My name is David Sláma and you may know me by my nickname Locklear from social media and the Bohemia Interactive Forums. Before I joined the Project Argo team as a Designer, I had been part of the Arma 3 Encoding Department (also known as The Department of Black Magic, which the Arma modders out there will surely understand). As someone who has been responsible, among other things, for characters, gear, and weapons configuration, the majority of my work on Project Argo involves weapons, damage, and soldier protection. This proved much more complex (and honestly, fun) than I expected. It's a pleasure to share some more details about our approach to combat in Project Argo in this devblog.
In one of our previous devblogs, you learned that it's important for us to provide an "Authentic Feel" and "Lethal Combat" in Project Argo. Luckily, as one might think, these two elements have always been a strong aspect of the Arma games, and since we're building upon Arma 3, our job in this field would already be done. However, it took just a few playtests to see that this assumption was utterly wrong. We found that the weapons indeed felt authentic and combat was lethal, but we had to remind ourselves that, in the end, Project Argo is not Arma 3, and what works in Arma 3, is not necessarily suitable for Project Argo. The gunfights and gameplay itself are different from what we were used to, and we needed to find synergy between our new game modes and combat.
To describe the whole problem in a nutshell: Arma 3 weapons were way too effective in Project Argo – especially in Clash and Raid, where there are no respawns. Compared to most first-person shooters, the famous and authentic Malden locations are much more open and less linear, which meant that players were often killed from unexpected directions, without any chance to react. Constantly dying from a stray bullet simply wasn't fun. Also, I must admit I was completely freaking out about the office playstyle back then, which was to randomly empty a whole magazine into a building or a wall, hoping to kill an enemy hiding behind. Thus, change was needed.
In many games, characters have a certain amount of hit points, and weapons have a clearly defined damage value. This makes balancing relatively easy. However, this approach is also quite simplistic and abstract, so it is only natural for a military simulation game like Arma to offer something more complex, and more authentic. Of course, in Arma there is a damage value for every type of ammunition, but that's merely one variable in a larger equation. If you delve into Arma 3's damage model and soldier protection, you'll find that the result of a hit is affected by many things such as the bullet speed, impact area, depth of the wound, number of body parts penetrated, phase of the moon (just kidding!), and others. Various parts of the character's body receive damage separately, with a portion of that damage being converted into a so called total damage. Once that, or one of selected parts, reaches a certain value, the character dies. In total, Arma 3 divides the body into 11 separate parts which can be damaged, but also protected. Frankly, while it's great to have such level of depth in Arma 3, balancing something like that for a competitive shooter would just be insane.
The first step in reworking the combat for Project Argo was to clearly define what we expect from which kind of equipment. We already knew that we'd want to provide some sort of customization in the future, which meant we had to adjust the equipment to reflect this. In Arma 3, for example, CSAT soldiers wear armored uniforms, which compensate the absence of a plate carrier vest in their inventory. However, in Project Argo, we eventually want players to select how they look like based on their taste, not based on the number we associate with certain gear. Therefore, we adjusted the defensive capabilities of all uniforms to be the same. Likewise, we removed armor from helmets – a controversial move indeed, and a good example of when we need to decide between gameplay and striving for authenticity – yet this decision allows you to wear that lovely boonie hat without feeling like you're compromising your protection. Next we considered the role of vests, which have several categories in Arma 3 due to their various capabilities in protecting different parts of body. We felt that system would be overly complicated for Project Argo, so we opted for three categories instead, with just two (medium and heavy) providing additional armor.
Together with defining the roles of all types of equipment, we also made a significant change to the damage model itself. We decided to divide a character into just four hit zones – head, body, arms, and legs – and to base the damage model on these. Considering the 50 weapons a player can use, with many different calibers and three different levels of protection, we set up a baseline for how many hits to the body part from a certain distance are needed for a character to die. Obviously, hits to arms or legs should be less lethal, whereas a headshot should, under normal circumstances, always kill, no matter what weapon is used. This very simple idea was enough to change a remarkable amount of numbers in configuration files to come to a brand new damage model.
On the technical side of things, Project Argo uses Arma 3 configuration, so we had to make sure we properly overwrite everything needed. We changed the base armor values of, well, everything – plus the values that translate damage from body parts to total damage. That is very important for the game to determine if a character should be dead, which in Project Argo happens when either body, head, or total damage exceeds 1 (total damage also ensures that it is actually possible to kill someone just by shooting their arms or legs – even though the body part damage doesn't do the trick, the total damage will eventually reach the death threshold). The new "defensive" values meant there should be completely new damage values for all weapons. As you can imagine, our time with a pen, paper, and calculator was spent well, since the calculations worked mostly as expected in the game once everything was properly configured. Done, right? If only…
Understandably, when an implementation works as designed, it doesn't automatically mean the design is good. So we entered a phase of heavy playtesting and (sometimes funny) problem solving. For example, we found that we suddenly had very, very sturdy windows, which refused to break until they took many hits. Eventually, we were able to teach them that Project Argo weapons use much lower damage numbers than Arma 3, and as long as they don't mind, we'd like them to break a bit more easily (doing this without breaking anything in Arma itself is not as simple as it might seem). Fortunately, all this tinkering also helped to stop people from shooting through every wall, since bullets were now only able to penetrate thin surfaces. With most of these problems resolved, we could finally begin to evaluate the gameplay of gunfights.
In general, we liked how the game had changed. We had skillful snipers taking us down from afar with a bullet or two, we had front-line soldiers utilizing their heavy vests to overcome enemies with stronger guns in very short wars of attrition. And we had myself dying frequently because I was not paying attention but appreciating the view. True balance. Actually, not yet. We quickly realized that balancing damage will take quite a while and that there is no thing as the right one. No matter what the settings are, there will always be people who feel it's too arcade while others complain that it's too hardcore. And that's alright, of course, for no two players have the exact same preferences. However, we believe that we can find the balance between lethality of combat and fun which we feel is right for our game. We want you to be rewarded for smart plays and skillful weapon handling, which also means we generally want you to have a little bit of time to react when you are under fire, to possibly even turn the engagement around. Achieving this is an ongoing process, and we closely monitor all your feedback together with a lot of data, which provide both subjective and objective grounds to base our decisions on.
During the first week of the Open Prototype, we received a lot of feedback from you telling us the weapons are underpowered. Based on the hard data, on the other hand, it seemed that the damage was not as low as it felt. So we had two sources of information, both similarly important, but contradicting each other. This might seem impossible to solve, and it meant we first needed to make sure we understood all the feedback and data correctly. Watching people play and analyzing our own experience showed us that players quite often thought they had hit a target with more bullets than they actually had. The damage statistics, on the other hand, showed that vests probably had lowered the damage somewhat more than we'd like for the kind of gameplay we intended. For the second week of the Open Prototype, we tried to address both issues: we made on-hit particles more pronounced, so hits would be easier to notice, and we slightly lowered the armor of vests to lessen their impact on gunfights. And we continue to listen to your feedback and we're analyzing the data to be able to make informed decisions about future tweaks.
Even though, eventually, we plan to let you unlock equipment and define your own loadouts, this isn't the case yet. So for the moment we had to find an alternative method. We started with the possibility to pick different weapons, but after playtesting for a few weeks, we noticed that everyone was running around with the Zafir. We accepted that weapons on their own are not balanced (and shouldn't be), and so we started to look for the balance elsewhere. As the first drafts of the metagame begun to shape up, we found that we can try to simulate the possibilities of unlockables via predefined loadouts. The advantage of this approach is that it's possible to balance sets of weapons, vests, and grenades as a whole. Plus, the intentional side effect of having various loadouts is the viability of these loadouts in Project Argo's game modes; Clash, Link and Raid each place you in different situations, where different loadouts and playstyles shine.
The question then was how to easily communicate to the player what to expect from each loadout, especially if they didn't play Arma 3 before, and the weapon names sound completely alien to them. Thus, roles, or classes, were born. While you might have no idea what "MAR-10" or "SPAR-16S" stands for, we were sure that keywords like "Marksman" or "Machine Gunner" would ring a bell. In the end, we created five roles: Rifleman and Grenadier with less powerful weapons and better protection, Marksman as a glass cannon, Scout being somewhere in between, and Machine Gunner as fire support. You can find a bit more information about the roles in the Field Manual in-game.
Even though we intentionally created the same number of roles as there are members on a team, we never expected to often see all five roles represented in a team. Each role has a purpose in all of the modes, and it's only natural for some to be preferred, but fortunately the data so far shows relatively balanced preferences (except for LIM-85, which we honestly have had some difficulties to find since we pulled out the Nerf Hammer). Anyway, this is also a great opportunity for us to clear up some confusion: the roles are not intended to stay, or to be build upon. They're a way for us to let you test various weapons and equipment, and to explore some opportunities for the upcoming metagame. Once that's finished, you will eventually be able to tailor how you look and what you use to your liking – with some limitations regarding your specialization, so you won't need to be afraid of people with the strongest weapons and heaviest vests running around and ruining your day.
What started as a few new scenarios with a lot of untouched Arma 3 mechanics resulted in a significant overhaul of the game's damage model, weapon handling, equipment configuration, and some other changes. We're happy to have actually been able to achieve the gameplay we want, but we must also appreciate the immense flexibility of Arma, which allowed us to basically make a new game within the original one. Revamping the existing equipment to create a whole new feeling of the game has been an enjoyable and rewarding experience – and shall remain one for some time, at least until we really nail it. In the meantime, we appreciate your feedback on our forums, Feedback Tracker, Twitter, Facebook, and most importantly, your presence in the game itself. We hope you have been enjoying your time on Malden, and that you're looking forward to all the news and improvements as much as we are.
Now, fancy some sector-capturing, node-linking, or Intel-downloading?
David Sláma (Designer)