-
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.
Yes, I agree that now it´s not historical/realistic (too few sinkings) and it is dull. SF got it better, IMHO.
However, I never had any problem sinking CVs in SF, as far as you attacked them with planes and in 3 or 4 consecutive engaments. Surface ships sinking CVs would be totally ahistorical (just two happened during WW II and one of them was a CVE)
Thats also right. Sinking a CV with something else then Planes should be very very rare. Maybe also only 1-3 ships in whole war, for all nations together.
About what you was able. We shouldnt forget, that a human Player will always play better as any AI. He will always focuse on this battles, and do much micromanegement, to save his irreplaceable Ships. The AI is usually much more indifferent about that. So humans will always sink more as a AI would with same number of ships, in the long run.
I remember the time in sooner patches, where the human player could move his ships out of battles very soon..and that way could evade the sinking of his fleet very often. Then Paradox made the rule, that there are minimum times of battles, so you couldnt easily pull back. Such rules make the Sea Warefare much better. Then even the Ai gets equal chances. Maybe we would need even more such rules, to help the Ai to get nearly even with a human.
Not true - some people ( I know 1 such person) were even banned shortly after SF release for bitching about Naval Combat System still being crap and calling out whoever designed it for not knowing anything about naval warfare in WW2.
I find some opinions in this thread interesting, there seems to be a view that WWII ship warfare was not lethal because not many capitals were sunk ** Gasp **. Wrong wrong.
What the admirals in WWII learned real fast was that if you sailed your fleet any where near an enemy coast without air superiority you would end up with your ships either sent to the bottom or badly damaged and a lot of injured / killed crew. Those Navies without air superiority kept their capitals skulking in ports protected by AA and sub nets and moored in such a way to minimise a torpedo hit.
When navies without air power (either protected by land or sea) sortied out they took a huge gamble and basically got punted you can see all the examples *Bismark / Force Z / Japanese fleet in 1944 - because when they were spotted, all sorts of badnews was vectored towards towards them, surface groups; subs; and planes. Navies without air power /sea power skulked in their Fiords or ports avoiding battle.
What the admirals in WWII learned real fast was that if you sailed your fleet any where near an enemy coast without air superiority you would end up with your ships either sent to the bottom or badly damaged and a lot of injured / killed crew. Those Navies without air superiority kept their capitals skulking in ports protected by AA and sub nets and moored in such a way to minimise a torpedo hit.
When navies without air power (either protected by land or sea) sortied out they took a huge gamble and basically got punted you can see all the examples *Bismark / Force Z / Japanese fleet in 1944 - because when they were spotted, all sorts of badnews was vectored towards towards them, surface groups; subs; and planes. Navies without air power /sea power skulked in their Fiords or ports avoiding battle.
You forgot submarines
You´re right, I think that´s the other main divergence from reality (apart from the huge number of engagements among major surface ships which we see in this game while IRL those kind of engagements were very rare): subs are useless against combat ships, especially against major (CVs, BBs and BCs) combat ships, when IRL subs sank ships in numbers far bigger than surface attack ships.
I also think (taking real life engagements) that encounters among major surface attack ships were quite lethal (in fact 50% of the times they were lethal for one of the major ships involved) BUT if you made them so lethal in this game as they were IRL then people would find very soon that they (players) or the AI (or both) have almost no ships. Why is so? Because in this game there happen far more engagements among big ships (BCs or BBs) than IRL. IRL (and not counting the 3 engagements among the British navy and French one) we got 16 engagements in total where both sides had one or more BB or BC among their numbers (8 of these engagements ended with one of the major ships involved being sunk). 16 engagements IRL means an average of 2.6 engamenets PER YEAR of war. In this game you can easily get that vey same number PER MONTH because AI usually sails its ships very happily. So, to keep number of sunk ships in historical numbers we can teach the AI not to sail its ships so happily (something that would be difficult) OR make BBs and BCs not as deadly as they were IRL, or make them more difficult to sink than they were IRL.
as you said, maybe tweeking this values a little bit??
NAVAL_COMBAT_ORG_DICE_SIZE = 1,
NAVAL_COMBAT_STR_DICE_SIZE = 10,
NAVAL_COMBAT_CRITICAL_HIT_DAMAGE_MUL = 10,
NAVAL_COMBAT_CRITICAL_HIT_DAMAGE_CHANCE = 10,
NAVAL_COMBAT_ORG_DAMAGE_MODIFIER = 1.0, -- Average damage is the same, but it will be more consistent with less randomness.
NAVAL_COMBAT_STR_DAMAGE_MODIFIER = 0.05, -- Average Strength damage is a bit higher and there will be much more randomness in it!
or:
tweeking the "distance" values and the "positioning" values.
Maybe make positioning for capitals ships much better then for screens??
Very nice work X_MasterDave_X. I think it should be used as a basis to rebalance the Naval Combat System
The goals should be that sunk ship statsistics is similar to real WW2 in a hands-off game:
- The total capital (and escort) ship losses
- The ratio of ships sunk in service and in-port -> Something similar to the massive Pearl-Harbour surprise attack should be possible
- The ratio of ships sunk by planes and sunk by enemy fire
I think also the role of intelligence is not taken into account well-enough -> The present model cannot reproduce the better results that the Allies obtained in the Pacific or Mediterranean combat thanks to that.
I have no understanding of how the naval combat is caclulated, but if someone can point me towards a good explanation, I could do some testing by tweaking the variables.
The goals should be that sunk ship statsistics is similar to real WW2 in a hands-off game:
- The total capital (and escort) ship losses
- The ratio of ships sunk in service and in-port -> Something similar to the massive Pearl-Harbour surprise attack should be possible
- The ratio of ships sunk by planes and sunk by enemy fire
I think also the role of intelligence is not taken into account well-enough -> The present model cannot reproduce the better results that the Allies obtained in the Pacific or Mediterranean combat thanks to that.
I have no understanding of how the naval combat is caclulated, but if someone can point me towards a good explanation, I could do some testing by tweaking the variables.
Nice suggestions. above.
Dear Paradox,
can you please improve re: balance naval combat we give you the following suggestions:
1) If AI navy has 33% or less hull strenght add up all ships of at war enemy faction (keep AI surface forces in port). The current AI of all ships need to be on a sortie mission all the time is dumb.
2) Increase naval combat lethality over current system if there is a fleet encounter... so that capitals can be sunk.
3) Rebalance current sub destruction / anti convoy effectivness rates - to previous SF / Hoi 2 versions tweaked versions.
4) Increase BB to historic realtive speed to other classes of ship. (my hobby horse).
Thanks heaps - from a player that wants a slightly better naval simulation in a great WWII game.
Dear Paradox,
can you please improve re: balance naval combat we give you the following suggestions:
1) If AI navy has 33% or less hull strenght add up all ships of at war enemy faction (keep AI surface forces in port). The current AI of all ships need to be on a sortie mission all the time is dumb.
2) Increase naval combat lethality over current system if there is a fleet encounter... so that capitals can be sunk.
3) Rebalance current sub destruction / anti convoy effectivness rates - to previous SF / Hoi 2 versions tweaked versions.
4) Increase BB to historic realtive speed to other classes of ship. (my hobby horse).
Thanks heaps - from a player that wants a slightly better naval simulation in a great WWII game.
There is an underlying problem that will not be solved - maybe not even in HoI4. The Naval units need their own command structure - with objectives, stances and action limits. As mentioned above, fleets would operate in areas where they had air superiority or else they suffered (lost units) - a good thing for some action limit would be if (BB strength < action limit, all BBs stay in port) and that action limit should be set per national level for majors and group by mid and minor nations. The real programming question is "How does the Naval unit know where areas of air superiority exist?"
That requires some major work in the exe and then creating specific hooks for Land/Air/Sea commands AND how those separate Commands can work together. I'd say that Land/Sea/Air should all go beneith the same Theatre Command but I can see the potential problem of combined arms actions - coordinating the Amry Group and Air Group to attack the same sector consistany either requires the ability of Air Groups to be defined as TAC or SAC (with some special tech tree research to increase the abilites of each). I believe it is not an easy fix, the coordination issue is probabaly insurmountable without major exe rewriting if the example of Minister control of trade/production license can be extrapolated - so instead of having any non-human nation basically not able to use the production license - it would ruin the game if air units could not coordinate attacks with land units as much as they currently do.
Another approach which might work (but again - in some future reworked game) would be Specific Responsibility Theatres (which would overlap geographically) and solve so many other problems
Garrison/Suppression Theatre: defined genearl area with stances and important objectives
Combat Land Theatre: with subordinate Air and Land units - like the current Theatre works
Combat Air Theatre: basically like the SAC or Bomber Command of US/UK
Combat Sea Theatre: Nimitz Command
Combat Land/Air/Sea Theater: MacArthur's Command
Then you might get Naval units behaving in a more historical way with periods of probing and recon (small engagements), then a brief very bloody battle and then another period of gathering forces to prepare for the next cycle.
That requires some major work in the exe and then creating specific hooks for Land/Air/Sea commands AND how those separate Commands can work together. I'd say that Land/Sea/Air should all go beneith the same Theatre Command but I can see the potential problem of combined arms actions - coordinating the Amry Group and Air Group to attack the same sector consistany either requires the ability of Air Groups to be defined as TAC or SAC (with some special tech tree research to increase the abilites of each). I believe it is not an easy fix, the coordination issue is probabaly insurmountable without major exe rewriting if the example of Minister control of trade/production license can be extrapolated - so instead of having any non-human nation basically not able to use the production license - it would ruin the game if air units could not coordinate attacks with land units as much as they currently do.
Another approach which might work (but again - in some future reworked game) would be Specific Responsibility Theatres (which would overlap geographically) and solve so many other problems
Garrison/Suppression Theatre: defined genearl area with stances and important objectives
Combat Land Theatre: with subordinate Air and Land units - like the current Theatre works
Combat Air Theatre: basically like the SAC or Bomber Command of US/UK
Combat Sea Theatre: Nimitz Command
Combat Land/Air/Sea Theater: MacArthur's Command
Then you might get Naval units behaving in a more historical way with periods of probing and recon (small engagements), then a brief very bloody battle and then another period of gathering forces to prepare for the next cycle.
Change this:
NAVAL_COMBAT_STR_DAMAGE_MODIFIER = 0.05, -- Average Strength damage is a bit higher and there will be much more randomness in it!
to this:
NAVAL_COMBAT_STR_DAMAGE_MODIFIER = 0.5, -- Average Strength damage is a bit higher and there will be much more randomness in it!
and naval combat is sorta like before ftm again with regards to sinking a few heavies and the assorted clusterfark surrounding them (screens) during legendary battles
...in defines.lua
NAVAL_COMBAT_STR_DAMAGE_MODIFIER = 0.05, -- Average Strength damage is a bit higher and there will be much more randomness in it!
to this:
NAVAL_COMBAT_STR_DAMAGE_MODIFIER = 0.5, -- Average Strength damage is a bit higher and there will be much more randomness in it!
and naval combat is sorta like before ftm again with regards to sinking a few heavies and the assorted clusterfark surrounding them (screens) during legendary battles
...in defines.lua
Last edited:
This is a good point. The AI does not understand where it can assert dominance (sail freely), where the enemy can assert dominance, and where dominance is contested. And among the contested areas, it is not good at telling which areas are important enough to risk losses and which areas are not.
Frankly, the AI is dumb. And the only reason that surface combat works as well as it does is because it can be reduced to a numbers game and the front is defined by where units are placed, so a lot of the intricacies are paved over with the simplifications of the engine.
To a large degree, the biggest rough edges of the engine exist to accommodate the AI. And this is not the fault of Paradox -- AI is really, really hard. Most publishers never attempt games like this because the AI is so hard!
What I'd really like is general moddability of the AI, similar to the Brood Wars API for Starcraft (see for instance the Starcraft AI competitions). But the reason combat AI is not exposed and in Lua is primarily that it has to be fast (HOI3 is more complex and has far more processing power devoted to AI than Starcraft), and modifying the HOI3 engine to expose those internal bits in a stable, reliable fashion without Lua is an incredibly difficult engineering challenge that I'm sure Paradox is not enthusiastic about undertaking.
Besides which, even if modders do fix individual elements of the AI, then they'll of course want to sand off all those rough edges that exist because the AI is broken... It never ends.
You have to begin with a tiered approach to the AI strategy, much like Chess AI's or the new verbal AI which won that jeopardy match in February.
The first tier is rating your naval strength and weighing probable enemy strength. Much of that can be done by out of game analysis, so that it becomes a set of baseline numbers rather than a set of intricate calculations.
that gives teh data (weighted relatice strength) to determine sea dominace decisions.
Okay, here is just one of my many Naval Combat complaints -- When two opposing naval forces meet, combat between the two should be solely at the descretion of the faster of the two ships/task forces. Remember, speed kills, and this was the whole thought process behind building Battlecrusiers in the first place -- whatever they could not out-gun, they could out-run.
But as it is, my BC's are forced to engage in combat opposing naval units that are far slower. And as far as I know, this has no effect on the disengagement timer either.
I can possibly understand the concept of positioning before the advent of radar, but afterwards, it became almost impossible to just happen upon an enemy fleet and not realize it had happened until both parties were at a medium or closer range (20,000 yards or thereabouts).
But as it is, my BC's are forced to engage in combat opposing naval units that are far slower. And as far as I know, this has no effect on the disengagement timer either.
I can possibly understand the concept of positioning before the advent of radar, but afterwards, it became almost impossible to just happen upon an enemy fleet and not realize it had happened until both parties were at a medium or closer range (20,000 yards or thereabouts).
Yes b/c two ships could never collide in a thick fog when they both are equipped with radar. (Andrea Doria)
the other BB was Tirpitz. after the Bismarck got sunk, Hitler was scared of losing that one too, so he set it to coastal patrols aorund Norway, where she was sunk while at anchor by a superbom dropped by the RAF. Graf Spee was a pocket-BB. A mix between a BC and a BB. Turned out to be a flawed design. (not heavy enough to outfight BC and not fast enough to outrun BB)
"Yes b/c two ships could never collide in a thick fog when they both are equipped with radar. (Andrea Doria)" Loatze, you might want to re-read the different versions of the sinking of the Andrea Doria. Both ships involved "saw" the other on via radar and therefore were completely aware of, the other. In my mind, this alone would support my theory that once radar was in widespread use, the faster of any two ships or task forces involved in a potential engagement had the sole descretion as to whether or not to engage the other.
The case of the Stockholm ramming the Andrea Doria is much more one of human error -- there is a fundamental difference between attempting a safe passage of another ship in a fog within the confines of a busy shipping lane and avoiding it altogether in the middle of an otherwise empty ocean.
I see this as a false analogy and still would like to know why my faster ships are obliged to engage slower ones -- sometimes significantly slower ones. This is simply not realistic.
The case of the Stockholm ramming the Andrea Doria is much more one of human error -- there is a fundamental difference between attempting a safe passage of another ship in a fog within the confines of a busy shipping lane and avoiding it altogether in the middle of an otherwise empty ocean.
I see this as a false analogy and still would like to know why my faster ships are obliged to engage slower ones -- sometimes significantly slower ones. This is simply not realistic.