Loading...
  OR  Zero-K Name:    Password:   

Why Rogue is 1v1 God-Tier

29 posts, 1486 views
Post comment
Filter:    Player:  
Page of 2 (29 records)
sort
Rogue has some interesting qualities that give it more advantages than almost any other combat-focused unit. Top players are growing more aware and dependant on rogue spam in order to gain an advantage. Personally, I think it results in games that are currently too frustrating to interact with. I'm going to break down the issue with rogue as I see it. Here are some of it's wonderful traits:

Problematic high range

Rogue is one of very few skirmishers that enjoys impunity from LLTs. In frontline combat, this means that reinforcing your position with them is useless. It also means that they can take pot-shots and kite other skirmishers relatively cheaply, gaining a huge retreating bonus in the process.

The only other skirmisher with competitive range is recluse, who does not enjoy the same low cost, low weight advantages. Moreso, Rogue is far more maneouverable than recluse. 57 speed vs 45. 391 deg turn rate vs 276 deg. This is an important detail, because when it comes to attack walk stand-offs or even retreat commands, recluses lack of mobility allows it to take hits more easily. A hit from a rogue is worth 345 damage, which is more than half of a recluse's 650 health. This means that a dancing rogue only needs to hit a recluse a single time to make more than it's cost of 120 in value (recluse costs 280, so a hit is doing more than 150 metal cost in damage here).

Low cost

At the cost of 120, Rogue achieves comfortable entry into the high granularity bracket, particularly when it is paired with dirtbag, bandit and vandal. This is a serious advantage because high granularity affords inherent protection against almost all long range attrition. An example - A lance has a cooldown of 20 seconds. At the reserved average 1v1 mid game income of 40-50 metal per second, a rogue builder can produce roughly 5 rogues in the time that it takes for a lance to reload. Getting 5 lances fielded (for efficient kill rate) 100% comes at both the loss of holding the frontline and with new jarring weaknesses that the enemy can exploit. Phantom suffers from the same theoretical issue to a lesser degree, the only upside being that it is slightly cheaper and has a 17s reload.

Emissary exists in the anti-rogue sweetspot, with just enough (600) damage to one shot a rogue (540 hp) if the shot is direct enough. Okay, now we might actually kill 1 rogue per 7 seconds. Which ironically, is enough time to produce just over 2 rogues in response. Another downside exists, with it's stationary nature and Rogue's decent range/speed, and rogue forcing all front-line defenses into uselessness, what is going to stop the rogues bridging the 590 elmo gap and abusing emissary's fixed position?

Bombing them with ravens is even more inefficient/costly. You will lose the battle if you try to make that trade on even or even advantageous footing.

Burst

With the two problems above combined, it becomes apparent that stinger exists neatly in the sweet spot of being able to efficiently fight rogues. It has the correct tier of damage and reload time to be the perfect anti-rogue rate-of-kill. Except everyone and their grandmother knows at this point, that 2250hp / 345 damage = 6.5 -> 7+1(for sacrifice) rogues to instantly burst a 450 metal structure at the cost of 120 metal. There may exist a theoretical stinger-forest creep approach to dealing with rogues, backed by enough welders or convicts, but that opens you up to arty. Those stingers also aren't going to go raid the enemy.

Dance on the spot AI, retreat advantage, AoE

We've already seen that rogue has a good time against recluse. But it gets worse. I have watched a single rogue dance against 6 recluses and kill 2.5 of them before it died. That's nearly 6 times cost made. Though no sympathy for spiders against shields.

Let's switch to more conventional skirmishing, like against ronin. The combination of increased range and AoE mean that ronin need to advance to shoot rogues. Advancing can often mean tightening proximity to eachother, which rogue's already convenient arc salvo AoE can exploit. If you run some attack move tests in single player, you'll see it plain as day. I've done same cost 33 ronin to 25 rogue on attack move, and the rogue have won 21 rogues to zero ronin. A profit of 2520 metal. Stationary bulkheads and buoys fare little better, as our friend rogue weaves in and out of shots at the low, low cost of no micro. Thanks to endless dancing they become silly hard to land a hit on in a timeframe where 2-3 more have not come to replace the one you did hit. Games become a frustrating and desperate search for something, anything, that can actually lock onto the little bastards in a timely and accurate manner.

Sight range

Just a small edit to mention a subtle detail not often considered. Rogue's sight range is 583, whereas Ronin's is 523. This subtle detail means that unless you have some other reason to have good vision, fighting into rogues can often feel like you're fighting blind into something that knows how to target you very well.

Paired with the current #1 combat raider

Well, why not raid them like you're meant to do against skirmishers? Simple and easy answer. Bandit currently dominates other raiders, particularly in high density where micro-plays aren't as easy to pull off. Kodachi might have some chance here? But otherwise, rogue comes equipped with the perfect high granularity sidearm. It's not possible to effectively raid them whilst bandit stands buffed.

Possible counters

Firewalker. That's it. And that's if you don't break the front getting one. Though I'd wager purchasing an aspis in response to the first firewalker shot soundbyte isn't that hard of a counter-ask.

AUrankAdminGoogleFrog I hope you'll take this more nuanced breakdown into account and consider rogue for rebalancing, sooner rather than later. A 120 metal unit shouldn't have this much influence over the game.

Here is one replay: Multiplayer B1393994 2 on Scaryland v1.02

There are many more. Maybe others who play more than me can contribute some.
+8 / -0

21 months ago
I did a rouge switch recently, because I saw some high tiers 2nd fac into rouge fac. So in a tank mirror I did the same. Multiplayer B1393926 2 on Red Comet Remake 1.7
+0 / -0

21 months ago
1v1 players only switch to Shieldbot because of one unit: Rogue, something seems wrong.
+0 / -0
Even I noticed how often PRO 1v1 players start with shieldfac nowadays, when spectating - and that it is mostly because of rogue.
I agree that it could definitely use some nerf.
However, I would like to add that shieldfac does not have very good riot units. Sure, they got felon, but felon is fairly expensive, and needs thugs to act as a battery (and even then it takes time for it to recharge, and thugs make it even more expensive). Which means that unless a second factory is built, viable strategy might be to sneak up on rogues with (cloaked) raiders, preferably some heavier raiders like kodachis or bolas, glaives would die to the felon.
+0 / -0

21 months ago
quote:
Firewalker.


In some games, I got an interesting tactic, only 1 rogue is displayed against the Firewalker, who tries to attack the firewalker and he fights 1 on 1 with the rogue, with tight lines, the firewalker turns into useless shit xD
+1 / -0


21 months ago
Agree 100% with Dregs. Only counter to rogue is firewalker which is pretty hard to afford and get going in practice. And firewalker can be countered by many things (unlike rogue).
+1 / -0


21 months ago
That's a nice list of reasons pair of replays. How about this: https://github.com/ZeroK-RTS/Zero-K/commit/3b30db4cef0e7a1a6ebc17eaec3b230ccec64a3d

Rogue:
  • Cost 120 -> 125
  • Speed 57 -> 54
  • Turn rate reduced by 5%.

It can be tested by selecting the mode Zero-K: Dev Version.

Rogue has been trouble. It was unbuildable four or five years ago due to overall worse stats, the most notable being a 9% slower projectile. Since then there has been a lot of fiddling with speed, projectile speed, health and damage. So perhaps it is now time to try making it slightly more expensive.

From the two games linked so far it looks like Rogue is being made to counter the right things. Rogue should be beating Ogre, Cyclops and Thug. The troubling thing is that people are responding to Rogue with Rogue, rather than some other counter. I think this has a lot to do with the stage of the game, to extend the point about granularity from the OP. There are ways to make attrition against Rogue, and there are defenses that will slow them down sufficiently for this attrition to take place, but those aren't viable until later in the game than Rogue can appear.

Raiders are the intended early-to-mid game counter to Rogue. This is why I like nerfing cost and speed. Making it take slightly longer to build up, to reach the frontlines, and having it be a bit easier to catch with raiders seems like a way to address these issues. I know that Rogue-Bandit can deal with raiders, but a more expensive Rogue means fewer Bandits, and a slower Rogue is easier to punish when out of position. I'm not looking to remove Rogue-Bandit as it seems like a pretty obvious combination. It's hard to imagine a world where Rogue and Bandit are each independently viable but where Rogue-Bandit isn't strong.

I'm wary about touching range, and not just due to Quant's Rule. No other shieldbot can deal damage to Lotus without retribution. Rogue outranges low range turrets because every factory should have a way to do this. Now take stopping distance and Skydust overshoot into account, and Rogue can't have much less than 530 range. It currently barely outranges a Stardust raised by 100 elmos, an elevation that corresponds to a Stardust at the top of a veh-pathable ramp. I also suspect that Rogue needs to at least hold off Recluse, or shieldbots will completely lack answers to Spider. A small range nerf would probably be felt most against Recluse, rather than against Ronin.

Tweaks to damage have happened before, but I don't think they are needed on top of the proposed nerfs. I'm fine with Rogue bursting down Stinger as Shieldbots otherwise have few good answers to a supported Stinger. The speed nerf already makes Rogue worst at setting up to snipe Stinger, giving an opponent more time to respond.

It would be nice to see Ronin in the Cloaky vs. Shield matchup. That said, I don't think we can let Rogue lose to Ronin as Ronin-Reaver might not have other answers. Having Ronin fight Rogue to a near-stalemate seems like the safest option.
+1 / -0

21 months ago
*sniff*
+4 / -0
21 months ago
Two games on test version https://zero-k.info/Battles/Detail/1395174 https://zero-k.info/Battles/Detail/1395170
nothing terribly conclusive given the elo difference and all. They do feel comparatively tanky, which is maybe odd for a longer-ranged unit that's safer than eg. ronin.
+0 / -0


21 months ago
quote:
I'm fine with Rogue bursting down Stinger as Shieldbots otherwise have few good answers to a supported Stinger.


Surely Thugs (a core shieldbot unit that's generally useful and has one of the best cost/hp ratios in the game, even ignoring the fantastic hp sharing and regen) counter Stinger incredibly well? A few thugs and something else (rogue, bandit or felon) kill stingers with zero attrition.
+0 / -0
AUrankAdminGoogleFrog - Thanks for the update. The cost increment seems a little small, but let's see what happens with this. I don't expect much at this level.

quote:
No other shieldbot can deal damage to Lotus without retribution. Rogue outranges low range turrets because every factory should have a way to do this.


That's not entirely true. Felon toasts them, Racketeer disables them and Thug honestly loses negligible effect from dealing with them. Porc busting is not shieldbot's weakness, though I will concur that recluse is (thus my earlier suggestion of a lighter weight racketeer on par with recluse).

However, if every factory should have a way to do this, why do moderator and scalpel stand under LLTs eating damage for days after their attacks? If the counterpoint is that lance and firewalker are the intended tools, then the arguement for rogue being the ranged element gets superceded by the existence of felon.
+0 / -0
quote:
then the arguement for rogue being the ranged element gets superceded by the existence of felon.


Felons are outranged pretty hard by llt. A pack of Lotus can most certainly ward away a small shield ball, especially if it is supported by units.

Racketeers do outrange most porc, so even if Rogue's range was nerfed, shieldbots would still have a way to outrange porc. Perhaps it would not be efficient, but it would probably be equally as inefficient as using Firewalkers or Lances to clear out basic porc lol.
+1 / -0


21 months ago
Again when have you ever seen an LLT with a felon and presumably accompanying thugs and thought, crikey, I wish these boys outranged those.
+0 / -0


21 months ago
Dealing damage to Lotus without retribution is an easy to state metric, not the only goal. I'm not fixating on Lotus (or Stardust), it just would seem bad for a factory to have no absolutely way to deal damage beyond 460 range. Yes, shields and Racketeer do something, but there a real benefits to dealing actual damage without exposing yourself to any potential damage in return that seem too important to miss out on. Moving a Thug/Outlaw/Felon into a supported pack of turrets has, if not more, then at least different associated risks.

Scalpel and Moderator are indeed able to have less than Lotus range because of Lance and Firewalker. These don't need to be early game units, but the options are there. Requiring a factory switch to safely poke at a forest of Lotus, Faraday and Stardust would suck.
+0 / -0
Rakketeer is an absurdly lopsided unit and I cannot fathom how having it be the answer to recluse would be a healthy state of things.
It kind of makes units that cost more than ~800 metal and do not outrange rakk flat out unbuildable, leaving very little non-firewalker type answers to rogue.
(Pre-nerf) Rogue kind of outbrawls grizzly as is.
+0 / -0

21 months ago
There's this one useability issue that I have with rouge, a group of them closely packed together will friendly fire when on a slope and firing up/down the direction of the slope, sometimes. I'd claim colvolume, but I've been wrong about that in the past. Bots B1395231 1 on Alphabet Siege Dry v1.20
+0 / -0

21 months ago
quote:
Moving a Thug/Outlaw/Felon into a supported pack of turrets has, if not more, then at least different associated risks

Aren't "not more but different" risks actually great and desireable from a uniqueness standpoint (Quant's rule for factories)?
+1 / -0


21 months ago
I think the threshold of uniqueness that makes for playable factories is at "can outrange HLT", not "can outrange LLT".
+0 / -0


21 months ago
Sprung wanted me to put these here:

Two games that are the same but inverted, where comm/rogue forces the opponent into submission.

Multiplayer B1395888 2 on Trojan Hills v05
Multiplayer B1395914 2 on Trojan Hills v05
+1 / -0


21 months ago
This one I want to add for myself: Multiplayer B1396000 2 on FrostyCove v1.13

Bandit put in most of the work here, but rogue was great at hitting my glaives too
+0 / -0
Page of 2 (29 records)