Note: Due to the length of this thread I have provided a extremely condensed summary of the thread in a separate post: here.
PART ONE: Infantry Weapons, Stats and Their Balance
Foreword:
A little under two weeks ago, in the early hours of Easter morning, I was buried deep in spreadsheets, pulling out numbers and turning raw data into something usable, something I’m pretty proud of. It spawned my first thread on this site, Why Reservists are great and Deltas aren’t (MG’s are broken). In many ways it was something of a breakthrough, but it was not without flaws. For all the good information and hard data I had, the analysis was too simple, a number of the assumptions were flawed, I didn’t have the knowledge I have now.
Since that thread I’ve spent an inordinate amount of time scraping information from the game through empirical testing, raw game files, researching the work others have done into how the game works and some very appreciated help from FLX . I've done this in order to piece together a far more detailed understanding of infantry mechanics that I suspect has few rivals. I freely admit, that I still don’t know everything, but this is probably as close as anyone is going to get.
I would like to give special thanks to FLX, of Eugen for diving into the depths of the game engine’s code and pulling out some important equations, Vasily Krysov for sharing his wealth of knowledge about the game and even more thanks to SomethingAwful.com’s “Power Crystals” who took the wargame modtools and turned it into a mass data exporter for us – this would have been exceedingly tedious without and probably wouldn’t have happened at all.
Let’s Talk Weapons:
All units in the game have three weapon slots. For the vast majority of infantry, this is distributed as 2x anti-infantry weapons and 1xanti-tank weapon, or 1xAnti-infantry weapon, 1x Specialist weapon. Fortunately, anti-tank weapons are pretty transparent with regards to their performance, but rifles and machine guns, not so much.
-Primary Weapons (Rifles etc)
The rifle is the primary weapon of every infantryman in the game and it comes in three distinct classes and one subclass: Battle Rifles (with a long rifle subclass unique to militia), Assault Rifles, and SMGs. These also come in separate tiers, for each of the training levels: Militia, Regular, Shock and Elite. Finally, when entering Close Quarters Combat (hereafter CQC), different versions of the weapons are used with different baseline accuracy. With few exceptions, weapons within a bracket all perform basically the same. There are minor variations on many of them, especially with suppression, but they are very small; you can mostly expect a particular weapon-type to perform largely the same no matter what it is.
-Machine Guns (SAWS, LMGs)
Infantry units also frequently carry with them a machine gun of some kind. The game categorises them into two types, “LMG” and “SAW”. The first, LMGs are your plain old static machine guns as found on most line infantry. They cannot be fired on the move, and they cannot be fired in CQC situations. SAWs on the other hand are machine guns that can be fired on the move and do not stop firing in CQC combat. They perform exactly the same regardless of the training level of infantry they are attached to and in CQC there is no hidden accuracy change – they either stop working or do not.
Nearly Number Time
So how can we compare them? The most effective way to do this is to look at everyone’s favourite video game statistic, “Damage Per Second” (hereafter DPS). In infantry combat, DPS is a measure of High Explosive per Second or HE/s, where 1HE does 1hitpoint of damage. The game only displays integer hit points, but “under the hood” hit points are actually calculated as decimals, and a fraction of a hit point is a real thing (You can sometimes see this even when the number of hit points does not match the number of men in a squad).
Technically speaking, all HE weapons, including infantry rifles, are small explosions, and indeed each shot has a splash damage associated with it. However in the case of small arms weapons, the splash is so small that it’s effects are rather binary, that is to say it either hits for full damage or misses. For that reason I have chosen to model infantry weapons as non-splash because it produces much more useful information. Suppression behaves similarly but has a much larger splash radius. Where I have included Suppression, it will be measured in HE/m^2/s, which is a measure of the mean suppression, per square metre. I confess, I have no idea what “1 suppression” damage actually equates to and how these values interact with morale states, but I believe it’s fair to assume that bigger is better. Suppression is tied to the calibre of the weapon, with outputs between “40” and “55” suppression damage points. Generally these differences are made irrelevant by rate of fire.
SHUT UP ALREADY! JUST TELL ME HOW THEY COMPARE
NOTE: At the bottom of the thread there will be a link where you can view/download the spreadsheet these are taken from. The spreadsheet contains much more detail than I felt necessary to present here. NOTE: The following graphs represent weapons at “rookie” veterancy performance. Increasing veterancy will increase effective accuracy – this will move their starting DPS to a higher point, and let them reach the DPS cap from hitting 100% chance to hit at a greater distance, but does not change the rate of scaling, nor the total achievable DPS (except in weapons which could not otherwise reach the cap)
Primary weapons:
Prior to DLC2 of ALB, primary weapons got a major redesign. Since then, they are all based around standard forms for each bracket and then slightly tweaked beyond that.
NOTE: For the sake of being able to see the differences between weapontypes at each training level, I have left the DPS axis unnormalised. This means care must be taken to observe the relative values on the axis when comparing graphs of different training levels. In the spreadsheet, you can find both these graphs and axis normalised graphs for better comparison.
NOTE: Due to the limitations of my knowledge on excel graphs, all weapons have a leading trail up to the point where they start. They actually instantly jump from 0 DPS to their starting DPS at their maximum range.
In my previous thread, I asserted that Long Rifles were the best Militia weapon and that SMGs were the best all round weapon. As we can see, at all ranges outside of CQC, the long weapon does perform the best, closely followed by the battle rifle. This makes it very effective when fighting in forests or shooting at infantry in the open, but a poor choice in block-to-block fighting. SMGs on the other hand perform exceptionally well in urban combat, and surprisingly, better in than the Assault rifle in most situations - in CQC, better than the regular level assault rifle even. There is some weirdness due to their much higher RoF that gets exageratted later. I would therefore recommend SMG militia and Long Rifle militia once more as being very excellent choices where available, so long as you play to their strengths. Other militia don’t really stand out but, they’re still militia – cheap and plentiful, so they don’t actually need to.
Once more we see battle rifles outperforming the other options, but it's all very tightly packed until you get into CQC. Nothing really distinguishes itself here, except the Battle rifles are best outside of ~100m and SMGs are best in CQC ranges.
Things start to get a bit surprising here. As you can see, Battle Rifles flatline at about 245, having hit 100% Acc and subsequently maximised their DPS. Assault Rifles over take them at about this point and continue to grow. The weird part however are the SMGs - because of their much higher rate of fire, (BR=17.1rpm, AR=22.5rpm, SMG=32rpm), the SMGs actually outperform all other weapons from their maximum range onwards. I've actually gone and tested this in game. It's really WTF. Suffice to say, SMGs are goooooood.
Finally we come to elite weapons. Battle Rifles flatline earlier, Assault Rifles flatline at around 175. SMGs behave much more normally here, performing close to, but slightly under the others, until both BRs and ARs have reached their maximum potential and they continue to grow.
I should note that there are a number of weapons that stand out as performing differently than their bracket would indicate, like the G11, the Type 85 SMG, AK-74SU. You can view the full spreadsheet later to get an idea of how different weapons perform and which ones stand out from the pack. On the whole, we can say that Elite weapons perform approximately twice as well as Shock weapons, which in turn perform twice as well as regular weapons, which in turn perform twice as militia weapons… well approximately. The numbers appear unchanged between ALB and RD generally and on the whole I have to say the rebalance was reasonably well done but the rather odd inconsistancies in SMG performance should probably be addressed. Previously I concluded with "good job Eugen", but uhhh maybe not so much in light of the new more accurate results.
Machine guns
I should note that machine guns remain unchanged throughout ALB and RD. They were not touched by the infantry rebalance. I will also note that there has been much speculation as to the performance of machine guns, the value of extra range and RoF. Finally we have the hard numbers to truly compare them.
I have taken a small selection of the machine guns, of high-performing, middle-performing and poor-performing flavours, from both LMG and SAW categories. As you can see, the differences are pretty huge. The Bren L4 is the worst Machine Gun over all at any range except CQC, at which point it beats out all LMG type weapons despite its pathetic damage by having any at all. The M60 and PKM are both abysmal and render their infantry, primarily USA and Warsaw Pact forces, much weaker than their competition.
So let’s look at them in a bit more detail.
In the above image the results are not scaled to ranges, so it’s a bit deceptive, but still useful. All of those represent DPS at max range. DPS 100%CTH represents DPS at 100% chance to hit (shocking I know) which happens at about 175m for all machine guns. We can also see, that besides the armoury lying about the rate of fire significantly (a depressingly common thing), true RPM is one of, if not the most important statistic because it defines the maximum DPS. This is why the MG3 and L.MG3 are at the top of the list. We can also see that the RoF completely over powers any differences in suppression values for weapons making the weapons with high RoF both the best at damage dealing and the best at suppressing.
I previously mentioned that the previous table is not the full story. So what happens when we look at them with range scaling taken into account?
It’s different, but the key messages stay the same. Note how, for example, the L86 LMG, which previously seemed to be the best static LMG, is over-taken by the MG3 before the L86 even gets into range - putting it squarely in the middle performance bracket. The M60 and PKM variants do get a range advantage, for what little that is worth, but once more they are completely and utterly outclassed by the MG3.
-What about suppression? The M60 and PKM should be good at that right?
In short: No. Suppression is tied to calibre, and ranges between “40” for 5.56mm weapons to 55 for 7.62mm weapons. We can pretty easily see that Rate of Fire generally compensates for the lower baseline suppression and then some. The cherry on top of course is the MG3 and LMG3 are both 7.62mm, so they not only have the best DPS, they have the best RoF and best Suppression.
So what do I think about Machine Guns?
I feel it’s pretty fair to say there are problems. The MG3 and LMG3 variants seem pretty Over Powered on their own, and in truth, they probably are. That they are almost exclusively paired with battle rifle infantry helps keep them pretty fair with other primary weapon types at least. The MG3 accentuates the strengths of the BR, but leaves open their weaknesses. In fact it probably makes West German Jager line infantry among the best at block-to-block combat of all nations most of the time (generally ~100m+, occasionally less, CQC kicks in at about 80m, cutting off the MG3 and penalising battle rifle accuracy), and kings of the forest fight, which tend to happen at around ~300m.
On the other hand, I believe many machine guns are just of plain old bad value. Many infantry suffer unfairly from bad machine guns, particularly the M60 variants and PKM variants which make up much of Warsaw Pact infantry and SA infantry. Raising the minimum performance also allows the MG3 to remain kick of the pack, which I am fine with, but not so overwhelmingly so.
Finally, where BLUEFOR gets several high end machine guns, mostly paired with shock and elite infantry, REDFOR is largely left in machinegun mediocrity, making Morskaya Pehota for example, significantly worse than US marines (of bother ’75 and ‘90s varieties).
I would strongly recommend leaving top end machine guns where they are, and compressing the low/middle end upwards in performance so the gap isn’t so huge. Suppression probably shouldn’t be tied solely to calibre either.
So is the current meta the results of the basic infantry stats? Do Special forces need a nerf etc
Honestly? No. The machine gun problems exaggerate some of the issues a little bit, by reducing their value efficiency of their wielders, but on the whole it’s not the source of Red Dragons infantry meta problems. How can I say that? Let’s take a look into the past.
Back to the Past :Airland Battle
Not so long ago, many of us were playing a game that used lots of reservists, lots of line infantry, lots of shock infantry and lots of elite infantry, but at the core of each infantry row, was the regular rifleman. It wasn’t always so however, it took many balance patches, including the great rifle rebalance of 2013 which gave us the exact stats we are using today. Okay fair enough I haven’t actually gone and checked every single unit, but the ones I have are the same and there’s no indication that they’ve gone and reinvented the wheel here. Some of you may remember a time however when it wasn’t quite so diverse. Every “competitive” deck consisted of Mixed-NATO, Probably Cat B or C for the availability bonus, because of all the damn reservists you could get. Someone had caught on that in large enough numbers, these guys outperform line infantry for price, and had incredibly good AP efficiency. It created a meta based around meat grinders on a colossal scale. First to run out of infantry loses. There was a pretty simple fix though; they made militia units national only. Prototypes. No longer could they come in the numbers to overwhelm line infantry and a sensible infantry meta developed afterwards.
This is actually pretty reminiscent of the situation we have now, except with the availability bonuses being tied to nation choice, no longer do you have to downgrade category. I’ll touch more on this later but as in ALB, Militia units are now more cost efficient and activation point (AP) efficient than line units. So when you think about it, the popularity of reservists isn’t all that surprising.
About the data
Using a modified version of the wargame modtools, I am able to mass-extract the raw data tables from the game. In particular this information almost entirely comes from the TAmmunition table and the English localisation files. After importing the data into excel, there is an awful lot of copying, sorting, replacing and typing of values to produce useful tables. Although I have made a reasonable effort to be as accurate as possible, there is a non-zero possibility that small typo’s or other errors exist within the spreadsheets. If you notice something is wrong, please let me know. You can get a copy of the spreadsheet used to generate these graphs from my dropbox here:
infweapons.xlsx
