XCOM 2 Alien Analysis: Codex

HP: 10/12/12/13
Armor: 0
Defense: 0
Dodge: 20/25/25/25
Aim: 75/80/75/75
Mobility: 12/12/14/14 (8/16 on Regular, 9/18 on Commander/Legendary)
Damage: 4-5 (+2)
Shred: 0
Crit Chance: 0/10%/10%/10%
Will: 100
Psi: 100
Tech: 150

Yeah, the Codex has more Aim on Regular. Probably another oversight, like they nerfed Codex Aim and didn't catch that they hadn't applied it to Regular.

Susceptible to Bluescreen Rounds and EMP Grenades/Bombs, but is not actually a robot otherwise, and can in fact be psionically assaulted. Primary weapon has unlimited ammo.

The susceptibility to Bluescreen Rounds and EMP Grenades/Bombs without being a robot is, surprisingly, not unique to Codices, and is why I tend to describe Bluescreen Round and EMP Grenades as effecting 'digital' enemies instead of robotic enemies: because it's not just robots, and I don't have a better way of summarizing the rather odd set of non-robots that are susceptible to them. Though more on that as we get to these other non-robots.

And it is specifically those two (And a half) Item types. Codices don't take bonus damage from Combat Protocol or Capacitor Discharge, nor from the Skirmisher's Whiplash ability. They similarly aren't treated as robots for miscellaneous other purposes, like how one possible Hacking reward is to get control over a random robot enemy: this will never give you control of a Codex.

This whole dynamic could've been communicated better...

Immune to all damage over time effects. Also a partial immunity to Stun.

By 'partial' immunity to Stun, I mean a Codex will reduce any amount of Stun to a duration of 0 turns. On most enemies, that would be very nearly the same thing as actual immunity to Stun, as it means a Codex will un-Stun on its turn and have both action points. Normally, the only time such a partial immunity would be distinct from true immunity would be when an activating pod member gets hit by Overwatch fire from the Bolt Caster, since a Stun triggering would interrupt their attempt to dash for Cover. In the Codex's case, it matters surprisingly strongly, because Stunning the Codex will prevent Clone from triggering for the rest of your turn. As such, you can't buy a turn by Stunning a Codex, but you can at least avoid your attempts to kill it making matters worse.

Or, well, you could if there was any way to 100% reliably inflict Stun. (Aside Stasis, which ignores Stun immunity and doesn't let you attack the target, so it doesn't really count) Still, you might as well line up Stun-capable options on a Codex in hopes of getting lucky.

Codex immunity to damage-over-time effects, meanwhile, is honestly a mercy. No multiple triggers at the beginning of the enemy turn on Clone, after all.

The Codex transports itself to a new location an enemy can currently see, bypassing intervening terrain and not triggering Overwatch or other reaction fire. The Codex will always have 1 action point after teleportation, no matter how many it had beforehand. This has a global cooldown of 1 turn, meaning only one Codex will teleport in a given turn.

Interestingly, the SoldierSkills config file indicates that at some point in time the Codex Teleport would've had a small area-of-effect damage effect attached to one or both ends of it. Particularly interesting is the file listing environmental damage, which makes me lean toward suspecting the Codex would've done damage in its wake, since having it do damage on arrival or at both ends would mean risking blowing up its newly-taken Cover. I suppose it's possible at some point it didn't use Cover, but then a teleportation ability seems a lot less useful so that seems a bit unlikely.

Also, I've commented on this before, but Teleport's mechanics mean that Codices are semi-immune to Tactical Analysis. After all, if there's one Codex, and only one, it will use its one action point to Teleport, then magically still have an action point, same as if you didn't have Tactical Analysis. I'm pretty sure this is an oversight, though I'm genuinely curious as to why somebody decided Teleport should arbitrarily leave the Codex with one action point instead of actually spending one action point. Were the mechanics different at some point in development such that this made sense, and then when the original reasons changed nobody noticed the issue because prior to Tactical Analysis AI behavior meant it worked out to basically the same thing regardless?

In AI hands Teleport doesn't have a local cooldown; it has a global cooldown, which is to say that after a Codex teleports none of the other Codices running around are allowed to teleport. This is important, as Codices will obsessively teleport if the option is available, preferring to flank your entire squad if at all possible (They don't need personal line of sight on their destination, so they'll basically always pull this off) before doing anything else.

Conversely, if you Dominate a Codex, the cooldown on Teleport is local. It also instead demands your squad has line of sight on the location, though otherwise basically anything goes: there's no range limit or the like. Though you should be cautious about going for super-flank Teleports, since you're risking activating a pod with the Teleport, same as with Icarus Armor. The weird action point mechanics also remain true, which is mostly just sort of whatever but does mean you can have a Dominated Codex move to trigger Overwatch fire, then Teleport elsewhere and attack or use Psuonic Bomb just fine.

Teleport itself is also notable for completely bypassing reaction fire, even reaction fire backed by Covering Fire, including that a Codex is able to Teleport out of reach of Bladestorm/Retribution without triggering an attack. A lone Codex is thus essentially immune to Overwatch fire post-activation unless you Disorient it or something due to their obsession with Teleporting for a flank. So, uh, don't try to fish for Guardian triggers or whatever.

Teleport also lacks any kind of range limit. It's easy to think it has a range limit due to the AI always Teleporting to somewhere at least one of your non-Concealed squad members can see, but nah, if a Codex is active on literally the other end of the map it can appear next to your squad instantly. The only qualifier here is that the AI does make some effort to pretend its knowledge of soldier position is actually derived from enemy line of sight. So for example if you active a pod, kill everything except the Codex, and then arrange for your entire squad to fall back out of its sight, the Codex won't Teleport to a flanking position; it will instead run toward your squad's general location, and once it has line of sight go for a shot or Psionic Bomb or go into Overwatch.

This itself touches on the fact that the AI doesn't normally abuse the weird action point behavior of Teleport. They'll neuter Tactical Analysis' effect, but in situations where a Codex walks from out of line of sight and finds itself with one action point remaining, even though it could Teleport and act, it won't, at least not normally. (I wouldn't be surprised if there's some edge case, but I haven't run into it yet myself)
Overall, Teleport is slightly less relevant and threatening than you might intuitively expect, though, as Codices are obsessed with using Psionic Bomb if possible and it doesn't particularly benefit from Teleport. So if you're reliable about ensuring Codices are never present two turns in a row, the potential for Teleport-based flanking shots and whatnot isn't actually a concern. Of course, if you do let a Codex get a second turn, they're alarmingly likely to pull off a flanking shot... so try to not let them get that second turn. Killing, Disorienting, or disabling a Codex are basically the only ways to prevent a flanking shot on that second turn.

When the Codex takes damage, it immediately teleports itself away to a new location, leaving a clone of itself behind at its old location. Its HP before Clone triggered is split between these two Codices. If the Codex has only one HP left when it attempts to Clone, it will simply teleport itself instead.

The HP split attempts to be even, but if the Codex's HP is an odd number then the teleporting Codex will have the extra hit point. The distinction of which Codex is the original is mostly not very important since Codices are immune to many status effects and several exceptions will prevent them from cloning themselves (eg Disorientation, Freeze, Stun), but there are a handful of effects -such as Holo Targeting- that don't block Clone and work on Codices where it's useful to know ahead of time which Codex will be suffering from the effect after Clone triggers.

It's also important for two meta reasons, both of which are a lot more relevant to War of the Chosen. The first of these points is that the teleporting Codex being the original means its active state is tied to its pod. Thus, if a Codex becomes damaged in the fog without activating its pod, but then teleports to within sight of your squad... its pod will promptly activate, no matter how far away the other members are.

In the base game, it's pretty difficult for this to happen, but there are a few edge cases, such as stray weapons fire causing a car far in the fog to start cooking off, and coincidentally explode on an inactive Codex. It's also more likely than you might intuitively expect, in that the reactive teleport aspect of Clone endeavors to drop the original Codex in a nearby flanking position... without any regard to the Codex's known information, or even the entire red team's info. As such, generally if an inactive Codex ends up injured but not dead without being activated by the injury, its pod will end up activating anyway by virtue of the Clone-port landing the original Codex in sight of your squad.

The other way it reliably matters which Codex is the original is exclusive to War of the Chosen, and more specifically to the new form of Retaliation mission: the first group of survivors is rescued, and their guards triggered to run on ahead, when all members of the two pods assigned to attack them are dead... all original members. If a Codex ends up cloned, and the clone ends up the only survivor, the Resistance forces will clear out anyway.

Incidentally, the first scenario also crops up very readily in these War of the Chosen-added Retaliation missions due to the Resistance fighters being able to attack enemies in the fog without activating them. If Codices are part of the second set of attackers and the game does its clearly-not-intended weird thing of the second set of Resistance fighters taking shots while you're still dealing with the first set, this can go very badly, very quickly. Thankfully Codices are uncommon in Retaliation missions in general, but yikes is this one of the more unfortunate oversights in War of the Chosen.

Less common but also new to War of the Chosen is the possibility of Lost activating, then ending up attacking an inactive Codex in the fog. Normally active Lost attacking enemies you can't see won't activate them, but in this case the Codex will probably activate as a result of the teleportation. If you're eg in a Supply Extraction mission and perpetually lurking in squad Concealment, waiting for your enemies to fight each other to death, this is distressingly likely to result in squad Concealment breaking and very possibly the Codex getting a free shot in, since it tries to teleport to flanking positions. So, uh, don't do that, particularly if you know Codices are present in the mission. It's technically optimal by certain metrics to wait for them to kill each other, but this kind of risk torpedoes how useful a plan it is. (In addition to it being very boring to just mash the end turn button until one side is dead)

Returning to less edge case-y points, the newly-produced clone Codex is, it should be noted, always immediately activated, even if your squad can't see it. (eg you blind-fired a grenade over a wall) Said activation doesn't involve a cinematic or anything, including there's no free scramble move, but it's ready to go right away.

In the base game, this is actively one of the more dumb, RNG-screw-no-playing-around-it bits of the design, as said freshly-generated Codex will get a full immediate turn if you caught the original Codex with Overwatch fire during the enemy turn. There's really nothing you can do to play better regarding this: trying to catch enemies with Overwatch fire is broadly good, it's just in this one case it can backfire. Before you get the Shadow Chamber online, you can't even make a policy of just avoiding going into Overwatch on missions with Codices predicted! (Which would be hamstringing yourself otherwise, mind)

This is particularly egregious if you have Alien Hunters, since one of the main ways to make the Alien Rulers more manageable is to catch them with Overwatch fire, and they're so threatening it just makes no sense to avoid Overwatch because it might backfire when Codices get involved. It's so bad a problem I prefer to actively put off triggering Codex spawns in the base game, though there's other reasons for this preference I'll get to later in this post.

Thankfully, War of the Chosen fixes this twice over: Clone triggering during the enemy turn does not allow the new Codex to instantly act, and furthermore in the base game Clone triggers instantly in response to Overwatch fire, ensuring only one shot can hit them at all, when it's borderline impossible for individual Overwatch shots to one-shot a fresh Codex on higher difficulties. (Really, even on Rookie it's not very practical...) In War of the Chosen, an inactive Codex that gets caught by Overwatch fire during its scramble move will trigger Clone only once it has finished moving. Assuming it survived, which is a lot less likely than in the base game.

Also note that Clone triggering doesn't affect the global cooldown on Teleport, even though it's conceptually and mechanically more or less the same outside the piece of splitting HP with a new Codex. Conversely, Clone-triggered teleportation must place the new Codex at least 6 tiles away from its original location, and at most 18 tiles away, where manual Teleports have no such restrictions. (Though their preferences end up pretty similar in practice, admittedly) This helps limit the danger of the new Retaliation mission situation, incidentally: you have to be moderately close for a Codex being shot by the second set of Resistance fighters to lead to a pod activation. It still happens, mind, but it at least means you generally won't have a pod activating 3+ turns away from the squad, rampaging through civilians.

So, all that fiddly mechanical stuff out of the way...

Clone is unique to Codices, and one of the game's more experimental, weird abilities, with an unusual, interesting set of implications. First and foremost it means the Codex is an enemy you should only put attention on if you're willing to commit: normally it's fine to eg kill two members of a pod and then have your final soldier take a potshot at the surviving member of the pod even though it can't kill. Against a Codex, the shot hitting -or missing, but with a Stock attached- is actively making the situation worse if you can't follow up and kill at least one of the two resulting Codices.

The second interesting aspect of Clone is how it interacts with the Codex having a good Dodge stat. On most enemies, Dodge is an uninteresting stat; if it triggers, yeah, they live longer, but it's only different from eg increasing their HP in that it's RNG-based, not necessarily applying to any given case. On a Codex, though, it changes the risks of taking a risky shot: against a Stun Lancer, taking a shot that will kill if it hits properly has 'Graze' and 'miss' as basically the same result if eg you're firing with your last soldier in the turn. Either way, the Stun Lancer is still alive, and the distinction between a Graze and a miss is only significant if it's backed by Dragon Rounds or Venom Rounds.

For a Codex, though, if that's your last shot and all a Graze is actively worse than a miss, adding an extra Codex to cause you trouble. That's actively punishing you for taking the shot, where against other enemies a Graze is only worse than not taking the shot in that you spent ammo, a largely-unimportant, renewable resource. As such, against most enemies with a Dodge stat, inaccurate shots aren't a reason to not take the shot, they're reasons to consider aiming at other enemies that lack Dodge, at most. Whereas the Codex Dodge stat actively discourages taking risky shots you can't follow up on, full stop. This is completely unique to Codices until arguably War of the Chosen, and is still mostly a Codex thing even in War of the Chosen.

It's too bad the design space doesn't really build on this idea more. The Bolt Caster from Alien Hunters and the Skirmisher's Justice and Wrath special actions are the only examples of being able to make an inaccurate shot with zero chance for a Graze outside the specific point that firing from Concealment ignores Dodge. And honestly I suspect in all three cases it's something of a happy accident that these specifically have anti-Codex potential, where if a full hit is going to be a kill but you're not guaranteed to hit then the shot is riskless. (As far as Clone goes, anyway)

Still, it's the one case of XCOM 2 really trying to do anything with the Dodge stat, and it works decently. At worst I could complain about it being unintuitive and opaque, but in the context of what the narrative is going for with Codices I don't necessarily consider that an actual flaw, and the opacity is less about Codices in particular and more about the fact that the game essentially hides the Dodge mechanic in a more general sense. If Dodge's existence and mechanics was more clearly communicated to the player, these Codex-specific issues would go away.

The third implication of Clone is that it heavily encourages leading with strong hits, rather than weak hits. This is backed by the strange implications of Dodge interacting with low odd numbers I went over with Vipers, but even without that it's just a bad decision to eg Quickdraw at a full health Codex twice and then start trying to mop up the three resulting Codices with Cannon fire. Better to open with a Cannon shot and then, if it survives, mop up the resulting frail Codices with weak Pistol shots.

This is particularly pertinent in War of the Chosen, where all three Resistance classes have fairly poor per-hit damage, at least at base, and make up for it in other ways. You'd rather have a Skirmisher mop up after regular soldiers do damage. Similarly, you'd rather have a Templar waiting in reserve to finish off one of the Codices after Clone has been triggered by another soldier, rather than having the Templar open with a Rend that can't possibly kill the Codex, taking advantage of how Rend can't miss. And of course a Silent Killer Reaper should be assigned clean-up duty, instead of performing the opening shot.

Though on that topic, it's worth noting that multi-hit abilities complete their fire chain before Clone has a chance to trigger. Banish won't fire, trigger a Clone, fire again, trigger a new Clone, etc. Rapid Fire fires twice, then Clone triggers. And so on.

There's one other interesting aspect on a design level to Clone, but that really requires discussing the Codex's final ability...

Psionic Bomb
Area-of-effect attack that immediately attempts to empty ammo from the weapons of all soldiers caught in its blast radius. This is a Will test, and so higher-Will soldiers are more likely to have their weapon unaffected by the Psionic Bomb. The turn after Psionic Bomb was first used, it will hit all units in the same radius for 4-7 damage, ignoring Armor, even if the Codex that initiated it is dead. Psionic Bomb has a global cooldown, and so only one Codex may use it per turn.

For the record, the +1 chance on Psionic Bomb's damage is the rather unusual value of 62%, making it slightly more lethal on average than you might expect from the obvious numbers.

Psionic Bomb is interesting, in that it's the most blatant and by some metrics earliest example of a design concept I've alluded to but not explained: complicator enemies.

In an immediate sense, a Codex is Not A Threat. 99% of the time, a lone Codex will Teleport to a flanking position, and then instead of taking advantage of its clean line of fire to try to kill someone, it will toss down a Psionic Bomb at a clump of your units, which for the moment just... eats ammo. That's inconvenient, but not really problematic on its own.

But then Psionic Bomb will do damage in its area at the start of the enemy turn, so you can't just reload and take shots. Your soldiers need to move, and they need to move now, or else they're going to take a pretty big chunk of damage. 4-7 per soldier is kind of bonkers when you realize a Codex can catch literally your entire squad with the Psionic Bomb if you've mildly clumped them, and also consider that Codices can show up very early in the game if you want. (More on that later, though)

Still, by itself, against a lone Codex, this isn't really a problem. At worst it just delays the fighting a turn, most times. And given there are multiple ways to dish out damage even when the squad has no ammo, it's easily possible for it to be actively a waste of the Codex's turn, where you shrug and have your Rangers Slash and your Sharpshooters duck out and fire their Pistols and so on.

But a Codex operating alongside other enemies is a lot more of a problem. If half your squad is scrambling for safety and reloading their weapons instead of taking shots or unable to eg fire Bluescreen Rounds-backed shots at deadly digital enemies, or otherwise you're being prevented from taking the actions you want to take to wipe out key enemies, that gives other enemies more time to act instead of being dead before they can do anything. The more other enemies there are, the bigger a force multiplier the Codex stalling your squad with a Psionic Bomb is, too! Whereas if the Codex did just take a shot at a soldier while being backed by three other enemies, even if the Codex killed the soldier outright... that would be less of a problem for you in an immediate sense.

This further intersects with Codices having Clone, where you have to think hard about whether it's even worth trying to attack the Codex at all. If you might end up with more Codices than you started... probably better to focus on other threats, honestly. This actively encourages giving the Codex a turn to toss down a Psionic Bomb, by virtue of focusing on other threats.

Thus: the Codex is a 'complicator' enemy. It doesn't, in an immediate sense, try to kill your squad dead, the way a Basic Trooper just takes shots turn in and turn out. Instead, its preferred course of action is one that isn't immediately threatening, but which changes the shape of the battle where options are being closed off and certain kinds of decisions are being forced, and depending on how your squad is set up and what other enemies you're fighting and so on, that can create a lot worse of a situation than just being shot at.

This is one of the big conceptual improvements to XCOM 2's design! The prior game's design suffered from the game not knowing how to create problematic situations other than 'enemies try to kill you', which limited the array of possible situations to resolve the game could produce. One of the consequences of this was that enemies tended to blur together in practical terms, where you cared more about overall 'tier' of threats (Heavy Floater is more dangerous than regular Floater) than you did about kind of threat. (Mutons and Floaters were depressingly similar, mostly distinguished by Mutons being the bigger threat) It was also a big part of why the game was so pressured toward 'kill enemies before they can act' as the optimal player course of action, because by default enemies didn't maneuver for advantages you could then maneuver to take away from them or the like, and the only way to prevent the enemy from rolling dice to try to kill you (Or using dice-ignoring ways of killing you) was to make sure they never got the chance to act in the first place.

Complicator enemies in XCOM 2 instead present situations that require different ways of approaching solving the situation, and the Codex in particular is very heavy-handed about communicating to a learning player that 'just fire at big threats' isn't always the best answer. A pod with a Codex and two other enemies, where your overall firepower can't reliably kill all three of them, is a pod where you should probably kill the Codex's escorts and not touch the Codex that first turn, or fight it carefully. (eg if you take out its two podmates and end up with two active soldiers still ready to go, it's acceptable to shoot it once if you have a guaranteed means to finish off one of the resulting Codices via the remaining soldier) And also you should be trying to spread out the squad in response to Codex presence, to reduce the impact of the Psionic Bomb, making that form of maneuvering important in a manner other than trying to reduce the odds of enemy shots landing.

This is great stuff, and indeed connects directly to two of my criticisms-that-implicitly-praise-the-game: the first being that XCOM 2 doesn't take it far enough, with the base game in particular having only a few such complicator enemies and not having them appear consistently enough, and the other being that this ends up whoops running contrary to the other smart overall design decision of the game of embracing the 'solve game by killing everything before it acts' aspect of the prior game. This is part of why, back in my War of the Chosen Specialist post, I commented I'd be potentially okay with XCOM 3 embracing a more give-and-take approach to combat: with complicator enemies existing, designing the game where you're expected to take a bit of back-and-forthing to kill an enemy or lose one of your own soldiers is a much more interesting and nuanced prospect. (Whereas such design is just slow and dull if combat is blandly 'enemies roll dice to try to kill you, with minimal maneuvering, and you do the same to them'. Might as well make it fast and lethal at that point)

Notably, War of the Chosen is biased heavily toward complicator enemies for its additions. Lost are not designed to be a threat in their own right: they're designed to put pressure on you with ammo-and-action-sinks, where the real threat is going to be the ADVENT/Alien forces the Lost are distracting you from, or the mission timer itself. Priests heavily prefer to create problems through actions other than trying to damage your soldiers. Purifiers try to do damage, but it's oriented toward delayed damage, and with the side effect of shutting off abilities and pressuring you to sacrifice a turn somehow or another. (Medikit use, Hunker Down to clear the fire, whatever) And we're not to it yet, but the final regular enemy added by War of the Chosen follows largely in these footsteps.

Even the Chosen fall under this banner, albeit much less obviously and completely so, though again I'll be delving into them in detail... a lot later.

Point being, War of the Chosen makes me suspect the devs are thinking more towards using complicator enemies as a foundation of the design space, instead of remaining committed to the 'kill everything before it acts' approach base XCOM 2 ended up more defined by. This is further supported by how the non-Lost additions are all durable enough that simply removing them from consideration via damage is difficult, with Priests outright having Sustain to make it extremely difficult to reliably prevent them from acting.

That would be very much a good thing, and I'm genuinely interested in what XCOM 3 seems liable to look like.

So all that big-picture design stuff out of the way... Psionic Bomb itself.

I've already touched on this a bit, but Codices heavily prioritize using Psionic Bomb. More precisely, the first active Codex in a turn will basically always Teleport and then use Psionic Bomb on your squad. If they don't do that, it's probably because they're Disoriented or something. I've... seen a Codex just take a shot a tiny handful of times, but it's stupendously rare.

Later Codices in a turn will not do this due to Teleport and Psionic Bomb both operating on global cooldowns. Instead they'll do the usual thing of trying to move for a flank and take a shot if possible, though in practice this comes back to the complicator enemy design point in an... unfortunate way, because actually enemies tend to react to a Psionic Bomb being put down by going into Overwatch.

In theory this is an intelligible design decision that feeds into the complicator enemy point, and I want to like it as an attempt, but in practice it's hamstrung by Overwatch design still not making much sense. Enemies going into Overwatch in hopes of catching your soldiers as they flee from the Psionic Bomb is actively counterproductive: it has lower accuracy than just taking a shot, it can't crit, and it gives the player control over who risks being shot. It's only possible for it to work out as an okay decision if a player unit happens to be out of sight of the Overwatcher, such that they'll have moved out into the open before being shot at.

Even there, the math doesn't add up, even aside the crit point. Overwatch shots are at 70% of the base accuracy of the shot, and the vast majority of enemies in XCOM 2 have 65-75 base Aim, with no range modifiers or the like to complicate this discussion. (With Codices in particular being 75 Aim, keep in mind) 65 multiplied by 0.7 comes out to 45.5: if I assume the game rounds up (I'm pretty sure it rounds down, but work with me), that ends up with 46% accuracy, exactly 1 point better than a non-Overwatch shot on a target in Low Cover. 75 Aim being higher makes things worse: it ends up 52.5, so a loss of 22-23 accuracy depending on rounding, ie guaranteed worse than non-Overwatch on a target in Low Cover. This is made even worse by the Aiming Angles effect being present, as a shot on a unit in Low Cover is only rarely going to be suffering the full -20 to accuracy.

So this only works out if firing on the open via Overwatch is happening in place of non-Overwatch on a target in High Cover, where eg 65 Aim is talking the non-Overwatch shot is 25% accuracy vs the Overwatch-on-in-the-open shot is 45.5% accuracy.

Notably, the AI tries really hard to always have its units end their turn with visibility on your own units, bar a tiny handful of specific exceptions I'll cover as we get to them. (Or that I've covered but have obvious reasons for technically being an exception, like Turrets being unable to move) This makes it extremely difficult for them to ever be in a position for Overwatch fire to make sense.

This whole aspect of Psionic Bomb would only hold up if Overwatch fire was designed to be better than regular fire by default, such as always ignoring 100% of Cover bonuses and having no accuracy penalty and having the same crit odds as a non-Overwatch shot.

... anyway, back to Psionic Bomb.

One mildly surprising aspect of Psionic Bomb to keep in mind is that it doesn't actually do terrain damage, and by extension can't damage assorted objects. (ie it won't get a car ready to explode) This is one reason why it's often okay to leave a Codex alive for a turn on the idea it will go for the Psionic Bomb: you don't have to worry about it wrecking Cover you expect to use soon, or detonate explosives you were hoping to use as Cover for the moment, or vaporizing supplies in Supply Raids.

Also, to be explicit: Psionic Bomb will disable Overwatch on affected soldiers. And both it and Teleport ignore even Covering Fire-backed Overwatch. So if for some reason you were hoping Overwatch was going to help somehow, a Codex being present will often completely ruin that plan. This is particularly important in War of the Chosen, where Clone's trigger behavior in relation to Overwatch was altered: you might be tempted, knowing about that change, to eg put your last soldier on Overwatch on the idea that they'll get safe damage on the Codex. Nope, sorry.

Note that Codices have both a global cooldown and a local cooldown on Psionc Bomb; the global cooldown is 1 turn, while the local cooldown is 2 turns. As such, if you've got multiple Codices running around and are worried about Psi Bomb, it can be worthwhile to target whichever Codices didn't use it. Or if you're more worried about being shot, focus on the one that used Psionic Bomb last.

Also note that Psionic Bomb is absolutely capable of friendly fire on both stages: enemies caught in its radius initially will end up losing their ammo, and enemies that are still in its radius on the following turn will end up taking damage. This latter point is especially useful to keep in mind: instead of trying to finish off a nearly-dead melee enemy that ran inside the Psionic Bomb's radius, just let it die to friendly fire and focus your efforts on other problems. By a similar token, don't be careless when tossing out Psionic Bomb yourself if you Dominate a Codex.

As for the ammo-wiping point itself, while it is a Will test and it is possible to pass it, you might as well always assume all your soldiers will fail it, particularly in War of the Chosen. As yet I haven't dug up the actual formula, but my personal guess is that it's the game's usual Psi+50-target Will formula, or possibly a larger bonus than +50: in War of the Chosen, where it's a struggle to have any soldiers particularly above 50-ish Will, I basically never see a soldier get the 'weapon unaffected!' result, and in base XCOM 2 even my endgame soldiers get their ammo wiped more often than not when targeted. Either way, treat it as a pleasant surprise if any soldiers pass the Will test, but never try to plan around it being a reliable thing unless you've, I dunno, deliberately used Covert Op grinding to get a soldier to 200 Will just to prove you could?

That said, even though Templar will get the usual announcement about ammo being drained, Psionic Bomb's first stage doesn't actually do anything to them, since they don't actually use ammo in the first place. By the same token, any enemy that ignores ammo -like Turrets and dedicated melee enemies- doesn't care about the first stage of the effect, though this is primarily something to keep in mind when Dominating a Codex. (Or playing multiplayer)

And as I've noted elsewhere, there's teambuild stuff you can do to minimize the impact. Auto-Loaders can let you reload without using an action point, allowing you to then move and shoot as normal. Run And Gun can allow you to perform a normal reload and still move and then shoot. Rangers can Slash, potentially even reload and then Slash if something is close enough. (Make sure to check their blue movement zone first) Sharpshooters retain full use of their Pistol even when their Sniper Rifle is out of ammo, including of course all Pistol specials. Specialists with offensive skills can move and use those. Grenadiers can fall back on their grenades after moving. Psi Operatives generally laugh at the idea of their gun mattering. Skirmishers and soldiers in light armor can Grapple out of the blast zone instead of manually moving. Soldiers in heavy armor can move and then use their Heavy Weapon. Soldiers with Implacable can potentially reload, take a shot, and then use Implacable's free move to escape the radius, assuming you can count on the kill. (Maybe the soldier has a Stock and there's a nearly-dead enemy, as one easy example. Hail of Bullets also works) SPARKs have multiple abilities they can use without ammo, with Strike even letting them pull off the Ranger trick of reloading and attack-moving out of the blast zone, and of course have a guaranteed Heavy Weapon. SPARKs can also, in a dire emergency, use Overdrive to let them reload, fire, and then move -maybe there's a Heavy Mec on Overwatch you want shot before anyone moves. (This is a thing that can realistically happen, not a weird edge-case hypothetical, to be clear)

There's really so many ways to minimize Psionic Bomb's impact that a decently-leveled, diverse team has surprisingly little to fear from it so long as you have a reasonably strong understanding of the game and don't get caught on the idea of needing to move and then reload. I wouldn't go so far as to call Psionic Bomb a newbie trap, but it is the case that a learning player, or a semi-experienced player who is slow to re-examine assumptions, are the main cases where Psionic Bomb tends to be consistently a problem.

Mind, one qualifier I should attach to much of this is that a Codex getting off a Psionic Bomb goes from 'really not that bad if you know what you're doing' to 'hell' if an Alien Ruler gets involved in the situation. The Alien Rulers actively punishing movement actions while Psionic Bomb punishes holding still with severe damage is a miserable, awful combination with all good solutions being based on having key tools to cut the Gordian knot, like having a Psi Operative use Stasis on the Alien Ruler before continuing with the turn. If you don't have the right tools, and there's no qualifiers like the Alien Ruler being nearly dead so you can kill it before it does anything problematic... that is a miserable situation, and the potential for it to happen is one of several reasons why I think the Alien Rulers are an ill-designed mess in the base game.

Though that's for other posts.

Interestingly, the SoldierSkills config file indicates that Codices at some point had a 'Malfunction' ability, which seems likely to have been something along the lines of a projected wave of weapon-disabling given the info in the config file. I suspect Psionic Bomb originally didn't have the weapon-disabling functionality, and the two abilities got rolled together at some point in development. Possibly it was originally their primary attack?

The Codex is a weird unit all-around.

In addition to all the other weirdness I've already covered, Codices are also unique in that normally your first encounter with one will be in the form of a Codex spawning in response to you Skulljacking an ADVENT Officer. (Edge-case mechanics point: this initial Codex is not guaranteed to immediately activate. It always runs for Cover immediately, but if it spawns somewhere out of sight and its initial move doesn't put it in sight of your squad, it's actually an inactive pod!) If you then successfully defeat that Codex (Kill all its copies), this will then add Codices to the standard spawn tables, which is kind of plot-weird but I'll get to that later. Notably, you have to do this plot-gated routine for Codices to enter normal rotation: if you put off Skulljacking an Officer until the endgame, you still won't see Codices until you actually do the Skulljacking. (At least, as far as I've seen: I wouldn't be surprised if there's some edge case the devs missed that I've just never seen no matter how long I put off spawning a Codex)

Note that the plot-mandated Officer Skulljack is the only time a Codex will spawn in response to using the Skulljack. Skullmining random ADVENT troops doesn't merit further spawns (Even if you Skullmine an ADVENT Officer), and you're forbidden from using the regular Skulljack ability once you've enabled regular Codex spawning.

I've already covered the majority of Codex strategy, but it's worth pointing out that they're surprisingly durable and dangerous enemies for how early you can encounter them. They're something of a miniboss for the early game, yes, but they immediately enter rotation as pod leaders, and are all-around good enough that even into the endgame where you can see stuff like a pod of just Codices they remain a relatively problematic enemy due to sheer ability to absorb punishment and ability to get flanks. Their rifle's damage doesn't impress compared to eg an Elite Trooper, but that's small consolation if you screw up and end up turning three Codices into six and five of them take shots while the sixth put down a Psionic Bomb.

Note that if you Mind Control a Codex, Clone will simply refuse to trigger on your friendly Codex. This is probably for the best, both because the cloning behavior raises all kinds of problematic questions -which team gets ownership of the clones, for example- and because the reactive teleport aspect would be all kinds of problematic in player hands. If it was outside player control, then the player would potentially be getting punished for having Mind Controlled a Codex with pod activations occurring through no fault of their own. If it was under player control, that would require setting things up so the game can pause the enemy turn to let you reactively pick a teleport point, which would be rather a lot of effort to cover what's really an edge-case scenario.

In fact, in general Schism-backed Insanity is surprisingly decent against Codices if we ignore their solid Will stat tanking its actual odds. If it succeeds, they don't Clone, because they're Disoriented, Panicked, or Mind Controlled, and those all suppress Clone triggering in response to Schism. That solid Will stat does make it a bit niche of info, though...

Regardless, Codices are actually one of the better Domination targets about, so long as you're not careless with their Teleport. The AI's flat Aim effects mean it's virtually impossible for the majority of enemies to reliably bypass Dodge, making a Dominated Codex surprisingly prone to surviving for an extended period even if you have it stand in the open to draw fire, their access to Teleport makes it trivial to get to high ground flanking positions to achieve 95% accuracy against many enemies, and the 85% you'll end up on when firing against the somewhat common value of 10 Defense will still bounce right up to 100% if you Holo Target first, the AI has no ability to exploit their weaknesses (There is no enemy with Bluescreen Rounds or EMP Grenades/Bombs), and while their shooting damage is a bit low by the time you'll typically have access to Dominate, Psi Bomb is a very useful tool in player hands, if used on the right targets with decent planning. (ie don't throw it down where you're intending to charge a Ranger to) Dominating a Codex also has the benefit of completely bypassing all the hazards of Clone, which can be a huge benefit if your team is not equipped to reliably kill the Codex in a single shot. The primary caveat to all this is that the Codex's 100 Will means you're never going to be able to count on a given Dominate going off.

Really, I'd argue Codices are flat-out one of the best Domination targets in the game. The main other enemy I consider a legitimately great general-purpose Domination target has even more Will, and has a very valuable, rare corpse -one of the easily-overlooked problems with Domination is that if you win a mission with a unit still Dominated, you don't get its corpse. Codices technically drop a Codex Brain object, but if you're not messing around with mods the only one that actually matters is that first one generated by Skulljacking an ADVENT Officer. You can't sell Codex Brains, and they aren't spent on anything: thus, winning a mission with a Codex still Dominated costs you nothing except a little experience. Of course, this doesn't matter as an advantage on missions you don't loot bodies on -such as Avatar Project Facilities- but for missions like Retaliation missions and Guerrilla Ops, it's absolutely a relevant advantage, removing one of the annoying headaches with trying to fully leverage Domination.

As a plot-critical enemy, the Codex doesn't have a standard Autopsy, including that you don't get any technology unlocks out of completing its Shadow Project, and the Codex Brain object has the unusual quality of being collected even if you evac from the mission, no need to have anyone explicitly pick it up or anything. Instead, your reward is access to the Psi Gate mission, plus a new Shadow Project that will unlock the ability to Skulljack a Codex for even more plot progress.

Killing a Codex will also unlock the Alien Encryption research if you haven't already unlocked it by hitting the Blacksite. How relevant this is depends in part on whether you're playing the base game or War of the Chosen: in the base game, I'd generally argue you should hit the Blacksite early and put off Skulljacking an Officer, as Codices are nasty enough to be a real concern all the way into the endgame, and they enter regular rotation the instant you've killed your Skulljack-derived Codex, whereas the Blacksite is a fairly easy mission, with the only particular challenge being a Mec that's guaranteed to appear even if you're too early for them to be in normal rotation.

In War of the Chosen, I tend to feel it flips around: it's better to Skulljack an Officer early and hit the Blacksite later. There are two-and-a-half primary reasons for this.

The first reason is the Chosen themselves. The Blacksite will always have a Chosen spawn in to defend it unless you deliberately put off hitting it until you've killed the Chosen that runs the Blacksite territory, and while it's an easy mission in the base game, it's also long and unusually high on enemies, with terrain that makes any of the Chosen potentially very problematic to fight. What's normally an easy, predictable series of weak fights punctuated by a sneak peak at a slightly later enemy can turn into a squad-wiping disaster... potentially even if you drive off the Chosen.

The second reason is the Fatigue system. In the base game, there's no reason to not take a mission the instant its available, outside the possibility of not feeling your squad is ready to take the mission on and needs better gear or more experience first. In War of the Chosen, hitting the Blacksite with your A-team can lead to the next mission or two going horribly wrong because you had to fall back on Squaddies -or Rookies, worse yet- or deploy one or more of your elites while they were Tired. In the latter case, even if nothing goes wrong that mission, you can end up with a promising soldier marred for quite some time by one of the more problematic Phobias, either visibly causing trouble down the line when the Phobia triggers or semi-invisibly creating problems by making you more reluctant to take advantage of this otherwise-good soldier. This then intersects with Chosen making the Blacksite much nastier: if it was as easy as ever, it would be easy to justify throwing Squaddies at it early, maybe babysitting them with one good soldier to trouble-shoot. With the Chosen guaranteed, though, you really want your best soldiers...

... which brings me to reason 2.5: that the Fatigue system effectively slows the rate at which your soldiers level, since you can't send the same six troops into every single mission so long as no one dies or ends up injured for too long. In the base game on Commander difficulty, it's pretty easy to have Lieutenants running around by the time you've reached the Blacksite. In War of the Chosen, you're a lot more likely to be still struggling to get anyone past Corporal by that point on Commander difficulty. Which, again, intersects with the Chosen presence making things rougher: you want to bring mid-level troops into the Blacksite if you can, and you'll be waiting for them longer than if you waited for the same in the base game.

Meanwhile, summoning a Codex adds Codices to the rotation, and that's unpleasant, but it only very slightly raises your Fatigue load that one time and War of the Chosen's new tools slightly reduce the threat Codices present anyway: a Templar has no ammo to drain and so is completely unimpressed by Psionic Bomb, and a Reaper makes it a lot easier to avoid pulling multiple pods such that you can't deal with a Codex readily because whoops you pulled it with your last soldier's last move. Even Skirmishers have some relevant advantages, like being able to Grapple out of the Psi Bomb area and having Justice to hit a Codex with zero chance of a Graze.

I'm sure there are players who still swear by hitting the Blacksite first, but I feel it's pretty unambiguous that War of the Chosen tilted things toward Codex-first. I'm curious if that was intentional or a happy accident, because personally I feel this shift is a huge improvement to the design but it's not the kind of thing where it's obviously deliberately targeted.

On an animation note, the Codex has an unusual quality that nagged at me for ages before I was able to pin down: they almost never ragdoll when killed! This is striking, as while the game has custom death animations for enemies, you normally won't see them very often: shooting enemies to death will normally switch them to ragdoll physics with momentum from the attack that killed them, and you're going to spend the majority of your time shooting enemies with regular shots or special shooting actions that still trigger this behavior. (eg Hail of Bullets, Lightning Hands, Rapid Fire) Only splash damage attacks, melee attacks, and certain weird ranged attacks like Soulfire will cause an enemy to play their personal death animation. (Exacerbating the issue is that many personal death animations aren't very distinctive: Mutons just kind of rear back and fall over backwards, ragdolling partway through, which doesn't necessarily look any different from normal ragdoll death animations. Sectoids are one of the few enemies whose custom death animation is striking enough to be obviously different from ragdolling)

Unusually, XCOM 2 normally applies ragdoll physics even to enemies that cannot possibly be borrowing standard humanoid animations, like Chryssalids and Archons, which is part of why the Codex being an exception is subtly striking. It's even willing to have the Chosen ragdoll on death! (Though you could be forgiven for not noticing: they teleport away shortly after dying, leaving you little time to watch them ragdoll, and they usually will play a custom death animation when killed regardless of source. I honestly wouldn't be surprised if they're not supposed to ragdoll at all) There's only a bare handful of enemies that never ragdoll, and they're all not even slightly humanoid (eg Turrets), among other factors probably contributing to their failure to ragdoll that they don't share with the Codex.

The Codex failure to ragdoll is particularly conspicuous when catching them with Overwatch fire, as normally enemies killed in Overwatch fire will tend to ragdoll on courses appropriate to their own trajectory: a Berserker that dies to Overwatch while charging one of your soldiers will faceplant, that kind of thing. A Codex will instead take that last hit and suddenly come to a complete halt, all momentum vanishing, while it plays out one of its custom death animations, such as the 'clutch at head screaming' one.

In the case of Codices, it's not even entirely accurate to say they don't ragdoll. Their body doesn't, with them normally playing one of their death animations, but once the animation is finished their brain is left behind and it does obey ragdoll physics. (Though it's pretty easy to overlook the brain once they're dead...) This gives them a fairly unique set of qualities with death animations.

The most interesting aspect of all this is probably a happy accident: probably the entire reason Codices have such an unusual set of animation qualities is because they have the unique quality of having the majority of their body vanish when killed while still having to leave behind their brain because it's a plot-important object you're supposed to successfully collect. I think it actually works pretty well from an in-universe standpoint, though, since Codices are supposed to be a pretty weird being, with Tygan insisting that what we're seeing is some manner of projection, and even if his exact framing seems questionable it's certainly the case that what we see visually is that the brain is a persistent object and the rest of the body is somehow generated and sustained by the brain, and mechanically damaging a Codex causes it to reactively split, with the left-behind copy not leaving behind a Codex Brain when killed. (Loot-wise, anyway. They don't have different animations, so they still leave behind a brain that ragdolls, visually. You can end up with a hilarious amount of the things rolling around if you let Codices split a lot) Coming to a complete halt in contravention of normal physics when finally damaged enough to be unable to sustain their projection is weird in a manner that feels consistent with the exact type of bizarreness ostensibly going on with Codices.

I do wonder if the animators noticed that on their own and decided to run with it instead of 'fixing' it. It'd be pretty cool if that's what happened, or something like it. 

That said, every once in a while, a Codex will ragdoll on death, including that their body will end up lingering, laying about. I'm pretty sure this is some manner of glitch, but have yet to identify a pattern to when it happens to thus guess why it happens.


Going meta for a moment, the Codex is XCOM 2's Outsider. It's a plot-critical enemy you have to stun Skulljack to get access to the first major Ethereal facility, where the enemy in question is explicitly claimed to be some manner of projection instead of an independent entity in its own right with strong implications in each case that it's some manner of Ethereal puppet per se, repeat copies are worthless/you can't sell its corpse drop, and its spawn table is plot-controlled.

It thankfully is a much better execution of this general concept, plot gating adding it to the spawn table instead of removing it and being a much more inherently interesting enemy to fight, but yeah, Outsider 2.0.

I have somewhat mixed feelings about the dev team returning to this basic archetype, as I didn't find it terribly interesting or compelling in the prior game and don't really feel there's any particular intrinsic value to it, but on the other hand the Codex escapes a lot of the reasons I actively disliked Outsider design, so... whatever, I guess?


... the Codex is legitimately one of the neater units in the game, with a lot of nifty visual effects doing a decent job of selling that the Codex resembling a female human form is extremely misleading to what it actually is. Visuals jerking about, gaps you can see right through appearing in the model, sometimes the entire yellow surface effect will vanish in favor of a metallic-looking layer underneath, and myriad lower-key effects help sell the idea we're not really seeing a conventional sort of being.

Notably, it really feels like the Codex is an attempt to take the concept of quantum superpositioning and translate that into a set of visuals and game mechanics. (If you want a basic understanding of the concept, Schrodinger's Cat is probably more practically useful, to be honest) Science fiction loves to invoke quantum physics and has for literally decades, of course, but usually I come away feeling like a given piece of fiction didn't get a given concept at all, or did get it but was only vaguely invoking it as a fairly thin excuse for some other goal ('quantum' is depressingly prone to being scifispeak for 'magic', in that manner where a writer doesn't want to admit to writing anything so unscientific as magic, but still wants to write magic), or otherwise wasn't really trying to run with the concept.

The Codex visuals and gameplay is a rare case of really coming across to me like more or less everyone involved got the concept at hand and was trying to be reasonably faithful to it even while turning it into a gameplay piece. It particularly stands out to me how the game doesn't try to call your attention to it: I have issues with what Tygan does have to say about Codices, and I'll get to that in a minute, but there's nowhere where the game tries to explicitly compare the Codex to quantum superpositioning. This stands out to me because usually when quantum mechanics is used as inspiration, a story will go out of its way to draw attention to quantum mechanics being the inspiration, and often the final result feels more like either a performative attempt to convince the audience the creator Knows Things, or more generously like the creator got recently exposed to the concept, got excited by this weird and wild concept they're still coming to grips with mentally, and by extension doesn't really get it. It's much rarer for pop culture to just matter-of-factly include elements of quantum mechanics, and XCOM 2 doing so makes me think the devs legitimately got the concept and thought it would be cool to use it as the basis of a new enemy concept. That would be a very different, much more science-fiction mentality than the prior game trying to take classic X-COM concepts and make them 'grittier' and 'more realistic', so if I am correct on this count I 100% approve.

And to be honest I'm having a hard time imagining an alternate reason for the Codex to have ended up with the exact set of visual and mechanical qualities it got...

Regardless, the one bit of strangeness in this regard is their rifle, which has more conventional visual design (Very similar to eg a Viper's weapon) that places it more in line with the Codex's skull that seems to be the only 'real' part of their body... but the rifle will vanish completely when teleporting or using Psi Bomb, and even in some pod activation cinematics. I suspect the rifle was a late addition to a unit originally imagined as relying solely on its psychic abilities, with the visual inconsistency occurring because it was a bit of a slapdash effort. It's possible it's actually intentional, but I'd be quite surprised, even considering the rifle can be seen turning yellow and vanishing on death, as at all other times its implementation is visually weird.


... a tutorial-only cinematic depicts a Codex spawning in, and no firearm is visible at any point in the cinema. This is particularly interesting since various elements of the tutorial-only cinematics make me think they were made much later in development than most of the cinematics, which if so would suggest the Codex picking up a rifle happened really late in development.

Though having thought on it a bit deeper, I suppose this could be a deliberate 'the Codex simultaneously does and does not have a rifle until observed, by which I actually mean it's cheating and whichever is most convenient to the Codex will be true at any given moment'. I don't think it is, but it's not like it's impossible.

Regardless of the rifle's intended or original place in the Codex design, a different animation point I like is... thing is, back in the Sectoid analysis post I commented at the end that XCOM 2 approaches missed-shot-reaction animations differently from the prior game, that in the prior game a frustrating number of enemies were animated as if they were literally dodging bullets, where XCOM 2 defaults to having enemies react too late, and in a manner consistent with 'whoa, that almost got me', instead of 'watch me dodge your slow bullets, mortal'.

Codices are the primary exception to this rule, with most of their 'you missed me' animations being constructed as if the Codex saw the shot coming and successfully evaded it. You might expect me to dislike this given I disliked enemies doing it in the prior game, but in the Codex's case it fits naturally with their high Dodge stat and reads like it's an extension of their broader quantum superposition thing. I didn't like enemies dodging bullets in the prior game because it comported itself like it was a gritty, realistic game you were meant to take seriously and then refused to give its aliens viable explanations for how they could dodge bullets and lasers. In the case of the Codex, its specific type of strangeness really seems like a fairly natural explanation for such -particularly memorable is how one of their 'missed me' animations has them transition from standing to crouching without really passing through the in-between stages, visually. The whole thing makes me think the Codex is meant to be pretty literally in an uncertain superposition state with some ability to cheat and pick whichever state is most useful to it, and that's a pretty solid Crazy Alien Ability to justify literally dodging bullets.

(Also it helps that XCOM 2 doesn't have literal lasers in its weapon list, though beam weapons are all hitscan in XCOM 2 so this doesn't help much)

Indeed, Codices having high Dodge makes much more intuitive sense than, say, Vipers, where I honestly am not sure what is meant to be the logic behind their high Dodge stat. Which I pick as my comparison point not only because I've already been over Vipers, but also because Vipers have higher Dodge than Codices! Codices are teleporters with some uncertain positioning madness going on that are possibly quite divorced from conventional reality in a more general sense. What's the snake-lady supposed to have going on that justifies beating that?

At least Vipers don't animate as dodging bullets when missed. I dunno, maybe their high Dodge is meant to represent their scales occasionally causing a shot to glance off instead of hitting dead-on?

Narratively, Codices are... odd. I'll be getting into some of the oddness more in a later post because their oddness is tied up elsewhere, but in an immediate sense Codices end up raising a lot of problematic questions that the game has no interest in answering or exploring. A Codex teleports in through a psi gate because you've hacked the ADVENT psi network and the Ethereals are Not Amused, but then Skullmining won't reproduce this behavior and Codices won't, for example, psi-gate in as a special form of reinforcement in a more general sense. So, uh, what was with that first incident? It's not like it's a special super-Codex, where I can assume the Ethereals aren't willing to throw super-Codices away on network intrusions after X-COM proves able to kill them, but is willing to patrol with regular Codices. Why did it happen that first time and only that first time, aside the meta answer that Skulljacking to spawn a Codex is advancing the plot while Skullmining regular enemies is non-plot gameplay?

Then there's them entering the general spawn table in response to the Officer Skulljacking. I appreciate it mechanically, but narratively it's deeply confusing given the story implies Codices are collectively some super-secret black ops enforcers of the Ethereals' will. They're not even one of the enemy types that normally refuses to spawn in ADVENT city centers. So, uh, what's going on here? X-COM learned Codices exist, so the Ethereals don't care about keeping them a secret, I guess? Why were they being kept secret in the first place, if the Ethereals are so willing to stop keeping that secret?

Then there's the Tutorial mission cinematic, where after Bradford rescues the player character we see a Codex ominously pop in to look at the Commander's tube and instead of, say, teleport-pursuing the Skyranger it just... goes away. I didn't question it at the time, because it looked like this was some kind of foreshadowing, but, uh, no? It's this inexplicable thing that happens and is never followed up on.

It's not like there's anything clearly wrong with the Codex's handling as far as these narrative elements, but it's confusing and feels like the devs either had a specific idea they forgot to actually implement or changed their collective mind partway through and didn't adjust everything to fully fit their new concept. Like maybe at one point the idea was the Codex would be a semi-autonomous network intrusion countermeasure unit that only shows up in response to Skulljack usage, but then they just sort of shoved it into normal gameplay without coming up with a reason that would make sense. Which would be a weird reversal of the Outsider problem, where it's annoying and nonsensical that they completely vanish once you've hit the Alien Base.

So I'm not entirely sure how to take Codices, as a narrative entity. War of the Chosen just heaps on further confusion, too, as each of the Chosen actually has dialogue responding to you killing your first Codex, but said dialogue is vague and what it maybe-implies doesn't seem terribly consistent with what the base game seemed to be going for. I'm especially unsure of how to take the Assassin characterizing the Codex as a 'temperamental' being that is 'so unmoved by our pursuits'. So all-around I'm not entirely sure what the Codex is meant to be, as a narrative entity, and how much of its gameplay presence is meant to reflect its narrative position. The core plot threads never actually directly state the Codex is one-of-a-kind, for example, but a lot of little things make it seem like that's the intention... but the game doesn't clearly signal that regular Codex spawn routines are Pure Gameplay Of Zero Narrative Meaning. So I just don't know.

A less obvious thing that... mostly stands out to me due to how the Forge mission is handled... is that the entire flow of events is unclear in X-COM-the-organization's intent. Tygan wants to hack the psi network to accomplish... I have no idea. This results in the Codex appearing and attacking, which the narrative is inconsistent about whether this is a surprise or completely expected. Then we kill the Codex and study the Codex Brain in the Shadow Chamber, which is useful and advances the plot but is never suggested to have been the goal behind hacking the psi network. So... what was the original goal? Did we accomplish our goal? Did we fail it, but take the Codex Brain as an acceptable substitute prize?

It's not like it's a nonsense flow of events or anything, and if you're not paying close attention it's easy to mentally fill in the flow of events so it seems perfectly coherent, but if you are scrutinizing the plot... it's just confusing what's going on here. The player is Skulljacking an Officer and killing the resulting Codex because those are explicit mission objectives (And also killing everything hostile is a default modus operandi), and the plot advancing as a result of these actions more or less makes sense, but 'let's hack the psi network' comes up in response to completing the ADVENT Officer Autopsy, which itself is a plot-critical Autopsy as part of examining the chip that was in the Commander's head. And that's where there seems to be a step in logic missing: what is expected to be accomplished by hacking the psi network? As far as I've noticed, nothing is ever suggested, even though it's presented as a targeted goal, rather than a more general goal of wanting to find anything ADVENT is hiding for no deeper reason than because they're the enemy and so anything they want hidden is probably something we want revealed.

I'm genuinely curious as to whether the plot chain was originally constructed differently somehow, or if the devs just never got around to coming up with a clear in-character reason for Skulljacking to be a solution. I'd absolutely believe the latter scenario, as the objective to Stun an Outsider in the prior game was really blatantly a working-backwards issue where if you know what X-COM is intended to need to do then of course it's going to be a link in the chain because it's necessary for the Skeleton Key, but where the characters themselves latching onto capturing Outsiders in particular (As opposed to, say, Thin Men, on the basis that their infiltrator status demands interrogation gets to skip the language barrier) is just inexplicable. But the exact way this detail is handled, I also would absolutely buy that the devs intended there to be an actual in-character reason and whoops just didn't communicate it properly.

On a different narrative-related note, the Codex is one of the more egregious examples of 'the developers have a concept of what's going on here, so the Scientist Character is going to spell it out even though that character has no way to know this is true and no reason to even suspect the possibility'. Like sure it's cool that the Codex is supposed to be some sort of psychic projection from another dimension or something in that vein, and the fact that this is the notion the devs have in mind fits reasonably well with the mechanics they came up with, but if I were Tygan I have enormous difficulty imagining why I'd come up with such an explanation on sight. Like yeah they've got the weird glitchy-looking visual effects, but it could just as easily be some kind of cloaking effect to obscure the identity of the Codex. It's not until they die and leave behind just the head that it's for-sure clear something weird is going on with the Codex in a manner remotely in the vicinity of Tygan's wild theory, and it's still a bit unbelievable it would be the first theory Tygan jumps to -especially egregious is that Tygan's response to a Codex cloneporting for the first time is to remark that he's never seen anything like this. So... uh... why does he know what it is on sight, then? 

Unfortunately, while XCOM 2 is less egregious about this kind of thing overall than the prior game, this is still a recurring issue; Tygan routinely jumps to wild, unreasonable theories and has the narrative (Implicitly, sometimes, explicitly other times) bear out his baffling beliefs, just like Vahlen before him. Just like with Vahlen, this is particularly frustrating coming from a scientist character, where such behavior is basically antithetical to the kind of person they're supposed to be and instilling anti-science notions into players that are presented as representative of science. I'd honestly be less aggravated if XCOM 2 had a wizard doing this nonsense. At least then it wouldn't come with the 'teaching players unscientific principles while making them think that's what scientists are supposed to be like' horribleness.

War of the Chosen has me wondering if the dev team is maybe finally cottoning onto what Actual Science looks like, but I'll get into that in a later post.

The especially frustrating thing in the case of the Codex is that Tygan was an ADVENT insider of mild importance. It would be so easy to have him allude to having heard rumors when he was still working inside ADVENT, and use that as the basis of Tygan explaining to the player the narrative concept the team has in mind. Or have Tygan indicate he was colleagues or friends with someone who helped develop the Codex, or helped integrate it into Earth's psi network, or whatever, where Tygan's insider status gave him a more direct line in this case to the relevant info -this would also help smooth over the issue where your support staff doesn't act surprised when the Codex spawns in, and in fact Tygan straight-up refers to it as a Codex, which as-is just comes across very strangely. (It also, if you take it completely serious, casts your support staff in a bad light; if they were expecting this, why didn't they mention it before you triggered it?)

Either way, there are very obvious, easy ways to get Tygan delivering this info that make far more sense than him jumping to an essentially arbitrary theory off nearly no evidence. So I just... don't get why they went with Wild Theories From The Scientist Character in this case, above and beyond my general principle reasons for disliking doing so.

Still. I do overall like the Codex, it's just Tygan's response to it is... so unnecessarily bad.


Next time, we cover the ADVENT Mec.

See you then.


  1. I understood the "projection from a psychic dimension" to be the explanation for the teleportation, (i.e. occupying two "separate" spots at the same time), and phasing in and out of view. A being that can move in additional spatial dimensions than the 3 we are used to would apparently do all those things constantly. In my mind, the "hair" was the connection to the "rest" of its body in the additional dimension. It doesn't explain why it consistently appears like a female humanoid (it should be constantly changing shapes in ways that aren't really recognizable), and leaving behind a corporeal brain is odd to say the least.

    Anyway what I am trying to say is that I did not think of them as "quantum" before this post. Thinking about it, it does work reasonably well too (ignoring that something that size should not be "quantum", but that's beside the point). The visuals (a "glitching" figure, rather than a shape in continuous flux) is more in line with something like that.

    Never noticed that the "skulljack an officer to gain access to the psychic network" never actually goes anywhere, narratively. Huh.

    I was expecting a little bit more on the shadow chamber projects, because even with a hefty dose of suspension of disbelief vis-a-vis Tygan and Shen's capabilities as geniuses, that was some mighty jumping to conclusions. I assume there will be a post on that later on.

    1. It could be that the projection stuff is intended to serve the purposes you lay out, at least partially, but there's really blatant Codex/Avatar connections that are why I take Codices the way I do, especially in conjunction with the Outsider parallels. (Which were also implied to be remote-controlled Ethereal puppets or something of that sort) I just haven't covered them in this post because I'd rather cover those connections in the Avatar post.

      Yeah, something at the size of a Codex shouldn't be quantum, but I'm willing to go with that for two reasons: first of all, Schrodinger's Cat was constructed in an attempt to illustrate quantum uncertainty being a ludicrous theory by imagining a scenario in which quantum uncertainty clearly interacts with the large scale in a nonsensical way. Actual physicists get headaches trying to reconcile quantum uncertainty with large-scale reality -it's why we even have quantum mechanics as its own field, whereas we don't have, I dunno, 'star system mechanics' as a field covering the a scale larger than what we're used to. (Because overall the really large scale is perfectly consistent with our own scale) And why scientists have been obsessed with inventing a 'unified' theory of physics -the unification in question is 'how do we get conventional physics and quantum physics to fit together as one system?'

      The other reason I'm willing to go with it is that, realistically speaking, if highly advanced aliens showed up, they SHOULD be doing things that violate our understanding of physics. I might object to specific choices for any number of reasons, but it would be tremendously silly of me to pursue scifi and then take the attitude that nobody is allowed to violate our current understanding of physics.

      Yeah, I'm intending to go into Shadow Project stuff more in other posts. There's enough jank with how the Shadow Chamber was handled it... really sprawls across multiple topics. 'Skulljack an Officer to accomplish ???' is just the earliest example of Shadow Chamber jank...

    2. We do have "star system mechanics" as its own field. We call it "astrophysics". General relativity does strange things once we start throwing around enough energy and gravity to create black holes. Also, there's a certain amount of specialist knowledge involved in working out physics under very limited information; we have a theory about, say, the Fermi Bubbles, but if we could get information from a telescope up in the galaxy's center, we'd probably have to completely revise everything.

      With respect to some of the Codex narrative weirdness, I'm inclined towards the theory that "Codex is a singular creature, and all the Codices that show up in gameplay are the projections of the same creature, using its Clone ability"; the Codex doesn't know about you, you do the skulljack, the Codex notices you, the Codex starts trying to hunt you down.

      (Presumably, under that theory, the Codex is still out there after X-COM 2; you never really kill the true Codex, only banish many, many projections)

      There's a lot of narrative weirdness that that doesn't explain, but I'm content enough with that as an explanation for why there are Codices everywhere, but it's sometimes referred to as a singular creature.

    3. Well, my point was that quantum mechanics and 'conventional' physics are... basically irreconcilable with each other on just very basic mechanics, as far as my understanding of current theory goes. Whereas at the scale of star systems you don't go 'now toss out just about all physics rules, to all appearances'. (Last I read, admittedly; it took me a while to get a proper explanation of what 'dark matter' and 'dark energy' were about, for example, so I wouldn't be surprised if I've missed a more recent 'this is a big deal' revelation in that realm)

      The Codex being referred to in the singular is pretty clearly the devs not knowing about -or not liking, or suspecting most players won't know about- Codex pluralizing as 'Codices'. My early drafts of this post from over a year ago started out with 'Codexes', then switched to using the singular because 'Codexes' just sounds wrong, and then at some point I learned 'codices' is the plural of 'codex' and started reworking myriad lines to suit it. I suspect my own historical unfamiliarity with 'codex' having a plural is pretty normal. And Chimera Squad makes it unambiguous there's meant to be multiple Codices in existence: the Progeny are describing as getting a hold of 'a' Codex to power their efforts, not 'the' Codex. (With it then being implied that every gameplay Codex is a Clone of the one main Codex they're using)

      Mind, this still leaves questions about what's going on with Codices, but at this point I suspect the dev team's concept shifted and that XCOM 3 may well touch upon them anew with a clearer explanation that's very different from XCOM 2's apparent model.

    4. Just for the sake of precision, the problem with quantum gravity is that the theory of gravity breaks when trying to apply it to the quantum scale, not the other way around. Quantum effects at macro scales just “wash out“ and become imperceptible, so for the most part there’s no concern in that direction. Not that I think this is important, if the devs said the codex is a macro scale quantum being, I would believe it just as much as the humans with latent psychic powers.

      The projection from additional dimensions doesn’t preclude codices being ethereal puppets, and in fact would maybe even do the exact opposite, in that it provides a natural (if unexplored) explanation in how they’re controlled from afar. In case the projection from a higher dimension concept I had in mind wasn’t clear, it’s the same notion explained in this Sagan video:


      To be clear, I don’t have a strong preference for either interpretation, I’m just mentioning the one I came up with my first time through the game.

    5. I wasn't thinking of gravity, as I wasn't even aware of there being a gravity-specific issue, and from what I'm reading the gravity+quantum issue currently existent is 'we don't have the tools to test/observe gravity's effects at the quantum scale'. I was thinking of how a lot of underlying elements of quantum mechanics range from 'extremely counterintuitive' compared to how things operate at a more human scale, to 'no seriously, how DO you make these both simultaneously true? They gotta be, we can see they are, but HOW?' Stuff like 'spooky action at a distance', or 'quantum mechanics really seems to have cases where effect happens BEFORE the cause of the effect does'.

      I wasn't so much disputing 'Codices as projection from another dimension' (That would be silly, given the devs-via-Tygan explicitly say that's what's happening) as I was disputing the idea of there being a singular Codex creature. In the base game, I'm pretty sure Codices are meant to be... basically a fancier version of Sectopods in classic X-COM: a technological apparatus puppeted psionically by Ethereal masters, with the Codex Brain being a receiver unit that doubles as the vector for projecting the 'body' of the Codex.

      Then War of the Chosen has the Chosen implying the Codex is a distinct being in its own right, and Chimera Squad has Codices persisting even with the Ethereals apparently all gone from Earth, so... like I said, I'm pretty sure the dev team's concept of what a Codex is changed pretty substantially after the base game's release, and I'm interested in seeing if XCOM 3 will bring them back in a radically different form. (I really wouldn't be surprised if XCOM 3 makes your own theory, or something extremely similar, canon)

  2. I would also like to see more complicator enemies in XCOM (and low-scale tactic games in general). I think I brought that up in the post about the purifiers already, although not with these same exact words. When both sides have low numbers and can only interact by dealing straight damage to each other, the gameplay stays a bit one-dimensional.

    In the case of XCOM, you also have to solve the alpha-strike issue, since it doesn't matter how interesting your enemy effects are if they're never alive to use them by the time their turn comes around.

    1. Complicator enemies are pretty normal for low-scale tactics games that have a fantasy aesthetic, though I don't think it's a particularly consciously-implemented idea. Spellcasters tend to be given buffs, debuffs, summons, and other effects that are, in an immediate sense, less of a problem than trying to kill someone is, but which can be maneuvered around ("I'll just keep out of reach of your buffed guy until the buff times out, or keep them stunned, or whatever") and/or answered in entirely different ways. (eg dispelling buffs, which is a special capability that costs a limited resource, as opposed to just attacking the individual) XCOM 2 is noteworthy in this regard primarily because it has a modern-to-futuristic aesthetic, and game devs pretty consistently struggle to make those more interesting than 'I shoot them, hopefully with my weapon having a specific range it's most effective in or something'. Classic X-COM was, similarly, noteworthy in no small part because it managed to take 'mostly you just shoot at people' and still make that a reasonably tense, nuanced game design.

      Yeah, the alpha strike issue is very much why I'm potentially interested in XCOM 3 returning to an Enemy Unknown/Within-esque assumption of give-and-take combat. I like Chimera Squad, but it mostly changes the nature of the alpha strike issue, where your goal is still 'try to kill or disable all enemies before they move', and it's just its turn order mechanics force you to prioritize differently. ("The ADVENT Mec is going first out of enemies, so I HAVE to target it first if I want no enemy to ever act.") Though it at least is trying to experiment with a more give-and-take design, which is part of why I'm leaning toward XCOM 3 probably moving away from the alpha strike design; War of the Chosen already made strides in that regard, and Chimera Squad is an experiment at trying to move away from that design in a different manner.

    2. IMHO, the biggest move away from alpha strike in CS was making death non-permanent. This severely reduces the long term consequences of getting shot at, which means the game can be balanced around enemies getting to act. The timeline mechanic is almost secondary in that sense, although it did make it harder to always prioritise taking down the highest threat (if they were high up in initiative). I do agree with your general point though!

    3. I've played all the way up to the highest difficulty in Chimera Squad, and soldiers being downed happened extremely rarely, especially past the very early game. It was only in my first run that people went down semi-regularly all the way into the third Act, and I felt the 'Bleed Out finishing equals an instant game over' element of the change more than offset the part where agents hitting 0 HP was never lethal, since it meant you 100% HAD to scramble to save an agent, no option to write them off. (Where in XCOM 2 you can go 'eeeeh, this is literally my third mission, that's a Squaddie, it's not worth trying to save them when the Chosen Assassin is still in the middle of trying to take apart my squad') Especially because Chimera Squad retains the part where explosives will instant-kill a Stabilized individual; enemies with explosives aren't a given in Chimera Squad, but they're not exactly rare. So even Stabilizing someone hasn't necessarily prevented the game over.

      Conversely, I found that, aside some specific agents being consistent underperformers (Most glaring example: Zephyr) to an extent that was seriously problematic, I could pretty reliably minimize the opportunities for enemies to get turns, especially in regular missions made of 1 or maybe 2 encounters. (Because then I could freely toss around Items, without having to worry I might need them more later) This itself in part came back to... janky design elements not really working as intended/sensibly, with the big offender being that during the Breach phase it's best to maim, but not down, your targets, since the game constructs the timeline AFTER the Breach phase is done and does its best to ensure your team alternates with enemies: thus, knocking an enemy down to 1 HP could result in them being slotted into the first enemy turn, and now you trivially finish them, whereas if you'd taken them out during the Breach phase some other, healthier enemy would get slotted in. Damage over time effects are particularly abusable in this regard: shoot an enemy down to low health backed by Venom/Caustic/Dragon Rounds (Or have Torque use her Toxic Greeting, in the early game against frail enemies), and they'll instantly die on their turn, turning it into a dead slot of a turn.

      Even when my squad was outnumbered 2-to-1, I rarely let enemies get turns once I was modestly experienced at the game. (Or perhaps more precisely 'turns that accomplished anything' -I'm perfectly happy to let a far-away Faceless get a turn, since it can't do anything relevant with it, that kind of thing)

      So like I said: Chimera Squad shifts around the nature of the alpha strike issue, but unfortunately it's still very much optimal, achievable play, in spite of the game clearly wanting a more give-and-take approach.

    4. Interesting, I hadn't thought about the *necessity* to rescue your operators as a way in which you still want to alpha-strike, that's a good point.

      You're right that exploiting the timeline, especially with spreading Breach damage around, means you can reliably avoid enemies attacking. Only way to prevent that would have been to hide which enemy is the one that gets to act next, I don't think any other way to adjudicate timeline turns would have prevented that. But that sort of goes in the direction of my previous comment, in that the Timeline rather than faction turn-by-turn gameplay does less to solve alpha-striking than balancing health pool and damage around the idea that you will be getting shot at. But maybe achieving that balance is just really hard.

      I found Zephyr to be decent enough with the right support. While her damage doesn't scale very well and she's a bit short on utility, parry means she can effectively attack one enemy while shutting down another (by standing next to them and baiting their shot). She's arguably the best target of Terminal's cooperation, since you still get to do both even with a single action, and it's not on cooldown.

    5. Well, one way to make the timeline not lead to alpha strikes per se would be to hard-enforce alternating teams. It's a rare model in games -and with good reason, as it has a lot of strange implications- but if killing whatever enemy was going next just changed WHICH enemy went next, that would make it so you flat-out couldn't alpha strike to prevent turns. You'd have to rely more on disables, make judgment calls about whether the next enemy is the worst possible thing to act next or if a replacement would be even more problematic, and just generally get reduced benefit out of focusing on killing things. But yes, tuning durability and damage so it's unrealistic to alpha strike is really the main solution that's both practicable and... well, unambiguously a good one.

      I don't really want to get into the whole thing here given I do intend to do big 'ol posts for the agents down the line, but Zephyr is burdened by a variety of flaws no other agent can hope to match, and suffers heavily from being a dedicated melee attacker in a game where everyone has a backup melee attack... where said backup melee attack isn't actually clearly worse than Crippling Blow. (Subdue ignores Armor and KOs targets: you have to spend level-up picks on Crippling Blow to get it to have those effects) In conjunction with assorted other decisions working in concert to hurt her case in slightly less obvious ways, there's... really no question that Zephyr is the worst agent in Chimera Squad by a wide margin. You can get through a campaign using her, absolutely, and I enjoy the Templar-like experience, but using her on higher difficulties, later in the game, is absolutely going to lead to agents going down that wouldn't have gone down if you'd just used someone better able to contribute.


Post a Comment

Popular Posts