At the risk of this piece turning more into a reflection on myself than on Obsidian Entertainment’s Pillars of Eternity II: Deadfire, I should note that I am one of the many people who is completely unable to play Baldur’s Gate 2 anymore. This does not necessarily reflect poorly on Baldur’s Gate 2.

That game was an icon of its time. It combined the Real Time with Pause mechanic first introduced in the entirely forgettable Icewind Dale games with immersive and effective storyline crafting. The original Baldur’s Gate wasn’t bad either, but it wasn’t Baldur’s Gate 2: Shadows of Amn. Part of this was a conceit of the system: you were playing a low-level character in the first Baldur’s Gate. The entire game system was built off of a modified version of the Advanced Dungeons & Dragons tabletop ruleset, which leads to fun observations: the “level cap” in the first Baldur’s Gate is actually an XP total (16100 for the Enhanced Edition available on Steam), such that you could only reach somewhere between level 8 and level 10 in the first game. Baldur’s Gate 2 let you import that character and start wherever you had left off, or create a new character expressly for this game. Then you had to spend an hour in the introductory dungeon, as the elf mage Jon Irenicus explains why he has captured you and your companions, and why he is torturing and murdering them.

I can no longer play this introductory dungeon. I know too much about it now, from the sneaking bits with the traps to the djinn in the upper right hand of the map to the distressing harem of wood nymphs that Irenicus keeps and abuses because he’s long since lost the ability to understand his own feelings. I’ve spent probably over 100 hours in that dungeon since age 13, and there’s no way I can ever make it through it again — especially not while playing Advanced Dungeons & Dragons.

This brings us roundabout to Pillars of Eternity II by way of lineage: the Pillars series is directed by Josh Sawyer, easily one of the best design leads in the business. He was a designer on all of the original Icewind Dale games as well as Neverwinter Nights 2 (and he was lead on the two best “shooters” of their time, Alpha Protocol and Fallout: New Vegas), so he knows more than a little bit about the sort of game he’s making here — enough to know that D&D is no longer the proper system to be setting as the baseline. Something like it, though — that is something people have been striving for: a system with a similar cultural resonance to Dungeons & Dragons without most of the terrible play inefficiencies of the system (Wizards are too good; Fighters are laughable trap options, and you’re likely best off playing a Bard). Sawyer and his team mostly succeeded with the first Pillars game, but it’s the second game that shows the true power of a good, experienced tabletop design team combined with the power of the modern computer.

This comes with the admission that I actively dislike the Real Time with Pause mechanic for RPG combat. It’s not great! The best thing about it is that its status as a compromise between a turn-based system and a full-on active battle system means that, should you be of a high enough level to disdain the system entirely and wish to do so, beating any encounter is as easy as pressing the button to unpause the game and watching your little minions work. This means you do not have to pause. That’s kind of important for a system whose selling point is “Real Time with Pause.”

2nd level paladin Jordan of Crane and the remainder of his Dyrwoodan crew, shipwrecked shortly after their arrival in the Deadfire Archipelago.

Deadfire understands this. It gives you a magnificent suite of active powers, passive abilities, and most importantly a framework of command instructions that builds very well on the gambit system developed by Final Fantasy 12 — a system that was not available in the first Pillars of Eternity game.

Real Time with Pause has always been something of a compromise of a system: before Deadfire, the idea was that you, the player, would run through the isometric game world with your party of characters until you ran into a combat, and when you ran into a combat you would choose to either allow the characters in your party to attempt to win it through blunt force or you would manage the characters to victory by choosing how they would use their powers and abilities. There were always simplistic AI settings that allowed you to set if, for example, your Paladin or your Priest would use offensive abilities or support abilities, but you could never rely on those settings to actually represent a response to the world that the game presented — this was true from Icewind Dale to Baldur’s Gate to Neverwinter Nights. Either your characters outclassed your foes to such an extent that it didn’t matter how you played them, or you had to dig in and play the fight — and when you played the fight, you generally found yourself wishing that you were actually playing a turn-based game rather than one based on the nebulous premise that you wanted to buy into a turn-based system, but weren’t willing to sit through all of the math.

For years, the best benefit of the Real Time with Pause approach to an RPG framework has been that at some point — at the end of the day, and when you have achieved enough system mastery — you get to ignore it. Combat begins, and you no longer have to press pause: the game runs itself, and you win. If that sounds scandalous, consider how many combats you have to play in the average CRPG; once you’ve demonstrated system mastery — whether through character build or succeeding at much more important non-combat questing — it’s not actually a good thing if you, as a player, are forced to spend five to seven minutes resolving one of the more basic fights in the game. It is actually a good thing if your party just runs out there, does a slaughter, and resolves the situation. Contrast this with strict turn-based systems: even if you’re level 20, and your foe is level 1, if a fight begins you must resolve it through a turn-based system unless the game gives you an auto-resolve. And auto-resolves have their own complications, especially if there is a chance of failure. Instead, Real Time with Pause allows you to just watch for 30 seconds as the enemies explode while the game speeds through its automated progressions.

In the first Pillars of Eternity game, this sort of auto-combat could flip from merely boring to actively punishing in a matter of seconds — the AI scripting was generic, baked into a couple of unhelpful “stances” such as Aggressive, Defensive, or Support, with occasionally class-specific behaviors such as Crowd Control for Wizards and Shapeshifting for Druids. Pillars of Eternity II opens up this auto-combat programming into a full suite of situational, if/then statement behaviors. This is hardly innovation qua innovation; among others, Dragon Age: Origins was a real time with pause game in this same genre space that had this sort of scripting in a 2009 release, while Final Fantasy 12’s gambit system was one of the first major forays into player-scripted party AI back in 2006. But given the conservatism of the CRPG landscape — for all their quality-of-life improvements and pretty effects, it’s unavoidable to see the Pillars games as anything but a love-letter and spiritual successor to a kind of game that hasn’t dominated the RPG marketplace in almost two decades — it’s important that if it is necessary some percentage of the time for the game to play itself, it should feel engaging and good for the player as the game plays itself. The way to do that, of course, is to give the player maximal control over how the game plays itself, and at this Pillars of Eternity II is an unqualified success.

6th level Brawler (fighter/monk) End of the Word and her party in the slums of The Gullet, a thin lattice of tunnel-connecting catwalks overlooking Neketaka’s ruined Old City.

This is all so important because more than just about any other video game genre — and certainly sooner than them — the computer games that trace their lineage from Dungeons & Dragons recognize that the only reason the system exists is to mediate the real attraction: the collaborative story being told. Tabletop D&D might have its roots in competitive wargaming, but the reason the system was able to make the leap from one offbeat product in a niche hobby into the common ancestor of an entire genre of tabletop gaming was how, through the rules of the system and with the guidance of a central, reactive and (hopefully) intelligent authority — the Dungeonmaster — friends could get together and tell stories in ways that they couldn’t before. Sure, in the beginning those stories were generally constrained to “how well did we loot this elven ruin, and how many of us survived?” But over the years, the rules in both D&D and the tabletop RPGs that came into being in conversation with it have grown to support so many more kinds of stories than these; even stories without fighting in them, if you so chose. And with the mainstreaming of the RPG thanks to videogames, the expectations of quality for these stories has risen as well.

No matter how good the Real Time with Pause system might be, it is wasted if it only exists in service of a poor narrative. How does the story of Pillars of Eternity II stack up then, both against its own expectations as well as against its spiritual predecessors? Very, very well…with caveats.

There are three main components to examine in a modern CRPG’s story, and they’ll be fairly familiar to anyone who has ever, say, read a novel. The first is the plot itself — how the game moves from A to B to C, the freedom it gives its player in making those moves, and the core competency it shows in handling them. The second is the larger thematic narrative — is there a coherent theme to the work? Do both the setting and the things that happen in that setting reinforce that theme, and is the work’s argument convincing on its merits? The final component is characterization; while novels and films might get away with having a small cast when appropriate — occasionally, casts of only one — games tend to have a wide array of non-player characters because the point of a game is interactivity, and in a story, interactivity requires agencies other than those of the player character to be present.

The plot of Deadfire starts out rather simple: you are the Watcher of Caed Nua, who can see and shepherd the souls of the dead through the great Wheel of soul reincarnation. After proving victorious in the first game, you retire to the manor estate you spent the game building up from a forgotten ruin into a minor local power in the region of the world of Eora called the Dyrwood. Then one day, the god Eothas — the god of light and a savior figure in the Eoran cosmology — incarnates himself inside the body of a mighty colossus buried beneath your manor. He stands up, kills you, kills all your subjects, and tromps out to sea. The god of death, Berath, plucks you from the Wheel rather than let your soul reincarnate, and sends you back to the land of the living to track down this renegade god and…well, Berath — and you — will figure that part out later.

The game is set in the Deadfire Archipelago, a loose collection of island chains that resemble little so much as they do a fantasy Caribbean during the age of exploration. Our age of exploration, however, was also the age of colonialism — and the age of pirates. The easiest way to set this stage, then, would be to have factions in the game that represent the historical colonial powers that actually conquered the Caribbean and the New World; one each for the Dutch, the British, the Spanish, and the French. Deadfire takes a slightly more nuanced approach. The first of the two principal colonizing powers are the Vailian Republics, which players of the first game will recognize as the not-quite-Italian-speaking federation of merchant duchies that, yes, invoke the merchant city-states of Italy, but more strongly in organization and economic power, the Dutch Republic which predated the modern Kingdom of the Netherlands. The second colonizing power is the Great Kingdom of Rauatai, a militaristic naval power whose fleet and fort designs might recall Britain and France, but whose politics more clearly match Imperial Japan of the Shōwa period: a rising traditionalist, xenophobic faction led by the heir to the throne has begun to promote the racial superiority of Rauatai’s aumaua majority (aumaua are basically sea orcs) and calls for further expansion of the Great Kingdom.

Into the Deadfire Archipelago these two powers send their Vailian Trading Company and Royal Deadfire Company respectively to oversee their colonial ambitions, and these companies are what they sound like — the Eoran versions of the Dutch East India Company and British East India Company, with all the slaving, violence, land theft, and general misery that entails. Caught between these two powers are the native Huana people of the Deadfire, a loose confederation of islander tribes that take inspiration from indigenous Polynesian and Maori cultures. And then, of course, there are the pirates, who have their own rough governing body in the Príncipi sen Patrena. This four-way cold-but-warming war is what Eothas stomps into in his massive new body, terrifying all the major players and opening room for you, the Watcher, to work.

If you’re looking for heroes, or a clearly correct faction to support, you won’t find it in the Deadfire. The Vailians and Rauatai hate each other with a contempt that can only come from seeing themselves mirrored in their enemy, and despite the roguish airs put on in the Principi court in Dunnage, the pirates are, well, pirates: aimless slavers and killers loyal only to their plunder. The Huana, whom you might expect to be the most sympathetic faction, can be the most infuriating: these islands are “theirs” only by dint of their own conquest, and they continue to enslave, murder, and even eat the tribal fishpeople and snakepeople whose lands they’ve taken. It is not a particularly new tack for a fantasy game to play with the distinction between who gets to be a person and who is labelled a monster, but Deadfire does it well and with subtlety. The Lagufaeth and Naga are sentient, social creatures with their own culture and history in the Archipelago, but they are not recognized as one of the enlightened kith races, and are therefore treated like even less than barbarians — as far as the kith are concerned, they’re just particularly dangerous wild animals.

The lack of a clear external cause to fight for in the Deadfire means that the Watcher, then, is fighting for him or herself. The cast of characters circling around the Watcher underlines this in the way that CRPG party members usually do, as one by one they’re yanked out of their orbits by the gravity of the player-character’s protagonism. Three of the seven main party members return from the first game — Eder, Aloth, and Pallegina — while the remaining four are new. There’s an art to designing a game’s companion characters that has more or less been codified in single-protagonist, party-based RPGs over the past 20 years: you want one companion character for each general perspective or approach that the player can take towards the plot of your story, you want them each to have a personal link towards your player’s character stronger than any they have towards each other, and you want to cram them all together in an enclosed space until the sparks start to fly. Not only does this approach creates a bunch of interpersonal drama to soak up the player’s time (Mass Effect 2, for instance) but at its best it externalizes the conflicts that the player and their player-character experience when they try to answer the game’s internal philosophical or political questions (Knights of the Old Republic II). Deadfire isn’t quite KOTOR II, but it executes this formula very well: the Watcher has two friends from the old days in Eder and Aloth, then Pallegina representing the Vailians, the orlan rogue Serafen stepping in for the pirates, the marine godlike Tekehu speaking for the Huana, the aumaua sharpshooter and spy Maia working for the Rauatai, and the naive young priestess of Gaun, Xoti, standing for the setting’s newest and fastest-growing faction — the followers of Eothas, the guy currently upsetting everyone’s pots and pans.

13th level transcendent (monk/cipher) Stranger Saw God, his party, and the tiny cat Gosha in the court of the Principi at Balefire Beacon.

Deadfire plays with this conceit, though, and tries to step away from the idea of your companions as different pieces of the main character arguing with each other. In addition to tracking how each of your companions feels about your Watcher, the game also tracks how they feel about each other, positively or negatively; if a relationship gets too bad, the two companions will have it out on the deck of your ship, and if your Watcher isn’t careful, one might end up leaving the party for good. If a relationship gets too good, on the other hand…

Now is probably a good time to mention three things about Deadfires’ supporting cast: most of them are dateable now, they’re either mostly or entirely bisexual (Pallegina seems like she only prefers women) and they’re incredibly, incredibly horny, especially the newcomers. Just about every figurative expression out of Serafen’s mouth is about dicks; Tekehu propositions you for sex within a few in-game hours of meeting you; Maia is only a bit more reserved — she insists you get to know her animal companion first; Xoti makes it clear she’s near-terminally thirsty for Eder before you even get out of her introductory scene. In most RPGs with dating-sim elements like this, this means that you, the player character, get to choose the person you want to sleep with out of a pool of your companions, and the companions you don’t pick just remain permanently single and available in case you change your mind. In Deadfire, if you decline to pursue a relationship with Maia, there’s a fairly good chance she’ll be sharing Xoti’s bed instead in a couple of weeks.

Regardless of how you feel about the odds of your full party being bisexual — given the sexual politics of the world of Eora, it seems quite plausible — this solution is much more elegant than having the companions’ orientations shift to make them attracted to whatever the player character’s gender is, and most everyone being willing to sleep with most everyone keeps the possibility of other interparty romances open no matter who the player chooses to woo (if anyone). But again: incredibly horny.

These three story components come together to form a doozy of an experience, and fans of the continuing plot will be glad to know that Deadfire is just as concerned with the goings-on of Eora’s ever-bickering pantheon of gods as the first game was. Deadfire earns its ending, but to say more would constitute spoilers.

20th level rogue Sion at the helm of the galleon Fear Has Wings, preparing to chase down the walking god Eothas one final time. Yes, that is a Xaurip crewmember in the lower left.

The game isn’t perfect, of course; as fun as running around the ocean map is on your fully-customizable privateer ship, ship combat itself is a clunky, tedious subsystem that recalls the frustrating plot duels from Suikoden II: not quite a stats check and not quite combat, it unsatisfyingly splits the difference with a heavy dose of rock-paper-scissors. Even once you figure out how to win ship combats you’re still probably going to find yourself immediately sailing right up to the enemy ship and boarding it so you can play the game with the combat system they spent hundreds of hours of development time on.

Fast-traveling is much improved in terms of going to places — you can now choose to warp inside notable buildings in the game’s cities rather than having to load the district’s map, walk to the building, and go through another loading screen to the building’s map — but you still must Gather Your Party Before Venturing Forth when you want to leave; at this point it might be time to retire that and just allow access to the map from any point in a district.

And there are some bugs in the most important part of the game’s systems: the quest progression. Progressing too far in the main quest can lock you out of side content, and the old maxim of CRPGs remains true: do all the optional content before any of the main content, and you’ll be fine. Hopefully this will be patched eventually; patches have already addressed many of the other balance issues in the game, such as level scaling on release not quite doing the job.

If a wonky new subsystem and some fast travel gripes are the biggest design problems your CRPG has, however, you’ve done an excellent job. I put some 300 hours into the first Pillars of Eternity; I’m around 90 on Deadfire now, and see myself playing through it again a time or two even before any DLC or expansions are released. I suspect by the time I’m truly done with the Pillars of Eternity titles, I’ll be able to bestow upon them the highest honor I can give a game like this: always skipping the damn intro so I can get right to the good part, where the world opens up and the story stretches out in front of me.

Liked it? Take a second to support No-Cartridge on Patreon!