XCOM 2 Analysis: General Enemy Intro


XCOM 2 is, in many obvious ways, operating on a broadly similar system to the prior game: enemies come in pods that are 'inactive' normally and activate when they see your squad (With the new caveat that they don't activate in response to Concealed troops), at which point the pod scrambles for Cover as a free action but doesn't get a chance to shoot at your forces if they activated during their own turn, etc.

There's some key details that radically change things, though, some not very obvious if you don't play the two games close together.

First, the relatively obvious: pods are now mixed by default, instead of mixed pods being restricted to specific cases like Cyberdiscs always coming with Drones in the prior game. You can still encounter pods of only a single enemy type, and indeed some enemy types will normally refuse to share pods with other enemy types, but it's much more typical for a given pod to be two or three different enemy types. This has a variety of implications, such as different enemy types being able to synergize with each other and this actually matter without requiring the player having pulled multiple pods at once, but one of the more subtle implications that the game uses to good effect is that enemy durability and overall threat level can be much more uneven.

As a concrete example, on Legendary Sectoids will have 10 HP, and can show up from the very beginning of the game. That would be completely insane to have in the prior game, as three such Sectoids (ie a pod of standard size) would be literally as durable as a Sectopod, an enemy you're really expected to fight with endgame weaponry. In XCOM 2, however, an early-game Sectoid won't show up with a buddy or two of its own species, but will instead be leading one or two 4 HP ADVENT Troopers. That's a much more reasonable 14-18 HP to deal with -especially since XCOM 2 has raised your damage a little, so that an early-game four-soldier squad will output a minimum of 12 damage if every shot hits but does not crit, where in the prior game their minimum damage would be a meager 8. Thus, the Sectoid can be quite a bit tougher than a basic ADVENT Trooper, show up just as early, and yet not lead to insane swinginess the way it would in the prior game's model.

On the note of leading pods, there's always a member of the pod that is considered the leader. There's a bunch of internal stuff to do with how the game organizes pod generation involved in this, but the primary reason I mention it is that pod leaders are always prioritized for Overwatch fire when the pod activates, and thus will usually absorb Overwatch fire before their podmates. This is a point that's somewhat opaque in actual play since the game is perfectly happy to animate the resolution of multiple simultaneous Overwatch shots in a completely different order from how the game expended them internally, and furthermore it's not necessarily obvious that the reason an ADVENT Trooper got shot instead of the ADVENT Officer by exactly one of your squad members is because that soldier never had a line of fire on the Officer. Regardless, typically, albeit not always, a pod leader will be the most durable member of their pod, potentially by a wide margin, and thus it's surprisingly uncommon for you to completely eliminate more than one enemy with Overwatch unless your firepower is complete overkill for the pod. A secondary implication of this is that it's often smarter, when initiating an Overwatch ambush, to actually have the first shooter target one of the weaker pod members, particularly early in the game when most of your classes are guaranteed to kill ADVENT Troopers with a successful hit.

This also means it's much more important to pay attention to whether the pod leader is in Cover relative to the rest of your squad or not when considering whether to initiate an Overwatch ambush or not. If the pod leader is in High Cover relative to your squad, it's alarmingly likely that you'll waste a lot of your shots on misses, even if your squad should be able to cleanly kill the other pod members. If time pressure means you can't actually wait a turn, it may be best to simply take the first shot and then fight the pod as normal, instead of going for an Overwatch ambush.

On the topic of pod activation and Overwatch, there lies one of the more subtle-yet-significant differences between XCOM 2 and its predecessor: in the prior game, when a pod activates it normally flees away from your squad, seeking Cover somewhere in the shadows, and once a pod is activated it's frustratingly likely its members will stay in the shadows, perpetually Overwatching so you're forced to close with them and risk being shot in the process.... unless there's a point at the very edge of your squad's vision they can pop into and take a shot from without triggering Overwatch fire, in which case they will happily pop in and kill one of your squad with absolutely no risk to themselves, because the AI in the prior game blatantly cheats.

An extension of this blatant cheating is that when a pod initially scrambles in the prior game, they know exactly where all your soldiers are, and move to Cover to ensure none of your troops will be able to flank them -even your Squadsight Snipers on the other end of the map.

This all served to suck out a lot of the relevance of tactical skill in the prior game, particularly in the early game when your options are very limited. You can't advance and chuck a grenade to remove Cover, because the pod fled too far. You can't fire a Rocket into their midst to kill them and/or smash their Cover, because Rockets can't be fired after moving and the pod fled too far. You can't flank them except maybe with an Assault, because they magically know exactly how far every one of your soldiers can move and shoot (Even though the player doesn't get this information clearly communicated to them!) and use their free move to make that impossible, the only exception being that they don't 'see' Run & Gun's effects -and even that helps less than you might expect, because them scrambling into the darkness means they're often not flankable by an Assault anyway.

A further implication is that Overwatch was surprisingly underwhelming: generally a pod entering line of sight from the edge of your vision (As opposed to eg busting open a door from much closer to your squad) would only trigger Overwatch from 1-2 of your troops, even if your entire squad was ready to go. Sometimes nobody would take an Overwatch shot, due to the pod's closest members entering the very edge of your vision and then turning right around, thus leveraging the 'have to be seen for more than 1 tile to be Overwatched' rule. Which was hugely frustrating because Overwatch being that ineffectual against activating pods didn't actually mean perpetually crawling forward and Overwatching was bad play: it remained optimal play you 100% should be doing in most missions, it's just the payoff was depressingly limited for how much effort you as a player had to burn on it.

Another, non-obvious issue with how this affected the early game is that terrain damage mechanics heavily favored the AI in the prior game. Since the AI cheated to ensure your tactical skill is worthless, you were largely stuck throwing dice at the problem, and aside the lower difficulties secretly weighting the dice in your favor the dice were stacked against you, not only in terms of eg an Impossible Sectoid having more Aim than your poor Rookies, but also because a pod of Sectoids missing would inevitably wreck your Cover, resulting in an easy killshot on a suddenly-exposed soldier. Meanwhile, your Conventional weapons were basically unable to destroy Cover, so even if you managed to persistently ensure you outnumbered the enemy at any given moment (Unlikely, but work with me here) you weren't liable to win by throwing dice at the problem.

Then, of course, this all combined with the pod activation system in a noxious way, forcing you to advance toward the enemy and risk activating still more pods. As the prior game was terrible at spreading out pods and had a large number of very tiny maps unsuited to how many enemies the higher difficulties were willing to cram into them, the very process of trying to engage the current pod was, on any difficulty the game wasn't actively cheating for you, alarmingly likely to activate still more pods. On top of all that, close-range bonuses were surprisingly difficult to come by against most enemies, making it sort of irrelevant that a Rookie could achieve a 100% accurate shot on a Sectoid by getting in its face, and further restricting your ability to tilt the odds in your favor through judicious tactical decisions.

The overall result was that player skill was a near-non-factor, particularly in the early game, and you were largely stuck rolling dice and praying the RNG didn't hate you, which was a futile prayer because the actual game design most certainly did hate you.

In XCOM 2, instead pods charge your position on activation. (You'll occasionally see a member scramble backward into the darkness, but this is usually because there's inadequate Cover to advance toward while staying spread out, and in XCOM 2 pods try to spread out on first activation so you can't mass-kill them so readily) They still endeavor to select positions that none of your squad members can flank with one move, but this aggressive behavior right away brings many of the mechanics that existed in the prior game into much greater relevancy: usually at least one enemy is in easy grenade reach, if not multiple, close-range Aim bonuses are actually possible to get in the opening stages of a fight, and you don't need to advance dangerously far and risk pod activation just to improve your chance to hit a little bit.

Furthermore, the AI doesn't know where every member of your squad is at all times. They're pretty reliable about ending in positions that aren't currently flanked, but soldiers hiding inside buildings or up back atop high ground or otherwise not visible to any of the final positions will fail to be accounted for, making it actually possible to move-and-flank-attack on the very turn a pod activated without specifically needing abilities the AI doesn't 'see'. Similarly, Sharpshooters that are far enough back won't be accounted for, potentially allowing a fortuitously-placed Sharpshooter to flank-snipe a target.

This behavior continues once a pod is activated: bar a handful of exceptions, enemies will continue to advance aggressively and pursue flanks whenever they occur, instead of standing back and rolling dice at you until somebody dies. This not only reinforces the implications of the initial activation aggression and reduced cheating, but also adds in the further implication that in XCOM 2 deaths are vastly more likely to be actually your fault, where in the prior game deaths were pretty much always down to the godawful game design/the RNG. (Depending on whether you take the long view of blame or the short view, basically)

Less subtle in its existence but still easy to overlook its implications is that the Aiming Angles Second Wave option has been standardized. No longer is an almost-flank completely worthless and liable to get you counterflanked for your trouble; now it cancels out half their Defense from Cover, encouraging you to keep moving to threaten flanks of your own. Even if you never quite manage a flank on a given target, you've at least raised your odds. Note that enemies benefit from this mechanic as well, making it so you're punished for being careless with your soldier placement, even if you never get completely flanked. Also note that High Cover and 'extended' Low Cover prevents Aiming Angles from triggering; if an enemy is leaning against a piece of Low Cover that's part of a larger strip of Low Cover, you'll have to fully flank them or break it to bolster your Aim. Conversely, all else being equal, it's better to put your soldiers in extended Low Cover than in non-extended Low Cover, so enemies can't get partial flanks on them. (This extended Low Cover behavior is never hinted at by the game itself, though you can recognize it in action readily enough if you know it exists)

XCOM 2 is also better at spreading out pods across a map, and notably does not increase pod numbers or pod size as part of increasing difficulty (Except a small spike on Legendary in specific, which... does cause some problems, but nothing as drastic as Enemy Unknown's issues with pod counts), and just generally trends toward larger, better-designed maps. (Among other points, you no longer see maps with variable terrain height where it's more or less impossible to get height advantage, which was extremely common in wilderness maps in the prior game. Similarly, while cars are still all bombs, ready to explode on a moment's notice, there's no longer maps in which the vast majority of Cover is exploding vehicles, and therefore you're no longer routinely forced to risk being blown up by cars) It's still the case that the late game increases pod density enough that it's not that unusual to have two pods alarmingly close to each other, but until somewhere in the midgame even Legendary difficulty is pretty reliable about spacing out pods so it's unlikely that even fairly aggressive play will pull another pod.

Unfortunately, this brings me to decisions I have mixed feelings about.

The first of these is 'downthrottling': if at least four enemies are currently fighting you, the game engages downthrottling, which actively pushes nearby pods to move away from the fighting instead of piling in. I'm broadly sympathetic to this decision inasmuch as I suspect it's more a just-in-case measure, meant to help if all the other design measures end up failing to properly pace pod encounters, but I'm not a fan of how it's antithetical to expected behavior in a manner that feels like it's cheating for the player. On top of those points, it doesn't seem to be very effective -I've repeatedly had cases where I got into a more extended firefight with 4-6 enemies and another pod piled in or at least hung out so close that going for a flank or advancing or the like activated it.

The second is Overwatch behavior: where in the prior game, the AI would tend to move once and then Overwatch anytime your soldiers weren't visible, in XCOM 2 enemies virtually never go into Overwatch if they can't currently see your troops, even though that's precisely the moment they should be going into Overwatch. Yes, it was frustrating and bad design that the AI would stand back and Overwatch in the darkness in the prior game, but going into Overwatch was a problem because they perpetually tried to bait you into their Overwatch with their free activation move you couldn't do anything about, not because they went into Overwatch in the shadows at all. This shift is basically low-grade cheating-for-the-player behavior.

The third is fleeing: in the prior game there was a lot of dialogue and enemy animations/sounds premised on the idea that Aliens could freak out and run, but this behavior didn't actually exist. It does exist in XCOM 2, and it's specifically that when you wipe most members of a pod the final member has a chance to decide they're done with this, turning and running toward whatever is considered to be the mission objective area. Normally at some point they'll encounter a pod and join up with it, de-activating until you activate this new pod. For reference, not all enemies are allowed to decide to run, though the list isn't necessarily what you'd expect: the Chosen and Lost are completely unsurprising, Berserkers, Chryssalids, and Psi Zombies all make thematic sense, and Sectopods and Spectres make sense for being robots... but the only other enemies that never retreat are Faceless and, weirdly enough, ADVENT Stun Lancers. All the other robots (Well, aside Turrets, because they can't move) actually are allowed to retreat, as are eg Gatekeepers, even though these are enemies you'd expect to never retreat.

Anyway, I have mixed feelings about this mechanic because it's very random -it's a 50% chance to happen- and even though it could have been tuned to be a nuanced mechanic, in practice it leans heavily toward being low-key player-favoring because it means there's a chance that almost wiping a pod is basically just as good as wiping it, and the fleeing unit joining up with another pod rarely pushes that pod over into being meaningfully more of a problem. Plus it's mildly disappointing that there's no mechanics for inducing this behavior deliberately -if the Flamethrower had high odds of pushing affected units to run, that would actually make retreat behavior an interesting mechanic with depth and nuance and I'd be less bothered by the innate potential for enemies to run. Among other points, it would end up serving as an implicit teaching moment, letting a player know what 'Flamethrowers make enemies run' actually looks like before you actually use it. (Not even getting into the part where it would help bolster the Flamethrower's relevancy...)

The fourth is AI aim climb -more precisely, their complete lack of aim climb.

On the one hand, in some sense this is actually a pretty minor change in real terms. Yes, ranged enemies in the prior game could get up to +35 Aim from getting in your face, but most enemies either didn't use the normal Aim mechanics at all (eg Ethereals, melee enemies) or had a marked preference for keeping back from your squad and so would rarely get the bonus. If you went back and modded the prior game so enemies didn't get more Aim from closing in and then had a player go through the game without informing them of this change, it's entirely possible even an experienced veteran of the game would completely fail to notice you'd done it.

On the other hand, it gets tied into broader decisions I'm not comfortable with because they feel constructed to be cheating-for-the-player mechanics.

The big one is enemy base Aim; most enemies in XCOM 2 have Aim ranging from 65-80, and they have very limited tools for raising their Aim externally. They can benefit from height advantage and Aiming Angles, but aside ADVENT Officers having Mark Target... that's it. This means that, even more so than in the prior game, you can outright make a critical error and then nothing bad happens because even though an enemy had a clean shot on one of your soldiers it completely missed.

It's not just that this situation can happen, it's that XCOM 2 has gone out of its way to embed this circumstance into the game in a one-way manner: your own melee units can achieve 100% reliable hit chances, up to and including that Rend on Templar doesn't even bother with Aim and Defense, whereas enemy melee attacks cap out at about an 80% chance to hit. (Aside the Assassin being unable to miss with her melee, anyway) That's pretty blatant and unambiguous, but the ranged combat design is the same thing, just less obvious. By comparison, in the prior game, yes, if a Thin Man took a flanking shot it probably wasn't going to be a guaranteed hit, but only because the AI didn't attempt to pursue guaranteed hits in the first place; if a Thin Man did end up with a close-range flank, that was a hit. By contrast, in XCOM 2 even the most accurate of enemies can mess up a point-blank shot, and indeed the majority of enemies -even on Legendary- are capable of missing even if they have height advantage and their target is standing in the open. And since enemies aggressively pursue point-blank flanks in XCOM 2, you're quite likely to end up seeing enemies missing point-blank flanking shots!

If it weren't for it being forced onto melee enemies, I might assume this was an unfortunate implication, rather than a desired end goal. If literally every enemy had 100 Aim, then the game raising Aim on higher-level enemies would be less meaningfully a part of their threat progression, and it would result in Cover's benefits being vastly lower for the player; currently, being in High Cover is generally slashing the chance to be hit by more than 50%, whereas if every enemy had 100 Aim it would be, you know, only a 40% reduction in their chance to hit. Point being, there's actually other possible reasons to not give ranged enemies enough base Aim to ensure a hit on unprotected targets, and having the AI not have Aim climb of their own, while an ugly solution, does in part serve the goal of making it so the gameplay is more fundamentally oriented toward aggressive play instead of standoffish play by making it so that getting closer isn't so much of a mixed bag: you get Aim, your enemies don't.

But melee enemies having an innate chance to miss is just randomly saving the player from their own mistakes, serving no other design purpose, suggesting that's really the primary motive behind enemies pretty much always having a miss chance on regular attacks.

I'm particularly suspicious of this probability because the secret cheating-for-the-player the lower difficulties of the prior game have returned and been extended. In the prior game, it was only the lower two difficulties that cheated for you with secret Aim bonuses and whatnot. In XCOM 2, only the highest difficulty doesn't, which is particularly frustrating since the game experience is pretty clearly tuned first and foremost under the assumption of playing at Commander difficulty -I've already been over how Legendary difficulty is like a completely different game, and correlated to that is that in the base game Legendary difficulty has frustratingly dubious tactical tuning, particularly if you've got the Alien Hunters DLC. (Base-game Legendary difficulty is borderline-unplayable with Alien Rulers running around) War of the Chosen admittedly seems to be tuned with more of an eye toward Legendary difficulty in specific, as I noted with Resistance Order design, but that doesn't undermine the point that the base game clearly had Commander designed as the baseline 'true' experience and yet insisted on secretly cheating for the player anyway.

On a less irritating note, a curious point is that there's entries for having Lost and ADVENT/Alien forces have secret modifiers to their hit chances against each other... which are entirely unused. Strange.

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

Overview done, next time we get started on specific enemies with the ADVENT Trooper.

See you then.

Comments

  1. I’m enjoying these posts! I read the ones on XCOM EU/EW, and now the ones about XCOM 2. I noticed you were overall kinda negative on the previous instalment, but more positive about this one.
    What do you think about the pod system, compared to the more free form deployment of aliens from the old games? In this post, you praise the improvements over the pod system in EU/EW, but you don’t compare it to the alternative.

    ReplyDelete
    Replies
    1. Well, I don't compare XCOM 2's pod system to the classic games because XCOM 2's usage of the pod system is purposeful in its own design in a manner that makes it difficult to make a useful comparison. XCOM 2 is very fundamentally designed around the pod system, in both underlying code terms (eg the pod leader/follower mechanics providing a mechanism for putting reasonable limits on mission makeup while still varying encounters) and in terms of many of the design space decisions. (eg 'boss' encounters heavily preferring to hover nearby the objective, in missions that have a clearly-defined objective location, or the Forge's Sectopod being specifically by itself to make for an easier first Sectopod encounter) And more precisely XCOM 2 is trying to be a fundamentally different kind of game than the classic X-COMs, such that comparing XCOM 2's pod system to classic X-COM's lack of a pod system is a bit of an apples-to-oranges comparison.

      Which is a good example of why I'm negative on EU/EW and positive on XCOM 2: many aspects of EU/EW are, on a design level, the devs attempting to broadly replicate the classic X-COM experience, but with a bunch of changes that aren't for the better and don't stay true to that experience. XCOM 2, though it has some cute shoutouts to Terror From the Deep, isn't trying to cleave to classic X-COM as its design target, and instead looks at what EU/EW's mechanics support and refines that into something clearly inspired by classic X-COM but still very much its own thing. The pod system in EU/EW was a bad mechanic, and if you'd asked me how I felt about it back before playing XCOM 2 I'd have vehemently insisted it should never have been allowed to exist. The pod system in XCOM 2 is a good mechanic, partially because of internal improvements to it (mixed pods as standard, for example), but mostly because XCOM 2 is designed so the pod system is essential to its design, where in EU/EW the pod system is tacked on and actively harmful to the kind of experience EU/EW was attempting to produce; the tension one experiences during Alien Activity in classic X-COM simply doesn't exist in EU/EW because pod activation mechanics mean you're in zero danger if you haven't already activated a pod. (Aside that one glitchy Cyberdisc in Slingshot...)

      That said, I do overall prefer the freeform deployment of classic X-COM. XCOM 2 is a very good, tightly-designed experience, but I would've liked to have seen something a bit more like classic X-COM. Still hoping Phoenix Point is good when I get to it.

      Delete
    2. That makes sense. Personally, I appreciate that the pod activation mechanic reduces the "swinginess" of the game, since you almost always get a chance to act before the aliens do. Given the huge difference between a Rookie and a Colonel, getting a soldier sniped out of nowhere would feel pretty bad. But I dislike how it turns uncovering the fog of war into a negative (which feels *so* wrong to me), and that it allows the game to be balanced around the idea that the player never gets shot at (since you can kill everything before it gets a chance to do anything).

      I have to assume that Chimera Squad is Firaxis admitting that there are issues with the mechanic, and experimenting with ways around them. I'm curious what XCOM 3 will look like!

      Delete
    3. Chimera Squad was handled by a different team, with minimal overlap with the other two games' teams, and was pretty explicitly a bold experiment. I won't be surprised if XCOM 3 pulls some value from it, but far more likely is that it will be more a continuation of XCOM 2-style design.

      Pod activation protecting your elites is important, but mostly because the games force you to operate at such a small scale. Losing a soldier or two out of nowhere is a survivable issue in classic X-COM because your squad size starts from 14 and goes up from there, where in the Firaxis games 2 soldiers is half your starting squad and a third your maximum squad size. Which itself is an example of why I look at XCOM 2 very differently from its predecessor; the prior game clearly intends and expects regular casualties like classic X-COM, but is set up so that's tactically insane. XCOM 2 shifts to a more appropriate 'if more than one guy does down in an entire mission, it's almost certainly something you could have avoided by playing better, and even losing one guy is probably still your fault'; much more fitting to a squad size of six.

      But yeah, I don't like how pod activation actively punishes advancement. It at least makes slightly more thematic sense in XCOM 2; you're ostensibly sneaking around, after all... but even in XCOM 2 this is a flaw with the design, exemplified by how Rangers being defined by their Swords falls away as pod counts rise to the point the game struggles to space them out adequately. A Ranger somehow implemented into the Gollop game's engine would be in severe danger, but wouldn't be endangering everyone else by doing their job!

      Delete

Post a Comment

Popular Posts