don't have the ability to directly set production. We might be able to compensate by cheating to give the AI things it seems to need from time to time.

Having written this stuff down, I think we could make an AI that would at least give the flavour for the game. I'm not sure that we could make a game that feels that it was designed for single player, however.

 

I have the events.

 

CHANGELOG (only in the scenario as you have the events)
-I restored Neutral territory everywhere it should be.
-I added Neutral Territory unit to the water around the Russian Front so that the Germans can't attack units there too soon. Can you please add a unit deletion event to when the Front opens? Any neutral units in the box 403,71 --> 407,71 --> 407,77 --> 403,77 should be deleted now so the front can open up.

I think another way we might make things fun for the Germans (and probably the Allies as well) is with the buildup of the Allied forces via events. Right now I just have a simply "If the Allies have X amount of points, they get a new Army Group" but what I'd like you to add is another condition where "And the turn is equal to or less than THIS" then they get the new Army Group. This would put a time crunch on the Allies marching through their points system and would give the Germans some hope because if they can have a good run for a number of turns they might be able to stave off some reinforcements and be more likely to prevent a successful invasion.

 

Events are available.

Changelog

Convoys can appear in boxes
specialNumbers.convoyArrivalBox1 = {xMin=1,xMax=7,yMin=1,yMax=13}
specialNumbers.convoyArrivalBox2 = {xMin=1,xMax=7,yMin=59,yMax=71}
specialNumbers.convoyArrivalBox3 = {xMin=1,xMax=7,yMin=105,yMax=117}

improvementAliases.cityI = civ.getImprovement(8)

Retreating and re-creating battle groups.

--Battle Groups and depleted battle groups can manually re-deploy to nearby cities, and defeated battle groups will regroup to a nearby city if it is available. The process for finding the cities is as follows:

The nearest 5 cities to the unit are put into a short list. If the battle group is already in a city, that city is excluded (airfields are also excluded). Cities on a different landmass (i.e. Britain vs. Continent) are also excluded.

City X is removed from the short list if there is a city Y such that travelling from X to the unit takes no longer if one must first pass through city Y.

City X is also removed if it is more than 7 squares farther from the unit than Y, and the detour of travelling from X to Y to the unit is no more than 10 squares more than the distance from X to the unit itself.

Finally, all enemy cities are removed from the short list.

If the unit is making a voluntary redeployment (by pressing backspace), all cities still remaining in the short list are an option. If the unit is defeated in combat, it travells to the nearest one, to respawn as a depleted battle group.

For depleted battle groups, backspace is also the key to consolidate forces back into battle groups (an option box is created). Consolidation requires as many trains as full strength battle groups to create, and one more depleted battle group than that. Veteran depleted battle groups become veteran battle groups. The option tries to create the maximum number.


--Turn deadlines for reinforcements added. All turn deadlines are currently set to turn 90, but they can be changed in the special numbers. Not tested.

--Function to delete neutral units near Russian Front when the front opens up added, but not tested.

specialNumbers.alliedScoreIncrementFreighter = 2

--Removed negative increments to German score (did a search for incrementCounter("GermanScore", so I think I found everything)

 

The events are available. I won't be able to do anything until next weekend as I'm off to the lake.

I tweaked things so that the Allies have to keep roughly 9-10 points per turn pace to receive their reinforcements. This is about what I averaged early game. I didn't want to go too crazy because I'm not sure how the payload change will affect this.

I'm leaning towards a 120-125 turn game. I think that 90 will leave things a little rushed when so much is cat & mouse and I'm concerned that the payload will change things up to require more time. 120 turns is still about 4 months to complete the game if both players are active and complete 1 turn done per day on average. I think that's more achievable even late game with the changes to the 1,000lb bombs and army groups. It shouldn't take 45 minutes of attacking cities with 100 aircraft to reduce defenses any more.

I allowed the Germans to earn their points whenever, because I'm not certain how many points they'd earn on average per turn. I'd need McMonkey's saves opened up to try and figure that out. Again, however, payload would affect this (the Allies would theoretically try harder to keep aircraft alive).

I've modified the Ju-87, Fw190F, Do335, Hurricane, Typhoon, and Tempest to have 15x4, 20x4, and 25x4 MP. As they are close air support (CAS) and carry the most powerful munition, a short range is justified. The 15 MP only leaves Cormeilles and Manston in range of each other - all other airfields are safe. The 20 MP opens up a few more bases, but will leave the CAS exposed to counterattack by secondary/deeper airfields. The 25 MP isn't enough for the Allies to reach several of the Dutch airfields from England, but again, you'll be susceptible to counter attack. I think that these ranges work and will help the overall balance.

 

Note: I forgot to download your most recent event change before making my changes. I used a program to merge both our changes into the same file, and it appears to have worked fine. However, if you notice something missing/wrong, that's probably the reason why.

Changelog

Introduced warning when munitions are generated and treasury is "low." Low is defined as the upkeep cost of all structures (calculated after production, so testing in the "workon" save file will consider 10000 to be "low", which is the initialization value for the counter used) for that civ.

After the warning, there are 4 options. The first is to suppress the warning until another 100 fuel is spent. The next is to suppress warnings until the game is saved and loaded again. The third is to suppress warnings for the remainder of the turn. The last is to suppress warnings for the player for the remainder of the game.

Note: if upkeep costs are changed in the rules.txt, the file upkeep.lua must also be changed, since there is a table of improvement costs there. Strictly speaking, the file upkeep.lua could just be shoved somewhere in the main events.lua file, since there is so little going on.


console.computeCosts(tribe) can be used from the console
console.setSubQualities() can be used from the console to trigger the submarine flag addition/removal

setSubQualities had to be put into onScenarioLoaded (which was added) rather than onLoad. It appears that onLoad happens before the rules.txt are loaded, while onScenarioLoaded happens after.

The hard to spot units must be added to the tables


local tempSubFlagAlliedUnits ={
local tempSubFlagGermanUnits = {

areaReactionDamage entries for splash damage reinstated.

Flak (or any splash unit) now simulates the damage that will be done to friends and foes, and will only attack if the simulation shows friendly damage at most 25% of damage that is done to the enemy in the simulation. Since only one simulation is done, there can be "mistakes," especially with only a couple planes involved, but you're not likely to lose several aircraft due to one or two flak batteries. Also, a single plane won't prevent massive splash damage attacks agains the enemy.

specialNumbers.maxFriendlyFire = 0.25 -- Friendly fire damage can be at most this fraction of damage done to enemy (predicted) or unit won't react. Set to 0 to never fire when friendly units could be caught in the crossfire.

 

I'm making more changes to the events. You'll still need to download the above file, since it has an extra module.

 

changelog

Reinforcement locations now in tables
local reinforcementLocations = {}
reinforcementLocations.AlliedBattleGroups = {{110,72,0},{139,65,0}, {140,44,0}, {137,11,0}}
reinforcementLocations.AlliedTaskForces = {{110,72,0},{139,65,0}, {140,44,0}, {137,11,0}}
reinforcementLocations.GermanBattleGroups = {{293,59,0},{299,63,0}, {324,50,0}, {347,55,0}}
reinforcementLocations.GermanTaskForces = {{293,59,0},{299,63,0}, {324,50,0}, {347,55,0}}

When reinforcements are to be received, a message appears explaining what the cities are that must be held (regardless of whether the cities are held or not). If none of the cities are held, the unit that would have been created is lost forever. The tiles chosen for receiving reinforcements must be city tiles.

Added reinforcements for all four types on the first turn, so that the message explaining the recruiting cities will display.

 

Are the events available or are you still working them? I have tomorrow off and would try and close the gap and finish some stuff up.

On that note, do I have to do anything specific to get clouds working again?

 

You can have them. I'm trying to figure out how the game applies the movement penalty for damaged units. The only thing I could find seems to hold only for a road multiplier of 3. I don't need the events to work on that, and I need it before I can implement the amphibious movement penalty.

Use the following command in the console to re-introduce clouds:

(this used to be attached to 6)

console.updateAllWeather()

How are we going to do the German Occupation Bonus? With only a few battle groups, it doesn't make a lot of sense to count how many are in France. I was thinking along the lines of counting the number of quartermasters that have been built, and only allowing construction/completion of the quartermaster building in occupied cities if a battle group is in the city. That would force the occupying forces to go around "occupying" the country. However, once the buildings are complete, only invasion would act as a way to reduce the bonus for the Germans. Perhaps this is acceptable. After all, the German player will probably be playing defense for most of the game, so it might be a desirable bonus for the German player to not have to contest the occupation bonus the way the Allies must for their convoys. Alternatively, the occupation bonus could be based on Quartermaster plus Railyard. After all, there is little value in plunder if you can't transport it.

Do you still want a start of game warning about fuel levels? At the moment, the player will be warned if reserves go below the upcoming turn's costs.

 

I think we're OK with just using the battle groups and adjusting their value. I would hate to complicate the scenario more than we have already and this system is already ready to go. Also, the Germans are going to start with a heck of a lot of these (I have 15 full battle groups and 12 depleted battle groups at the start). I need them to have several more than the Allies to make an early invasion problematic for the Allies.

This probably isn't going to be necessary given the new warning.

The only change I made was to add the beaufighter and remove the Ju88C from the sub issue given that German aircraft can attack by day and night and we probably don't want them to have sub attributes by day. I added the freight train because I wasn't sure if the nil value would creep up without anything in there, and I thought this might make those a bit more survivable anyway. I can add "can spot subs" to the P-47s to make them more useful in the trainbusting role as well.

I didn't do the clouds yet as I wasn't sure if I'd be changing around more text on the map.

The events are available.

 

OK, I didn't know how many you planned to have. I should note that at the moment, my code to re-form full strength battle groups "consumes" one depleted battle group and makes the rest full strength. That would mean that the 12 depleted battle groups could be formed into 11 full strength battle groups (with 11 trains). It should be easy to set a maximum number of battle groups that can be combined at once (i.e. a max of 3->2 or a max of 4->3) if you want it. I didn't do this since I figured that amassing a larger number of depleted battle groups would keep them out of combat longer, and so offer a trade off in that sense.

Perhaps depleted battle groups could give more points during the occupation, so the German player won't be quick to re-form full strength groups.

We might want the event that opens up the Russian Front to produce some German Battle Groups on the Eastern Front as well (and reduce the number of Battle Groups Germany starts with). Otherwise, the German player is essentially keeping his troops in the West until the Russian Front opens up, which is sort of the opposite of what you'd expect. That is, the Germans would want to keep as many troops in the East as possible until the Western Invasion.

We can give the spot subs flag to all Allied day aircraft, and then make the German aircraft all sub (unless there are some that are easy to detect at night). That way, Germany can Wilde Sau without being at a disadvantage at night. I suppose that would make Allied planes better at finding German Wolf Packs, but we could probably find some way to compensate for that.