Long War mod discussion

[Amineri]

We definitely always appreciate feedback. I'll take your long post and raise you another. ^_^

 

The Long War "8 class" system grew sort of organically in fits and starts, and is actually largely based on player feedback.

 

Some history:

When I created the first "Expanded Perk Tree" mod for Enemy Unknown, I resized the perk tree sprite to handle up to a 3x7 grid. I didn't really think that people would fill out the whole 21 perk grid, but wanted to provide flexibility to configure whatever shape of perk tree was desired. A rank could have 0, 1, 2, or 3 perk choices at any rank. At that time the perks were defined via hex edits, which was a bit of a pain, but Bertilsson put together a nice web tool to easily create the hex mods necessary. Putting a 0 value in a perk slot would leave that perk slot empty.

 

At that time we were also laboring under the restriction that hex changes had to preserve the upk file size, so to support the increased perk tree I byte-packed 4 perks per rank into a single 4 byte integer. This allowed in theory the definition of 4 perks per rank, but of course the perk tree sprite only supported 3 icons displayed.

 

So around the time preceding Long War v1.9 I had the idea to create "specializations". These weren't really full new classes, but rather a slight extension of the perk tree. The concept was that the first perk choice in the tree would provide different access within a full 4-wide perk tree. The class would have access to either perks 0 through 2, or perks 1 through 3. Perks in the slots 1 and 2 were shared, while perks in slots 0 and 3 were specialized. Eventually JL and I cooked up the current subclasses.

 

In Long War EU, v 1.9 through v2.12 final, these subclasses shared parts of the perk trees, as well as their stat progressions. The only differences were access to different weapons, and some different perks. I always thought of them as specializations or subclasses, but in practice most people just referred to them as 8 classes. That's what happens when your creation gets out in the world.

 

As time went on we got feedback which was good but difficult to implement because of the limitations of the shared perks and shared stat progressions. The 8 different classes become more cemented in people's minds as a feature of Long War.

 

When Enemy Within come out, it moved quite a few things into the config files, including the perk trees. With a few adjustments I was able to define completely independent perk tree for each of the "subclasses" in order to accomodate the previous feedback, and also created separate stat progression tables for each of the subclasses. Functionally each subclass was now de facto an independent class. The only remaining vestige of being shared is that one of the original 4 base classes is randomly rolled and the player can then choose between the 2 subclass alternatives. In theory I could create 3 separate subclasses based on each class, but our perks are stretched thin enough as it is.

 

That's basically how we've gotten to the point we are now.

 

---------------------

 

Each of the classes does have some unique traits, but is not completely 100% distinct from the other as in vanilla. In vanilla there are 48 perk slots distributed 12 to each class, and every perk shows up exactly once in each tree. Also in vanilla each class gets unique access to one weapon, with the sole exception that Assaults can use rifles.

 

Long War deliberately blurs the line a bit between the classes -- classes can fulfill slightly overlapping roles depending on how they are set up, which we view as generally a good thing. With the longer injury times, fatigue, and vagaries of chance in random class assignments, a bit of blurring between roles seems generally to be a good thing, and allows for some quite creative thinking by the players.

 

The unique/defining attributes. as of beta 9a :

- Sniper : only class to get Squadsight

- Scout : Only class to get In The Zone, one of 2 classes to get Lightning Reflexes (and gets it early)

- Assault : Only class to get Run and Gun

- Infantry : Only class to get Light Em Up

- Rocketeer : Primary Rocket Launcher using class, and only class to get specialist Rocket-based perks

- Gunner : Only class that can use LMGs

- Medic : Only class that gets Field Medic and other specialized Medikit-related perks

- Engineer : Only class that gets Sapper and Bombadier

 

Each of the 8 subclasses (I stubbornly stick to that! ^_- ) does have some unique attributes. But the roles overlap a bit. Rocketeer and Engineer can fulfill some of the same role. Some player prefer one, some another, which we take to indicate they are reasonably balanced against each other. An Infantry and a Gunner can both lay down Suppressive fire, but the Gunner is better at it because the LMG has more ammo. The Infantry is more tactically flexible though with the very flexible Light Em Up (renamed because we combined 3 different vanilla perks), which we justify because the Rifle is a more "handy" weapon than an LMG.

 

Recall that in addition to the perk trees each class also has a separate stat progression and access to different weapons. I'm personally in favor of a bit more restrictive weapon-to-class assignments (I'd rather not see Engineers, Scouts and Medics have access to Shotguns, for example). But some are still quite specific. Only Gunners get LMGs, only Rocketeers get Rocket Launchers, and only Snipers get Sniper Rifles. In beta 9 JL experimented with giving Scouts access to Sniper rifles to try and simulate Marksman Rifles, but it really didn't work.

 

Starting in beta 10 we're going to be more than doubling the number of weapons, introducing entire new classes of weapons, which will exist (with one exception) at all five tech tiers.

 

Pistols:

Pistol (baseline, as in vanilla)

Machine Pistol (heavier, slightly more damaging)

Sawed off shotgun (heavier, quasi-melee weapon) only at tech 1

 

Rifles:

SMG

Carbine

Assault Rifle

Battle Rifle

 

Support weapons:

SAW

LMG

 

Sniper-scout weapons:

Marksman Rifle

Sniper Rifle

 

Assault weapon:

Shotgun

 

I won't go into too much detail about each weapon, but the concept is to give each class a few more options within it's own unique space.

 

For example, for Snipers the choice will be between the MoveLimited but unlimited squadsight range Sniper Rifle and the non-MoveLimited Marksman Rifle which supports Squadsight ranges only about 5 tiles beyond visual range. The Marksman Rifle can also overwatch at this limited squadsight range, which the Sniper Rifle cannot.

 

The Scout's choice will be between the Carbine/Assault/Battle Rifles which work better at close range but whose burst fire mode drains them of ammo faster than the precision fire from the Marksman Rifle.

 

The Infantry's choice will be between different classes of Rifles featured ease of handling (improved accuracy and mobility) vs higher damage.

 

The Gunners choice will be between the MoveLimited LMG, which features the same extended squadsight range as the Marksman Rifle (built in, with no perk needed), including the ability to Suppress (but not overwatch) units at the extended range as well as increased damage but at the cost of reduced mobility. The SAW can be fired on the move but deals less damage and cannot be used beyond visual range.

 

Anyhow, what I'm getting at is that the perk trees aren't the only thing defining the different classes. Perks and weapon choices are the most obvious two defining elements, with stat progression a factor that is less visible to the player.

 

I would agree that I feel that in some cases extra choices are added into class perk trees in order to "fill them out" to 3 choices per level. Sometimes there are three good choices, but sometimes there are only two "optimal" choices. However as we've learned time and again, sometimes those alternative choices end up with a good use as a clever player comes up with something we didn't think of.

 

-----------

 

On the slight nitpicky technical side, I think you undercounted the number of permutations. You came up with "2*3*3*3*3*3 or 486 for each class", but each class actually has 3 choices at each rank, so 3^6 = 729 combinations of skills. For the 8 base classes this results in 8 * 729 = 5832 total "choices" just for the 8 soldier classes. The 8 MEC classes I don't view as distinct classes, just variations on a theme. There were only 12 MEC perks defined in vanilla, and there wasn't any way to stretch those across 8 truly unique MEC classes, so they are really more "flavors" than actual classes in my mind.

 

There are two ways to look at this situation :

1) There are too many choices, so players don't know which are the most optimal

2) There are enough reasonably balanced choices that there aren't just a few best choices, enhancing replay value due to exploration of alternatives

 

One of our goals is to add "depth" (whatever that means) in order to improve replay value, so that different campaigns can turn out in different and novel ways. I wouldn't say that we're there, but it's what we strive for.

 

------------------

 

Let me address one of your specific points ... the difference between Infantry and Gunner. Gunner has been, arguably, one of the more difficult classes to bring into balance with the rest and provide a unique role to. We split the Heavy into LMG and Rocketeer in large part due to our difficulty in accepting a single soldier toting around both a heavy machine gun and a rocket launcher. Unsurprisingly it was the Rocket Launcher that was the more powerful/unique of the two weapons.

 

We slowly been making some headway in providing the Gunner a unique role. In beta 9a we introduced a special ability to the LMG which makes its Suppressive fire inherently superior. In addition to the 6 shot ammo clip (compared to a rifle's 3), LMGs automatically grant the Opportunist perk while Suppressing, making Suppressive fire from the LMG more effective. Of course an Infantry can also get Opportunist as a perk, so it's not perfect by any means.

 

The next step will be in the next beta 10 release, where we're adding a second Gunner specific weapon (the original "LMG" will become the SAW, and the new weapon will take over the LMG name). This will give the Gunner the ability to shoot with limited Squadsight and provide Suppressive power at squadsight range, something no other XCOM unit can do.

 

 

---------

 

In closing I want to say I definitely appreciate the feedback, and your taking the time to write out such a post. We do the best we can, but certainly don't claim to be infallible in the decisions made :D

[Hobbes77]

Amineri, thank you for the detailed explanation of the history of the current perk system, your answers to my questions and the future plans. I realize it may not seem fair for you guys at this point of LW development to hear someone saying: "the class/perk system is too complicated" and I appreciate you taking the time to read this. :)

 

I'll be brief on my reply since you mentioned right away the main issue behind all this when you said: "The Long War "8 class" system grew sort of organically in fits and starts, and is actually largely based on player feedback."

 

My question here is: but when does it stop growing?

 

If you look at it, there will always be 1 more perk/class/weapon/rule that could be added. There will always be another player with a great idea for something new. And if you're an artist and love what you do, nothing is ever perfect - there's always something that could be changed.

 

But, in life, there's always a point when creating great works where you need to decide: it's enough. Usually there's a manager and/or deadlines to help with that. Here you guys are working on LW as a hobby and you probably haven't considered that there will be a time when you'll want, or have to move on and do something else in real life or because LW isn't so appealing anymore. Isn't it better to start thinking and planning for this moment by setting up goals? To have a time to stop adding new features and focusing instead on balancing and fixing bugs on what exists and removing unnecessary features?

 

To me replayability isn't just about adding new features (and LW already has a ton of those, I'm loving the additions to the Strategic Layer) but as a player to feel comfortable with the system and know the rules and that requires time to try everything out. If new features are constantly added then I will not have time to try everything and I'll have to be constantly relearning the ground rules.

 

This said, I don't have a magical solution of my own to the class/ability/weapon system but I do feel it has gotten a bit complicated. I think the most important things to consider are: when do you stop adding new stuff? And is it really needed to add weapon/ability xyz? Does it really add something new or is it merely a variation of what's already present. If you have the time, read that Civilization modders manual I've posted before, specially the section Design Philosophy. It has a ton of interesting insights about mod and game design (they even interviewed Civ 4 and 5 developers for it).

Edited May 29, 2014 by Hobbes77

[ElementUK]

I'm just wondering how many dozens of times this has to be played before anyone is supposed to figure out the 'right way' to go about things... I.E; Which is the most advantageous research order, the most needed facilities etc etc.

 

 

'Friends in Low Places' mission: Lol what a manic mission this is now but hugely fun. The first 90% of the mission is all about surviving the wave after wave of skinnies coming at you. You need a Gunner with suppression, an assault for taking every opportunity you can for a good run 'n' gun with a shotgun, a scout with lightning reflexes, a damn reliable sniper with low profile (maybe two) and a good Medic or two with two kits each. Having to hunker down to avoid taking damage from Acid becomes a real pain on this mission. Panic is deadly because they can run 8 squares etc. Playing with the 2nd wave option of Critical Hit is nastier for you than the enemy in this mission.

 

PS for the Developers: You ask for specifics regarding possible changes. I notice that when you have the 'Officer training School' Lead by Example, it can be a bit weird as to who gets to be leader. Quite often the ability goes to a Squadsight Sniper which is useless as he often also has the 'Lone Wolf' perk. Obviously there is little point in them being the Squad leader as he is not anywhere near his troops. It has to be Positional so we can choose the leader perhaps.

 

One bug I often come across is not being able to use the 'escape' key to get to the main menu.

Edited May 29, 2014 by ElementUK

[GAJackdaw]

First off, I just wanted to applaud you for all the hard work that's gone into Long War. I love the interception mechanics, appreciate the depth of the new class system, and am astounded by the way you constantly have to think about logistics in order to preserve your forces. Realizing you've walked into an ambush and having to scramble to evac your troops isn't something I thought I'd appreciate, but successful "aborts" (in which I manage to save the majority of my troops) feel just as satisfying as victories. As someone who thrived on Classic Ironman in vanilla, I appreciate the added challenge.

 

Installation was a snap, and I did a full purge and reinstall of XCOM (including all the config files that lived in My Documents, just to be sure that I had no conflicts with the mod--everything went swimmingly.

 

Unfortunately, there is one thing that diminishes my enjoyment of Long War: Thinmen. I'm inured to their ability to projectile vomit on a sniper in full cover on a rooftop (their torsos are esentially pressurized gas bags, after all), and their aim was legendary even in Vanilla. Both of those points are fine--but the Thinmen in Long War are essentially immortal due to their armor. If you'll bear with me, I'll lay out the crux of the problem.

 

Playing Long War on Normal, the skinny folk have five or six hit points. I wouldn't have a problem with that, were it not for their armor. Ballistic weapons, such as assault rifles, only do a single point of damage on Thin Men (I've come to loathe that blue armor icon popping up after a successful hit), which would require a total of six shots over the course of a round to kill a thinman (barring the presence of gunners or snipers, which can deal 2HP and 3HP worth of damange on critical hits with lasers, respectively). 3 AP grenades can be used to bring one of the suckers down, if you're in a pinch. Beam weapons are marginally more effective, as a laser rifle can deal 2HP worth of damage to these humanlike aliens (on a critical hit).

 

Here was the last straw for me: after losing ALL of my highest ranking troops on the first mission of Slingshot (due to the inability to abort), I went into an abduction mission with my well equipped C-Team (What was left of the B team was in the hospital ward)--everyone had laser weapons, and I had a medic loaded down with medkits ready to rock. I set up a chokepoint in an alley, and prepared for the onslaught. My team took down 3 Mutons, 10 sectoids, and 5 seekers. In the process, they lost a scout due to a lucky shot from a sectoid and a critically wounded gunner had to be stabilized. My medic ran low on supplies, but everyone on their feet was still kicking. Then 2 Thinmen arrived. My squad managed to survive 2 rounds against the Thinmen before the death of 2 soldiers sent the remaining 4 into a panic-chain that could not be recovered from. Combined fire from 6 soldiers over two rounds with laser weapons was not enough to take out these two enemies.

 

To sum up, as of Long War 3 EW Beta 9a, a single Thinman is more difficult to deal with than a squad of Mutons with flanking advantage or a pair of Cyberdiscs. I love every aspect of this mod, and it's an absolute shame that a single enemy type has rendered such an ambitious project unplayable.

[Amineri]

This really isn't the right place to post feedback of this nature. The correct place is over at : http://www.nexusmods.com/xcom/mods/88 in the beta feedback forum.

 

This forum, being part of the Mod Talk, is more about technical aspects of the mod (for other modders that want to discuss Long War-related features/code/etc).

 

We're happy to accept feedback, but it's better in the right forum :)

[GAJackdaw]

My apologies. I'll be sure to head over there for any issues I have in the future.

[ellatan]

I've experienced a weird bug on Ironman mode normal difficulty. The bug happened during first month. A medium sized UFO showed up at one time, and I tried to scare it away with fighters. However it ended up above a north pole while it claims to be in Germany. It doesn't leave or land and just hangs there. I can engage it with Ravens but there is no way I can destroy a medium sized UFO that early. So now it's a permanent fixture on my mission control until I get enough firepower to bring it down. I wonder if anyone else had that happen.

Edited June 5, 2014 by ellatan

[prune1]

Could you tell me which of the following mods have been rolled into EW (e.g. enemy randomisation I believe is one?) and which ones would be compatible if that is not the case?

Basically just a few ease-of-use and small tweaks I'd like:

From XCOM Interface and Gameplay Tweaks:
No Disappearing Corpses
Hunker Down and Overwatch Delays
Display Soldier XP, PsiXP and Mobility in Barracks and in Battle
Remove Death Animation Delay
Geoscape Pause on Notifications

From Random Alien Pods:
Maps Reused (a config file basically that works alongside Random Alien Pods).

[johnnylump]

We don't test other mods in ours, so we can't say which woudl work. THe following are already in LW:

 

No Disappearing Corpses
Hunker Down and Overwatch Delays
Display Soldier XP, PsiXP and Mobility in Barracks

Geoscape Pause on Notifications

 

We have our own map config, so you'd mess up LW by using someone else's.

[ElementUK]

Respirator Implants NEED to make your troops 100% Immune to Acid. It's bad enough that you get 13 hp Thinmen that not only have Lightning Reflexes but regenerate 2 hp turn.

The slot is way to valuable for something that works....a little. Where is the 'Tactical Rigging'; it needs to come into the game much earlier. Exalt are way to powerful in their later stages. 22HP Rocket troops that have Double Tap - sheesh! Injury time needs to be reduced imo, 42 days to have your best troops layed up....

Alien Metallurgy (the one that increases the Alloys you get) takes far too much Meld. 60 when you desperately need it for troops.

 

I'm not even playing Ironman and it's still pisses me off )

Anyone who says they can get through this in Ironman mode is to my mind, simply taking the piss.