Thought I would put this up to get some feedback and discussion going on changes I've been working on over the past 1.5 years for the second generation of Alternate Washington Naval Treaty scenarios; some of these will no doubt find their way into an adaption for Scen 1 and 2 or DaBabes.
Note that all the changes here are broad concepts, not specific change lists, which would be impossible given that the project constitutes a ~75% reconstruction of the database (only thing not really touched is bases and leaders/pilots so far).
The project is based on DaBabes scenario 28 as a baseline, and has evolved from there. Many of the things here were made possible thanks to the efforts of JWE and michaelm, so many thanks to them for that.
There is currently a 'test' scenario in development consisting of an alternate history Marianas battle, which will hopefully be released later this month. This scenario feature most of these changes, and has been used to fine tune many of them. I will look at following it up with more ‘small’ scenarios, both historical and alternative, in hopes of letting people get a feel for the changes as I work on finishing the campaign versions.
Theres a lot to go through, so Ill probably do one section at a time whenever I have time for it.
The core of the changes, devices have seen probably the most work, which is reasonable seeing as they make everything tick.
Some of you may have seen my post in radar gunnery thread where I outlined these changes, but I will do so again here as briefly as possible.
Every single naval and coastal defence gun has been redone, with a consistent formula and system used for calculating both accuracy and penetration values; my plan is to release the spreadsheet I have used for this in due time in case anyone wishes to scrutinise it or use it for their own changes.
Almost all the devices are now named in a format as follows;
(Bore)/(Length) (Model) (Elevation/Purpose) (Modifiers)
For example, the United States 16 inch gun on the Iowa's would be;
16in/50 Mk7 45* ++
Whereas the Japanese 12.7cm Type 89 would be;
12.7cm/40 T89 AA -=
(Bore) is the bore diameter of the weapon, in appropriate units for that country.
(Length) is the length of the weapon in calibres.
(Model) is the designation used for the device.
(Elevation/Purpose) is the maximum elevation of the weapon, or if it is a DP or AA gun.
(Modifiers) are a series of symbols at the end of the device name, with different meanings; there can be up to 3 of them, and they are described below. These are included so the player has some feedback on his weapons systems; they could easily be omitted from the final display.
The first is the 'fire control quality' modifier, and represents the overall quality of FCS associated with the weapon. This has pronounced effects on the weapons accuracy for both anti-air (AA) and anti-surface (ASu) purposes. This can vary a lot depending on the platform (merchants tend not to rate too highly, for example), the date and the nationality of the system.
The second is the 'rate of fire' modifier, and represents how the weapon ranks up against others of its kind in terms of RoF. The easiest example here is the 5in/38 Mk12 in its many variants; the BB EBR mounts have the best cycle and acquisition times, followed by the slightly slower DD and CA/CL mounts. Even worse than these are the open mounts used on some of the older DDs. This modifier along with rate of fire has a very significant impact on AA accuracy, and a lesser one on ASu accuracy.
The last symbol if present is always hash '#', and represents a weapon that has undergone an ammunition upgrade. Examples of this include the new shells issued to US battleships in late 43/early 44 which increase penetration of those weapons by ~10%.
The first and second symbol's use the same system for representation, which is as follows;
^ is Excellent
+ is Good
= is Average
- is Poor
. is Terrible
So if you see a weapon rated as a ^=, you can easily deduce that this is a weapon with an excellent FCS and average ROF for its model (for example one of the main battery guns on one of the USNs rebuilt standard BBs).
As mentioned before, anti-air and anti-surface accuracies have been split, as has the effect value for AA, which is now based on explosive volume rather than direct shell weight. As a result, some weapons have been brought more in line with their historical capabilities, and I illustrate this with the destroyer guns of the IJN and USN.
In the stock scenarios the IJN 12.7cm/50 in its various forms rates an accuracy of 59 for both ASu and AA (same for all the versions); historically this was a rather inadequate weapon, with only the Type D being a solid DP gun. Its anti-surface capabilities were not particularly great either. Under the new system, the Type B merits a 39 for ASu and a 38 for AA accuracy. The later Type D gets a 44 for ASu and a 48 for AA.
Contrast this to the USN 5in/38 Mk12 EBR, which as a result of attempting to model its AA capabilities tends to suffer from being too good at anti-surface work, particularly in the very early AE databases where it had an accuracy value of close to 200. In the current database its rated as 62 for both AA and ASu, making it only marginally better than the IJN weapon. Under the new system the 5in/38 Mk12 += which is a common destroyer variant, rates a 58 for ASu and a 74 for AA accuracy. This is more in line with its historical capabilities, and it shows in tests.
Penetration values have been redone using values from Naval Armour and Ballistics (which is in turn based on Nathan Okuns work) as a baseline, with minor alterations as needed. They are all calculated against the same standard, which is US Class A armour. Armour quality modifiers have been moved to the actual armour values of ships.
In general penetration values tend to be slightly higher than in stock, which serves to both make larger ships more vulnerable (a Kongo class BB getting shot up by USN cruisers was almost impossible in the original system) and make surface combat always carry a risk if facing equal forces.
As an example, the US 16in/50 Mk8 has gone from 828 in stock to 966 here (before the ammo upgrade), while the IJN 20cm/50 3YT-2 has gone from 291 to 342 here.
I'm happy to provide details on any specific weapons people want to look at.
Lastly, CD guns, which use the same system, though their designation replaces the modifiers with a 'CD'. These weapons were rather varied in stock, and in many cases with inconsistent data; this is now corrected, and in general these weapons remain something you want to avoid, particularly anything in the 6 inches and upwards category.
Like Naval guns all AA weapons have been revised; this includes both the large caliber land-based AA guns, which use a formula similar to that used for ship-based DP and AA weapons, as well as the smaller shipbased and land-based automatic weapons like the 40mm and 25mm. For these weapons, I used a formula similar to DaBabes calculation for them, along with adapted values for anti-surface actions.
Like Naval guns, ship-based light AA devices include modifiers, though the system is much simpler; the weapon name is followed by a +/=/- to indicate the quality of the mounting and its associated fire control. This only applies to mountings that used fire direction, pedestal mounts like the HMGs and 20mm do not use this system.
In addition, the Japanese 25mm has been split into 3 different versions for the single, twin and triple version with appropriate accuracies to reflect the inefficiency of the triple mounting.
In general, AA is much more effective than in stock, particularly for the allies, though this is also largely caused by the better AA accuracy and effect values for DP guns in addition to changes to light AA. As a recent example, a carrier clash in the test scenario saw me loosing ~130 aircraft to AA as the Japanese compared to some ~15 for Allies. This is in August '43.
A conversation with csatahajos earlier this year about the incredible strength of aerial torpedoes against even very large battleships led to an experiment where I cut their effect values in half. This has yielded very good results, as the smaller 18" aerial torpedoes are no longer as devastating as before against large targets, while the larger torpedoes retained most of their power.
However, recent results have led me to experimenting with using 75% instead of 50% of effect for these weapons, but I am unsure if this will be final. I may opt to use 50% for aerial and 75% for surface and submarine weapons.
Regarding surface launched weapons; like guns these will come with a rating for accuracy using the simplified +/=/- system to represent quality of torpedo directors on the ship (mostly old DDs getting a -, most getting the =, and USN late war systems with radar used for torpedo direction getting a +). Like the older AltWNT scenarios, wakeless torpedoes like the Long Lance and electric torpedoes receive a small bonus to their accuracy values.
I was experiementing with some changes to these earlier in the year, but JWE's work for DaBabes forms the basis for the current changes here; for those unfamiliar with it, the changes cut bomb effect values in half, along with changing the formulae for anti-soft and anti-armour values considerably.
My own tests with these changes gave me the feeling that while the air-to-ground issues were resolved, the effect of these weapons against ships was now too low; as a result I am now testing a system of effect at ~60% for GP, ~45% for SAP and ~30% for AP of bomb weight, compared to DaBabes original values of ~50%, ~30% and ~15% respectively. Anti-soft and anti-armour values will continue to use DaBabes formulas.
These have seen 2 major sets of changes; the first was the recalculation of all effect, accuracy, etc values based on my own formulae, from basic data like shell weight, explosive filler, muzzle velocity, etc. Accuracy is a function of ballistic properties and rate of fire (the 'lead hose' effect), while effect is calculated as 'weight of fire', a function of shell weight, rate of fire and HE content. Like with naval changes, the goal was to create a framework that would calculate values for these that did not fall too far outside the original games data set. This way, any new fictional weapons or weapons not included in the original database could be added and would fit right in.
As an example of changes, here are a few weapon comparisons. Changes are shown as X->Y, where X is the stock value and Y is the new value. Values with no arrow remain the same;
.50 Browning MG
7.7mm Type 89
20mm Type 99
20mm Hispano (MkII)
Once again, if anyone would like to see a certain weapon compared, just ask.
The second change I mentioned is to seperate out defensive weapons from others, mainly as a way to deal with the excessive capability of heavy bombers to shoot down large quantities of fighters, even at night.
Defensive variants of a gun are marked with an * for those mounted in powered or ball turrets, and a *- for those mounted in flexibile positions. Those in turrets have their accuracy value reduced to half, and those in flexible mountings have it reduced to a third. This has given very good results in testing, as bombers can still drive off attackers by weight of fire, but shooting down fighters is less common, though it still happens, especially using capable bombers like the B-17 or B-29.
One of the first things I did when I started on Mk2 was creating a spreadsheet to handle aircraft data, and to automatically calculate things like Extended and Normal ranges, Endurance, as well as useful non-game data like fuel efficiency and drop tank range effects which I could use to make sure all the data was sensible.
A lot of ranges were tweaked a little, and of course any new aircraft or non-historical models just had them extrapolated. The general pattern for ranges in stock is that Normal is 75% of Extended, which is ~70-95% of Maximum depending on aircraft type (though note that the game automatically halves Ext and Normal ranges since they involve a return trip. I chose to keep these ratios, though I did consider dropping the Ext/Normal ratio to 60% or so since it would push up Ops losses as a result, but so far there does not seem to be a need for it. As a general rule Extended was set as 80% of Maximum, with the except of recon aircraft which are at 95% and bombers with space for internal fuel which are at 70-75%.
Service ratings were completely redone, with all seeing increases. An interesting discovery I made while working using the .csv files is that although service rating is capped at 5 in the editor, if a value greater than this is used on a .csv and the scenario repackaged, the game will happily recognize that value. Im still doing tests on this, but if it works as advertised, then the likely values for service ratings will be something like below. Note that these are just guidelines, every plane will be asessed individually.
3 - Radial Single Engine (F4F, etc)
4 - 'Difficult' Radial Single Engine (A7M, B5N1), Radial Twin Engine (A-20, Ki-48), Inline Single Engine (Spitfire, Ki-61)
5 - Large Radial Twin Engine (B25, G4M, Ki-21), Inline Twin Engine (P-38), Jet-Prop Hybrids (FR-1)
6 - Four Engine (B-17), Jets (P-80)
I should note that the differences between 4 and 5 for example are much lower than those between 2 and 3, because of how SR works. What happens is that the SR is used to divide 100 to work out how much wear the aircraft can take before it needs maintenance; this is visible on the Planes tab of a squadron. So SR3 is 33, SR4 is 25, SR5 is 20 and SR6 is 16/17. As a result, I might even push up the upper end of the scale to SR8, which is 12/13. Probably something that needs a lot of testing.
My ongoing game with csatahajos uses the early version of this with everything rated from 3 to 5, and it has a noticable but realistic effect on sortie tempo; its possible to stage strikes with 100% operational squadrons, but dont expect to do so for several days in a row, availability will quickly drop to ~70-80% for single engine bombers, and less for the bigger stuff. Couple this with Babes style AV support limits for real fun.
Another thing that I had issue with was the marked difference between the way MVR values for the allied and japanese planes varied. Take for example a comparison of the F6F-3 and the A6M5. The F6F-3 is rated as 17-17-17-17-14, with the A6M5 as 29-29-29-23-17. This just seems off, especially as even though both aircraft had superchargers, decreasing air density would result in reduced maneuvering ability even if engine power was unaffected; instead the values are constant until the critical altitude is reached and then drop off.
The same 2 aircraft under the new system I use are as follows; 24-22-20-18-14 for the F6F-3, and 31-28-25-20-13 for the A6M5. These are probably the most fluid part of the changes to aircraft right now, and have gone through about 6-7 iterations so far, with more to come as I get a chance to do more testing. In brief, I try to account for wing loading and general manuverability (the base value), the effects of super/turbochargers (fall off with altitude, point of steeper falloff), along with the effect of reduced air pressure resulting in less manueverability, with acrobatic aircraft suffering more.
An intesting side effect of testing this was that the stratospheric sweep effect is much less noticable, especially in the early war; even an unlayered CAP will do resonably against them, though layered is always better. I have yet to see how it plays out with late-war aircraft like the P-51H, etc.
Production and Research
Without going into too much detail on industrial changes, suffice to say that as a result of Japan having less HI overall, each point invested into aircraft will be that much more 'expensive'. This, coupled with the fact that the repair cost for each engine factory type can be different (from ~1000 supplies for Ha-33/35, to ~3000-4000 supplies for late war versions), should allow a more realistic representation of the limitations faced historically.
When it comes to R&D, I identified 2 main things that caused unrealistic results as Japan; the first was the sheer number of R&D centers, meaning you could easily afford to use 5-10 on an aircraft you really wanted, each at the 'optimum' size of 30-40. By cutting the number of R&D centers to two dozen or so, the player is forced to be much more selective about what to research, assuming realistic R&D is on of course. The second was the crazy ability to 'upgrade' an R&D plant without damage to any plane in an upgrade path. This was dealt with by removing the natural upgrade paths for all IJ aircraft; the ones ones where a shift like that is permitted is within subvariants (Ki-44-Ia to -Ib and -Ic), or when moving to an 'alternative version' (G4M1 to G4M1-L).
Thats it for today; Ill look at aircraft related devices more when I update this, followed by land and industrial changes. That will probably be followed by changes to ship classes, and lastly aircraft.
< Message edited by JuanG -- 1/26/2012 11:56:34 PM >