Thoughts: Pillars Of Eternity 2 – Deadfire


I must admit to approaching Pillars of Eternity 2: Deadfire with an unaccustomed mix of resentment and resignation. It’s a feeling that reflects Obsidian’s fall from grace; after Pillars 1 and Tyranny, both extremely flawed games with only the Obsidian-brand reputation and faction systems to really make them stand out, in my eyes they’re no longer the accomplished masters of quest and mission narrative who came up with Mask of the Betrayer, Fallout: New Vegas and Alpha Protocol. Partially this is to do with other RPG developers raising their game, but there’s very little separating Tyranny from something like Torment: Tides of Numenera and Obsidian games have to do rather more to sell themselves to me these days. I had all but ignored the Deadfire crowdfunding campaign, was not really up for another game set in the rather tedious Pillars world, and only really bought it because I was done with BattleTech and Thrones of Britannia really wasn’t grabbing me.

I was therefore quite surprised to discover that Deadfire is… well, not great exactly, certainly it didn’t enchant me the way as Original Sin 2 did last year, but it is a game that, for the most part, represents a very respectable return to form for Obsidian. They’ve achieved this through first having the self-awareness to recognise what the first Pillars didn’t do so well1, and then coming up with a single (albeit very major) change to the game setup that addresses the majority of them in one fell swoop: you’re no longer some characterless bum mooching around a dreary, samey set of fantasy cities and villages doing fetch quests for entirely forgettable NPCs and organisations, but instead are captain of a pirate ship sailing around a fantasy analogue of the Caribbean (or the Pacific Islands, it could go either way) and dealing with interfactional squabbling between various colonial groups, including an alternate version of the East India Company.


Swapping out your pre-existing world and gameplay mechanics for a more piratey incarnation is not quite the entirely innovative change it would have been a few years ago, as Assassin’s Creed did almost exactly the same thing with Black Flag to wondrous effect. As Deadfire isn’t doing full 3D pirate ship shenanigans — instead choosing to engage in a much more Pillars-esque (not to mention cheaper) method of portraying its ship-to-ship engagements via a dedicated turn-based system inside the existing text adventure interface it uses for its narrative vignettes — it has a ready-made blueprint to turn to for the other bits: Sid Meier’s Pirates!. The Deadfire overworld plays out in a very similar fashion, with you gazing down upon your ship as you sail it around the Deadfire archipelago in real-time. If you sail into another ship you’ll start one of the text-based ship battles, and if you find somewhere to land on one of the many, many islands scattered around the archipelago you can disembark to explore it on foot. You have a crew of anywhere between 5 and 20, each of whom has their own (basic) set of skills, such as Expert Cannoneer, and each of whom can be assigned to man specific stations on your boat matching their skillset. Your crew will also unfortunately require food and drink, as well as a nominal amount of wages, and if their morale gets too low (if you make them spend a month eating nothing but hardtack, say) they’ll eventually mutiny and try and take over the ship. You can upgrade your ship’s cannons and sails and also buy bigger, better models of ship that can both take and dish out more punishment. While you’re sailing around the map your crew will sing in-universe sea shanties to pass the time, and some of them aren’t half-bad — this is another idea that’s been purloined from Black Flag, but it worked awesomely well there and is no less effective in Deadfire.

Really, replacing Pillars’ moribund stronghold of Caed Nua with a pirate ship — a tangible thing that you’re constantly staring at and interacting with and which has a huge impact on the game — is by far the biggest improvement Deadfire makes. I have some issues with the economy surrounding privateering, which we’ll get into later, but even with its flaws it’s about a million times more engaging than repeatedly repairing your curtain walls after they get wrecked by bandit groups who are all inexplicably packing siege weaponry. Deadfire even tacitly acknowledges how awful Caed Nua was by having it get smashed to pieces in the opening narrative by the awakening of the giant adra statue buried beneath it. Said statue is now inhabited by the god Eothas, who was supposedly killed in a much more interesting-sounding game that took place twenty years before Pillars 1, which Obsidian inexplicably decided not to make in favour of incessantly reminding you that you weren’t playing it2; Deadfire lets you import your save from Pillars 1 so you have the same player character, and its excuse for starting you back at level 1 again is that Eothas sucked out all of your soul juice and then accidentally stepped on you on the way to wherever it is he’s going. Some godly intervention brings you back from the resulting coma to find that returning companion Eder has put the pair of you on a boat bound for the Deadfire archipelago so that you can chase down Eothas and get your… soul… back…

…yeah, now that I’ve written it down I’ve just realised that this is almost word for word the plot of Pillars 1, except with nobody antagonist Thaos replaced with a titanic, nigh-indestructible god. The difference between Deadfire and Pillars stories lies almost entirely in the execution, as Deadfire handles the scenario far more deftly than Pillars managed, and while I did end up ignoring the main plot for most of the game again, this time it was at least because the not-main plot bits of the game — the sailing, exploring and inadvertent murdering of anyone who happened to be standing within a hundred feet of my quest objective — were such good fun to interact with rather than the main plot being so dreary I couldn’t stand to interact with it unless I absolutely had to. Obsidian have invested in making almost every line of dialogue in the game fully-voiced and I was actually somewhat surprised at how much difference it made to my suspension of disbelief; they’ve also done rather more with the little story vignettes Pillars uses in place of cutscenes than I was really expecting them to. The plot is still somewhat perfunctory in content, relying on its four squabbling factions and their associated sidequests to give it a bit of flavour, and the ending was disappointingly open-ended and resolves next to nothing, but it’s at least been beefed up enough that it just about holds up under the weight of what it’s being asked to do.

Which, as it turns out, is basically “give the player an excuse to be on a pirate ship so that we can work on the parts of the game that are actually going to be good”. I keep referring to it as a pirate ship but you don’t have to play it that way if you don’t want to; you can just use it as a method of getting from point A to point B and not really bother engaging with the ship-to-ship combat at all, although this does mean you’ll have to occasionally flee from bona fide pirates who do embrace that kind of lifestyle. Most of the ships populating Deadfire’s world map are named targets for some form of bounty quest, however, so if you’re a completionist you’ll eventually end up attacking somebody from one of the more law-abiding factions, and it’s a short step from there to indiscriminately slaughtering anyone unlucky enough to run across your path for the experience and loot. There are also land-based bounties available, as well as uncharted islands to explore and even some medium-sized dungeon experiences full of puzzles and traps that aren’t related to any quests, but are nevertheless thrown onto the map to liven the place up a bit. And it really does work; the sailing and the variety of encounters you can run into make Deadfire’s world a hell of a lot more compelling than the one in Pillars.

Still, while the breadth of content is very nice the thing that impresses me most about Deadfire is its freeform structure, in that there’s next to no quest gating present on that content. What I mean by this is… say there’s an ancient underground temple, and there’s an NPC in the world who wants you to fight your way through a couple of dungeons’ worth of nasties to retrieve something from this temple. Usually you would need to talk to the guy first, and he’d give you the Arse of Nebuchadnezzar (or some equally absurdly-named Macguffin artifact) required to open the door; that or the relevant bits of the dungeon simply wouldn’t spawn in until you’d talked to him, locking it off until you’d gone through the games’ content in the order that the developers prescribed. Deadfire doesn’t do that. Nearly every single dungeon, item, and bounty target is present in the world from the moment you get out of the tutorial sequence, whether you have their corresponding quests in your quest log or not. This is a trick that Original Sin 2 also pulled, and it works equally well at making Deadfire’s world feel far more organic than a typical RPG; more than once I ended up doing a quest backwards, completing a dungeon or killing a baddie, getting some loot that was tagged as a quest item, and then later randomly running into an NPC who wanted me to do that dungeon and handing them the quest item with a jaunty “Oh yeah, I did that already” accompanied by a suitably smug expression.

This appeals to me particularly because I enjoy min-maxing my way through games to exhaust the majority of the content as quickly as possible. By that I don’t mean I like out-and-out speedrunning (although there’s already a 26 minute playthrough of Deadfire on Youtube), but rather that I try and experience as much of the game as I can while optimising my path through it. Deadfire is a particularly good game for this; it’s so inefficient to do a bounty, go back to the quest giver, get another bounty, do that bounty, go back to the quest giver… It’s just so much quicker to board every ship you see and hunt down every combat encounter on the various unexplored islands littering Deadfire, and then go to the bounty NPC, dump a giant bag of heads out in front of them, and let them sort through the pile like pick ‘n mix to find the ones that they want. There’s far meatier examples of sequence optimisation than this to be found in the game, and it really made exploration feel worthwhile and useful knowing that I could tick islands off the list without having to worry about having to retrace my steps back there later to do some special quest-related objective.


Speaking of boarding actions, I should spend a paragraph talking about the naval combat — because I don’t really think it works all that well. As stated earlier, it’s done by way of a turn-based text-interface. Each turn is split into three actions, alternating with the opposing ship each time, and your actions consist of stop, sail forward, turn the ship to present a broadside (which can take multiple actions if you’re sailing a big one), brace to reduce injuries from incoming cannonballs, and fire port or starboard cannons. Your ship equipment and crew skill at various tasks affects things like ship speed, turn initiative and cannon accuracy, there are choices to be made in ship outfitting over whether you want cannons that can stand off and engage at long range or more quick-firing variants that are designed for short-range brawling etc. etc., you can fire different types of shot to wreck enemy sails or target their crew… it’s clear that a lot of thought has gone into designing this system, for all that it’s handled entirely via text. It might even be a good system, if I ever chose to engage with it.

Unfortunately that well thought-through system is rather undone by some simple cost-benefit analysis on the macro level. If you choose to fight a ship battle, you will take damage from the enemy ship, even if you’re in a giant galleon and you’re taking on a piddling two-cannon voyager. Most of the damage will go to your hull and there’s also a chance that your crew will be injured and have to spend time recovering. Damage and crew injuries take multiple days to make good, during which time taking on another naval action is riskier, not to mention the fact that you have to rotate any injured crew out of their duty posts to rest. In short, fighting ship-to-ship has a real cost in terms of time lost, and it is far easier (not to mention quicker) just to sail straight at the other ship at full speed, take a single raking broadside that will nevertheless inflict less damage than slugging it out ship-to-ship, and then initiating a boarding action. Boarding actions transition the fight to the decks of both ships and have both crews fight each other to the death using the regular Pillars combat engine — your ship type and equipment have no impact on your ability to win a boarding action, and as long as you’ve built your party halfway sensibly and aren’t taking on a crew that’s several levels higher than you you’ll win all your boarding actions quite handily. Importantly there’s no fallout from this that lasts past the conclusion of the boarding action — unlike the ship battles, any downed crew members get straight back up once it’s done with no lasting consequences.


This renders the ship battles somewhat pointless — you can do them if you want, but the outcome is inferior in every way to simply boarding the enemy ship — and I would be far more annoyed about this if it wasn’t for the fact that the boarding actions are by far the most fun fights in all of Deadfire. Not challenging perhaps, but definitely fun; you’re facing an enemy crew numbering anywhere up to twenty, but any benched companions who aren’t currently on the party roster will pitch in on your side along with a subset of your own crew. Boarding actions therefore end up being massive rucks involving thirty or forty combatants, with people leaping from one ship to another while spells go off everywhere accompanied by an appropriately swashbuckling soundtrack. Some might call them a confusing mess, but I thought they were absolutely glorious.

The last word before I wrap up should probably be about the factions: simply put, the factional rivalry in Deadfire reminded me very pleasantly of Fallout: New Vegas. There are four major factions present in the Deadfire archipelago: the Vailians, who are basically colonial Spain despite being a nation of Italian stereotypes; the Royal Deadfire Company, who are a state-sponsored concern of the Rauritai with a lot of cannons and guns; the native Huana, who are trying to resist the colonisation efforts of the previous two groups; and the pirate Principi faction, who exist to cause a lot of pain for everyone else but who are afflicted with their own internal ructions. All of them want to exploit Deadfire for their own ends and, as usual for RPGs, treat the player character as a sort of combination handyman/exterminator for cleaning up their rivals. Throughout the game they give you side quests to further their interests and work against those of other groups, and these quests inject a huge quantity of flavour into a game that’s otherwise a little light on story meat, fully-voiced dialogue or no. In fact I’d say probably 50% of your time in Deadfire is spent working for the factions, with another 45% exploring and the remaining 5% actually chasing that damn statue across the Deadfire. It’s important that they be well-written and characterful, so it’s a good thing that they are… mostly, anyway; I was a little puzzled that blowing up Vailian ships didn’t piss the Vailians off any (this goes for all factions), and in fact it appears to be impossible to annoy them to the point where they’ll turn hostile, since this is reserved for a rather artificial pick-one-fight-the-others plot mission towards the end of the game.

(As weird as these little structural tics with the factions and story are are there’s at least nothing as lopsided as Caed Nua or the second city in Pillars, both of which were included because they were promised as part of the Kickstarter rather than because anyone thought they were a particularly good idea and which ended up being a waste of development resources. Obsidian are learning here too; I think the most ambitious thing they promised in the Deadfire crowdfunding campaign was to include backer NPCs, who are at least relatively inconsequential and easy to ignore.)

Deadfire’s a dramatic improvement on the relatively humdrum Pillars. Obsidian have kept what worked — it’s the same engine, but prettier; the same combat mechanics, but streamlined; the same reputation system, but overhauled; the same totally broken economy where you store a thousand sets of bloodied armour and weapons in your bottomless stash to fence to merchants later… actually, maybe they should have fixed that — and haven’t been afraid to throw things out and make sweeping changes to try and address the worst of what went wrong in the first game. Yes, these changes aren’t entirely original; we’ve got the obvious inspirations of Pirates! and Black Flag, and I also suspect Original Sin 2 of having had a non-trivial influence on the more freeform philosophy behind Deadfire’s design. Despite that, it still takes talent (and not a small amount of courage) to weld together so many of these new elements that haven’t been tried before in this kind of RPG, and to have them work as well as they do. It’s also unsurprising that some of them (ship combat) have been flawed on the first try, but I’m willing to forgive that because it was the first try; I’m far more annoyed by the story that, despite being well-told this time around, is still frustratingly insubstantial, when a storming narrative is what Obsidian is supposed to be good at. Still, Deadfire at its worst is a far more confident game than Pillars or Tyranny  were at their best, and restores some of my confidence that Obsidian can crank out a good RPG when they want to.

  1. To be fair, most game developers do not set out to make bad games and are painfully aware of their games’ flaws. Few of them get the chance to address them in the way that Deadfire does, however.
  2. I’m still bitter about this, if you can’t tell.
Tagged , , , ,

9 thoughts on “Thoughts: Pillars Of Eternity 2 – Deadfire

  1. ilitarist says:

    Happy that alt-text is conquered again. Now to the review itself.

    Wow! I made a first day purchase anyway – really liked PoE1 with White March and Fallout New Vegas fame still forces me to consider everything Obsidian – but I’m glad you confirmed it is indeed even better. You didn’t say much about combat system but I understand it’s good. Still going to wait for some patches; just recently it payed off greatly with Mass Effect Andromeda: I would never guess it’s not considered the best Mass Effect game if I hadn’t seen reviews and circlejerk hate before I played it after the final patch. I’ve tested if PoE2 works and it suddenly finished intro battle with a pirate 1 second after it started and proceeded to the CYOA thing and it looked like a bug. So I’ll wait for a patch or five, or maybe those season path DLCs.

    My playstyle is similar to yours, it seems, and usually it means that early game is hard but later everything is boring and straightforward so I just run to the end without proper thematically appropriate preparation (only Mass Effect and Dragon Age games seem to make endgame sound menacing so that I tie all loose ends before getting into a final fight – even Andromeda worked for me). Did you play on Path of the Damned, did you use upscale for enemies?

    This looks like I have another RPG I’ll replay several times. First one since FNV, it seems.

    • Hentzau says:

      Played on Normal (or “Classic” which is the middle of the five difficulty levels and is, curiously, one higher than the default one) and deliberately turned upscaling off because I have my fun in these games breaking the levelling power curve, and that’s no fun if the enemies power level adjusts to compensate. I was impressed that they included the option, though.

      The combat is *eventually* fine, although I find I’m becoming less and less a fan of its complexity after playing more elegant and easy to understand systems like BattleTech. Despite having a pretty good grasp of the PoE 1 system, and despite the Deadfire system being mostly the same — they’ve gotten rid of endurance and streamlined wounds and spells — I still found it very overwhelming to start with. Eight different damage types! Four different save types! An armour penetration system that, while mechanically interesting, is entirely hidden from the player! Fortunately they include party AI this time, so at the start of the game I could at least micromanage one party member while letting the rest of them do their own thing.

  2. ilitarist says:

    Also I’ve looked into the comment problem. Making iframe bigger doesn’t work well because it would still make it possible to enter too much text to be visible and it creates a lot of empty space if you answer to somebody’s comment. Comment textarea size is constantly updated by some evil script, so the ugly and dirty solution is adding this to any available css:

    #respond #comment-form-comment textarea, #comment-form-share-text-padder textarea {
    overflow-y: scroll !important;
    height: 200px !important;

    Changes comment section to fixed height and adds a wonderful UI invention: vertical scroll. Seems to work without issues.

    • Hentzau says:

      Nice, I’ll have a play with that this weekend to see if I can get it to work without blowing up the site.

  3. As someone who quite liked the first Pillars, I’m very pleased to see that everyone seems to agree that Deadfire is the superior game. I get the impression that Josh Sawyer, probably most famous for overseeing New Vegas, had a much greater presence in Deadfire than in the first game, which likely explains many of the dramatic changes. Tyranny, meanwhile, was clearly somebody’s pet project and not really on par with Obsidian’s usual output. It had some really good ideas, but the systems were a mess and the game felt padded.

    Sadly, my motherboard died two days after it released, so I barely got to play Deadfire before having to seek out repairs.

    • Hentzau says:

      Josh Sawyer spent an awful lot of time on the combat system in the first game, which perhaps explains why the story structure was more than a little generic. Also Obsidian have always been at their best when taking an existing property and putting a new spin on it — I’m a bit surprised that this applies just as much to their own IP as it does to other people’s, but Alpha Protocol aside that rule seems to be holding up.

      Man, Alpha Protocol was great. I might go and play that again.

  4. Kellogg says:

    The Arse of Nebuchadnezzar? That takes me back.

    • Hentzau says:

      I forget exactly who came up with it — either Kieron Gillen in the Diablo 2 CYOA preview or mad genius Jonathan Nash on one of the PC Gamer back pages — but 20 years later it’s still my go-to for an example Pointless Artifact Name. Plus it means I’ve gotten really really good at spelling the word “Nebuchadnezzar”.

  5. David Glover says:

    “There’s far meatier examples of sequence optimisation than this to be found in the game,”

    Meatier than sack full of severed heads??

    (Also, while I was enjoying PoE, I never finished it. Glad to know that it sounds like I could skip straight to this one for some pirate-y fun.)

Leave a Reply to Kellogg Cancel reply