Pillars of Eternity

Modified slightly for Patch 1.05.

Few game designers really appreciate the power that systems wield in influencing the behavior of players, even when those players are cognizant–and resentful–of the coercion. The people at Obsidian have got this more or less figured out by now, which accounts for some pretty smart design in Pillars. There’s no “dump stat” and no stat that gives you extra skill points to spend. Experience points aren’t rewarded per kill, but are strictly measured out, and usually received by finishing quests, which helps players roleplay without seeing each NPC as a cut of meat. It also seems to be understood that randomization plus quickload leads to the worst incentives of all, which is why traps aren’t disarmed by a roll of the dice, and won’t explode when the player tries. The stealth doesn’t leave much up to chance, either: it’s not thorough enough to be what I’ve previously described as the best way to handle stealth in a CRPG, but the genre has never really done it well, and as systems go, it’s a good start.

The camera control seen in Wasteland 2 has been traded for beautiful environments designed from one angle, which works so well in Pillars’ favor that it’s sometimes hard to believe that the two games were both made in Unity. Combat is tactical and satisfying, and the choose-your-own-adventure-esque D&D skill challenges might just be the coolest thing in the game.

Whoops, this is a screenshot from the 1994 game Realmz. How did this get in here?

Whoops, this is a screenshot from the 1994 game Realmz. How did this get in here?

Therrre we go.

Therrre we go.

But there are also a hundred things that probably could’ve been handled better. The game raised something like 4 million dollars, which is great, but probably less than the Baldur’s Gate games. A lot of work clearly went into the stronghold, but still ultimately feels shallow, even if it’s at least a lot better than Norende–it doesn’t have the visual or mechanical customization of something like Breath of Fire 2’s Township, or the reward factor of Tales of Symphonia’s Luin. The game has a lot of great voice acting, but it’s marred by technical problems, sometimes cutting out, sometimes poorly compressed, with raspy sibilants that make me want to grind my teeth. And while crowdfunding has undoubtedly done great things for the revitalization of this and other genres, it sometimes seems like gamers might be the worst publishers of all. Publishers wouldn’t veto a working 5-attribute system just because their favorite games have always traditionally had 6. They also wouldn’t make hundreds of thousands of dollars conditional upon the inclusion of hundreds of fanfictionish characters. Obsidian made it work, but I have a number of notes.

Combat and abilities

Combat is cool, but there are some issues with pathing, particularly with characters getting in each other’s ways in narrow passages. In time I got around these issues by not casting spells where I saw the “footsteps” icon, instead manually moving my spellcasters into range, and by luring enemies into choke points instead of starting battles with my own party halfway through. But it would have been nice to have an autopause setting that kicked in whenever a character gave up trying to path to some place and just went idle.

Some skills felt largely overpowered in the ongoing balancing of the game post-release, and I relied a lot on Firebug, petrification spells, and confusion at various times. I wished that the charm category of spells worked a little differently, as it would totally convert the charmed character to the other team as far as systems were concerned, which was really irritating for micromanagement. I wouldn’t be able to use “target ally” skills, even ones that reduced the duration of ongoing domination spells. And when the characters still ostensibly under my control would deviate to attack a confused ally, I had to wonder which of them was truly confused.

It may be that rest-based health and magic recovery systems have never been better. Campsites are a consumable item, and the resting bonuses at inns are significant and long-lasting enough to be worth the trouble, which was a nice touch. Conservatism influences tactics more than I think is ideal, but players can go all-out against bosses, enemy parties in bounty quests, and the like. But it’s still not too difficult to get through Normal difficulty by spamming the good spells and resting more than once on the way through some cave or dungeon. In D&D, players are kept from resting as much as they like by a DM who can creatively punish players for taking too long to achieve a goal, and that’s not really replicable here. Maybe you could have a system where once a player reaches the starting point for certain quests–the mouth of a cave where a hostage is held, or the rush to catch up to a villain before they make their escape–players can’t rest more than once until their objective has been completed, or else the quest is failed and can perhaps be retried from the entrance.

But I think cooldowns would have simply worked better than per-rest and per-encounter spells. You could reset cooldowns between battles, so people aren’t encouraged to wait around for a minute before stepping into the next room. You could also not start certain spell cooldowns until a battle begins, so the most destructive abilities only become available in longer-lasting fights.

I liked the way the classes were handled–that you can build some silly muscle-wizard if you want, but more importantly that you can still build a bad muscle-wizard, and consistency is what’s important to get right. But I wished there had been a greater range of class-specific talents, as it doesn’t feel satisfying to level up and only get a small passive boost to damage, defense, or anything else available to any character, at least not when the max level is 12 and the rare level-ups feel like they should be all the more impressive for it. Some of the magic classes have levels where they’ll suddenly gain like nine new spells all at once, but others aren’t so lucky, and it would have been nice to see some more fabulous, game-changing perks on characters like my fighter or paladin.


There are a couple things in the UI that are really nice. For instance, I wish Wasteland 2 had the two tabs for the messages window, and had put the whole party’s inventory on a single page with no tedious space management the way Pillars does. But there are a number of problems, too. Choosing powers off the bar in combat can be iffy, both by hotkeys or mouse, and sometimes I just had to escape out of a spell placement, click the little Cancel button, and reselect a spell to unambiguously know what my character was doing.

Here’s a few other things I’d change:

  • Taking something back out of a shop’s sell window puts it in my party inventory, not back in the stash I sold it from.
  • Special properties of rare items are often communicated poorly and need to be googled to see what they do.
  • I can’t view a character’s existing abilities/equipment from the level-up screen, which often inform the talents and powers I want to take.
  • Suppressed effects from other equipped items aren’t immediately conveyed, so any equipment change requires a lot of trial and error.
  • The character sheet could make better use of space. I’d rather not have to scroll around so much to see how many points a character has in athletics, etc.
  • A character’s power bar hotkeys aren’t remembered if they’re removed from the party.
  • Chants were getting reset for me, or losing their names, so either there’s a bug somewhere, or the chants window can be closed without applying changes, in some unintuitive way.
  • Level-up and “party member wants to talk” icons on character portraits are hard to notice, as is the red-handed cursor if I would get in trouble for stealing from a certain container (I was many hours into the game before I caught on to the cursor thing).
  • The Notes tab doesn’t let me click to a point I want to type from. There also doesn’t seem to be a way to delete notes.
  • Some of the default keybindings are a little bizarre (I ended up rebinding the time controls to the thumb buttons on my mouse so I could pan the camera with WASD).

Other mechanics and design

The way the player character can gain a reputation for certain behavioral traits, like honesty or cruelty, is pretty cool, especially for players who roleplay certain ideologies, doing a no-stealing playthrough and so on. Also neat are the ways the paladin & priest oaths to their gods and orders can mean consequences for deviating from these behavior roles, in the form of passive character buffs and the like. It reminds me of the way Planescape: Torment let you make certain vows in conversation, and I’ve always had an interest in systems like that: the gamble of making promises when you don’t know precisely what you’re walking into.

Reputation also exists in terms of a faction’s opinions of the protagonist, like in Fallout: New Vegas before it. But this part of the system feels somewhat toothless and underutilized. For example, being the Champion of Defiance Bay didn’t let me get into the animancy hearings on the weight of my own name. I don’t actually want there to be an unequivocally superior “I Am The Best” solution for every problem, but if it doesn’t pay out at moments like that, there’s little point in even having your character sheet tell you that the people of a certain city think you’re the bee’s knees. I seem to recall that in New Vegas, reputation could mean groups like the Powder Gangers not shooting the player on sight. That doesn’t really seem to be a factor here.

I felt that it was encouraged to play too much of the game in Scouting Mode, which made it take longer to get around, which also just meant playing the game in Fast Mode arbitrarily. I didn’t cover every square inch of forest with Scouting on, and a lot of traps and secrets were in places where you’d expect to find them, but there were always the few exceptions. I think the idea of slowing down was somewhat unnecessary. Maybe they could have made Scouting necessary to find traps, but allowed other secrets to be revealed simply by passing the perception check, no matter what mode the player was in. Or maybe Scouting could’ve worked at regular speed, at the cost of faster fatigue generation, as a result of characters having to concentrate harder.

The stronghold’s use of Turns for some events, which advanced as quests were finished, and Days for others, which advanced when the player rested or travelled, which seemed a bit strange. I liked the idea of locking some aspects of progress behind quest completion and thus not allowing the player to earn infinite money by quick-travelling around, but it seemed strange that they didn’t go all the way with it. It probably would have been better if the number of building improvements were locked behind progress in the main questline, so that every time a major quest was finished, the player could build or customize two or three more things, and by the time they were 75-90% done in the main questline, the stronghold would be completely restored.

Rather than the stronghold’s focus on fiscal management and hirelings, shallow or not, what I really wanted to see was utility benefits. Thankfully, Patch 1.05 apparently allows you to access the stronghold resting screen from the campfire button, which reduces the amount of loading screens and time spent running along the same maps, but even then you aren’t getting as good a resting buff as you could find in the inns of cities. If you could solve a quest for a city’s innkeeper, or otherwise pay or persuade them to share their feng shui secrets (or whatever it is that makes their rooms so good), that would’ve been cool. Or hiring on some kind of teleportation wizard who can get you to a distant quest hub in less than an hour, so you can make the most of those resting bonuses.

Most of the game’s quests flowed pretty well, but a handful lacked satisfactory choices or player representation, resolved in unclear ways, or locked you into some hit to your reputation in the area.

Writing and atmosphere

I really was reminded of some old-school gameplay experiences while traipsing deep underground, getting assaulted by spiders, and being charmed/dominated by giant plants every goddam second. The Endless Paths was one of the more fun things to take on.

I did already allude to the “vibrant soul” NPCs, but if nothing else, their overly fabulous lives do help to build a world that’s a lot bigger than the player’s quest. That said, you’re still pretty much the only person on the task of solving the world’s biggest problem, and at the end of the day all these godlikes are just loitering in town. It was an awkward situation, and Obsidian tried to make the most of it. Patronage has, historically, made the world turn. To their credit, these soul-vignettes could be pretty interesting–it just wasn’t where I would ideally have wanted so much of the game’s writing work to be focused. It was a struggle sometimes when I’d walk into a room and see a half-dozen more of those NPCs, knowing I was going to read them all even though I didn’t really want to.

High reactivity and a specifically “game”-unique story aren’t really on the table here, but I don’t know anybody who went into this expecting different. The story does pontificate about some Big Questions in its endgame, and the player gets to participate in the choices made, and I always go in for that sort of thing. But for the most part, it’s not a particularly engaging narrative, apart from some moments where the player’s purpose is clear and the villain is showing what he’s really capable of (as in the sanitarium). The central hook of the plot, while good, didn’t stop the story from stumbling or getting a little dull at times, and a few companions had the Chris Avellone effect of showing off some cool lore while lacking a human touch, or else requiring a little too much patience to find it. But there were also characters like Aloth and especially Eder, who were easy to relate to from the start (and had some great voice acting).

I talked to a few too many representatives of factions outside of White People City that gave me exposition in the form of “My tribe believes that souls are…” instead of just “Souls are…”

The expository dialogue trees were too much, especially Durance’s, and my heart sank whenever I saw multiple “I want to know about…” conversation options that each lead to a half-dozen sub questions. Apart from one lore enthusiast in the Dyrford inn, whom other characters directed me to if I had questions–and I knew exactly what I was getting into when I started talking to him–it tended to feel weird and videogamey to get so much out of strangers. It’s typical of the genre, but I would have liked to see more conversations where the train just keeps rolling while you pick only one ever from a list of responses at each juncture. That is, until the NPC says “Any questions?” The Exposition Guy in the inn was alright because he wasn’t there to tell me critical info, just bonus stuff about the history of the town for nerds to nerd out over. But for the average NPC, minor and major, the sense that “Less is more” could’ve been better realized.

Hopes for the future, closing thoughts

Overall, I think Pillars of Eternity is an promising milestone in the genre, particularly as an important step in exploring how we can continue to introduce modern design sensibilities to the classic CRPG. It meets expectations in most respects, surpassing some and falling short in others, but I wouldn’t call it the apex of anything, as it stands now. I hope its inevitable expansion content can move things in a good direction, and I’d be pretty sad if the upcoming Torment: Tides of Numenera couldn’t learn at least a few lessons from Pillars. In either case I’m thrilled that Obsidian has released something that is being received well and that they have tools and a strong intellectual property to call their own.

Pillars succeeds most at combat and systems design, and is tactically interesting. It might have come closer to something truly eternal if it had been built as a great game first and foremost, rather than an Infinity Engine throwback above all else. Also if it raised ten billion more dollars–I mean, it’s your fault, really.

This game was thoroughly enjoyed by the reviewer. It is an excellent game that may be too simple or not ambitious enough to be a 5, or there are design flaws meaningful enough to prevent it from enduring as something truly beloved. Highly recommended.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s