Jagged Alliance 3 from Haemimont Games

[udm]

I always just chalked up enemies not dropping all their shit to "oh well, the weapon is so crap and broken anyway, might as well just leave it".

 

[Alienman]

I always just chalked up enemies not dropping all their shit to "oh well, the weapon is so crap and broken anyway, might as well just leave it".

Yeah, I mean, it's combat after all. Some stuff probably broke in the fighting. The game gets way too easy if the enemies drop all their stuff.

 

[potatojohn]

I don't know why they keep wanting to remake JA.

- It doesn't have any brand recognition with normies

- All the previous attempts failed miserably

- The JA manchild fans will hate it no matter what

 

[oasis789]

I always just chalked up enemies not dropping all their shit to "oh well, the weapon is so crap and broken anyway, might as well just leave it".

its explicitly this way in battle brothers. if you can kill the enemies before they break their weapons/armor you can loot them.

 

[ArchAngel]

I always just chalked up enemies not dropping all their shit to "oh well, the weapon is so crap and broken anyway, might as well just leave it".

That is OK, but in videos they showed us you get zero from them, true nuXcom shitty style.

 

[MadMaxHellfire]

Don’t let anyone convince you to play the v1.13 mod for your first time.

i did. best choice in my life.

 

[Ghulgothas]

Dev Diary 6 - Combat, Part 1

Hello there! I am Boian Spasov and it is my pleasure to welcome you to a DevDiary on a subject that I’ve been wanting to write about for a long time – combat! Yeah, it’s a big one - there is so much to talk about that a single article won’t be enough and you can expect a second combat DevDiary down the road.
As stated in our first DevDiary “Game Vision” the combat is one of the pillars of Jagged Alliance 3. It is a deep, involved and realistic experience and is the single aspect of the game that we iterated on the most during development.
Combat in Jagged Alliance 3 is turn-based with your entire team taking a turn, followed by the enemy team. During your turn you are free to activate your characters in any order and intermix action between them.



A typical character turn in many tactical games can be described as simply as “I move to this position and attack with this skill”. This level of abstraction is perfectly fine for these titles and we have seen how it can create deep and engaging gameplay, but for a simulative game like Jagged Alliance 3 we wanted more precise level of control over your character’s actions. How exactly do you move - will you hustle recklessly or carefully crawl to the target location? How exactly do you attack? Do you take your time to aim carefully? Will you attempt to cripple the target shooting a burst at their limbs or gamble for a killer headshot instead? This is achieved with several game mechanics working in concert, the most important of which are the Action Point system, the movement stances, the weapon firing modes and the body parts targeting system.

Action Points​

All actions that a character takes during their team’s turn are limited by their available number of Action Points (AP). A simple action like crouching may cost only a single Action Point, while a more time-consuming action like a carefully aimed attack with a rocket launcher may consume most of the characters’ AP for the turn. Attack actions may be modified by spending additional AP to aim more carefully, increasing the chance to hit precisely with the net benefit from Aiming also depending on the weapon and the character stats.



An average rookie merc has around 10-12 AP per turn. This number is increased for veteran mercs and when conditions are favorable, like at high morale, but never too much. We intentionally kept the numbers relatively low to ease the mental calculations related to Action Points that players do each turn. However, even though the available number of AP is always displayed as an integer, it is internally stored with higher precision and certain very simple actions like moving at a short distance effectively cost only a fraction of an action point.

Stances​

Characters are always in one of the three movement stances – standing, crouching or prone. Movement actions have different costs based on the chosen movement stance – crawling takes significantly more time than running the same distance but will realistically hide you from sight when you are behind an obstacle and is generally safer against firearm attacks and explosives. Conversely, if the enemy will attack you with a melee attack you will be at a disadvantage if you are crouching or prone.



When moving you can always lock your chosen movement stance, manage stances manually or let the game manage them automatically, switching to standing when this will optimize AP usage while moving but still ending the movement in your desired stance. This approach is not without risks – your characters are more exposed if they are running around standing between safer spots and if you expect to provoke an enemy attack it might be better to move crouched or prone.

Firing Modes and Body Parts​

You have three important decisions to make when attacking – how many additional AP you are willing to spend aiming, what firing mode do you wish to use and a what body part to target. Firing modes are pretty straightforward - an automatic weapon, like an AK-47, is able to attack not only with single shots but also with burst an auto-fire attacks, shooting more bullets at the expense of accuracy and AP cost. Since bullets are simulated individually this also tends to create more chaos on the battlefield, but I will talk more about the bullet simulation further down.



With a double-barreled shotgun you can offload both barrels with the same attack, but you will have to reload afterwards. A dual-wielding character may alternate between firing with both weapons or just one of them by selecting the appropriate firing mode.
Body part targeting allows you to try to hit a specific body part and inflict additional effects with the attack. Headshots are often difficult to pull of but deal massive damage, while arm and leg shots are often useful for crippling enemies that you will not be able to finish off during the current turn. Melee attacks may be targeted at the enemy neck, inflicting various crippling effects that depend on your weapon of choice.
(Note that some of the following screenshots demonstrate some debug functionality available only to developers. These shots are marked with “Dev mode enabled” in the bottom left corner and are not representative for the game visuals as seen by the players.)



Firing at a particular body part is only possible when you have a clear line of fire to it – as determined by the geometry of the level. Some body parts may be armored, presenting interesting moment-to-moment tactical choices during the battle.
Body part targeting is never possible when you don’t have clear sight to the enemy like for example when you are firing at an enemy behind a wall…




Bullet Simulation
Hitting someone behind a wall – what kind of sorcery is this? I apologize for getting a bit ahead of myself here, but I will explain immediately. Hitting enemies through walls and even through other enemies is possible in Jagged Alliance 3, thanks to our bullet simulation logic.
The bullet simulation logic involves a set of calculations for each individual bullet fired, based on the caliber and type of the bullet as well as the materials encountered along its path (armor, bodies or environmental objects). Both accurate and inaccurate attacks may have various unexpected effects because of it, like penetrating an enemy body to hit another enemy, grazing an ally by accident or destroying some of the environment on the bullet path.

The bullet simulation and the destruction system took considerable amount on effort to implement and support but all the effort was worth it because at its core combat in Jagged Alliance 3 aims to be a realistic experience, one that would not be possible without a realistic simulation running behind it. Which neatly brings me to the final, and perhaps the most important, point that I want to discuss in this DevDiary…

No Visible Chance-to-Hit​

Each time you are setting up an attack in Jagged Alliance 3 you will see various factors that affect it both increasing and decreasing the chance for the attack to be accurate. What you will not see is an exact, precise chance-to-hit percentage number.



During the early years of development Jagged Alliance 3 displayed visible chance-to-hit, just like XCom and many other tactical games do. What we observed time and time again during our playtest sessions was that people were focusing on this number to the point where they centered their entire gameplay style around it, like never attacking when it is below a certain threshold. It also created moments of frustration and disappointment as in-your-face randomness sometimes tends to do.
We don’t feel there is anything wrong in principle with visible chance to hit. There are many immensely successful tactical games out there that play exactly like this and CTH was present even in some of the most popular JA mods. It is, however, not the kind of a core experience we had in mind for Jagged Alliance 3, a game meant to represent firefights in their entire chaotic and messy glory. We wanted an experience that allows you develop a sense for certain situations, a game that makes you focus on your surroundings and the unique combat situation instead of a number in the interface. That was our reasoning when we decided to experimentally hide the chance-to-hit number in the interface and observe if the players will approach the combat situations differently afterwards. The first confirmation that we were on the right track came from none other than Ian Curry, the creator of Jagged Alliance, and many more followed in the months after – players were more involved now, found the situations more unpredictable and the game more unique and distinctive. Encounter after encounter, they were gradually developing a sense of mastery and generally had way more fun this way!
We are fully aware that the decision to remove chance-to-hit will never sit right with some players but still feel that it is the crucial design decision that made our combat “click” and feel right. There are many tactical games with perfect and detailed CTH information out there, but too few where you play “by feel” as was the case with the classic Jagged Alliance!
Thank you for reading the first combat DevDiary. Here are some of the subjects we might explore in the next one – Weather Effects, Night and Darkness, Stealth and Overwatch/Interrupt Attacks. If you are interested in any other aspect of the combat gameplay, please suggest in in the thread below.

 

[whydoibother]

I don't know why they keep wanting to remake JA.

- It doesn't have any brand recognition with normies

- All the previous attempts failed miserably

- The JA manchild fans will hate it no matter what

Worked for XCOM.

 

[whocares]

I will not eat the bugs. I will not live in a pod. I will not use 2FA.

 

[agris]

that is so unbelievably stupid

 

[whydoibother]

I will never understand this ass-backwards logic. We wanted to make things clear and simple, so we obscured information from the player.

It sometimes makes sense. Skill says 25% change to trigger. Its actually 21% on the first hit, then 27% on the second hit if the first didn't trigger, then 33% on the third if the second didn't trigger, and so on. As in, its not 25%, but its something more complicated that feels like 25%. Because actual 25% doesn't feel like 25% to people.
So its complicated and obscured, and the information you are given is objectively wrong, but this has made it actually simpler.

 

[Tyranicon]

It's a game, not a politician. It doesn't need to lie to me.

 

[whydoibother]

It's a game, not a politician. It doesn't need to lie to me.

I was just looking it up, and I lowballed it. For DOTA2, its very heavily weighted towards the expected average.

A 25% chance to trigger event has 8% chance to trigger on the first hit, because it feels wrong if it does. Its weighted towards 3rd, 4th, 5th hit.

 

[Harthwain]

I will never understand this ass-backwards logic. We wanted to make things clear and simple, so we obscured information from the player. And now instead of knowing right away at a glance how much an action costs you have to guess how much it costs, all because numbers higher than 10 are spooooky.

Having the game do the calculations for the player, while keeping it under the hood, is not that outlandish. Although I do agree with you that a centimal system is not really complicated nor should it be of any concern if calculations were to be done by the computer anyway, and there is absolutely no harm in allowing the player to do the math himself, if he so desires.

That said, I do like what they did to hit-to-chance. It encourages people to judge the accuracy of the shots by the eye. As a result people will be more likely to take chances and won't fall victims to the gambler's fallacy, because they aren't guaranteed to hit every second 50% shot they take.

It's a nice surprise for me to hear the game has the bullet simulation, by the way.

 

[Zombra]

This. In JA2 all enemies worked together on one sector (in 1.13 they even could support from adjusted sectors). A stupid nuXCOM system (even if enjoyable by itself) would spoil any feel of a real battle.

Citation required: quotes from real life war veterans saying that playing JA2 felt like a "real battle".

 

[whocares]

I will not eat the bugs. I will not live in a pod. I will not use 2FA.

 

[Taka-Haradin puolipeikko]

This. In JA2 all enemies worked together on one sector (in 1.13 they even could support from adjusted sectors). A stupid nuXCOM system (even if enjoyable by itself) would spoil any feel of a real battle.

Citation required: quotes from real life war veterans saying that playing JA2 felt like a "real battle".

Ask them about cinematic entry poses that hostiles do while you're at it.

 

[Norfleet]

That said, I do like what they did to hit-to-chance. It encourages people to judge the accuracy of the shots by the eye.

That would be true if you were actually taking those shots yourself and thus could see any of this. What is actually happening is that you're feeding a meaningless number into a sausage grinder to produce an unknown outcome, with no means of judging whether this actually did anything because you only see the end-product.

The only way to gain any meaningful data out of this is thus to take yourself out of the game and conduct controlled tests by repeating the same test over and over to get a proper statistical sampling of what any of that actually means, controlling for each variable until you reverse engineer the system for yourself. Or decompile the code. Otherwise you've learned absolutely nothing. All you really get is that it didn't work so you have to use MOAR.

Ultimately, this simply adds a lot more workload you have to do before you can even play the game, and a lot more extra work to be doing while playing the game.

Concentric circles giving a rough idea of how likely a shot is to land are great.

That would be nice, and would avoid the need to barrage the screen in obscure numbers, if the concentric circles actually represented anything, such as a visual indicator of the shot spread cone. Then you wouldn't need a number, you would just be able to visually assess how much your enemy was in the cone and whether you needed to tighten up. Particularly if the ballisto-dynamics is fancy to the point where giving an exact shot probability would be infeasible anyway.

 

[whocares]

I will not eat the bugs. I will not live in a pod. I will not use 2FA.

 

[Harthwain]

That would be true if you were actually taking those shots yourself and thus could see any of this. What is actually happening is that you're feeding a meaningless number into a sausage grinder to produce an unknown outcome, with no means of judging whether this actually did anything because you only see the end-product.

Not sure what you're on about.

You can see the circle getting bigger or smaller, which should be good enough visual indicator of how accurate your shot will be. Hell, you had the same system in Jagged Alliance 2, with the aiming reticle getting smaller as you spent your APs.

And the same criticism [of "feeding a meaningless number into a sausage grinder to produce an unknown outcome"] can be easily applied to %-based chance to hit.

 

[Norfleet]

You can see the circle getting bigger or smaller, which should be good enough visual indicator of how accurate your shot will be. Hell, you had the same system in Jagged Alliance 2, with the aiming reticle getting smaller as you spent your APs.

In JA2, the only thing that actually indicated was how many AP you spent, though. The size of the circle was in NO way correlated to any stat that actually translated to accuracy. This was very apparent when you had 1.13 actually displaying CTH: A huge circle could be translating to 95% CTH, a small one could still not be worth shit. Because that circle only told you how many extra AP you had put in...not what that actually would do for you.

And the same criticism [of "feeding a meaningless number into a sausage grinder to produce an unknown outcome"] can be easily applied to %-based chance to hit.

Sort-of-true, but at least with an explicit CTH, you're given some idea what to expect. Honestly, I'm not a fan of that, either. It's better than nothing at all, but seems too trivial.

Consider, for instance: I aim at my opponent's head, it says I have a 50% CTH. Well, at least I know that's definitely my chance of hitting his head, assuming the numbers aren't just total lies, but is that all? Maybe even if I miss, I might hit him in the torso or something. Maybe I might hit someone else slightly behind him. All of these possibilities would make it worthwhile to attempt, whereas knowing only a 50% chance of success, all I take away from that is "Don't bother, that means you miss 90% of the time". But at least I know that much.

A better way of showing this might be, say, an actual scatter ring, like that "circle" above. Only instead of just being a functionally useless indicator of how many AP I've spent, forcing me to need to save and then repeatedly re-test the scenaro just to get some idea WTF is going on in such a situation, the circle woud actually be showing me something. Is the circle mostly inside my enemy? Excellent. Is the enemy only occupying a small portion of the circle? Not so great, but two of his friends are partially in the circle, so I'll take it. No explicit CTH needed, but I'm still getting meaningful information out of this. A circle that shows me where my bullets will land? Useful. A circle that shows me that how many extra AP I am spending...totally meaningless and useless.

 

[Harthwain]

A better way of showing this might be, say, an actual scatter ring, like that "circle" above. Only instead of just being a functionally useless indicator of how many AP I've spent, forcing me to need to save and then repeatedly re-test the scenaro just to get some idea WTF is going on in such a situation, the circle woud actually be showing me something. Is the circle mostly inside my enemy? Excellent. Is the enemy only occupying a small portion of the circle? Not so great, but two of his friends are partially in the circle, so I'll take it. No explicit CTH needed, but I'm still getting meaningful information out of this. A circle that shows me where my bullets will land? Useful. A circle that shows me that how many extra AP I am spending...totally meaningless and useless.

Again, not sure what you're on about, because it is possible to hit an ally by accident or another enemy. It's not like XCOM, where your attack is relevant only to that one enemy you're targeting.

 

[Krivol]

This. In JA2 all enemies worked together on one sector (in 1.13 they even could support from adjusted sectors). A stupid nuXCOM system (even if enjoyable by itself) would spoil any feel of a real battle.

Citation required: quotes from real life war veterans saying that playing JA2 felt like a "real battle".

Yesterday I had a great war with my wife, so I am kind of a veteran and playing JA2 is precisely the same.

Does it count?

 

[whocares]

I will not eat the bugs. I will not live in a pod. I will not use 2FA.

 

[ArchAngel]

I will never understand this ass-backwards logic. We wanted to make things clear and simple, so we obscured information from the player. And now instead of knowing right away at a glance how much an action costs you have to guess how much it costs, all because numbers higher than 10 are spooooky.

Having the game do the calculations for the player, while keeping it under the hood, is not that outlandish. Although I do agree with you that a centimal system is not really complicated nor should it be of any concern if calculations were to be done by the computer anyway, and there is absolutely no harm in allowing the player to do the math himself, if he so desires.

That said, I do like what they did to hit-to-chance. It encourages people to judge the accuracy of the shots by the eye. As a result people will be more likely to take chances and won't fall victims to the gambler's fallacy, because they aren't guaranteed to hit every second 50% shot they take.

It's a nice surprise for me to hear the game has the bullet simulation, by the way.

Which is bullshit because experienced soldiers know well what they can hit even under stress. At best numbers should be hidden based on some stat of soldiers that goes up based on exp