StarCraft II 5.0.11 PTR Patch Notes

Yeah, I’m still confused by the Raven, and I fear they removed the Corvid Reactor because they forgot to add it back or they’re worried about turret spam in the endgame. I fear the professional players will use the excuse of removing Corvid Reactor because “we don’t use it”, but it’s sometimes used in lower leagues and I fear it will remove future diversity where a professional will research it as part of their build, if just to prepare for the endgame.

I heard one problem is that TvT endgame armies tend to be Marine/Tank/Viking, with the marines slowly being replaced with Ravens, and that apparently needs to be discouraged. I would argue for reverting the Raven’s changes but then changing the Auto-Turret’s weapon with a projectile-based one so there’s more overkilling and so marines still have a purpose with their hitscan damage.

ADDENDUM: Regarding the Observer buffs, it’s reverting a change where Blizzard slowed down the Observer because of the ‘disappointment’ of just failling to kill one when you scan it because it managed to dart outside of the zone where it could be seen.

1 Like

I think you are wrong. The reason you would use an auto-turret in tvt is because your marines can’t engage, or they would be melted into pieces if try to. The auto-turrets are being thrown right in front of the tanks, forcing a friendly fire and killing the tanks. So marines can’t replace auto-turrets at all. They can be used in combination with interference matrix, which would be still the option after the patch.

The issue with terran’s late game is that it’s dominated by the siege tank. It’s impossible for ground units to attack, so they move up to massing air units. The siege tank / liberator are the real issues. One or two or three are fine, but 15 of them completely stalls the progression of the game. There need to be timing windows where the tanks / liberators become obsolete. They can only stall for so long. They buy you time, but can’t last forever.

This issue extends well beyond TvT which is why, I believe, they are nerfing the sensor tower range. The sensor tower allows the Terran to always know where enemy attacks are going to hit, which makes it too easy to move your tanks into position to defend. Reducing the range means the terran will be slightly slower to notice the attack, and the tanks won’t be able to move as far forward before having to siege. So it’s equivalent to pushing the siege tanks backwards a bit. This effectively reduces the area that a siege tank can adequately defend. In late game scenarios where you have to take far-off bases, the siege tanks are greatly weakened. That means they have a power-curve. They are strongest in the mid game, but once the terran has to expand to far off expansions they become a lot weaker.

1 more suggestion:
Can you make units that are on HOLD POSITION or PATROL commands be excluded from being selected with F2? This would make holding watch towers, keeping medivacs in corners, and keeping zealots in the wall a lot easier.


my proposed changes actually fixes the game, not this garbage.

In order to fix it we must tackle it from the ground up which is what my changes does.

1 Like

I can’t say I know for sure what’s happening internally, although I get the sense that a lot of the patch proposals, in both Co-op and Versus modes, are being performed by third parties, with Versus in particular being done by high-level professional players. The problem with this is that at the professional level, there’s a large emphasis on putting on a good show, wishing to enforce games that are fast-paced and dynamic. And while I don’t doubt the skills of the professional players, I find a lot of their games play in a very particular way.

I suppose some things should be considered from a game design perspective rather than “what do the professionals currently do”? The Raven, having been brought up so much, is being looked at as if you’re building an Oracle… you build one, maybe two, when you first get your Starport, then hotswap it back to a Reactor instead of the Tech Lab, never to use it again… at least that is what is trying to be encouraged, it seems. The problem is that professional players have the micro mechanics to keep their Oracles alive for the rest of the game, so it can keep building value with Revelation and Statis Wards. All of the current proposals to Ravens are about getting it out as fast as possible, but then tone down its power so it’s not unbalanced in the early game, which then harms its usefulness in the late-game when you have more Orbital Command energy available for scans (hence you don’t need them for mobile detection all that much), and yet Auto-Turrets and the Anti-Armor Missile are being weakened. Maybe Ravens are rarely, if ever, used in match-ups outside of TvT, and a fast Raven may help more in TvZ to clear creep, but harming its late-game potential is the wrong approach, I feel. I think it would be better to see if its abilities can be made more useful. e.g. Having Interference Matrix also reduce the unit’s vision to 1 and allowing it to target Overlords and Overseers (give them the Psionic tag) could produce some interesting engagements when you disable an enemy’s detection or high ground vision (or just a unit that’s further ahead of the tank line acting as a spotter).

I get the impression that some decisions are possibly a little narrow-minded. One that’s been around for a while now is the Tectonic Destabilizers for Tempests. I can understand the reasoning for it… PvZ matches had devolved into “Spore Forests” and Vipers that slowly creep forward and in the right hands are very difficult to break through, and this was compounded with the nerf to Feedback that prevented High Templars from outright killing full-energy Vipers, so with needing a long-range unit that could pick apart the forest while being able to keep a relative safe distance from enemy Vipers, the Tempest was the logical option (Carriers are a bit too slow and a large number of Spores, coupled with Fungal Growth, would shoot down Interceptors). The problem though is that this caused an overlap between Tempests and Carriers, with Carriers supposed to be assault units while Tempests are a strange mix between a siege unit and siege breaker, and more importantly, it adversely affects other match-ups that rely somewhat on structures, such as mech. In PvT(mech), Terran players rely more on Missile Turrets, Sensor Towers and Planetary Fotresses because of mech’s relative immobility, but the Tempest already counters most mech units and trades fairly evenly with Thors, but at the same time, a group of them will mincemeat an expansion very quickly and there’s not really mucn you can do except to overproduce Vikings, which come with its own problems (vulnerable to Blink Stalkers after kiting with the Tempests and eats into supply that might be better suited to Siege Tank, Thor and Hellbat production). More than anything it’s just something that feels a little unfair.

I do have my own suggestions for the Tempest problem, but I’m not sure if it will be liked all that much, and it’s hard to suggest anything without being called biased (I main Protoss and off-race as Terran).

  • Bonus from Tectonic Destabilizers reduced from +40 to +30 vs. Structures (will be on equal footing with a Siege Tank’s damage output, and will still 6-shot a Spore Crawler… reducing it to +25 vs. Structures would require 7 shots, and may be an option if it needs better balancing with the second change below).
  • New upgrade at Templar Archives named Twilight Focus (150m 150g, 100 seconds to research) - increases the damage of Feedback from 0.5 HP/energy to 1 HP/energy (back to its original damage, which will permit a well-placed Templar to kill, not just injure, a badly-placed Viper!). Being behind an upgrade, it will only become prominent in the late-game since, unless you have a very good reason for rushing the stronger Feedback, you’d be researching Psionic Storm first (I can possibly see it as being useful in PvP if the opponent has a large number of Phoenixes).

In the end just another nerf for unit who already had no role in game cause overkill and slow rate of fire

Honestly, the best use I find Tempests personally is in late-game PvT(bio) where the Terran player has ranged Liberators in play (Advanced Ballistics will zone out Disruptors) and you’re otherwise in a position where you can’t engage without taking heavy losses (especially if there are also Ghosts out) - using an Oracle (which will probably get destroyed) to cast Revelation, you can force a more favourable engagement by having the Tempests fire upon the LIberators using their superior 15 air-to-air range. Marines absolutely shred them though even with their +2 base armour, so you have to keep them out of harm’s way. Notably, their precise damage output doesn’t really come into play (hence neither does the upgrade)… it’s about the range and doing token damage to force the engagement before the Terran player loses all of their Liberators over time.

Tempests are also otherwise very good against enemy Carriers, Colossi and Battlecruisers (not so much Brood Lords because Corruptors tend to support them). Against mech units they are quite strong all round because Siege Tanks can’t shoot up and are rather immobile and Cyclones die in 3 shots. They are fairly even cost-for-cost against Thors, and equal each other’s attack ranges.

1 Like

Auto turrets have the same issue that locusts do, albeit on a smaller scale due to their cast range, that is that they’re free. Neither the auto turret or the locust require the caster to stick around, and both being ranged means that they can to some extent they can stack DPS.

The broodling has such a short lifespan that it forces the broodlord to stay in a somewhat dangerous range, and being melee it also can’t stack damage the same way ranged DPS units can.

Frankly speaking, I would like to see AT and Locusts removed the same way IT was, and see them replaced with something thats not a free unit, or alternatively push a mineral cost on them the same way Interceptors do.

Because the observer was already hard enough to spot and it was pretty damn quick. When that patch was in, there were multiple instances where players dropped multiple scans centered on the obs to kill the observer and failed to do so - without needing the obs upgrade.

Having said that, the raven changes may mean that this reversion might be okay because it makes the raven significantly more accessible in opening builds with cost and build time reduction.

No. This really isn’t needed imo.


This opens the topic of how much mundane micro do you want to have in the game?

The more such micro there is the more mechanically demanding the game is, but it gives player less time for “strategic” decisions.

For example, what if we give “select all production” button to everybody like there is “select all Gateways” button for Protoss? (if I got a penny for every game I lost because I accidently unbinded all my production…)
What if your current production is displayed in game like it is when watching a replay?

What if we make it possible to use abilities/spells for all units selected in a group (like for Tychus in co-op), removing Tab-Tab-Tab from group micro requirement?

What if we add option on what alerts Space button move camera on? (currently this feature is completely useless)?

What if there is a warning when you are close to supply block but not there yet?

What if…

Control group your units and you’ll be fine.

But it does become a somewhat slipery slope because some of those micro things are part of the challenge of the game and also part of the fun of the game too. So it becomes a question of what is fun what is challenge and what takes away from one or both?


That’s the direction the devs have been pushing towards to since the SH’s redesign : locusts duration was reduced so there would be offtime, infested terrans were removed. Units that could globally induce a resource-less game were either removed or toned down, in order to reduce unnaturally long stalemates games issues.

Yet, they have not been removed entirely :

  • Zerg currently have two free units : the SH’s locust (18s), and the broodlings (which are spawned by broodlords but also by buildings’ destruction) (5.7s).
  • Terran have one : the autoturret (10s)
  • Protoss don’t have, though the interceptors could be regarded as semi-free considering their exceptionally low cost.

Now :

  • None of the two zerg free units can shoot air.
  • The terran free “unit” can shoot air but can’t move, and has to be landed on available ground space.
  • Only the protoss interceptor can both shoot air, ground and move, but it’s not completely free.

So conceptually, there’s a certain logic (the less advanced the species, the more they have free units) and ingame a certain balance (free units are limited compared to non free ones, either in mobility or AA capabilities). So I don’t see the issue there.

Second angle, are there situations where despite that the free units could be problematic ? Let’s cast aside the SH (for whom the prolonged offtime is the current balancing key) to focus on the raven.

  • TvZ : ravens are used for harass and decreep, then sitting ducks once mutas are out, and limiting the starport to one techlab thus produced in small amounts. The short range of the autoturrets’ deployment (3) implies the raven themselves will often take his from the queens when approaching from the mineral lines. And ravens are 100M 200G units, so each hit taken has a repair cost. Is that imbalanced ? Don’t think so.
  • TvP : ravens could be used for turrets, but only occasionally, as their preferred use will be to prevent DT harass from getting out of control (and DTs were buffed with the blink), and to matrix large protoss units during timings (particularly colossus). AAM sees some occasional use against gateballs. Thus autoturrets aren’t a problem in the match-up, and even if raven harass was a thing, batteries can in a way generate free shield points to offset the harass.
  • TvT : that’s the only match-up were mass raven sees some use, with raven being powerful both in regular compositions (anti-tanks matrixes), in metagame openings (cloak banshee & BC vs cyclone raven defenses), and getting to being massed in lategame (yet 99% of the time with regular units to complement them). That because the detection abilities, Matrix, AAM, autoturret all see uses in that MU, thus making the raven a solid choice in a lot of circumstances.

So that brings the question, if an unit is only popular in one match-up, is it really problematic ?

  • It does not impact global match-ups balances (PvT, TvZ, ZvP) as mirrors are by definition not relevant to those. Though nerfing it could deter even from the occasional uses it sees in those.
  • Considering TvT now:
    • does raven has reliable counters in the current metagame ? Yep, air dominance (mainly vikings), and anything AA really (marines/cyclones being the best example) if ravens wander unsupported (harass situations).
    • Does a raven opening fare much better than other openers in the current metagame ? I’m inclined to say yes, as cyclone raven openers can virtually counter anything thrown at it but proxy raxes or eactored marine tank vikings allins.
  • Still considering TvT, does mass raven as an opener has reliable counters ? Well vikings outrange autoturrets (though not autoturrets + ravens), and ravens alone would make easy targets in empty aerial areas. If turrets are massed somewhere, just back for 10s and you’re good to go. Now if you did let the turrets being massed somwhere where you’re forced to defend, you’re in quite the trouble, but hey this strategy is still exceptionally rare, and more diversity is better isn’t it ? Why would we want the match-ups to be played only one way ? Wasn’t isn’t it other’s RTS’s (AoE…) main issue ?

So globally assessing the situation, the only points where current ravens do reduce the variety of the metagame is TvT openers diversity, and to a lesser extent TvT lategame compositions diversity. Could it be improved without hurting other match-ups ?

  • Matrix nerf : will impact TvP. Bad idea, they already came back from it.
  • AAM nerf : would impact TvP as well.
  • Autoturret nerf : probably the least damaging change, as it is the main tool in TvT mass ravens. Would still impact TvZ but marginally.


Personally if I were to reduce the impact of mass ravens in TvT while trying to preserve other match-ups, I would

  • Keep Interference Matrix duration untouched (already sub-par compared to Neural, and useful in TvP). -1 nerf on the range (9->8) up to discussion specifically to decrease the effectiveness against vikings/tanks ; but if retained I would compensate that by making it more useful on aerial targets (longer duration vs air ?) as BCs nor carriers aren’t perfectly balanced either.
  • Keep the AAM untouched (already sub-par compared to fungle, parasitic bomb, and storm. Already the least damaging AoE spell in the entire game, it does not need to be toned down again or it will get unused.
  • Keep the current stats of the autoturret, but reduce the cast range from 2 to 1. This way a raven would really need to get close to a target in order to damage it by himself, thus emphasizing ranged air counters (vikings ++) or anything ground already in placed to defend, unless the raven is supported, or the target immobile, unsupported and without AA.
  • If said changes are incorporated, the raven’s cost should decrease.

I know current balancers think they can offset the impact on TvZ and TvP by lessening the cost, but a 15% discount on a 33% nerf isn’t the way to do it. :mag_right:



Why aren’t the PTR balance changes being posted here or on the News blog? Most people don’t even know about the adjustments and those who are trying to keep track have to rely on Reddit posts.

Changes to the current Warcraft III Reforged PTR are being posted right on the game’s forums.

I do think the Raven should be reset to what it is currently (including the return of the Corvid Reactor) and then re-tune from there, since making its build time faster just so you can get one or two out in the early-game is not the way. Ravens aren’t Oracles so they shouldn’t be treated as such. Rather, make it more useful in other match-ups. I’d start with improving Interference Matrix so it has some situational uses outside of TvT and which isn’t performed better by Vikings or other Starport units. For example, have it so it reduces the target’s sight range to 1 while the effect is active - thereby removing an enemy Viking’s ability to spot for Siege Tanks or a Warp Prism’s ability to see onto the high ground - and add the Psionic tag to Overlords, Dropperlords and Overseers so there are more viable targets against zerg (while also allowing their transport and mobile detector to be disabled like when against terran and protoss).

Mass Raven is a niche build that is very annoying to fight against, but can be countered if you know what you’re doing, and the problem is that because you don’t often face it, players aren’t sure what to do.

  • If you’re terran, Ghosts will nullify Ravens easily, and Liberators are also a very good choice to hunt them down due to their speed and splash damage. For attacking the turtled bases, Cyclones and Marauders are best I find (Marauders can tank the Auto-Turrets and Planetary Fortresses and do a lot of damage to them, and Cyclones with Mag-Field Accelerator will melt them too while staying out of range after lock-on).
  • If you’re zerg, Vipers rule here, whether it’s from casting Parasitic Bomb on the Ravens or Blinding Cloud on the Planetary Fortresses, after which Roach/Hydra can handle the rest. Infestors aren’t bad if you can catch the Ravens with Fungal Growth, but are trickier. Mutalisks are okay too, but watch out for Widow Mines and Hellbats (for counterattacks). Mass Ultralisk can work later, but if scouted, the terran player can switch production into Liberators to defend the bases.
  • The build isn’t too good against protoss because there are so many counters, namely Blink Stalkers, High Templars and Phoenixes, while Immortals and Tempests absolutely wreck the turtled bases (and Immortals melt Auto-Turrets). Colossi and Disruptors can also clear out supporting Widow Mines.

I like your approach to analysis :+1:

You missed a few very important details though

Almost nobody is doing it outside of meme builds because instead of Raven you can build 2 Banshees, or Liberator+Medivac, or Viking+Liberator, or Viking+Medivac - any of which are better at both harass and defense against Zerg Roach push which is very popular these days and is in fact one of the main reasons why 3CC 2 Banshee build is meta.

TvP I guess?)
Normally Raven is built ONLY as a response to incoming DTs because the only thing it is good at is being $h!tty observer.
IM works only against one build in very small timing window and only if Protoss does mistake in micro.

IM force 2 Colossus or 2 Archons to chill for ~10 seconds which may (may!) give your army a chance to kill Protoss 3d base (expansion) if you arrive at exact timing when 2 Colossus or 2 Archons are ALL T3 units that Protoss have, while you already have 2-4 Medivacs, Stim, Shields, +1, several WMs and decent bio ball.

AAM DOESN’T work because all it does is forces Protoss army - that moves faster than Terran one - to chill for ~10 seconds. Oh and considering how slowly AAM projectile moves you are very likely to hit your own army with it and give Protoss more benefit from the spell than you would get.

Ravens effectively immobilize siege tanks making IM infinitely stronger than in TvP.

However Ravens only work until your opponent get ~6 Vikings or begin to turtle up with missile turrets or builds couple of Thors all of which will zone out Ravens very effectively.

Ravens will work against people who try to play like progamers with moving around a lot but most of people on the ladder - and that include Master league - don’t try to show off like that and just turtle, making Raven Po$ in TvT too outside of early game where double Raven opener is indeed good against many - but not all - builds.

Auto-turret spam is some silly meme done by players who troll people 1-2K MMR below them. It doesn’t work against an opponent of somewhat close skill.

1 Like

absolutely no one will be using Cyclones against Hydra :smiley: so seems kind of pointless. If the Cyclone had an attack similar to BC then it would be useful for large groups of enemies, at current… I guess Cyclones are good against single units

Cyclones are supposed to be good against large, single units that are relatively slow with a short attack range. On paper, they should be good against Ultralisks and Roaches as well as transports. Hydralisks not so much because they appear in large numbers and are pretty nimble. But now though they’re kind of a jack of all trades, master of none.

I’m just going through the extension mod again, mainly to update my own Raven Rework suggestion mod. I thought everyone might like to see the patch notes in the 5.0.11 extension mod (also shows an experiment they did with the Liberator that was reverted the very same day!)

1.9 (9th January 2023)

  • (BALANCE) Hydralisk speed upgrade buff reduced from 0.78 → 1.05 to 0.78 → 0.98 off creep
  • (BALANCE) Viper post Abduct delay nerf reduced from 0.71 to 0.57
  • (BALANCE) Disruptor AOE nerf reduced from 1.5 → 1.35 to 1.5 → 1.375
  • (BALANCE) Ghost’s Steady Targeting maximum range increased from 13.5 to 14

1.8 (3rd January 2023)

  • Fixed issues with delayed attacks dealing damage in unusual situations:
    • Fixed an issue where a Zealot’s second hit could damage units that were in stasis, invulnerable, lifted, or very far away. The maximum distance to receive damage is now 2.1.
    • Fixed an issue where a Queen’s second hit could damage units that were in a transport, stasis, invulnerable, lifted, or very far away. The maximum distance to receive her melee attack’s damage is now 5.
    • Fixed an issue where a Reaper’s second hit could damage units that were in a transport, stasis, invulnerable, lifted, or very far away. The maximum distance to receive damage is now 7.
    • Fixed an issue where a Thor’s second hit could damage units that were in a transport, stasis, invulnerable, or lifted … (truncated - not sure what the rest is).
    • Fixed an issue where a Colossus’s attack could damage units that were in stasis, invulnerable, lifted, or very far away. The maximum distance to receive damage is now 9 (+2 with upgrade).
  • (BALANCE) raised Cylcone’s Lock On’s auto-cast range from 7 to 7.5
  • Fixed an issue where Medivacs could heal units affected by stasis
  • Fixed an issue where Shield Batteries could recharge units affected by stasis
  • (BALANCE) Raven build time reduction reduced from 42.9 → 30 to 42.9 → 34.3

1.7 (2nd January 2023)

  • (BALANCE) Liberator
    • Rework removed
    • Cost reduced from 150/150 to 150/125
  • (BALANCE) Raven
    • Interference Matrix change removed
    • Starting Energy change removed
    • Auto Turret Energy cost change removed
    • Auto Turret duration reduced from 10.0 to 7.86
    • Auto Turret health reduced from 150 to 100
    • Auto Turret armor reduced from 1 to 0
    • Auto Turret is no longer affected by Neosteel Armor upgrade
  • (BALANCE) Carrier
    • Interceptor changes to shield and flying radius removed
  • Improved implementation on Lurker unburrow random delay change

1.6 (2nd January 2023)

  • (BALANCE) Creep Tumor respread cooldown further increased from 12.86 to 13.57 seconds
  • (BALANCE) Brood Lord speed increase reduced from 2.3 to 2.24
  • (BALANCE) Viking (Fighter Mode) damage point reduced from 0.12 to 0.04
  • (BALANCE) Liberator:
    • Health reduced from 180 to 150
    • Build time reduced from 42.9 to 32.1
    • Supply reduced from 3 to 2
    • Damage (Defender Mode) reduced from 75 to 50
  • Hydralisk - Morph to Lurker random delay removed and average delay added to build time

1.3 (26th December 2022) - no 1.4 or 1.5 listed

  • (BALANCE) reverted duration of Broodlings created by buildings
  • Shield Battery Restore’s tooltip now uses Blizzard time and shows its range indicator on button hover and in target mode
  • Fixed an issue where changing warp in speed did not change the animation’s speed
  • Fixed Shield Battery’s auto-cast targeting Dark Shrine
  • Fixed an issue where Changelings would permanently switch sides after Neural Parasite. Neural Parasite now disables the morph
  • Fixed burrowed Swarm Hosts and Ravagers colliding with burrowed Movers. Roaches change their collision after researching Tunneling Claws
  • Added Observer PlusPlus extension mod as a dependency to include its features
  • (BALANCE) Forge Level 1 upgrades research time reduced by 7.1 seconds to 121.4
  • (BALANCE) Forge Level 2 upgrades research time reduced by 8.9 seconds to 144.6
  • (BALANCE) Forge Level 3 upgrades research time reduced by 10.7 seconds to 167.9

1.2 (11th December 2022)

  • Archon and Disruptor minimap radius now matches the unit’s actual radius
  • (BALANCE) Ultralisk range slop increase from 1 → 1.4 to 1 → 1.25
  • (BALANCE) Gravitic Boosters move speed bonus increased from 1.31 to 1.41
  • This was missing from the previous version as the upgrade is listed as increasing movespeed by 50%
  • Fixed an issue where certain flying units did not cause water ripples when flying over water.
  • Standardized the height at which an air unit will cause water ripples.
  • Fixed a variety of incorrect Upgrade scores (how much score you get should equal the combined resource cost
  • Fixed an issue with the attack animation of the Thor when using the Tyrador skin
  • Fixed an issue where simultaneously using Mass Recall and Strategic Recall could move the recalled units to an incorrect location
  • (BALANCE) Sentry move speed increased from 3.15 to 3.5
  • (BALANCE) Cyclone lock on now prioritizes air units (who can attack the cyclone)

1.1 (10th December 2022)

  • Fixed a variety of Upgrade scores being incorrect
  • Fixed an issue where alternate ghost skins would not have the new feature to manually cancel snipe
  • Fixed an issue where ravager’s corrosive bile animation would override ravager’s burrow animation
  • Fixed an issue where lurker could damage units loaded into transports

1.0 (8th December 2022)

  • (BALANCE) Hydralisk speed buff now only affects speed off creep
  • First banshee missile is no longer delayed 0.11 seconds after banshee completes its attack
  • Widow Mine no longer targets units affected by Neural Parasite without a manual order (same behaviour as normal unit attacks)
  • Fixed an issue where Thor in High Impact Payload mode would not collide with Locusts
  • Fixed Ultralisk/Observer/Disruptor size changes with skins (observer size +17.5% → +10%)
  • Removed Enhanced Shockwaves icon from Ghost command card
  • Removed some unintentionally included changes
  • Stasis again stops ability orders – Too many bugs when allowing all orders

That’s fairly strong. Cyclones were already capable of locking onto a defensive structure and backing out before it could shoot; now they don’t even need to stray into the attack range.

1 Like

A big one with that is that they don’t have to venture into the range of a Planetary Fortress with Hi-Sec Auto-Tracking. They can lock on and still sit slightly outside of its attack range.