Discussing Unit Roles

[vestal horizon]

Who made this? It does not make any sense, units in BAR do not behave in ways that are similar enough to eachother to bunch them all up in roles. There are like 20 stats for each unit and every single one is unique, even faction equivalents are different enough to basically play out completely differently. For example Janus and Pounder, Sniper and Banisher - they can't be put in roles like this.

Also this game has too many complete outlier units that don't belong in any category even remotely - like bedbug, skuttle, LRPC, Nuke, etc, etc.

What is the point of putting units into roles like this anyway? Is it supposed to help from a game-design viewpoint?

[mortal cape]

This whole thread like a nested Google translate where u keep retranslating same roles over and over again but it makes different words pop out

It ain't that deep honestly idk the fuss to try to figure this out

[vale roost]

This graphic is OP, I love it 😄

We'd classify stuff a bit differently but it communicates the interactions so well

I really want to make a graphic to communicate the whole spectrum of archetypes and their subspectrums

I think I'd actually put assault under anti-unit and combine it with riot as a subclassification and then have raiders under anti-structure since they can evade statics and units to deal direct eco damage

Not to say assault isnt good against some statics, particularly light ones, but skirmishers are too and its not like we want it to be so black and white anyway. To me assault is anti unit

[fathom elk]

I still aren't 100% sure of what the role of the assault tank is but maybe i will be after i tried the rework 😉

[vale roost]

Main Battle Tanks are pretty generalist even though assault is in the name

[fathom elk]

that's doesn't help me much but okay

[mortal cape]

What is purpose of trying to define out roles exactly? It this for tech split reasons? Doesn't reducing the game simulation to archetypes for units kinda depreciate the value of unique use for those units? Like is this just for communication purposes or is this with the intent of changing units to be more like eachother

[proven trench]

imo assault is a pretty generalist role

[vestal horizon]

you are not going to get a serious answer, neb is not interested to talking to people that have a chance of disagreeing with him and you already seem skeptical from your previous messages

[proven trench]

roles like raider, skirmisher, riot and artillery are much more pure, and "minmax" but assault is pretty vague and versatile

[mortal cape]

Well i mean if its just for communication to new players i dont really see a negative thing about classing units so people that have never built them before can get an idea about what they do but since neb is spearheading the tech split agenda it makes me suspicious that the intent is to define archetypes and then change units to fit those archetypes and at that point u basically creating new bar fork since simulation is already in pretty good state

But if I recall right it already has descriptors for each unit when u click on them, like reaper heavy assault tank or roach rolling amphibious bomb etc.

[crimson lance]

RED alert 3 has very cool unit descriptions with videos, if that was done for bar that would be enough.

[fathom elk]

It's like in books, picture helps when you start learning to read?
hmm maybe im wrong actually idk

[open abyss]

Lrpc area control icant

[mortal cape]

Honestly the unit descriptions in bar are already badass, really got me interested in the game initially

[vale roost]

Therxyy is here, that's OP 😄

The way we were using assault its pretty broad but so is skirmisher, raider, and artillery. The way I see it is they are broad spectrums on an even broader spectrum of unit balance and theres tons of subtypes. The language could maybe be more precise

[raw falcon]

Why is it impossible for you to acknowledge or accept the design that assault is often generalist and that Brute Stout Bull Tiger are these things?

I've seen it argued, very well, in this thread a number of times, and it's just... ignored afaik

[proven trench]

i just think that assault in of itself is pretty versatle

[vale roost]

Yeah it is very versatile. I guess if you extend it to include MBTs then its fairly general, but its not more general than skirmishers, raiders and artillery in the way we use them for BAR

[raw falcon]

If you need more specificity then they're 'breakthrough assault' easiest subtyping ever

[vale roost]

Its not that I wont ackowledge it, its just that I was seeing MBT as the one exception, which it still is. I think MBTs are as close to skirmshers that an assault unit can be without becoming an outright skirmisher.

Its a good point though that MBT should fall under assault and that does broaden the category. Im not seeing it as an exception anymore

[crimson lance]

Infantry can be assault unit same with tank. It's just a role. But infantry is squishy irl and dies to small arms while tanks need large calibre AT weaponry only recently drones are introduced...

[fathom elk]

What MBT has to do with it? blopson is his new name anyways

[crimson lance]

Infantry still has same weaknesses. Tanks have same weaknesses across their respective "platforms", just that they are equiped for diffrent roles on battlefield. There could be AT weapon which is mounted on tank, infantry, light armor vehicles, helicopters, aircraft, does not matter which platform. Each platform has it's own weakness. I don't think making tank platform be squishy makes any sense visually and gameplay wise

[raw falcon]

Also re: assault being 'anti-unit,' it's time to call Starlight Assault Siege Artillery

[vale roost]

I mean this is all just descriptive language to describe a unit. The more descriptors that you have, the better. But some descriptors are intentionally more vague than others because of the level of design/function that they are describing. Thats why we've focused on the very broad categories of: Skirmisher, Assault, Raider and Artillery, for game design's sake the way we use these terms is a bit different from what they mean in real life.

So we don't use infantry but its not like you couldn't describe some units as infantry. Its just a matter of what you are trying to describe and how you are trying to describe it

The archetypes we've been talking about mainly just describe the tradeoffs a unit makes and very vaguely what interactions you expect it to be useful for, and this is really effective for game design.

[crimson lance]

Same way if you really want to draw IRL corelations tanks had different guns for different targets and even different shell types.

[raw falcon]

Lrpc is deep building toucher

[vale roost]

Like skirmishers can describe a unit like a sheldon which can fire safely from a distance and is maneuverable enough to kite while dealing damage, but it can also describe something like a janus which has to dip in closer to an enemy, deal damage and retreat.

[fathom elk]

Realism mod when

[vale roost]

Artillery we just use to mean units that have exceptional range, assault units are units with durability in combat, and raiders are fast, but there are endlessly granular ways to describe a unit.

And theres a good design reason for this, but I don't have time to get too deep into that right now

[vale roost]

I feel like raiders in BAR aren't raiders enough. They have so much HP and even a T3 raider like mara is still not really that fast and the DPS is also low, but their HP is fairly high too for a raider.

I wish T3 raiders were even lower HP:cost than T1 raiders, but could have crazy high speed and DPS

[fathom elk]

That sounds like a lot more fun than current BAR, great success will come eventially with enough hard work😉

[crimson lance]

That's terrible for game, fast units can choose engagements they would be obnoxious to push into, just look at ticks openings before grunt accurate lasers. There is a reason sprinter has to be weak for its cost due to having highest in class speed.

[vale roost]

Why is it terrible to design fast units to be able to avoid engagements? That's already part of their design intention.

[misty yew]

the fire support category makes no sense. rocket bots can't shoot over allied units afaik so they don't form the "long range dps" that you put behind your front line
swap rocket bot with banisher

tac launcher is area control, but LRPC is not really area control in the sense of "stopping someone from winning this territory" LRPC is mostly shoot bases and squishy high value units

[vale roost]

Fire support feels like a flavor of skirmisher to me

[misty yew]

what they are saying is backline dps unit

Both skirmisher and artillery can function as long range dps tho the best ones will be the slow ones that are meant for straight up fighting (starlight, medusa arquebus, catapult)

The reason why you have backline dps is there's only so many spam/medium tanks you can have on the front before they get in each other's way

[vale roost]

Skirmishers is a broad category and includes backline DPS as well as hit-and-run type skirmishers and many other varieties. I've been trying to find the simplest way to describe a skirmisher in the sense that we use it, but since its such a wide spectrum its hard to simplify

[misty yew]

isn't the definition of skirmisher hit and run? If you plant a starlight and use it as a mobile laser tower while defending it with units, that's not skirmishing

hound, sheldon are clearly skirmishers. Banisher is skirmisher. Arquebus has better stats but slower -- it's harder to harrass people with arquebus so on the spectrum arquebus is closer to a (slow-dedicated long range dps). Sniper is the same, while tries to attack while not being hit, it's not like snipers can walk on the map and hit-and-run (enemy unit just can back up)

Skirmisher: Hound, Sheldon, Banisher
Hybrid: Arquebus
Immobile long range DPS: Starlight, Sniper, Mauser/Quaker (if the speed was reduced)

[vale roost]

Not in all cases, its a really broad spectrum. Something like a sheldon is more of a kiting skirmisher in the way we've been using the term

[misty yew]

yeah you could change the hound to a lower range unit that gets in range of enemy t2 fire but uses mobility to get favourable trades. Still an in and out skirmisher, and kiting skirmisher vs most t1

[vale roost]

Thats what we were trying to do with the split changes, but I wasn't super satisfied with them when I tried them

You'll notice they have only 500 range now, compared to sheldon's 650, but theyre more maneuverable and accurate. But they felt a bit weak to me

[misty yew]

well you could just buff their throughput (dps) by 10%. hound never had good dps for cost, so people used recluse instead which hits a lot harder

[vale roost]

I mean also recluse were more of a skirmisher too so you basically had two units doing the same thing. That's why we are playing around with making recluse more of an artillery unit that is just barely not a skirmisher and hounds are a skirmisher that leans more towards assault

Its cool continuity too because bots basically have a t1 skirmisher thats leans heavily towards artillery (rocket bots) and then in T2 it crosses over into the realm of artillery but still leans towards skirmishers

[misty yew]

the reason why i call even the lower range hound a skirmisher is it's trying to make favourable trades over time, and harrass. Even if inside enemy range = use mobility to dodge shots and pick off weak angles.

A skirmisher doesn't want to fight a 20 second done and done battle, whereas a raider or assault unit is perfectly fine with that

[vale roost]

Yeah Ive been saying skirmishers basically use some form of evasive maneuvering to deal sustained damage and that feels like a sturdy and simple definition to me

[royal cedar]

Marauders are squishy for t3 units, and they are tanke enough to go through t1 static defense wich they absolutely should, there's nothing to fix about marauder it interacts perfectly with other units in the game

[vale roost]

This was a post I made a week or so ago for gdt

The current definitions are too narrow and don't capture the meaningful trade-offs that make roles distinct. For example, calling raiders just 'weak, fast, easily killed' misses that they should excel at DPS and favor speed over range.

The purpose of defining unit roles is not to restrict unit designs or to introduce rock-paper-scissors interactions. Rather, it's to ensure that units fulfill a clear purpose and don't overlap in a "grey goo" of balance. This greatly strengthens unit designs, unit interactions, and the available strategies for a player. In our testing, having stronger definitions for unit roles leads to more meaningful role-to-role interactions that incentivize interesting compositions rather than mono-compositions.

For initial changes I'd recommend focusing on the following archetypes; I've intentionally tried to use language that leaves the roles very open to diverse unit designs. Though a role may favor certain stats, individual units may make different trade-offs.

Here are my proposed revisions:

Scouts + Intelligence

Current GDD definition - Scouts

• High vision – cannot survive direct conflict
• Fast
• Weak
  Current GDD definition - Intelligence
• Provides knowledge on enemy positioning

Suggestion:

I think these should be combined. Scouts are more like a subtype of the Intelligence Archetypical Role, and all intelligence units tend to be very fragile and have high vision. So my definition would be something like
Intelligence

• Provides knowledge on enemy positioning
• Favors high vision
• Fragile
  Scouts - Intelligence Subtype
• Fast and evasive

Skirmisher

Current GDD definition - Skirmisher

• High range
• Ideally maintains distance from enemy

Suggestion:

Skirmisher

• Favors range and maneuverability
• Trades health, DPS
• Relies on kiting, dancing, or dodging to safely deal damage

Some subtypes are harassing skirmishers like rocket bots, hit-and-run skirmishers like janus, and aggressive skirmishers like the hound we are testing in the tier split

Assault

Current GDD Definition - Assault

• Frontline fighting
• Can dish out and take damage

Suggestion:

Assault

• Favors health and DPS
• Trades speed, maneuverability and range
• Good frontline units for their ability to soak up damage

Raider

Current GDD Definition - Raider

• Hit-and-run attacks – can survive incidental contact with the enemy
• Weak, fast, easily killed

Suggestion:

Raider

• Favors speed, maneuverability and DPS
• Trades health and range
• Relies on a combination of speed and evasion to surround vulnerable targets and deal massive damage

Artillery

Current GDD Definition - Artillery

• Targets far beyond its vision distance
• Requires spotters to find targets
• Bad at close quarters combat

Suggestion:

Artillery

• Favors range and high per-shot damage
• Trades health, mobility, vision and efficacy against moving targets
• Cannot effectively engage targets while retreating - must face targets
• Great for bombarding defensive positions from a safe distance
• Blind and highly fragile - relies on other units for intelligence and protection. Highly vulnerable to rushes due to inability to retreat

[royal cedar]

I think a lot of people wouldn't be so harsh with their feedback if you kept your scope on units that are unanimously bad, like hound, juggernaut, hover, but changing beloved units like marauder, thug, sheldon wich have very good places in the game allready is bound to upset people

[misty yew]

intelligence maybe is referring to radar, cameras, spybot, gremlin
scout being the visible, fast, cheap intel gatherer

[vale roost]

Yeah its true changes can be hard for people to accept but dont forget we love these units too and are looking to make them better in what theyre loved for. That's what detailed feedback is good for

Also yeah chrono ^

[misty yew]

Skirmisher:
Higher: range, maneuverability. Not neccesary but can have high effective burst damage. Uses these attributes to take positive trades over time.
Lower direct stats: hp, dps

[topaz lichen]

This is great although I wouldn’t say LRPCs are static defense, they’re static offense

[royal cedar]

LRPCs are like nukes

[misty yew]

LRPC is almost purely offense yeh

[royal cedar]

They make you build the defense thingy or kill you

[raw falcon]

The intention is not to restrict unit designs
suggests a bunch of additional definitions to roles that will restrict design
Well, it might not be the intention, but it's what it does.

[vale roost]

That's a strong claim to make in the face of all of the design ideas we have within the system 😄
They are intentionally very broad but Im sure I could clean the wording up.

[velvet valve]

Defense is like the 'department of defense'. All static defense are designed to attack, except walls.

[raw falcon]

It's not a strong claim at all since it's directly in the language of the new suggestions relative to old definitions.
I guess the "will" word is a bit too operative though

[vale roost]

But the fact is that the freedom we have in designing units under the system is directly contradictory to your claim. Ive always said these are just descriptors for broad spectrums on an even broader spectrum but units can exist arouns the boundaries/Grey areas

[raw falcon]

Oh? The impression I got through a lot of your messages was that you had problems with a lot of units that fell on the boundaries and wanted to change them.

[vale roost]

Perhaps imprecise wording on my part, its difficult to communicate abstract ideas about unit design. That's why we need OP graphics like what @velvet valve made 😄

My problem is that some units take up too much space in the spectrum of unit design/balance, and also that too many units fall in the middle of the spectrum. The work we are doing in the tech split right now essentially pushes units more into the extremes of unit designs, so that they are moving towards the extreme ends of the spectrum. That's why, for example, we gave T1/T2 plasma artillery even more range and compensated for it in other areas.

I think it's explained better in https://discord.com/channels/549281623154229250/1421227359398269090

Man I should just make a graphic

This is gonna be OP

[lyric night]

It was originally about the Mauser, and then the hound got thrown in arguing about skirmisher or assault and now Mauser is skirmisher, and that moves and shoot is a skirmisher making the term skirmisher irrelevant.

And part of it was that the Mauser and the Quaker are still labelled in gane as mobile artillery, and for a new player it doesn't really act like artillery more like a really light tank with a big long range gun.

[velvet valve]

To be clear. My graphic has nothing to do with balance or the tech split. But is a new player tool for people going wtf are these 200+ units.

Yes a good player should know that janus and pounder are completely different due to 20 other stats about handling and projectiles. But a new player doesn't know why they should build pounder vs when to build whistler.

[lyric night]

So stuff like aa missile truck, pounder ; riot tank seems to make sense. But if like the Mauser and Quaker it's mobile artillery but they're used to scoot and shoot its like wait what is this?

[timber hornet]

I still think there should be room in the "Artillery" role for anti-unit capabilities. Making units like the Mauser/Quaker have a mortar-like trajectory as its only option means that they'll hardly ever be used against units anymore, when they could instead be more of a middle-ground between Starlights (also considered "artillery" in the tech-split) and Negotiators. Granted, both of the latter units are being moved to T3, but the (currently unchanged) Rocket Hovercraft are being put in T2 next to the Mauser and Quaker now.

I'm not arguing that Mauser and Quaker are fine as they are, but I don't think they should be changed so drastically that they become useless for supporting a frontline against units.

[velvet valve]

An expert shouldn't be handicapped by the limitations of language and categorising. Arguing about it is pointless. They are tools for understanding.

But if you know the turn rate and speed and turret traverse, fire angle etc you should think of the units in that phase space not as 'skim' vs 'arty'.

[lyric night]

How can we as non devs play with unit stats and use them in a game to see how they all work? Is it really difficult and do you need coding skills?

[lyric night]

I just want a unit that is like a guardian a heavy plasma artillery unit. If the Mauser and Quaker aren't that it's fine. I would just like to be able to use an artillery unit that does what the heavy plasma turrets do. Trade fire with defences, disrupt grouped units and also allow area saturation with lots of them acting like modern artillery.

[vale roost]

Okay sorry to disrupt but I was making this:

This is the game right now

[timber hornet]

Is there any unit in the tech-split for each faction that fits that description?
Something just fast enough to keep up with the front line and shoot over them at enemies (and actually be able to hit them with volume of fire), while being vulnerable to raiders that manage to flank them?

[vale roost]

And this is the type of design we are testing right now, this diagram shows some hypothetical progression of raiders. Im sure you can even tell what role this would be describing. And its not like you need to be super strict about where there is and isnt overlap, theres so much room for design

[vale roost]

I think you are describing a type of skirmisher

[vale roost]

This is like what we envisioned for hounds and also turtles to some extent I think

[spiral aurora]

Make more spikes 😉
Don't make balls please, balls are all the same, easy to handle, but just a boring grey goo.

[lyric night]

Stuff like counter battery fire is perfect in bar, trading of artillery is fun (for me) and I loved doing it company of heroes. But most rts games the artillery doesn't feel long ranged enough, bar and TA, supcom it always felt good.

[vale roost]

When higher tech units take up less area on the unit balance spectrum like this, or maybe not less area per se but are shifted more towards their niches, you can really see how it leaves so much extra space for other unit designs to shine

[lyric night]

But the hounds role as fido is something the bots need imo, why not make a new unit instead of replacing an icon TA unit?

[vale roost]

I dont think we really changed the hound away from what it does in-game already, its just a bit more micro intense. But that's just my opinion

[spiral aurora]

The area depends on the positioning of the values around the circle. But generally the idea is sound.

[lyric night]

Gauss hounds are fun to use they just need stats tweaks to massage them into a spot that's not useless but not op

[timber hornet]

What I'm describing probably falls in the grey area between Skirmisher and Artillery according to your definitions. Something like a long-range Pounder, with low health and enough of a trajectory to shoot over allies in front of it.

[vale roost]

Yeah its all just to make a point not meant to be precise at all

Also meatpopsicle that's what we are trying to do, I guess everyone's definition of change magnitude is different though

[raw falcon]

wtf does this mean the stout is going to have more range and hp than a bull
Like I know that's not, but it's what the graph says

[vale roost]

I think you're not connecting with what I'm trying to explain, I'll try to think of a clearer way to explain it for you.

[lyric night]

When we played neb, the tigers vs your hounds. I could see that the hounds are just faster and have longer range so can be good to play into its benefits over heavy tanks. But when I heavy tank is pushing the hound isn't so strong that it can just hard stop a tank push. (which I think is the sort of interplay and trade off's with bots and vehicles that was alway intended by TAs designs. Even if in reality some of the units just didn't work)

[vale roost]

Yeah, and then you'd be encouraged to include a different unit into your composition, but when I was playing armada I felt like a response was missing. Welders are just so slow now, but they would typically be a unit thats good against tigers pushing into them

[lyric night]

I feel like welders are good with being slower chunkier and with their aoe damage like slow advancing front line

[vale roost]

Actually an argument could be made for gunslingers filling that role against tigers too because of their impulse. Maybe they have some room there

For the tiger vs hound discussion I mean. Welders do feel good in that slow chunky frontline role, so maybe gunslingers could be the option for dealing with tigers? Although maybe it should be hounds. Whatever it is it definitely feels like arm t2 bots is missing some power somewhere

Id maybe even want to say that a proper T2 raider sprinter should be able to surround the tigers while hounds also engage

[spiral aurora]

Don't you dare to give Armada impulse and let suffer Cortex from it. 😘

Armada is magically faster -> lighter -> suffers from impulse.
Cortex knows that -> makes impulse weapons -> slows Armada down.

[lyric night]

I think also not every unit should have a unit that's a direct counter, so mixing is encouraged. Like you said. Tigers and bulls crush walls don't they? I find I'm using walls more and more.

I also think static defence should be playing a role too, but all to often it's not put down or used - the pop up sabot turrets are pretty deadly still. Maybe static defence has become too weak to be a real deterrent? I find the heavy laser turrets are kinda a joke now. Every time I build one I think 'this could have been units'.

[vale roost]

Roberts thoughts on static for this are nice. Basically making some defenses that are sturdier and longer range but lower damage like the HLT for area control. I like that direction too.

Raiders could still run past some defenses but that would be by design

[lyric night]

It's even stuff you see in team games, no AA units or turrets and then the 'air' player taking the blame because the other players can't be bothered even planning ahead for a bombing run.

[spiral aurora]

It's the lack of information.
You "can't" afford to build useless AA towers. But when you know you need them it's too late.

[lyric night]

That's part of it though right? It's the downside of defending you're reactionary or have to have some prediction of what's to come (from scouting).

[topaz lichen]

Making Tech 3 units worse in range, alpha strike, and HP per mass seems like a huge mistake. Higher tech should be overall superior, sidegrades should exist within tech levels not between tech levels. And gating support units behind tech makes earlier game sterile

[vale roost]

This isnt for all tech 3. This is just illustrating how a raider tech progression could look, but you can do all sorts of stuff with different tech branches, factions, etc and the roles skew in different directions too. Im just illustrating how teching works with it being about stronger specialization and using three raiders as an example

[round cloud]

Graph might be missleading, the way i understand it is that proportionally other stats get way more vs some other stats at t3. Like, 10x more HP but not 10x more range, as an imaginary example

[spiral aurora]

Don't take all these values as a given. It's just a quick example..

Can you make a diagram that makes more sense, in your head, pls? Then we can compare it and understand your wishes more precisely.

@vale roost may be add that to the text.

[spiral aurora]

Feel free to scribble something for us, so we can get a deeper understanding.

[vale roost]

Yeah, exactly. T3 "skews" more to the edges of the graph because its more specialized - even stronger in specific niches but also much weaker in others. Currently theres really heavy stratification of the tech levels, and the diagram just shows what it could be like if there was less stratification and more specialization (one of the core design goals of the tech split)

Here is an example of what an artillery progression could look like for a branch/faction

This is an extreme example of a t1 > t3 progression you could make of 3 assault units where they are getting slower and tankier with more damage and less range in higher tech

You can visualize how roles are defined by tradeoffs that skew the graph into certain directions. Overlay these and you have an imaginary single branch of a single faction's tech tree

You could have some t3 assault unit with high HP and speed but low damage and range too and the graph would look different. There are infinite ways to design stuff with the specialization system we're using

And you can imagine how even if t2/t3 units have overlap with T1 units, if you were to average all the different roles out the average power levels of each tier are still stratified, so a complete T2 composition is stronger than a complete T1 composition.

(Averages of all roles in a tier)

[topaz lichen]

Oh I didn’t know that these were divided by unit role, that actually makes sense and yeah I support this. Make units feel completely different to handle. Although I would say for land units you shouldn’t make the turn rates super low or the speed super high (like ridiculously high, faster than fighter jets)

[vale roost]

I feel like for land raiders anything up to like, 160 speed should be fair game lol

But I also wouldn't mind T3 fighters with like 600 speed haha

[raw falcon]

You should really scrawl "Artillery" or whatever archetype you're drawing for when you draw these. There was no way to know you were talking about raiders last time. (I looked back and did you edit it in to the text or did I miss it the first time? I could've just missed it)

[timber hornet]

Artillery
Favors range and high per-shot damage
Trades health, mobility, vision and efficacy against moving targets
Cannot effectively engage targets while retreating - must face targets
Great for bombarding defensive positions from a safe distance
Blind and highly fragile - relies on other units for intelligence and protection. Highly vulnerable to rushes due to inability to retreat

Most of this is fine, but I think "efficacy against moving targets" should be removed from the role description. The archetype otherwise perfectly applies to both Sniper/Starlight and Ambassador/Arbiter. It also seems overly restrictive against in-between units like Mauser/Quaker.
*Snipers can engage targets while retreating, but not for long before they get overrun.

[misty yew]

if you added a 3 second flight time to starlight it would be shit, long range anti-porc is not the same as long range anti-unit

[timber hornet]

My point is that I think some units can be considered anti-unit "artillery" due to having long range and high per-shot damage, but not being maneuverable enough to be used as Skirmishers, which seem to be the only other long-range role defined in Neb's list.

[spiral aurora]

I still think "tank hunters" or something along those lines is a legit role, especially for the starlights.
Sharpshooters are insane alpha skrimishers? Idk.

[misty yew]

https://www.beyondallreason.info/unit/corshark

hunter = chasing down prey, or careful sniper. Starlight is neither.

Starlight is basically like a void ray (pre speed buff). It just stands it's ground and shoots.

[spiral aurora]

https://en.wikipedia.org/wiki/Jagdpanther

I always see a jagdpanzer, When I see a Starlight. E.g. Jagdpanther.

Just the fire angle seems too big with 180°

https://en.wikipedia.org/wiki/Jagdpanther

They have the same weight and speed as the tank variants. But they don't have a turret. They trade traversability for a bigger gun.

The same gun was used in the tiger 2.

https://en.wikipedia.org/wiki/Tiger_II

Obviously much heavier, slower and tankier.

[vestal horizon]

voidray is a flying, maneuverable unit that can't be targeted by half of sc2 and the starlight is a slow ground unit that cant turn and dies with no support

[spiral aurora]

Tempest is a bit closer to the Starlight.
Voidrays are more like flying beamers.

[vale roost]

Yeah, I think a different name could make sense than tank hunters but I know the role you are talking about. High accuracy single shot damage at long range

[brisk lynx]

i see starlight as more of a siege tank type unit as far as playstyle goes

[gritty basalt]

It can hit statics but its talent is in that you cant dodge it

[brisk lynx]

needs support and overcommitting it is very fatal

[vale roost]

I think we basically use artillery to mean siege units in BAR, high trajectory plasma artillery is also a type of siege unit isnt it?

"Siege breaker" is maybe more descriptive? Im not sure, but I get what you're thinking

[brisk lynx]

well it was a specific comparison to the starcraft siege tank, i agree tho high trajectory plasma is siege

my description for the role of starlight would be t2 ground superiority unit

[fathom elk]

why compare to starcraft when you can look at real life tanks and warfare

[velvet valve]

why compare to anything when i can compare to paper scisors rock

[fathom elk]

this graph isn't as pretty as nebs ones

[velvet valve]

i can add more scribbles

[vale roost]

Hey now colored pencils and highlighters are an aesthetic

[spiral aurora]

Fire support might be a good role.

[vale roost]

Fire support is basically what we'd call a skirmisher, particularly a long range skirmisher

[fathom elk]

Could add the supressive effect that then make units lose accuracy, this would be nice for fire support role

[vale roost]

Thatd be so interesting. I wonder how hard implementation would be

[fathom elk]

They do it in battlefield, it must not be that hard

[raw falcon]

But... I was promised there wouldn't be rock-paper-scissors unit design.........................................

[spiral aurora]

And there won't. But there will be noticeable (not as OP as currently) differences in efficiency against different types of units.

[gritty basalt]

Fundamentally, units are more useful against specific units

You want that design since the other option is everything has same efficiency than everything else.

[misty yew]

just the weapon. long time ago in WoL they were slow

maybe you can call starlight a single target siege tank without minimum range. Though it can advance while shooting

[topaz lichen]

And no splash damage. At that point the comparison is tortured

[misty yew]

starlight just trades maneuverability and kiting for range and raw stats (a bit tanky for a long range unit, very high throughput)
it's not an assault unit, but within the range of long range units it's assault-subtype

[topaz lichen]

Yeah I would say it’s a hybrid of Skirmisher, Artillery and Assault roles which is why it’s OP

[misty yew]

Low projectile flight time
Long range pure dps -- maneuverability and kiting for range and raw stats (Starlight, Arquebus)
Long range skirmish: accurate enough to hit units -- (Thanathos, Banisher, Medusa)
Anti-Ball: High aoe and medium flight time (Tzar and Fatboy). Only fast units can dodge it completely.
Dodgeable Artillery: Some ability to hit units, but microed non-slow units can dodge it. This is determined by flight time + aoe. (Mauser/Quaker with nerfed weapon velocity . Current Hound Plasma)

High projectile flight time
Area Denial / random fire: Very inaccurate fire but with high aoe (Tremor, Wolverine). Best against fragile skirmishers, Acceptable vs porc (Wolverine, Tremor, Napalm),
Siege: Very slow projectile shot and only moderate aoe. Making it primarily anti-building and afk units. (Ambassador/Arbiter/Shellshocker)
Tac Missile: Very slow projectile but very high aoe, despite very slow travel time can aim to hit units. (Tac launchers, Thor emp missile, nuke)

[rich cargo]

That's why I added a class of units like this - Slayer. Or Hunter, whichever would stick better. Basically focus on unloading on heavy units, but weak against numbers.

[vale roost]

This chart is interesting

[gritty basalt]

If you use per cost stats, you also need a total cost one so you also include amount of X combinations, since having 1k DPS on a 10k HP body is worth more than having a effective hp cost and dps but its a tick.
There is a difference to keep track there

[mortal cape]

in real life u almost never getthe actual dps of a laser unit so units like razor grunt etc do way less effective damage

also looking at dps is mostly wrong, depth charge units have it included even tho its like a totally different domain, better to look at burst, turnrates, weaponvelocity etc

[spiral aurora]

https://tenor.com/view/how-bout-both-both-is-good-i-want-both-gif-12224132

[gritty basalt]

thats a laser thing, so lasers need to be adjusted for effective dps, but as long as accuracy isnt a problem, units are equal to their dps.
Burst has its own value but thats basically like taking a loan on dps, and then needing to repay it later with reload.

dps cost is still useful, but a burst cost is useless, you need to compare the full burst value to other units with similar roles, or vs total hp of units it will face.

[vale roost]

You need a lot of stats to get a full picture of course. I think this chart is mainly missing alpha damage and a modifier for grunt DPS to take fall off into account, as well as turnspeed.

But its still generally useful to see where unit's efficiencies are

[crimson lance]

Burst is big reason why T2 claps T1 when numbers get big enough. You can't get in range and 30% of your army is dead before firing a single shot. T2 can be overwhelmed with T1 cause of raw stats of T1 units. if oyu are in range of T2 units T1 wins for same cost (medium tanks). Problem is that players let enemy hit them for free and allow T2 player to utilise range advantage.

I think range is very nice mechanic to allow higher tiers to scale better. Range scales with higher numbers. More of your units shoot more often and range is pseudo burst amplification to your army composition. Lets say tzar shoots 2 times before medium tanks close in. Thats enough to kill 2-3 brutes and remaning 6 brutes can flank and kill a tzar. When you get to 3 or 4 tzars you need a lot more T1 tanks to close distance, probably around 30-40 instead of 8 vs 1 tzar. Also lab cost of 2900 is lower % of total army cost comparing 1 tzar +t2 lab vs 4 tzar and t2 lab. There is a timing where T1 player (in equal skill scenario) can just crush T2 player who is teching badly.

Just removing range and adding raw DPS is actually making powerspike worse. 300 dps tiger has much more potent rush timing than 155dps tiger with more range. 300 dps tigerr is actually better to rush cause it's so efficient in its range. While 155 dps tiger (vanilla) scales with numbers due to how range contributes to burst.

So in essence current tech split encourages skiping T1 even faster casue T2 timing would be absurdly stronger to crush T1. Then at T3 we unlock normal range progression like we are used to where you benefit more from massing than just rushing those units out asap.

[vale roost]

The problem with higher range being the end all be all though is that higher range doesnt allow for any interaction at all. The lower range on the tiger means they cant interact without getting well within range of even a lot of T1.

High burst damage is also a way to balance units to waste DPS on lower tier units

[crimson lance]

It's not a problem, it forces you to mass and avoid enemy it's interaction in itself, changes how you have to approach the engagement, just afking your units is objectivelly bad play. Forcing constantly fighting and "interacting" is just wrong approach. Looking at stats of what you are doing with units is where its going.

In other RTS (sc2) you can say SIEGE tank is not interacting but it forces enemy to attack from multiple angles. Just afking your units in siege tank range is completely wrong way to play the game

[vale roost]

Thats the point though, its to look at more stats than range. And its not about making T2 units weaker, they still ought to be advantaged. Its just a matter of making their advantage less about universally more range

Some units need range of course, because they rely on non interaction, but not a tank...

[crimson lance]

So T2 rush is more potent cause now T1 can't do anything. When you just have better stats you can ALWAYS force an engagement and enemy loses, when you have to USE RANGE to win, that's skillful interaction, 1 guy just avoids, while other tries to hit for free....

If i just have faster and better units and you can't run away, it's over, everything comes down to who has more efficient units sooner. Current T2 units lose in head on fight vs t1 often if out of position. This will make interaction a lot less micro intensive and it's boils down to who can get more T2 sooner.

You did not make T2 units weaker you made T1 vs T2 even less skill expressive. T2 with less range and more raw DPS would just be more worth to rush compared to current T2... Combined with T2 con as well, just makes teching even more required play then before, makes T1 window even more obsolete.

[vale roost]

The point is not to give T2 better stats in every area. Its to make sure they are strong in relevant stats but not everywhere. If they are just better its an issue

[crimson lance]

Cause T1 had numbers advantage wen they are in range, now they have 0 advantage in any area, so in end everybody will just skip t1 asap if they are competitive and play to win.

Why would i ever build any unit when nothing can stop just better brutes (tigers) why do i need arty or anything when i can maintain army advantage with just better brutes while you have worse brutes and CAN NEVER actually engage on me due to losing stat check every single time. I mean you will realise that this is flawed system eventually.

Range advantage allows for interesting interaction. Forcing everything to shoot each other at same time constantly makes game very streamlined. It's just a numbers game who can get most efficient stat check units faster and then you can NEVER ever engage their army frontally.

I seen Tzar overextend and die to pawns, in higher skill lobby. Explain how anything can contest 300 dps tigers in their range? You can't, you can jsut saftely scale behind 300 dps tigers and it actually makes less interactive gameplay.

[vale roost]

Range advantage doesnt allow for interesting interaction in all cases, it often means the interaction is exclusively one sided. Units should not universally have the range advantage in higher tiers, or else there is just no interaction.

[crimson lance]

Why is that an absolute statement? Range is nice balancing factor if done properly and arguably it's already done properly, just that everybody can skip t1 cause of absurdly low T2 tech cost.

[vale roost]

Because it should be intentional, not universal. thats the issue

[crimson lance]

Why? I don't see any actual argument other than it's not interesting.

Range is very good balancing factor, it scales with numbers due to how burst works. Just reducing range and slaping more DPS on units is wrong approach and makes interaction WAY more 1 sided.

If my unit has same range as yours and just betters stats you have NO ADVANTAGE to play around. Right now there is advantage to having T1 over T2 but due to player skill issues and how easy it's to tech in team games it's hard to utilise that advantage properly

Banisher can die to 3 brutes which cost less (if they are out positioned). Tiger probably dies to 3 brutes. But when balled 10 tigers shred 30 brutes due to how range scaling works. But with these new stats tiger has advantage from moment you build it which just streamlines progression into more rushing T2.

So you are telling me that you want to eliminate correct positioning from game so that everything becomes stat check gameplay? Thats opposite of interesting gameplay. Tiger isn't objectivelly better than brute but scales better than numbers due to how range works in RTS unit interactions.

Forcing you to play optimally aka not letting higher range units hit you for free is form of interacting, higher range unit changes how you SHOULD choose engagements, it's interaction on meta level. What you want is to change combat pattern to just brainless interactions.

[vale roost]

You are right. Range is a good balancing factor, but its not the only balancing factor.

[crimson lance]

It's factor that makes most sense to BAR type of game where eco constatly goes up. In other RTS ECO STALLS OUT and you have to balance game around different factors.

I can hit you for free = i can use time to my advantage, you have to react or you lose. Just having better stats on units makes you have no CHANCE at any point in game. You just lose inside your own range and you can't choose engagements which favor you since there is no chance for you to win.

I don't see how is that meaningful interaction if you have no chance to win any engagement with lower tier of units.

[vale roost]

we agree that units shouldnt "just have better stats" but Im not sure why you'd exclude range from that statement

[crimson lance]

Range scales diffrently and i think you don't understand how it works in RTS terms. It's not just a better stat like HP or DPS or BURST. It changes how you engage enemy.

[vale roost]

The time factor of range is very interesting, but universally applying range to everything makes it less interesting when it becomes the only balance

[crimson lance]

Range is actually what defines unit TIER. tiger is just better brute if they have similar range but better stat, BRUTE loses it's relevancy.

Higher tier units scale better in numbers cause they have higher range. Brutes actually stat check any T2 unit in vacuum when it comes down to pure cost to cost interactions. What tiersplit will do is actually just make progressing even more linear and make lower tier units have overall WORSE efficiency in every area.

If my brutes kill your tzar on same cost, only saving factor you have is range and kiting, which is interesting interaction, i try to find angle to all in you, while you protect yourself from ALL in. This is very basic interaction.

[vale roost]

When every t1 vs t2 unit interaction is like that, though, its just not interesting design to me, but if you like it that's also valid.

[crimson lance]

Same way why destroyers die to friggets alone but when numbers go high range scales better, that's interesting.

If you just remove range, then it's binary interaction Tiger>brute at every range and metal invested. But now 3brute>1tiger but 10tiger>30 brute. POWER balance switches due to RANGE. That's good scaling and makes sense, you get out teched and you should lose

[vale roost]

Here's the thing, this actually flattens the t1 vs t2 unit interaction into a very singular thing. Its not about how certain units behave, but just about how any t2 unit interacts with T1. If you like that design, its totally valid, but I want something with more depth than "t1 has to all-in rush T2"

[crimson lance]

What 300 dps tiger does is this any amount of tiger> 3x any amount of brute UNIVERSALLY. It's terrible and makes gameplay very linear, removes any decision making, you just want to get tiger asap.

[vale roost]

If you are talking about the obviously overtuned tigers that got nerfed quickly in the split then we agree already 👍🏻

[crimson lance]

Well you can't want something but not understand implications of changing stats. I think more interesting change would be to reduce cost of T1 units by 30% when you have built some building, just doubles down on T1 unit identity

[prisma cape]

Hmm, if something excels in one stat but not others it cant stat check but still be op. That stat does not need to be range.
I am thinking of wc3 siege tank atm. So tanky it just ignores units and just drives past and kills your base. Also cant even damage units so will lose in battle.
Terribly unfun example but does maybe make a point.

[vale roost]

Exactly. That's all that Im challenging. Currently that stat check is range and HP for most higher tier stuff, regardless of it's intended use.

[crimson lance]

If tiger is same as brute in efficiency and has same range then there is no point teching, you just muddied unit potential. Why would anyone build tiger if there is no PAYBACK at any level of scaling. That's why range is best balancing factor. Same way tiger can stat check razor for same metal but quickly gets outscaled due to range difference.

I know nothing will convince you that this is failed approach and your underlying assumptions are wrong. You will just have to see faliure of this approach months down the line.

If your goal is to diversify unit roster what will actually happen is make teching unrewarding (if T2 is just less efficient) OR will make T2 over efficient and skipping T1 asap. Overall less diversification and more linear gameplay.

[vale roost]

What will convince me is real testing data, not theorycrafting, but you are missing that Im seeing your point, I just disagree that its interesting design.

[crimson lance]

Go test 3 brutes vs tiger and 30 brutes vs 10 tiger. this is your data

[vale roost]

Thats a very simplistic approach, but thank you for the feedback.

[crimson lance]

What is data that you require? From what sources will you gather data? Excel spreed sheets are not it for unit design.

[vale roost]

Replays, lots and lots and lots and lots of replays.

Except in some circumstances like the old split tiger its just obvious where there is an oversight and something got over tuned.

But without replays theorycrafting isnt really substantial, tbh.

[crimson lance]

My example actually proves that t2 scales better with numbers, try your tigers vs brutes at ANY amount of units, tigers are just straight up better or straight up worse if they ahve same range.

[crimson lance]

I am talking about design philoshopy is wrong. But nothing i can do anymore, it's just your own project you want to push cause you don't like status quo of current design. So i think it's pointless, i can give balance feedback but i don't want to contribute to flawed unit design system.

[vale roost]

To be honest, I think you have an incomplete understanding of our approach.

[crimson lance]

I think redesinging some underperforming units is welcome. I like T1 veh arty change very much.

[vale roost]

Definitely, and for implementing any change starting with the easiest and smallest ones that will be easier to have accepted is best. We are looking for which changes get the best response so thank you for the feedback!

If it makes you feel any better, the tiger change in the split is very low on my priorities for proposing cause I dont think we got it right yet, feedback reflects that.

[crimson lance]

You can't change units 1 by 1 when your whole balance philosophy and design is different. Right now units are balanced by range. What tech split is going do to is throw that out of window and start from scratch complete different design choice. It's not same game essentially if it changes whole combat system.

[vale roost]

Thats why we are testing the full system first and then pulling the most favored changes in small batches to see how they hold up on their own.

I think saying it changes the whole combat system is an exaggeration though. No mechanics are changing

[crimson lance]

That's very bald statement. Changing how every unit interacts is not changing whole combat system? Yeah unit skins are same but every unit is different.

[vale roost]

I dont see it that way. The combat systems are still in place, but we are shifting around how some units fit into it. The combat system itself isn't changing. Unit interactions arent really changing much either, they are being emphasized. At least that the goal, we arent gonna get everything right immediately.

If you like the t1 arty changes, hell yeah. At least we got something right.

[crimson lance]

It makes arty identity more interesting without making other arty obsolete (T2). Tiger will make brute obsolete in every way if you want them to have similar range as brute.

[vale roost]

Tigers definitely need work in the split, I can see that. Their current iteration is around 215 DPS I wanna say and 305 range?

I think tigers need a change to be able to stay in the 1600 metal T2 lab though, otherwise they'll be too strong. The reason we tuned range and DPS is so that they could maintain their high HP identity.

[crimson lance]

I know you are forced to change thins around due to range, that's the whole point, you want to neutralise range advantage so much that nothing will stay the same.

[vale roost]

What's a better alternative for the tiger in the split?

[crimson lance]

With your design philoshophy nothing. Tiger has to be better than brute at every stage and number. Cause you want to remove range factor due to it being "uninteresting" while it's actually only interesting component of teching up.

[vale roost]

Thats oversimplifying the reasoning for the changes they have

The reduced range also means they can be tuned to be more effective than medium tanks at cleaning up light spam

Current design philosophy for tanks is higher range at the cost of accuracy. I know Im not the only player who finds this troublesome

[mortal cape]

low range units usually giga shit unless u balance them with enough mobility + hp to make use of that low range

[crimson lance]

Well let me walk you trough. Since you boiled everything down to few roles, assault has to have low range, then you slap tigers and brutes into same role, so they have to follow same archetype of stats, therefore tiger has to be better if range is same between them otherwise why would anyone build tiger ever.

[vale roost]

Assault doesnt have to have low range in our philosophy. Mammoths are 750 range

The way you're thinking about roles is too narrow. They're very very broad spectra

[crimson lance]

Mamoth is not progression of medium tank. Tiger is natural progression.

[vale roost]

Agreed, but the idea for the tiger progression in the split is they become more effective against light fast units too but shouldn't be universally oppressive against t1. In theory, since they are the progression of the medium tanks, they should be a better option

[crimson lance]

why not give them laser then? Salamander is t2 brute, same stats but more anti spam weapon

[vale roost]

🤔 I mean I dont see any reason to have any sort of restrictions that mean we couldn't give them a laser, but minimal changes and all that...

I was thinking of tigers as the progression for tanks thats more like brute with some pounder crossover

[crimson lance]

Minimal changes, right... You are kidding me

[mortal cape]

i dont understand all the sweeping changes why not just split buildoptions with basic units and not try to balance buildoptions + unit stats

[vale roost]

The test can be less minimal, but when it comes to pulling changes out from testing for individual proposals, it should be. I mean again, tigers arent really something Id take out of the split right now because we didnt get them right yet. Getting them right means a minimal change that makes them work in the split and makes people who enjoy them already happy too.

[crimson lance]

They already work in base game, just you don't like how they work.

[vale roost]

I know, thats the issue. We are testing changes that address a problem in a way that has side effects. So, they might be OK in base game but not in the split. How do we preserve their desirable qualities in base game, but also make them work with the other changes that are so important like lab costs?

[mortal cape]

isnt the purpose of tech split to make lab decisions matter more? dont they already matter a lot in current game?

[crimson lance]

You can't due to different design philoshopy, i would like to agree but i just can't lie to myself.

Well tehnically you make tech split more balanced but you will remove unit diversificaton in process and actually make overall gameplay more linear. I want nothing to do with that.

[mortal cape]

yeah doesnt less buildoptions per lab mean i have less options tactically

[crimson lance]

if 1 unit is just better than pervious tier, it's no brainer to not build that unit. Tzar is not better at everything T1 can do for example... Tiger in tech split has to be better than brute due to underlying design philoshopy of nerfing range.

[mortal cape]

in tech split is like quaker in a separate lab from tiger?

[vale roost]

Yeah theres a core difference here in our design philosophy. I think for teching up to be meaningful, there should be some degree of invalidating the lesser tier unit in the progression when the higher tier one comes out. But it shouldn't invalidate all t1, and if it does thats an issue.

I.e. if tigers invalidate brute its OK and desirable, as long as they arent invalidating every T1 unit and have T1 counterplay options aside from brutes.

[crimson lance]

Your unit stats don't reflect your goal

[prisma cape]

I think there is some valid criticism here actually. More labs, same amount of units, you are only supposed to have one lab => less units to choose from at any given moment.
What do you think about that neb?

[mortal cape]

But you build tigers and brutes for different reasons - brutes soak up high reload shots and disperse hp for overkill and provide better individuals for flank and tiger is a high metal dense unit that has fundamental different role tactically especailly depending on scale of incomes

[vale roost]

It seems backwards to me if a T2 unit has to compete with the t1 version of itself. It makes teching up less meaningful on the unit side of things

If the stats dont reflect the goal, thats an issue we'll address, and I think we didnt get tigers right yet. I agree with you on that at least

[crimson lance]

With lower range tigers there is no reason to ever build brutes OR brutes, one or other will always be go to unit

[vale roost]

I think rosters are bloated now, and more units to choose from isnt always better. It creates a lot of redundancies and overlap. For example, hounds and recluse have been in a weird balance due to their overlap

[crimson lance]

They are just not differentied enough after gauss removal, another bad choice

[mortal cape]

yeah i mean it feels dumb if they are effectively the same what actual decisions are going into unit diversifgication? like right now there are actual tactics involved in mixing t1 and t2. Brutes leave wrecks to block unit pathing and I can do it a lot - I can soak up heavy hitter shots and cause other units to get flanking damage, I can force overkill etc, then I can sweep in with my t2 units and do the real damage etc but if they are basicalyl copies of eachother thats just dumb i lose that whole decision making

i mean tigers are already kinda meh and they have good stats lmao gg make them worse op

[vale roost]

I also have a follow up question for you, do you see the current issue right now with labs only having a few viable meta units despite their large roster sizes?

[crimson lance]

What you are focing is more units being vialbe but less time to do so... Cause higher tier just makes lower tier obsolete in every way.

[mortal cape]

its almost like they nerfed spybot rip

[vale roost]

It should only be the case if you build a robust composition, which takes time too. Otherwise there should be gaps for T1 to exploit. Can we agree that dominant monorushes are a bad thing?

[crimson lance]

Thats why you are essentially doing with reducing range of tiger and buffing dps, brutes have no window to trade vs tigers, in base game they can still overwhelm on same metal value if numbers are low enough.

So problem is actually worse

[prisma cape]

If you can make the unviable units viable do that first and then we can talk about bloated roster.
I just build the lab with the most meta units and will miss those that i cant access anymore. The useless units will most likely still be useless.

[vale roost]

The idea is that you shouldn't be trying to trade brutes with tigers, you should be able to find a different t1 counter though.

[crimson lance]

I think best tech split would be to just make T3 lab have all T2 units. So if you want tzars you pay 4000 metal for lab but you can still build tigers. And T4 is gantry

[vale roost]

I think thats a great idea for starting with minimal changes, target the stuff no one is using anyway first 🙂

As far as implementation of the split goes I think the roster changes/lab splits can come way later once it feels natural

[crimson lance]

Tzar powerspike fixed behind higher metal cost for lab, ez, no need to redesign whole game... Make T3 have tzar, starlight banisher etc and all lower tier units so you pay premium to be able to produce those units since they are more cost effective and scale more

[vale roost]

Yeah Im trying to figure what changes to propose from the split for a first test patch. My first thought was moving Tzar/Fatboy to T3 and buffing HLT

Changes that are controversial like tigers are obvious no-goes but I want to keep testing them in the split. The nice thing about testing is its non-commital and we can try anything out

Maybe HLT and T1 arty changes make sense to do together too

[crimson lance]

Why would anyone build tzar when they have shiva, i am assuming this is for base game? you just made units have overlap

[vale roost]

Yeah definitely, tzar would need changes to make them T3 too and not just a worse shiva. Thats why they have changes in the split. Especially so it doesnt feel like we are just removing an option

[crimson lance]

Just make intermediate lab T2.5 which has heavy hitters in it, and maybe move T3 to little higher price, like 11000 and this new lab around 4200 metal. So T2.5 has all T2 units and heavy hitters. And when you tech you obviously won't have tzar imidiatelly or fatboy but regular less "opressive" units for teamgames where range is king due to 8 commanders

[vale roost]

We tested that already, it feels bad to just have an intermediate lab plus that just means coop rushes will target the heavier labs and solo players are SOL

[crimson lance]

That''s coop issue, which you said won't be solved by this...

[vale roost]

True.

But regardless, 4 levels of tech doesnt feel very good to me anyway right now in the split. The gantry is pretty redundant for various reasons

[crimson lance]

this is not 4 levels its 3 and you pay extra premium for heavy hitters cause they disturb 8v8 noob lobbies too much

[vale roost]

Thats still just adding complexity to the tech tree without much benefit.

[crimson lance]

What do you mean no benefit? It nerfs most toxic aspect of T2 powerspike for teamgames, that's actually sole reason of whole thiss discussion

[vale roost]

What's the benefit for gameplay of having the same lab with a few extra OP units? Now you just complicate the choice of T2 or T2.5 labs and its really just choosing between unlocking a few extra units, no utilities or other interesting aspects.

I think the lab splits should be an emergent change, not a direct one.

[crimson lance]

Okay i don't get how you dont understand that. I will just add numbers then. Lets say you have 6000 metal to spend. You can have T2 lab and 5 tigers or T2 lab and 2 tzars. With 4200 lab you can have only 1 tzar vs T2 lab and 5 tigers

[vale roost]

It solves a problem but in a way that would be a design mistake.

You just highlighted my point. It adds complexity and theres not much of a gameplay benefit aside from gating some problematic units. Its just a bandaid and a poor one at that

[crimson lance]

gameplay benefit is that tzar hits later. That's literally what every 8v8 glitter or supreme player wants. How is that not gameplay benefit?

[vale roost]

It doesnt enhance the experience of teching, it makes it more frustrating. It does solve the issue by making tzar hit later but thats not my issue with it. My issue is that it solves it in a way that makes teching gameplay worse

[crimson lance]

It fixes balance problem, i am all for that. 8v8 balance problem

[crimson lance]

Yes cause you don't want to fix problems of majority of players. Which you should since they are majority of this palyer base 8v8 glitter and supreme players.

I don't have problem with tzar or fatboy since i just dislike 8v8 glitters. But this change would make more players stay in game before they tranistion to more competitive gamemodes.But every time i try it and enemy gets fatboy and i get T2 con i quit glitters for a week

[prisma cape]

Is 4 levels of tech including the gantry?

[crimson lance]

https://en.wikipedia.org/wiki/Lanchester's_laws

https://www.youtube.com/watch?v=wpjxWBwLkIE

[mortal cape]

i love when stellaris ur military strength the same but then ur torps get countered by ponit defense and its just gg

[crimson lance]

Everything in video apllies to every RTS. You can test it in full metal as well... Also encaptulates why i don't like proposed techsplit design choice. TELLS you how range hp and dps scale with math since you just don't take my word for it...

from zergling to juggernaut

[vale roost]

Yeah. I dont really like the gantry in the split to be honest.

[prisma cape]

I thought the entire idea of tech~split~ was having 4 tiers (counting gantry) instead of 3 as we have now.
I think moving tzar to gantry and making it essentially an experimental is just removing the unit. Why would i pay for a gantry and then make tzars?

[vale roost]

Its a bit more nuanced than this. Tzar are moved to the experimental vehicle plant in the split, where they arent competing for roster space with Shiva. And the exp labs give you more utilities than just the tzar, you get significantly better cons and other options

[spiral aurora]

I did read most messages, it's a lot so I wont reply to everything.

My main takeaway is:

1. Don’t make range a blanket tier perk. Let's reserve it for some roles. That allows for different kinds of skill expressions other than kiting OP.
2. Cutting assault range can indeed trade positional skill for raw stat checks. We should avoid that at all cost. We don't need another: 'rush better units sooner' and auto win!

[vale roost]

Key point: tie range more to role and less exclusively to tier

[crimson lance]

Good my arugments were not in vain, only this point was worth yaping around for 2 hours

[vale roost]

Haha, well I hope you see that we are not just looking to argue even when defending our positions, your feedback really is very appreciated 🙂

[crimson lance]

This is glitters situation ofc ranged units just add constant dps and sway outcome and it feels "non interactive"

ON flats and forest tzar is inferior to tiger unless you are protecting vital area... Cause every unit gets to shoot and be involved directly.

This is just non square form of trading, 1 on 1. But range is like a square law. Allows higher scaling with numbers, idk if it makes sense. Just nerfing range, makes units react linearly and removes scaling from them

[raw falcon]

SotL is the GOAT

[prisma cape]

Can you elaborate on better cons? I think t3 eco options are a very dangerous proposition.

Sorry, maybe i just should play techsplit instead of asking dumb questions.

[crimson lance]

It's not playable, numbers are way off, some units are interesting design changes like T1 arty which could go in to vanilla tbh, there is no harm in trying it though

[prisma cape]

Well when he says t3 con next we have t3 AAfus and a whole t3 economy and that sounds like disaster to me. So i panic!

[raw falcon]

Yeah I reaaaaaally hate the idea of "T3 eco" because the design of the scale of BAR is perfect atm.

[crimson lance]

Eco is whole another topic

there you go high range vs low range, literally now in tourny how it should be played T1 wins btw if they commit

[raw falcon]

yeah they're too scared, literally also watching. Ah there they go, they went in

[crimson lance]

It creates interesting dance. if they had same range, you just go in and game is over... Cause you have better stats on SAME range. I don't know any better data than this.... They are using lesser ranged units properly

T1 army won

[crimson lance]

T1 tonks used in conjuction with T3

[raw falcon]

I thought it was gonna be pure feed but they did good and helped break thru

[crimson lance]

cause that's how you use T1 in current meta, you can forget this play if we apply tech split philoshophy.

[raw falcon]

iirc it is part of their intention, mentioned somewhere, to make earlier tier units more viable in later stages
But it seems you're saying the particular stat and role changes suggested would go against that?

[crimson lance]

This works only cause medium tonk is best unit when it's allowed to overwhelm enemy. If you just inflate tigers stats cause you remove their range, all of these units would be tigers instead in addition to stomping brute at every metal count

[vale roost]

They are quite fast and current OP buildpower for cost so we have to tune them but I appreciate their speed a lot

[spiral aurora]

Your feedback really opens new ways to look at things, so thanks a lot, it really helps.

I think I understand what you mean. Scaling is a complex topic. Power density, damage before the enemy can engage, blocking line of fire of each other.

Two important things I want to point your attention to are:

1. Brutes and Stouts can kite Tigers now.
2. Armada's main battle tank (the turtle) has a higher range now and wants to keep its distance. While Cortex wants to get close and personal.

I personally like the flavor of this. What do you think?

[crimson lance]

In vacuum it sounds reasonable. Depends what happens when tiger closes distance. I think T2 player should be required to kite, it just seems more natural way of progressing

It just reverses engagment pattern and brutes will still get overwhelmed if tiger is balanced properly

To be accurate i would need to be in tiger vs brute stage many times to get any meaningful feedback

I think it will mirror problem brute has vs welder, just more pronounced, brute can't trade into welder, you have to avoid them or attack other parts of map

[spiral aurora]

Why? Like really, a why out of interest!
I think it's a nice idea to have faster T1 units that need to run away to actually have a chance.
Especially against the brute force, I fuck your face, Cortex units!
Armada may kite... They are weak and wanna win through 'tactics' or something.

[spiral aurora]

No AAFus! You may relax 😉

[spiral aurora]

T1 can still be the fast metal brick without much DPS that is easily repaired, tier.
Or what is the current unique value you are referring to that would get deleted from T1?
Tigers are good against T1 but not because of their HP/metal, nor the overkill potential that the smaller units always carry. So the T1 value stays throughout the game imo.
But I might very well have a blind spot.

[crimson lance]

Comperative value of T1 to T2 will just fall off. T1 has it's niche right now. Tigers are good now cause they can ball up and concetrate firepower due to range advantage. If you allow them to fight with stats only they lose unless numbers get bigger.

Also brutes won't be able to push T2 player ever again then cause tiger wall would just trade too effectievly, it would create much more defensive gameplay.

Same way you can't push into welder sniper, that's why it's meta play in FFA

[spiral aurora]

Ic, ic.
How it could end like that.

Perhaps new tactics could evolve, like Brutes would need to dive in, chip away some hp and run again.

Or the first 2 or 3 tigers arrive to the battle field and they can't run and generate infinite value.

[crimson lance]

t2 player just stands in porc wall and scales eco, while you are constantly outscaled and can't push into him due to getting stat checked, i played that exact scenario in ffa many times

[spiral aurora]

I don't say I am right, but I am not sure if stuff like that could make stuff more dynamic.

What porc? T1 Arty should have killed it by then.

Maybe t1 Arty buff?

[crimson lance]

with 1500 metal lab no, t2 scales too fast, but i need to actually see competitive games on tech split first

[crimson lance]

well tigers can just move in and out to pressure arty away, that could be possible but we need more tests to prove points

[spiral aurora]

Yeah, I hope we don't end up in another.
T2 sits in porc and techs, situation...

At least it's no instant auto delete for T1 units and defence anymore... If only this is achieved, it's a first small step into the right direction imo.

[crimson lance]

Cause range is less opressive vs mass T1 player. If you allow T2 to get critical mass you lose. Same way 1 colossus is not scary but 6 completely wrecks bio play in sc2. Now we are making higher tech unit more like immortal

[spiral aurora]

Same as 1 baneling gets sniped easily or even an ultralisk is useless.

[crimson lance]

I already explained why i like range model, i think going over it agian is not helpful. It allows for more unit stats variety in end creating multiple power switches with numbers invested in each type

[spiral aurora]

Yeah I understand your point.
Let's try it a bit and see if T1 can actually use its range for its advantage.
Banishers exist as well and might be needed in the composition.
That's what I hope. But if the approach is bad and plays out boring, we will make it better. Our minds are open.

Do you have concrete values for the tigers in mind?

A proposal from your side can help my imagination.

[crimson lance]

I don't have values for your tiger archetype. I could make it "balanced" around low range but i don't think its good. Arguably 300 dps or maybe little less could be balanced but in low numbers it stomps T1 but lacks scaling due to low range and becomes feed vs other T2 units when massed

[vale roost]

Why would it become a feed vs other T2 units when it still has exceptional health and speed? Its a tank, the lower range doesnt mean that it cant get into range of other T2 units to deal damage.

[crimson lance]

Cause of how range changes effectiveness of mass vs mass. In your model it probably won't cause everything is short range, except sheldon or hound and they numbers won't ever get high enough to threaten tigers before T3 comes out so in a sense they will work in tech split with small adjustments.

[spiral aurora]

The 300 DPS came from a typo. They should have 214 now.

[rich cargo]

If you'd set to increase Tiger's DPS, then you also have to cut its HP, because its Tiger's main point - an obnoxiously high HP per metal, that allows it to roll though most T2 defenses.

[timber hornet]

What about Sprinters? (Not just for the Tech-split, but in general.)
They're definitely a raiding unit with their stat spread, but they're so specialized into that role while only really having a speed and per-unit advantage compared to Pawns.

[misty yew]

they have more range than pawns and can kite them

sprinter is a decent unit rn it just takes more work then usually you feel like putting in.

they are also good for breakthrough and cleanup

[timber hornet]

That's what I thought. Probably one of those units that really rewards being able to micro them, but isn't that good if you don't have time to do so.

[vale roost]

Sprinters objectively have bad stats for their cost

[misty yew]

can't wait for sprinter +50% dmg mathematically approved

[velvet valve]

They'd be a lot nicer to use with +100 los.

[spiral aurora]

That sounds like a decent choice.

[topaz lichen]

Sight is an important stat. Intel wins wars

[vale roost]

Yeah, sight is very valuable. I think we undervalued it a bit

[heady star]

Calling Janus a skirmisher is a joke.

It's a specialized hit and run unit. It's not designed to take any hits and frankly they should be explosive and chain react like starlights to make them riskier to use as right now sticking them in the middle of your army has no negatives.

[vale roost]

It's a specialized hit and run unit
Thats a type of skirmisher
Its not designed to take any hits
like a hit-and-run skirmisher

[heady star]

Arguably gunslingers sorta fall into this category when against T2, but vs T1 they can brawl...

Which highlights how units can be in multiple roles depending on context and timing where they hit the field.

A Tzar at 5-7 minutes is basically a high aoe artillery unit. They are a win condition at this time.

At 15-20 minutes they are a solid threat and almost a "skirmisher" (really don't like this term) but realistically they are high burst hit and run or doom push.

A tzar at 30+ minutes is a metal donation and something to clog chokes with it's dead body most of the time.

[heady star]

A tigers SHOULD be a counter to Sheldon and Sharpshooters if you drive straight at them.

However it doesn't work that way.

The game in terms of phases/what wins or goals is basically like this:

T1 time phase: more units and snowballing wins. Macro and unit efficiency wins. (Final goal hitting buildings)

T2: range is the focus. Using range, radar, and vision to control and create gaps that Raiders can move through to hit buildings is the goal.

T2.5: spam mixed with high value/efficiency units with superior range wins. (You win by hitting buildings after creating a gap)

T3: ultra healthy or high DPS units that use whatever stat they are highest in to bully lower tier units wins (by hitting buildings that chain react)

Notice how at all stages fighting buildings is the core job of all units. Anything that lets you do that without engaging units is key

What's important to note here if how Tigers at T2 don't really fit based on their role. They are coming out at a time when things are largely decided by range but raiding and if you want raiders you're usually better off using cheaper raiders after you create a hole with ranged units.

Tigers and assault units leave big high value wrecks and also get stuck on their own corpses (which is terrible for raiders)

Jaguars by comparison are far more effective once the hole is created. They also counter air unlike salamanders. (Which usually try to be the fast response via shuri or gunships)

[raw falcon]

I think it's interesting to note that in TA the Core Storm and Arm Rocko (now Aggravator and Rocketeer) were labeled "Assault Kbot" under the unit name when you click on them. Very funny. Storms were generally better than Rockos and were CORE's best T1 Kbot, and the gameplay iirc was just "yeah shove these at the enemy base it's chill" not so much in-and-out gameplay as with modern pseudoassault skirmisher rocket bots.

[heady star]

Is this too say they were like t1 kargs, or that they were like the Legion medium rocket bots

[rich cargo]

Problem of Gunslinger is that they have no role. They do not have the proclivity stats that would inform their intended use case. Any combat T2 unit can brawl vs T1.

Tzar is an assault, though it bends towards artillery, due to range and deceptively low durability for its cost.

[rich cargo]

Tigers do counter Sheldons and Sharpshooters if you drive straight at them. Its just that people always prefer to fire from safe distance, rather than commit into assaults. Tigers have an absurd amount of armor for their resource value and easily brave the incoming fire from Sheldons/SS and statics both.

[rich cargo]

The difference is that Jaguars are raiders, and Tigers are assaults. Jaguars are extremely effective at dealing with spam and damaging the enemy fighter wall, but they fold easily to any meaningful defenses. Tigers at the same time can literally ignore any porc in their wake and wreck the economic backbone to the ground, even though they don't have the same oomph the Bulls do.

[oak python]

Unit roles/archetypes are a brief descriptor of what a unit is generally best at doing based on its stats and weapons.

I play near-exclusively Cortex, so this is how I see the units and their roles/archetype distribution:

Raider

 - The unit posseses high Speed which allows it to outmaneouver opponents and perform hit-and-run raids against sensitive critical targets. Raiders tend to sacrifice Health and Range which makes them ill-suited to prolonged engagements.

Raiders are outfitted with high Rate-of-fire weapons such as Light lasers, Flamethrowers.

Examples: Grunt, Fiend, Duck, Commando, Demon, Incisor, Garpike, Goon, Predator, Supporter, Brimstone```

---
## Assault

• The unit is durable with high Health that allows it to endure prolonged engagements and push frontlines.

Assaulters are outfitted with High impact weapons such as Heavy lasers, Plasma cannons, Heat rays, Torpedos.

Examples: Thug, Sumo, Mammoth, Termite, Shiva, Karganeth, Demon, Behemoth, Juggernaut, Brute, Pounder, Salamander, Tiger, Tzar, Poison Arrow, Cayman, Halbred, Orca, Riptide, Destroyer, Buccaneer, Battleship, Black Hydra```

---

(...)

[oak python]

(...)

Fire support

Fire supporter units are outfitted with Anti-Static Rockets, Long Range Cannons, Laser Cannons.

Examples: Aggravator, Sheldon, Arbiter, Lasher, Wolverine, Quaker, Negotiator, Banisher, Tremor, Mangonel, Catapult, Cataphract, Messenger, Kraken```

---

## Reconnaisance (Scout, Recon) 
``` - The unit posseses very high Speed and high Line-of-sight range at the cost of all else. Their main purpose is to support allied units with extended vision ranges and gather intel on enemy formations and structures.

Recon (Scout) units have no armaments or only light Beam lasers.

Examples: Spybot, Rascal, Supporter, Goon, Finch, Watcher, Condor```

---

## Specialist 
``` - The unit provides special support or situational function such as Radar and Jamming functions for intel gathering and denial, anti-air protection, behind-the-lines operations or minelaying.

Armaments are dependent on the individual specialist, but commonly are Missiles, Beam lasers or Laser Cannon.

Examples: Augur, Deceiver, Twitcher, Thrasher, Manticore, (Decoy)Commander, Bedbug, Skuttle, Commando, Trapper, Omen, Obscurer, Fury, Saviour, Death Cavalry, Herring, Pathfinder, Oasis, Arrow Storm, Phantasm, Birdeater

---

Aerial

Examples: Valiant, Shuriken, Hercules, Bat, Whirlwind, Hephaestus, Dam Buster, Cutlass, Monsoon, Nighthawk, Hailstorm, Wasp, Skyhook, Angler, Dragon```

In my opinion a more general Fire support is better as a combination of the OP: 'Skirmisher and Artillery' as Rocket units are ideal for countering static defenses, but they can also work well behind another unit as a shorter-range artillery role or even skirmish alone, but will still lose out to an Assault push as they don't have the speed to outrun a prolonged push.

Another note is that many units can kind of serve multiple roles, but I try to sort them towards what they do best.

Sometimes I may use Tzars as Fire support for my Tigers, but generally only Tigers/Tzars forward as Assault and Banisher/Quaker/Wolverine as Fire support is better/more-reliable.

Experimental units are also a bit difficult to sort as they tend to have multiple weapon systems and most of a unit's role comes from their weapon identity. I'd be inclined to maybe put all Experimental units into a separate archetype/role category of its own.

[spiral aurora]

Thx for the input. It will be read after I come back.

In another matter:

@vale roost will leave the BAR community for an undefined amount of time now.
Toxicity levels reached a critical point.
Eating shit for putting so much effort into this project and mapmaking and and spending 500$ for the omega cup (instead of his own tournament) and making tutorial videos and organizing the fight nights and reaching out in DMs to ppl is just too much.
I hope all the haters are happy now. I know some ppl didn't mean it and only wanted to stop the project. For all of them:
An apology via DM costs little and is of high value.

I myself will take at least one week of break now. Losing the person that was fighting with me the most is a deep cut.

I hope there is a rest for the project or whatever.

Until then I wish you much fun.

[river magnet]

Reading through this thread is frightening. Role based balance, is that what tech split was? Cause if so, that is poor direction.

[rich cargo]

I see no connection between the two

[river magnet]

you can search for 'role' in the tech split to find the connection

[topaz lichen]

It’s basic RTS game design

[raw falcon]

That don't mean it good :(
I'd be hard-pressed to call Starcraft role-based in its unit design tbh. Each unit is very unique and the playstyles of the factions feel so different.

[crimson lance]

SC2 units expand in capabilities laterally not linearly towards higher tiers. IN exponential eco game it makes more sense to keep current system. Does not mean "roles" should not exist though.

[rich cargo]

Most of all, Starcraft was originally designed to have completely different factions, finite resources, and limited army sizes, with very small magnitude growth between cheapest and most expensive units. Therefore, there's no requirement to maintain a level of parity between factions, which vast majority of other RTS require.

[chrome forge]

SC2 unit design revolves around 3 basic concepts:
Human tech tree -- all units have more than 1 range, zero units are melee. All units have a versatile role. Tempo is the key to success. Examples (which all have some equivalent across all factions):

• Marine - main DPS. Fragile, mobile, APM intensive. Good for almost everything
• Marauder - Main damage soaker. Keeps up with marines, has similar apm requirements.
• Banshee - Highly mobile, high dps strike craft. excels at raiding mineral lines.

Protoss tech tree: All units are specific tools for specific jobs. Building the wrong units is punished heavily.

• Zealot - Main damage soaker, also high dps. Good for almost everything.
• Stalker - Top level APM rewards use and allows both main tank and dps role. Otherwise is somewhat versatile ranged dps with high mobility.
• Oracle. Highly mobile, high dps strike craft. Has utility spells for scouting, crippling enemy forces, and doing damage. Excels at raiding mineral lines.

Zerg tech tree: Swarm, with high value units protected by lower value ones. Map control is key to success.

• Zergling - main damage soaker and dealer. fast melee unit. Good for almost everything.
  -Roach - main damage soaker. Strong defensive unit, can be powerful when added to swarms or in packs.
  Mutalisk - Highly mobile, high burst dps strike craft. Regenerates hp somewhat fast, allowing for high activity away from bases. Excels at raiding mineral lines.

If you noticed the pattern, the roles in SC2 aren't things like 'skirmisher' or 'raider'. They're 'this is the staple of every army', and 'this raids minerals'. Roles exist, but defined based on the needs of the game they exist within.

[chrome forge]

Current system very specifically IS an expansion of capabilities. t1 has limited capablities. you have a scout, a fast fighter, a slow fighter, a long range fighter, an anti-numbers fighter, and an anti-air fighter. Veh also gets a siege unit.
T2 is where you get specialized roles. sprinter/fiend/jaguar/ are all raiders super fast fighters. Tiger is half a raider and half a 'majority of the time is in your army' fighter, along with the Bull. You get the actual skirmisher role spread among a number of units. You get actual siege units.

[raw falcon]

Sure but the particulars are so diverse that you need to describe them more as the situations they're useful in than the 'role.' You can also make an argument that roles are just situations units are useful in, but shrug I feel like SC2 using role language doesn't really say anything meaningful.

[chrome forge]

Sc(all), WC (all), AoE(all), SupComm (all), Homeworld(all), the list of games that have roles into which their units can be defined fairly clearly covers every single RTS I know of.

[raw falcon]

I just don't agree that it's as applicable as you say for SC(all)

[chrome forge]

I'm honestly baffled at the pushback where people are saying 'hey roles are bad'. It's almost certainly among the most basic factors that provide identifiable interactions for both devs and players

[raw falcon]

Shrug If it's being boiled down into that phrase, it ain't what I'm saying, I'll say that.

[chrome forge]

'roles' are 'jobs' in SC2. as in 'this unit performs this job' within the context of the faction's toolkit.

[raw falcon]

You're just repeating the same thing and my earlier objection still stands.

[chrome forge]

SC jobs boil down to 'cheap main army unit, heavy main army unit, map control/base denial/mineral raid unit, flying map control/base denial/mineral raid unit, anti-army unit, heavy air unit, flying anti air unit

things like that.

But yes, I'm repeating it. Because to me it is blazingly obvious that roles exist.

[raw falcon]

Then you don't seem to understand what I'm saying, or you just like repeating yourself, because I understand your point. My contention there is that those are just too loose of "roles" to say anything meaningful about the units in question, so their categorizing usefulness doesn't apply and therefore calling SC1+2 unit design 'role-based' doesn't fly. It's too much of a heap to be a pile.

Whereas AoE2 does have very strict and clear role-based unit design (or it can be said to be so)

[misty yew]

roles are okay but in general if both sides don't agree on the same roles definition, roles are just going to confuse the discussion

[chrome forge]

That is specifically what im contending "the roles are too loose to be comparable". It literally is, just from the opposite direction role interaction is typically considered.

To apply SC logic to BAR, we need to identify critical jobs each faction needs to fill via it's roster, then determine to what degree each unit fills each job.
To apply other logic to BAR, you need to identify popular uses and movement patterns for different sets of units, then adjust those units accordingly within classes*

[raw falcon]

The point in dispute was whether it's appropriate to call SC1+2 unit design "role-based," and we're not agreeing on it.
Whether looking at SC units as jobfillers is helpful when designing or balancing BAR's units is another matter, but imo to unify the approach of understanding and design under the label of "role", like we use role in AoE2 and BAR (which is more role-based than SC1+2) just muddles things.

[chrome forge]

So what roles do you think exist in BAR then. And how would you define them

[topaz lichen]

I didn’t make this, and some units like the Grunt I would not put in certain categories, plus some static defense is actually more like static offense, but generally this is accurate

[raw falcon]

Where in AoE2 I can appropriately say that Skirmishers are Anti-Archer (role) and Camels are Anti-Mounted and Crossbows are Anti-Everything lol, and there's not as much to say about them outside of those labels,

I'll preface it with 'it's less useful (and good for fun design) to use unit role labels in BAR than AoE2,' because of the fluid & more expansive statlines and designs of units as well as the physics-based qualitative differences and impact of map and mode. So when I say that "Grunt is a raider" that straight up doesn't capture how they express themselves in 1v1: first they are "open-phase assault+raider+scout," then once the open phase becomes closed they lose the assault descriptor, then during late game spam they come back as a scout, and you can say much more besides.

[chrome forge]

From my perspective, there are a couple of different filters on both 'roles' and how those roles interact with each other. For the following I will ignore air units, bc they are roughly identical across factions with some notable differences, but lack at least one of the filters (eco strength).

As just noted, there is 'eco strength'. Every lab has at least 1 unit that you can build on basically nothing, relative to the eco necessary to build the lab itself. Gunboats on sea, scouts on land and hovers.
Next is 'unit strength'. This directly affects unit-to-unit interactions, but is flexed upon by the eco strength filter to enumerate equivalency of roles. Thugs require more eco to build as quickly as maces, so you are expected to have more maces than an opponent has thugs if you push into them.
Finally is 'Apm-influence'. This affects how much reward strong apm gives with use of the unit. Ticks, for example, have high apm influence.
There may be other filters, but i haven't identified them yet

These filters apply to roles and adjust their value across a timeline within a game. There is little APM reward for microing grunts most of the time during t3 phase, so using them as vision/cloak clear makes the most sense.

[misty yew]

roles are only good if they accurately describe the game

For example, one of the correct role definition improvements was redefining skirmisher as:

A unit that seeks to get favourable trades by using range, speed (dodging shots and finding angles), or burst dmg (fire and retreat)
These units sacrifice raw stats (hp/dps) and usually do NOT want to directly stand and bang, unlike say stout, welder, starlight

-Kiting skirmishers (sheldon, hound vs lower range units)
-Speed skirmisher (gunslinger, centurion vs t1. t2 example, leveled gunslinger vs bull. Or hound vs poison arrow.)
-Burst skirmisher (janus)

Even units that aren't dedicated skirmishers can kiting and dodge like skirmishers in certain matchups (e.g. a bull wants to kite stouts/tigers)

[chrome forge]

I mean just banisher, arquebus, and technically? sharpshooter work
Arq and banisher use hit and run and alpha, sharpshooter uses range and alpha, etc

One of the quirks of BAR is that their long range anti-unit often outranges actual defenses, which you rarely see in RTS games, which pushes those units into siege territory

[crimson lance]

Defenses buy you time, they don't need to directly retaliate, if you play around this concept its fine

[raw falcon]

Appropriately I could non-trivially call Grunt a speed skirmisher at some points. Like on Flats I used Blitzes as speed skirmishers. This is Diogenese holding a stripped chicken saying 'behold a man.'

[chrome forge]

not saying they do, just noting a quirk compared to other games

[chrome forge]

sure, but this is in t1 phase. After t2 starts you basically never build a blitz again, no? T1 units in general defy roles more strongly than t2 ones do.
I'll also note that essentially every counterargument you've made to say BAR doesnt have smooth roles has relied upon t1 unit examples. Which... sure, but they're almost entirely general-purpose units to begin with.

[crimson lance]

I don't think bots or veh have to be viable on every map. They can be interchangeble. They are only T1 units from which you transition. They should however be albe to trade fairly within limits.

[raw falcon]

Sure. Ultimately that's what I see as desirable, continued defiance of roles to the extent that BAR currently does. I don't want BAR to become more like AoE2 in its strict role adherence or SC1+2 in its extreme unit particularity.

[chrome forge]

I'd say that if this is what you want, then BAR has too many units.

[raw falcon]

That statement is either incorrect (as the current role defiance persists in spite of your claim) or claiming that my perception about the game currently is wrong. It's not the latter when it comes to this discussion, I can assure you.

[river magnet]

sorry i dont want low os to interpret what a units' intended role is and rebalance it based on their low os interpretation

[raw falcon]

I was burned long ago when Sakurai told me I was playing Smash Bros wrong and then made the shittiest gameplay imaginable (Subspace Emmisary was hot tho)

[topaz lichen]

Okay, who said anything about low OS people making the balance? Low OS people didn’t come up with the general concept of unit roles

[river magnet]

please review all the aforementioned contents and build a context window that includes the messages i have sent. then and only then, please respond with an ok and your full response

[topaz lichen]

Plus I think that’s a little bit of a hyperbole, just because you’re not good at the game doesn’t mean you’re gonna think a Mammoth is a raider

[raw falcon]

The contention is that design intent of unit roles was being used to justify said rebalancing. It's a little subtle but yeah if you broaden the context window it pops out.

[topaz lichen]

A well designed unit has one or more use cases, and also not too many. Whatever those specific units end up being and what their use cases are doesn’t matter

If not, you end up with shitty noobtrap units like the Hound

[raw falcon]

Again I think this is just overextending the argument for roles/intentional design and misapplying it. You'd have to make the specific case connecting Hound's current features to why it's not well designed and therefore becomes a noob trap. You'd also have to rebut more simple (thank you Occam) solutions to Hound being noob trap like "lol it's just bad balance. Buff the Hound."

[topaz lichen]

It’s not well designed because it’s a weird suboptimal mash up of a skirmisher and a raider and an arty that does none of them particularly well

[oak python]

SC is a poor comparison to BAR for unit roles and balancing as SC has damage and armor types that define their effectiveness against each other. Anti-Light vs Light armor, Anti-Heavy vs Heavy armour + upgrades of existing units rather than moving to a different unit tier.

[brisk lynx]

only unit role rts needs is: cool units that are fun to use

[rich cargo]

And that's why role-oriented approach has to be respected. The Hound is a noob trap because it tries to be too many things t the same time. It also should be noted, that Hounds were used on a regular basis with the abilty to switch to a more anti-unit Gauss weapon. Then this capability has been removed for the reason, that had nothing to do with balance at all (desired simplicity), which turned them into a noob trap.

BAR doesn't have the 'ability caster' type of units, nor does it have dedicated armor and weapon classes, so the entire framework of balance is an interplay between strong and weak sides defined by the stats and the parameters. The exclusions to the rule include Gunslingers, which are effectively unusable without the perks, that makes them broken, and Termites, which has no defining stats informing their use case and therefore ignored altogether in favor of specialists.

[rich cargo]

SC has a different approach to this. the unit roster is not per faction, but is shared, meaning that you can pack all existing units in a single table, disregarding the aethetics and faction-specific flare. They have a granular viability in cost and role, and then they're distributed between the factions in a way to make sure each faction has a unit, that cover a certain range of interaction.

Other than that, units are highly specific to what other units they're effective against, hence there's a certain range of units, that are not used against another faction altogether, and you typically choose which units to build towards based on what you expect your opponent to pick on their side. A given unit's role is defined by what other units it counters. New units and mechanics in SC2 have expanded on that principle without changing it in any meaningful way. Its a way to reinforce the low-scale, micro-intensive, and tactical combat model of that game, that is not applicable to BAR even remotely.

[chrome forge]

If you want units to have a harder time being identified by having role adherence or particularity then what you want is for units to be more generic. If what you want is for units to be more generic, then BAR has too many units. Generic units are units that are used for a variety of roles, and are useful for a variety of jobs. Grunts are a very good example of this.

As it stands, I think that's actually where BAR actually is rn, as roughly half the units you CAN build simply don't get built most of the time. Seaplanes, hovers, and roughly 1/3rd of t2 unit roster being incredibly rare sights in most games. This is because they are built for somewhat vague roles, and other units typically do those jobs better. The units that do those jobs better are either simply better stat-wise (or ease-of-access-wise) while having roughly equivalent vague roles, or are incredibly role-defined, making them so much better than other units at their identified job that there's no reason to make something worse.

[river magnet]

wat? seaplanes are op and hovers are op

[raw falcon]

Hound is noob trap because it has too many roles, but it was NOT a noob trap when it had even more (hidden) capabilities
Man. The first paragraph's logic just doesn't follow. Another contradiction, is Razorback a noob trap? That shit has so many roles and it's super cool, very useable unit overall, currently not OP.
Termites have a pretty clear use case afaik. Maps with important mountains and as a pseudo anti-spam unit akin to Salamander. DO YOU SEE WHERE THE ROLE LANGUAGE BREAKS DOWN IN FAVOR OF SITUATIONAL/PARTICULAR LANGUAGE LIKE IN STARCRAFT? This is a viable design path with fun units and it's worth ignoring role-OBSESSED design in favor of.

[raw falcon]

False dichotomy.

I had a huge paragraph that I was gonna post last night that basically went "look, using role language isn't inherently bad. But here's where it's useful and here's where it's not." I deleted it because it's too much yap, but suffice to say @rich cargo I respect it where it's appropriate. In fact the "help diagnose OP unit" was part of that paragraph. So I understand all. My whole argument is that it's at high risk overused and misapplied (AND WAS INDEED).

[chrome forge]

a convergent truth, i'd say

[rich cargo]

No, Razorbacks are Assaults. They do not have the speed for raiding, they do not have the range for skirmishing. They do not have many roles, but simply have meaningful utility that justifies its T3 tech grade, and since they're Armada cookbook and the Bot by unit class, they do not exceptionally durable, but also not slow.

[chrome forge]

To add to that, they do around 1k dps, do 25% damage to air units, and since most air units have between 500-2k hp, they are also incredible AA units.

[rich cargo]

Termites are virtually not used, especially compared to many units from the same factory, and from the other faction's analogues. That's because they do not have the speed, the range, and the durability that would give you a way to achieve a particular goal. They have the all-terrain ability, but that alone isn't enough to make them a viable combat unit.

[chrome forge]

There is also only very rare use of crawling bombs of all kinds, with skuttles being the only ones really seen due to cloaking

[rich cargo]

Yes, but they mechanically operate as AA mainly to protect themselves against gunships. Again, given they're not particularly durable, esp. compared to Thors, their strong suit is accurate DPS. Regardless, they're still Assaults, and there's no other roles for them to cover.

[raw falcon]

Ya just because Termite not used doesn't mean you can use the hammer of unit-role-obsessed-design to tighten down the bolt of balance. You could, but by the time you get the bolt tight it's misshapen and flat and more like a rivet. And that's LAME as FUCK.

[chrome forge]

Mobile AA, anti-spam, anti-heavy dps while moving fairly quickly for a t3 unit and having 13.3k hp seems to fill a lot of roles though. they even have minor splash damage lol

[rich cargo]

You're not tightening the bolts of balance. You're informing the approach to balance.

If you have an understanding of what the unit's intended role on the battlefield is, then you can make a decision on which stats are to be adressed, whether just buffing something, or buffing it more and nerfing other aspects.

This is the only efficient approach to this process, unless you're willing to implicate the unique features and abilities, which doesn't seem to be the desired direction for the balance and design decisions of the BAR development team.

[chrome forge]

I think basically everyone can agree that the lines of 'roles' should be blurred to some extent. It should be noted on both ends that having roles that are too clear leaves you with very obvious rock-paper-scissors boringness (which is why i don't play sea), but also that having lines that are too blurred both makes unit interactions hard to balance or tweak, and also leads to many units that could have a space to shine, simply being passed over for generic units that do more things better.

I personally feel that defining roles clearly, then blurring the lines they create, leads to the healthiest and most intuitive foundation for doing balance work, because that means you at least have some idea of what types of units can or should exist, and go from there to push units into variety and interestingness

[rich cargo]

They should not be blurred per se, but there should be a certain degree of leeway and crossfade. This is especially permissible with the more advanced a given unit is, since the part of the unit's tech value is that its not as strictly bound to the stiff role definitions like the T1 units are. This way you are not forced to design the units, which are flat-out better versions of the unit from the previous tier.

And that is already the case for a good portion of the units in roster - units typically given some sort of utility or secondary flexibility, that differentiates it from T1 lab analogues, rather than just having flat-out multiplied base stats.

[chrome forge]

I'm not against this... much. Let's take Mausers as an example. They're long range anti-t1/1.5/2.0 artillery, that also turn fairly quickly for what is essentially a siege unit, with a fairly accurate plasma shell that does just as well vs t1 units as it does vs structures. They cant shoot on full retreat, but they are still used as fairly effective skirmishers. They're an example of a good 'siege unit' that has enough range and flexibility to be used 'to skirmish'. I don't think this is unhealthy, but it does draw fairly clear lines as to what the unit does. It's certainly an upgrade from the t1 artillery vehicle

The problems I have come from when units that should be 'upgrades' to a unit type do too many things well (or poorly), or when a 'sidegrade' unit does an 'upgraded' units job so well that the 'upgraded' unit has no reason to be built

[rich cargo]

Yes, but you can also argue, that Mauser can be more differentiated from a Quaker, as to make it more of an anti-unit skirmisher, while making Quakers more of a AoE siege battery. This does not relate to the role-based balancing principles, but moreso to the distinction between factions, and how units of the same role can be designed differently by nudging them here and there to emphatize different use cases, and making faction mixes more interesting, without devolving into SC-like narrow specialty too much.

Like I said, my main gripes are with units, that simply do not have a role. They kinda have mediocre stats (Termite) or outright depend on one stat, that cannot work on its own merit (Gunslingers).

Another questionable case, is when the faction just flat-out does not have a unit to fulfill a role, i.e. Cor having no high-alpha units altogether, or lacking any sort of cloaked vehicles for vision. In part that also originates from more general problems with the game, like having no sensible radar/jammer interactions, and players having to rely on unit spam for vision.

[river magnet]

how are razorbacks not skirmishers by this definiton

[abstract drift]

My 2 cents: archetypes are useful to understand but eventually we must return to designing units on a case by case basis for their unique coolness, capabilities that drive fun RTS control of each unit. On that note, I strongly recommend we account for the archetypes of fun.

Why do players make rocket bots? So they can deal free damage to an otherwise frozen standoff where neither side wants to run into the enemy's porc/defenders advantage. It's tactically advantageous and fun to hit and run while making sure you don't overinvest in rocket bots such that the enemy can now push into you profitably with their 100% more HP/dmg efficient units.

Every unit should feel sufficiently unique, always fun, and we balance everything else around that. Let's make sure we have our priorities straight.

[royal cedar]

unit roles are kinda pointless, because the moment a unit gets outranged it either has to retreat or is forced to be a bad "assault unit"

sheldons can no longer skirmish when you encounter sharpshooters, at that point you need to desperately get in range of the sharpshooters or you just reclaim all your shelldons back at base

[crimson lance]

if there were damage modifiers, armor types, upgrades, abilities, game would be more interesting regarding unit designs, but with just stats and cost it's hard to get diversity. Best that can be done is diversity by projectile speeds, ranges, hp, etc...

[fathom elk]

I bet a lof of unit role definition gonna change with the new regime

[topaz lichen]

This is the most effective way of making a unit feel unique. Having different flags and attributes attached to units like an RPG game is boring

[crimson lance]

It creates stat check meta and monospaming, but makes sense for bar economy

[topaz lichen]

What is "stat check meta". What competitive game doesnt have people checking and comparing stats

[wise surge]

what new regime? 🤔

[crimson lance]

Modulo balance consultant, time to nerf tzar i think 😭

[fathom elk]

neb left i think

[mortal cape]

I think we need a unit role for commander

[rich cargo]

Because.... they don't have the range for skirmishing? Are you even reading what you're replying to?

[rich cargo]

which is 50 unit distance difference.

[royal cedar]

yes but it means when the sniper moves backwards and the spam moves forward the sheldons can no longer skirmish the trash units wich means they are forced into a brawl
even if you bring anti spam its going to be very messy

[rich cargo]

I dont understand your point. Does skirmishers supposed to win against artillery?

[royal cedar]

are snipers artillery? ok what about hounds vs banisher, both are "skirmishers" but realisticly hounds are not they are basicly just t2 assault tanks cause they allmost never outrange anything

unit roles like skirmisher or artillery are pretty pointless the units just end up behaving in the same way anyways and range is all that matters

[rich cargo]

They outrange 8 out of 10 units from the same tier. And Hounds together with word "tank" makes me laugh uncontrollably.

The more you press into range, the more you become vulnerable to raiders, and to smaller extent, assaults. Range beats shorter range in terms of non-commitment, but units with highest range are slowest, more brittle, and either less accurate, or have harder time dealing with sheer numbers. Skirmishers aren't expected to defeat Sharpies, they're intended to harass and attrit the assaults and static defenses by staying in their desired range, which artillery have harder time doing.

But yeah, if you don't get the point, then it pointless for you.

[chrome forge]

Comm is chess king :p

[oak python]

Some High Alpha damage Cortex: Behemoth, Cataphract, Arbiter, Banisher, Commando(?)

There are high damage units, but they typically aren't a tzap attack type like the Starlights or Sharpshooters

[chrome forge]

so 3 of Coretex's 'high alpha' units are t3, and you did not include the Juggernaut in this for some reason. Arbiter is a siege unit, so while yes, high damage, no, it doesn't really hit units unless theyre fully unmicro'd. Starlights and Sharpshooters have 900+ range -- firing well before any Cor unit can get into reach. Adding spam to them makes them nigh unkillable except using multiple tremors to simply deny the existence of units in the area. The chain of interaction continues in a long line from here, but basically BAR is what it is and is played how it is because of how the interactions work. modifying anything would change the whole system.
Thus the constant pushback on changing anything unless the change is so small nothing is different

[oak python]

True, the Juggernaut does have his laser, it does do high damage.

I was merely listing some of the high alpha damage units, if you wanted a how-to guide for Cortex: the Arbiter costs significantly less than Sharpshooters to create, have no cost to fire and have a decently higher range than the Sharpshooter.

By sending a spam unit to see Starlight/Sharpshooter weapon discharge in fog-of-war or otherwise obtaining vision or radar signature on the Sharpshooter/Starlight, you can safely fire a volley of Arbiter missiles at no risk to self.

Arbiter needs only 1 rocket to connect with a Sharpshooter for a kill. Starlight requires around 4 rockets, but is even worse at mobility and has limited firic arcs, so even if a Starlight dodges you're fighting the pressure a Starlight creates by forcing them to move away from rockets.

With proper micro on both sides, Arbiter vs Sharpshooter with spam support, it looks to me like a bit of a standstill, but Cortex has invested less resources. Sharpshooters will dodge rockets, but can't get into range to kill Arbiters.

With poor micro, Arbiters could get sniped if they are forgotten, or a single Arbiter rocket can take out a whole ball of 5ish Sharpshooters.

Negotiator is a similar unit to Arbiter, albeit with slower, but harder hitting rockets with slightly more AoE.

Tremors are another option, as you also mention, more expensive, but much more area denial.

[chrome forge]

Sharpshooter sees a unit and does 2.5k damage to it. Arbiter fires a missile, the enemy team takes a coffee break, and when they come back the missile is nearly landed. They're not even remotely comparable.

[oak python]

They aren't comparable, they are completely different units from different factions that do things somewhat differently.

I am arguing against your point that Cortex can't do anything against Sharpshooters, that Sharpshooters are unstoppable and cannot be outranged.

[chrome forge]

I never said that. I said there was a chain of interactions tied to them that cortex has to utilize well. Sharpshooters having no counter would mean all pros would only build those

[river magnet]

are you ok? they have speed. this is why giving people the idea of ''roles'' for balance is an issue. speed is range, if i can get in and out before units can hurt me, its pretty much like RANGE aka skirmishing

[topaz lichen]

Man some of the things people say in this thread make me scratch my head

[rich cargo]

I agree. I guess speed is range. Perhaps range is also armor, and armor is DPS. Nails are bananas. How one could be sure, am I right? It only depends on the perspective.

[chrome forge]

From what I can tell, very few people with purple names are worth listening to. They just troll and derail discussions for the most part. ignore them

[river magnet]

Cause an idea is being floated that skirmishing is just from range, anything that provides an advantage that allows a unit to engage allows it to be a 'skirmisher'. Survivability is not just range but flight ability, speed, tankiness depending on the match up.

Imagine if roles were being used to balance units. I believe it would make the units to one dimensional

[chrome forge]

Did you read the whole thread or just the last few bits? Because skirmishing is not 'range'. And I personally don't agree that skirmishing is just range and movement either. It encompasses a variety of traits and stats that are moddable, from turret targeting radius and turret turn speed to unit turn speed and LoS, to even the type of weapon used. Range and damage and fire rate are just the most basic stats that allow skirmishing to be possible at all. If you have low damage and fire rate and range, you can't really skirmish unless that stat distribution is the norm of that tier. Like-wise, if those stats are all high, that puts your unit into sniper role territory.
A lot goes into setting up a unit to perform a role. Just range is a baffling take for anyone.

[chrome forge]

Again, roles are not the be-all end-all of balance. They are a lens to use to look at how units perform and for what reasons they are used. Establishing a role for a unit is specifically how you modify a unit so that it doesn't lose it's current function

[river magnet]

rip cant even define the roles to see how units perform gg, i dont see it turning out well. i agree with others, focus more on stats than trying to cookie cutter units into roles people can't justify

[chrome forge]

I'll make sure to add you to the list of people with purple names that aren't capable of genuine conversation though, so that's a positive

[abstract drift]

It's fascinating how range and speed both matter a lot, but also don't matter much in certain cases. For example, when the enemy units have high HP (ex. Stout tanks), the advantage of outranging them may effectively only be 1 or 2 shots, which may be less than 10% of their HP. Range is much less useful when the enemy is semi-fast and tanky against your moderate outranging DPS. Conversely, when you have high projectile speed and damage-per-shot (ex. sharpshooter, starlight), fast enemy units don't matter as much because they just die to your alpha strike. 2500 damage is worth about 400 metal (assuming average 6 HP to metal ratio), so essentially any fight you start with a huge metal advantage as long as the targets are tanky enough to get full value out of your big damage shots.

Unit names and aesthetics are really important for driving unit designs too. "Pawns" for instance imply that word's function from chess. You put them in front of your valuable units to soak up damage (like the ablative shield on an orbital re-entry spacecraft). Perhaps this is why Pawns had unusually high HP per metal from the beginning. Maybe they should continue to have good HP/MC and be less of a raider (less speed, maybe DPS/metal ratio too)?!

[royal cedar]

this is actually not true in this case, because as long as you one shot a few tanks in the first row they will spawn wrecks wich will slow down the stouts massively, but allso youre going to either have LLT creep or dedicated frontline units for your ranged units like welders + snipers

my point was mostly that dedicated skirmishers are niche by design and only work as long as they dont get out ranged, if a skirmisher gets outranged by a dedidacet ranged unit wich allso has a dedicated melee unit in front theres not much left to do for them

[rich cargo]

If the tank formation is staggered and not cluttered, then it won't be much of problem, and that's how you're expected to handle vehicles in general, since none of them turn on a dime like bots do.

Everything works 'as long as lalalala", otherwise its imbalanced crap. Everything is supposed to have a counter-play for the same cost, or you have an OP unit and a garbage balance model.
A dedicated ranged unit with a dedicated melee unit is countered by a group of dedicated assault units for the same cost. If it doesn't, then your assault are garbage, and you need to balance further.

The core tenet of role-based balance, is that you have a response to any play given there's intel or informed expectation to deploy it in time and minimize the collateral damage.