• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

unmerged(14164)

Second Lieutenant
Jan 29, 2003
119
3
Visit site
First off, thanks for all the hard work the CORE team put into V0.3. This is truly a ground-up rework of almost every aspect of the game. I'm especially excited about the way CORE handles logistics and industrialization.

I am, however, concerned with one aspect of the mod: the way CORE models the navy. My experience is based on playing a game with Germany up until 1939--UK and Russia remain to be conquered.

Individual Ships

Already in 1939 I had to deal with 120 naval units. No fun. HoI2 made big strides in keeping the game streamlined and abstracted; this feels like a step in the wrong direction. Not to mention that separating individual submarines makes convoy hunting too easy--when my subs were in the water in September 1939, I was destroying 20-40 convoys per day.

(I noticed that Germany had 2000 convoys, was the fix to this problem just to massively increase the number of convoys each country has?)

I assume there is reason why it was modeled in this way. How does this logic stack up against the increased hassle for a player? It's hard to imagine that giving me four destroyers (all different models to boot!) really makes the game more realistic, on average, than giving me one destroyer flotilla with average stats.

Automated Navy

I was blown away when I started CORE and realized that the team had written events for every ship every country ever produced between 1936 and 1939. The amount of work and research to accomplish this feat seems staggering. Unfortunately it also doesn't seem necessary--and is actually quite frustrating.

I assume that the reason for this change was to force every country to construct its historic navy or suffer the consequences (massive dissent, slept ministers, etc.).

I have seen a set of events floating around that simulate the early game naval treaties limiting the size of Germany's fleet. What if the same triggers were used to check if the player is building the appropriate sized fleet?

If the player isn't, say, building two battle cruisers and a dozen submarines, his naval minister threatens to resign and dissent is given (FLAG1 set). If the player continues not to build his requisite navy, more dissent is given and the minister actually resigns.

Taken from TRPv91 (Naval_treaty.txt):
Code:
trigger = {
		NOT = {
			submarine = { country = GER value = 12 } # less than 12
		}
}
The benefit of this is that it gives the player more control and it uses existing game mechanics.

Again, an excellent first release. Keep up the good work!
 
Hi,

CORE made the decision to go "single ship" a log time ago. I probablly wouldn't have chosen it either (due to other concerns) but that is how it is here. This has reaped both praise and scorn from the public, but the general reaction has been positive and most of those truely interested in the naval side of things do seem to like it. So you end up in the minority here and there isn't much I can recommend as it will stay single ship barring a major change in direction. Comments:

1. Convoys are based on RW hull numbers. So you need to sink a lot of vessels with CORE as ENG has tns of them.

2. The single ship setup does allow a much more accurate representation of RW navies, particularilly smaller ones. When I was doing multi ship flotillas at HSR this was a real issue.

3. Naval builds: The purpose of this isn't to lock a player into a RW setup, but rather to give credit for production that is already partially complete. All of these ships are pretty cheap compared to building them normally, and some are extremely good deals. We used to have them in the build que but Paradox changes to the EXE ruined that setup and we switched to event driven. While not exactly what we want it is pretty good. In the long run we plan on adding the "naval Treaty" events we had back in CORE1.

mm
 
CORE made the decision to go "single ship" a log time ago. I probablly wouldn't have chosen it either (due to other concerns) but that is how it is here. This has reaped both praise and scorn from the public, but the general reaction has been positive and most of those truely interested in the naval side of things do seem to like it. So you end up in the minority here and there isn't much I can recommend as it will stay single ship barring a major change in direction.
Thanks for responding. As someone about to start a PhD in what amounts to "military reform" I understand the monolithic nature of vested interests and sunk costs.

In the first HoI there was a great deal of tension between specific historical details and underlying theoretical principles (generalization and abstraction). The tech tree, for example, allowed you to research very specific items: basic service rifles, advanced service rifles, etc. On the other hand it also attempted to work as a generic tech tree for every major power.

In the end the specific details conflicted with the underlying theoretical principles. There was a reason the Americans developed the Garand and why the Germans went on to develop outstanding automatic weapons. Countries ended up with abnormally strong divisions because they cherry picked all of the best techs. The specific details (an overly detailed tech tree), because of other limitations, actually made the game a less accurate simulation. When I refer to "tension," this is what I mean.

HoI 2 resolved this by abstracting the specific details into broader doctrinal paths. You had very large differences between armies based on the type of division they used and the doctrinal path they followed. Now, HoI 2 did flub on a number of crucial components--differing levels of motorization for example.

CORE looks to have corrected some of the glaring problems of the HoI 2 tech tree while preserving much of the emphasis on underlying principles instead of specific details. In particular the new motorization scheme and the differing levels of industrialization is revolutionary.

On the other hand, there are parts of the mod where the cult of specific details prevailed. Modeling the appropriate power and capabilities of a destroyer flotilla (the underlying principle) is far more important to making a historical simulation than making sure every model type of German destroyer is named and made present (specific details).

This tension is present in a number of other places as well. The naval doctrine tree has a beautiful set of techs identifying the various pieces that, when put together, differentiate British from German maritime commerce. Then there are a group of techs at the bottom that give every country different naval values.

Instead of determining the underlying principles CORE fell back to specific details. Blue Emu noted that German ships tended to get the better of British ships in short battles but not long battles because their range finders were more accurate but caused eye strain. So, instead of a hard coded command to give German ships better positioning, give a tech "Early Range Finders" that increases positioning, but lowers defense.

A similar problem occurs in some of the unit models. Instead of all models getting progressively better, some later models are worse than earlier models. Move from destroyer model 0 to model 2 and suddenly your range drops by half. Move from model 1 to model 5 and you suddenly lose 66% of your offensive power.

Instead of building 80 IC torpedo boats with 3 SA (model 1), the computer will build 415 IC destroyers with 2 SA (model 7) and get massacred. There are, of course, subtle trade-offs between the 8 different models of destroyer. Some have better range. Some have better AA. Some have better surface attack. The AI doesn't understand these differences and in the end this makes the simulation less accurate.

I'm afraid that this might sound more critical than is intended. Everybody recognizes that this is a work in progress. Different ideas need to be tested. Bugs need to be extinguished. I played a game until '39 without a single crash or problem--that's a huge accomplishment for a first release and a testament to the coders and beta testers.

Again, what this note is intended to do is to put the spot light on the tension between the desire to simulate every nuance of WW2 and the desire to create a reasonably accurate model that gets all the important parts right.

I hope all that made sense!
 
Last edited:
dec152000 said:
3. Naval builds: The purpose of this isn't to lock a player into a RW setup, but rather to give credit for production that is already partially complete. All of these ships are pretty cheap compared to building them normally, and some are extremely good deals. We used to have them in the build que but Paradox changes to the EXE ruined that setup and we switched to event driven. While not exactly what we want it is pretty good.
What was the change in the EXE? Does the AI now compare the queue to its build priorities (thus canceling all of its cheap ships)?

There must be a more elegant solution that doesn't involve building a navy by event.
 
baylox said:
No, the reduced costs specified in the scenario setups were 'reset' to the cost specified in the unit files after a while in the game, which means that these ships weren't being built cheap after all.

IIRC.

but the progress remains... right? so a battle-ship which costs lets say 2500 IC (IC-cost x days to build) is still cheap if its almost finished - even if its costs per day is as high as in unit-file...
 
If you don't like the naval construction event, why don't you just edit it out in db/CORE2_events like this

event = "mod-CORE2\db\events\other\CORE2_naval_construction_events.txt"

Put a # before

#event = "mod-CORE2\db\events\other\CORE2_naval_construction_events.txt"

Or will that cause alots of problems? Haven't tried it, so can't tell.
 
grendel2 said:
Instead of building 80 IC torpedo boats with 3 SA (model 1), the computer will build 415 IC destroyers with 2 SA (model 7) and get massacred. There are, of course, subtle trade-offs between the 8 different models of destroyer. Some have better range. Some have better AA. Some have better surface attack. The AI doesn't understand these differences and in the end this makes the simulation less accurate.

I'm afraid that this might sound more critical than is intended. Everybody recognizes that this is a work in progress. Different ideas need to be tested. Bugs need to be extinguished. I played a game until '39 without a single crash or problem--that's a huge accomplishment for a first release and a testament to the coders and beta testers.

Again, what this note is intended to do is to put the spot light on the tension between the desire to simulate every nuance of WW2 and the desire to create a reasonably accurate model that gets all the important parts right.

I hope all that made sense!
Well, in some cases the inaccuracies created by the model are actually not as inaccurate as you might think. For instance, while many nations had torpedo boats during World War Two, no nation used them to much effect against capital ships. The traditional 'destroyer' is in fact short for 'torpedo boat destroyer', because destroyers are intended to counter exactly the sort of threat that torpedo boats present. Compared to either submarine torpedo, air attack, or heavy gunfire from capital ships, the threat of torpedo boat torpedoes was relatively small.

So the fact that the AI in its innocence 'misses out' on the construction of torpedo boat fleets isn't as unreasonable as it sounds.

Also, here's another proposed fix. I'm not sure how the naval tech tree works in the new version, but there are 'national doctrine' techs down at the bottom, right?

Give some nations a 'Brown Water Navy' doctrine, and have that doctrine make the advanced destroyer types obsolete. This reflects that a nation whose fleet is intended purely to control its own coastline isn't going to build long-ranged oceangoing destroyers with a full suite of AA and ASW technology; it's going to build torpedo boats.
 
Hi,

The devil is in the technical details. With the current DD the AI always upgrades it's production to the highest model. This occurs even for the units that are currently under construction. so if the AI has a BB that is one day from completion it will magically convert to the better model once the new version is available. This is Paradox at it's worst IMO, as they just slapped an unrealistic band aid on to the engine, rather than addressing the fact that the AI poorly manages it's build que. Along with this fix we ended up with cost reseting. This killed our "production driven" naval builds as they were all using lowered IC rates to make them accurate.

Anyways, Matedow is the real Naval guy for CORE so any major issues need to be run through him. His availability is very spotty though, so I answered your question as best I could. If you want to take up any issues with him I'd suggest posting on the CORE board as he will definitely see it there.

BTW, Torpedo Boats are getting the axe in 0.35 for the reasons described above.

mm
 
dec152000 said:
With the current DD the AI always upgrades it's production to the highest model. This occurs even for the units that are currently under construction. so if the AI has a BB that is one day from completion it will magically convert to the better model once the new version is available. This is Paradox at it's worst IMO, as they just slapped an unrealistic band aid on to the engine, rather than addressing the fact that the AI poorly manages it's build que. Along with this fix we ended up with cost reseting. This killed our "production driven" naval builds as they were all using lowered IC rates to make them accurate.
I think I understand now. Various countries are building a number of older, cheaper ship models. The AI automatically upgrades these ships to the most expensive models thus massively increasing build costs.

Normal HoI2 has a relatively linear cost progression for divisions. The cost of a battleship doubles between model 0 and model 10. On average each model change only costs about 8% more. In CORE the increase is more like 3.8 times and closer to 16% per model.

This again relates to COREs attempt to model all the various ship types. Germany builds a ton of these cheap little torpedo boats. The AI suddenly converts all of these ship types to the highest model destroyer and increases its build costs by nearly five fold. A similar story with submarines (two fold increase).

Given the existing conditions you guys worked with (individual ships, very different model types, rapidly escalating model costs) there doesn't seem to be another great solution. To me, however, this indicates the distinct draw back of trying to model something the HoI engine can't really handle--individual submarines and destroyers instead of balanced flotillas.

I remember back in the first HoI when armor loss rates were being discussed, one of the proposed solutions was to model armored divisions as individual armored battalions. A good idea but it was mugged by the reality of the HoI engine.

I'll take this to the CORE board when I get the chance.
 
grendel2 said:
I think I understand now. Various countries are building a number of older, cheaper ship models. The AI automatically upgrades these ships to the most expensive models thus massively increasing build costs.

Normal HoI2 has a relatively linear cost progression for divisions. The cost of a battleship doubles between model 0 and model 10. On average each model change only costs about 8% more. In CORE the increase is more like 3.8 times and closer to 16% per model.

This again relates to COREs attempt to model all the various ship types. Germany builds a ton of these cheap little torpedo boats. The AI suddenly converts all of these ship types to the highest model destroyer and increases its build costs by nearly five fold. A similar story with submarines (two fold increase).

Given the existing conditions you guys worked with (individual ships, very different model types, rapidly escalating model costs) there doesn't seem to be another great solution. To me, however, this indicates the distinct draw back of trying to model something the HoI engine can't really handle--individual submarines and destroyers instead of balanced flotillas.

I remember back in the first HoI when armor loss rates were being discussed, one of the proposed solutions was to model armored divisions as individual armored battalions. A good idea but it was mugged by the reality of the HoI engine.

I'll take this to the CORE board when I get the chance.

Not exactly.
The automatic model upgrade is a problem in itself, you want nations to build *exactly* what they were building, not the better model they could have built.
The other issue is that even if you specify in the build queue a battleship at 1IC cost, it will revert to it's full cost in the first game hour, thus ruining all the calculations you've made to balance the build cost to the nations budget.
 
gosam said:
Not exactly.
The automatic model upgrade is a problem in itself, you want nations to build *exactly* what they were building, not the better model they could have built.
The other issue is that even if you specify in the build queue a battleship at 1IC cost, it will revert to it's full cost in the first game hour, thus ruining all the calculations you've made to balance the build cost to the nations budget.
Nations not having enough free IC to finish their historic production would seem to indicate a misalignment in either the unit or economic model.
 
You misunderstand the problem - the peacetime builds were often significantly cheaper in a cost/day, since they took 2 or 3 (or more, in some cases) as long to build a ship as the equivalent models built in wartime - however, you cannot model this in game, because your shiny new cruiser (that you are spending 6 yrs building, instead of 3) immediately reverts back to costing you the 3 yr build price, whilst still taking 6 yrs to build - that is done by the game engine, and can only be fixed by making it into an event, as we have done.

Tim
 
I do like the single ship approach. My only concern is, that the AI tends to send stacks with two DD's on ASW missions. Worse enough in the vanilla game, but here it means that it has two DD where a human player might send something like ten SS.

I'm only in '38 in my first game (as Germany), so I have no experience how the CORE AI handles it, but from the vanilla game I'd say the AI sticks to the 2 DD fleets even when it has more than enough DD's for larger fleets.
 
The single ship approach gives you a nice historical feeling, but is rather inconvenient practically due to the large number of units. I at least would not protest if it were removed .... whatever weight my vote counts :rolleyes:
 
HistoryMan said:
You misunderstand the problem - the peacetime builds were often significantly cheaper in a cost/day, since they took 2 or 3 (or more, in some cases) as long to build a ship as the equivalent models built in wartime - however, you cannot model this in game, because your shiny new cruiser (that you are spending 6 yrs building, instead of 3) immediately reverts back to costing you the 3 yr build price, whilst still taking 6 yrs to build - that is done by the game engine, and can only be fixed by making it into an event, as we have done.

Tim

Couldn´t you change naval construction cost drastically as part of the mobilisation effects? i.e. make it cheap and long initially and expensive and short in case of war?

You might even give an option for suspending BB construction, like the Japanese did with Shinano.

Regards

T.
 
Hi,

We've considered this. The problem is that modifiers to production time and cost now effect the ships already in the build que. So an event that cut time in half would immediately finish any ships that are more than 50% complete...not good.

mm
 
Snowmelk said:
I do like the single ship approach. My only concern is, that the AI tends to send stacks with two DD's on ASW missions. Worse enough in the vanilla game, but here it means that it has two DD where a human player might send something like ten SS.

I'm only in '38 in my first game (as Germany), so I have no experience how the CORE AI handles it, but from the vanilla game I'd say the AI sticks to the 2 DD fleets even when it has more than enough DD's for larger fleets.

Hi,

This is a very real issue as we have almost no control of the Naval AI and how it forms fleets.

mm
 
dec152000 said:
Hi,

We've considered this. The problem is that modifiers to production time and cost now effect the ships already in the build que. So an event that cut time in half would immediately finish any ships that are more than 50% complete...not good.

mm

Yes, that would be unrealistic - but would come in handy in some situations :rofl:

Of course it could be done in stages, with rearmament initiative, partial mobilisation and full mobilisation and a couple of steps after mobilisation. That would be dozens of events, though, so not really a viable option - although you are the mod of 200 Munich events, so why not 400 naval mobilisation events. :D