Pillars of Eternity II: Deadfire

Pillars of Eternity II: Deadfire

Not enough ratings
Most fun POTD MC build guide (built in 5.0 for Trial of Iron)
By BecauseHeLives
This guide is designed to explain a very fun build along with some important core principles, recommended party considerations and some helpful tips I've picked up along the way. In my opinion, a huge part of the fun of CRPGs is the strategy, which includes building, so this guide is more to help you build your own fun and effective character rather than just tell you what to do. Built in 5.0 on POTD difficulty for Trial of Iron.
   
Award
Favorite
Favorited
Unfavorite
Introduction
Hello everyone! Despite being a life-long gamer, this is actually my first guide, so I hope it's helpful!

The reason I'm choosing to create this guide is because I believe that Deadfire is a fantastic game with a ton to enjoy, and I found a particular build that I love alongside some helpful, seemingly hidden tips that I wanted to share.

I've tried a lot of builds in this game. I'm what I would call a "casual blind min-maxer," meaning I strive to theory-craft strong builds without cheesy exploits and without following guides from other people (other than to gain an understanding of core mechanics). So I'll talk here about the best build I've come up with that was able to carry me through my first ever POTD trial of iron run.

Why trial of Iron?
I've always been a fan of strategy games. My journey started with Final Fantasy Tactics back when I was too young to read. To be fair, I played it with the help of an old PlayStation Gameshark, giving me max damage and invincibility so I could get through without needing to read much—but that's beside the point. Over the years, I've grown to love the challenge of a great strategy game, with favorites like XCOM 2, Baldur's Gate 3, Wrath of the Righteous, and Pillars of Eternity II: Deadfire.

As I've played more of these games, I've noticed a recurring pattern: "strategy" games can often turn into "puzzle" games, where solutions become formulaic instead of requiring true critical thinking. That's why I've come to appreciate games with a "Trial of Iron" or similar no-reload mode. These modes add a sense of real consequence to your actions. Without any do-overs, every decision matters. Victories feel that much sweeter, and defeats, though painful, become part of the experience, reminiscent of the classic "game over" screens of old. If a game is well-designed—and aside from difficulties meant purely to make you suffer, like Wrath of the Righteous on Unfair mode—careful planning of your build and party should allow you to overcome even the harshest RNG.

However, no-reload modes also come with their own risks, like game-breaking bugs or sudden, unforeseen spikes in difficulty if you're not thoroughly familiar with the game. That's why I only recommend them for stable games with a smooth difficulty curve. Among such games, Deadfire stands out as a solid example.

All that said, the game is meant to be fun, so do what sounds best to you!
Party composition
For flavor, I'm choosing to run with only full companions. I really like the "sidekicks" and wish they had been fully flushed out, especially because some of them are quite strong, but oh well.

Regarding the actual composition, I like to create a party that has two dedicated front liners, one dedicated back liner, and two companions who will primarily stay back but can absorb some pressure if needed.

For this, I typically employ Eder (swashbuckler) and Pallegina (herald) in the front line. There is some flexibility on both of these companions, but I like the tank primary (fighter/paladin) and the niche secondary (rogue/chanter) for these companions, as I feel the classes synergize well for the roles I have them play. Additionally, they both have respectable stats for front line characters.

I equip Eder with a medium shield and Shattered Vengeance primary with whatever is useful as a secondary. I like clubs because the weapon modal reduces will by 25 which is huge for a lot of CC spells, and that particular club is just solid and comes early. I have him in medium armor for the early game (I like Bloody Links) and then I swap him to heavy armor after he gets the fighter ability to reduce recovery penalty of armor. There are lots of good heavy armors so I just use whatever my team needs. I build him with defense and disruption in mind, focusing on weapon and shield style with knockdown and accuracy buffs (so knockdown hits more). Note that the displayed hit chance is not accurate with this ability; it shows deflection chance to hit but, per the tooltip, the attack actually targets fortitude. Persistent distraction helps a ton because you just instaflank almost everyone, so you can bottleneck enemies rather than trying to get something behind them for flanks. Any skill points above and beyond that role can be used to help increase his damage output.

I run Pallegina with the Blade of Endless Paths in the main hand and a morningstar like the Saru-Sichr in the other. I think the Blade of Endless Paths may be the most underrated weapon in the game, because you get high accuracy, penetration, action speed, and a stacking deflection penalty on crit. The morningstar is my secondary because piercing immunity is moderately common and the modal reduces fortitude by 25, which is a huge boost to hit chance for some power packed spells and abilities like Disintegration or Blind spells. I give her the ring of focused flame because her fire attack ability gets a whopping bonus 20 to accuracy (in total) and her chanter flame damage/heal invocation gets a bonus 10. I also quite enjoy having her run Firebrand scrolls, but since it's a consumable resource it's harder to justify all the time.

My main character is going to be a dedicated backliner for the first half of the game with the option to do some off-tanking after power level 5, but we'll get to that later.

For my last two characters, I like to bring Tekehu (chanter) or Maia (either multiclass) as a damage dealer, and Xoti (priest) as a healer/buffer. I like Tekehu because his kit is more flexible than Maia's, but Maia is a fantastic option if you need some extra damage.

I take Tekehu as a pure chanter because his PL 8 and 9 abilities are fantastic, with special emphasis on the upgraded tornado for a whopping 13 bounces of roughly 50 high penetration damage (650 damage total) with some CC to boot, and the party bonus to penetration from both the passive chant and the active invocation. He also has some cheesy stuff at these power levels with the singing scimitar and the storm attacks but I'm not really about that life. His druid class is good as well, but I often find traditional casters a bit underwhelming due to the low stock of spells, even though Druid spells can be quite good. I like to dual wield the Singing Scimitar for party action speed and any useful ranged weapon in the other hand as my primary kit. Dual wielding in this way offers some interesting options because it basically lets you swing one weapon significantly faster and you can swap between near and far targets at will. I will include a shield in my secondary slot in case he needs to absorb some pressure. Note that the accuracy penalty from larger shields applies to spell casting as well, so a small shield with weapon and shield style is what I usually take. For chants, there are a bunch of chants I really like. In no particular order, I really like Ancient Memory, the 50% heal buff, the skeleton summon, the resistance to constitution and intellect afflictions (for situational use), the 15% fire damage on weapons, the -20% recovery/reload time with ranged weapons, the 10 deflection and 1 engagement, the +1 penetration and the 10 point damage shield. My ideal party would probably have 2 multiclassed chanters and one pure so that I could have 6 of those chants active 75% of the time and rotate through the others as needed, but that's not really an option with the way I'm writing this current run so we'll have to make do with 1-2. With the two chanters in my party, I usually find myself running the two heal chants and either the reload/fire damage ones (if I need damage) or the deflection and damage shield (if I'm getting overrun).

I choose Xoti because she has useful spells at every power level, and I go straight priest because her PL8 spell "Crowns of the faithful" is very good by itself and ridiculously good with the chanter summons that Palleginia will bring to the party. I give her a reload weapon of any kind because spells can be cast at any time during the reload, so that allows her to always be ready to cast a spell. A last-second withdraw or heal can turn a party wipe into a victory, after all. Spearcaster would be a good option here.

I immediately rule out Serafin because in addition to not liking his character, his wild mind passive can immediately end you. Not great on a trial of iron run. If I expanded the pool to include sidekicks, I would likely sub in Konstanten and Ydwin for Xoti and Tekehu.


General tips about stats
Let's talk about stat scarcity. For the main character, you can basically max 2 stats by default, then you can max a third if you dump a stat (give or take a couple points), so we need to pick how to prioritize.

Best stat

Accuracy is king. As in most games, many people don't really understand the true value of individual increases in accuracy. On lower difficulties, it's not all that important, but on POTD it's massive.

Let me put this into perspective. Let's say by default you have a 50% chance to hit (we will revisit grazes and crits in a bit). If you increase your perception by 1, most people think "hey 1% extra chance to hit, not really a big deal." While this is true from an absolute perspective, your actual change in chance to hit is 1/50, which means your increase from one point of accuracy is actually 2%. This impact gets exponentially larger as you have harder to hit abilities; if you move from a 1% chance to hit to 11%, you increase your hit chance by 1,000% for just a +10 to Perception.

Now, at very high percentages, ordinarily the value of each addition point would get exponentially smaller because 100% hit chance is already guaranteed. However, because graze->hit and hit->crit chance scale linearly with accuracy, you continue to get value with as much accuracy as you can obtain, since you'll never reach 100% crit chance on higher difficulties in normal circumstances. Additionally, penetrating enemy armor can be difficult on higher difficulties, and pen bonuses are not super common, but crits give a big boost to penetration so critting frequently also substitutes for low penetration on high DPS/low pen weapons and spells.

Finally, perception is directly tied to landing CC which can be huge.

I remember reading once that games with RNG are still generally designed to be beaten, so if you're going to lose, RNG has to be unusually bad. This means, all else equal, reliability is more important than high potential because the average RNG should be sufficient to win, and perception is one of the key "reliability" stats in the game.

So to summarize, perception means you deal more reliable damage AND your last-second CC lands more often, and when you can't reload a save, landing clutch CC is probably the single most important way to turn a total party wipe into a won fight.

Stat Tradeoffs

Now for the relationship between dexterity, might and intelligence. For a general purpose build, might is actually the least useful of these stats, due to availability of stat boosts as well as how the stats interact with actions and each other.

Let's start with a talk about DPS. With no special abilities, the first point in might and dex will give you mostly the same benefit. However, this quickly changes as more points are added. There are a few reasons for this.

First, damage bonuses usually stack linearly with each other and multiplicative with action speed bonuses. For example, if you have a base damage of 10 and a 10% increase in damage (11 damage), your next "10%" increase in damage is actually 1/11 which is only about a 9% increase for that point. On the other hand, a 10% faster action speed will allow you to hit the 11 damage 10% faster for a full 10% increase in overall damage per second.

Second, overkill is a thing. If you can only attack once per combat but deal a thousand damage to a single enemy, you're still going to lose, because every point of damage beyond their health is wasted and other enemies still exist. On the other hand, if we attack a thousand times in combat but only deal 1 damage each time, we will do the perfect amount of damage and can still win the fight. So action speed by principle is more valuable than damage per swing.

Note that action speed bonuses are kinda complex, and there are guides out there for that if you want to dive into the mechanics, but the basic thing to remember is that reductions in recovery time are usually more beneficial than increases to action speed. At its core, doing things "100% faster" means twice as many actions, but taking "100% less time" to do something means infinite actions.

Third, many sources of damage occur over time or in an area, and INT gives a better percent return on damage than MIG does (e.g. 5% longer duration instead of 3% more damage, plus the bonus is multiplicative similar to how dex works). If I'm not mistaken, there are also some bugs in calculations that make INT even more valuable in damage calcs due to having an extra tick on cast, but this guide isn't addressing those and the conclusion is the same anyway. For a build that exists solely to auto attack, MIG could be better here, but INT also impacts buff duration and such so even then I'd probably put my money on INT.

What about big aoe damage spells? Well, since they are usually limited in resources, there is an argument to be made that MIG has a place, but in my opinion, the fight-changing spells are buffs, CC or over-time damage rather than spot damage, meaning that PER and INT are going to usually be better.

The main place might shines is ironically not in damage, but in burst healing. Healing over time is still usually benefited by INT more than MIG, but for the priest AOE spot heals, a high might can make a big difference. In other words, if you want a dedicated healer then you might ditch PER or INT for MIG, but that's generally about it.

Finally, we're left with constitution and resolve. These are both "try not to die" stats, but my recent mantra in most games is "they can't kill you if they're dead," and the relative impact to you winning a fight is much higher for the other stats than for these two. You don't want to dump CON completely because dying in 1 hit instead of 2 is a sure way to lose your run.

Summary

My stat spread looks something like:

MAX PER and DEX
PUMP INT
DUMP RES
LEAVE MIG and CON (I do 8 MIG 10 CON or something close to it)

The reasoning is because PER and DEX let you use the "right" ability for a given scenario more often and actually land offensive ones when you do, INT increases your DOT, HOT, buff duration, AOE and CC duration, and the remaining stats are simply less impactful in general.


A note about stacking bonuses and inspirations
I'm not going into too much detail here, but it's important to note that passive bonuses typically stack with everything, while active bonuses typically only take the highest of that given type of buff. "Active" are considered abilities that come from the left side of the skill tree (which includes modal abilities like paladin auras and I'm pretty sure also competes with food and potions, but I can't recall right this second), and "passive" is anything on the right side of the tree plus most gear effects that are not triggered (e.g. anything with a per-rest charge is usually active). This means that chanter chants are considered passive(!), while chanter invocations are considered active.

Note that a buff that gives "20 to all defenses" is not the same type of buff as "20 to deflection," and so these two bonuses would stack even if they both come from active sources. However, a "10 to deflection" and a "20 to deflection" would only give you 20 total to deflection. The various bonuses from inspirations are each considered separately and considered active. This means that if you have the Brilliant inspiration and the +10 bonus to INT you can get from the monk modal, you would only get a +10 rather than a +15, because they are both direct buffs to INT. Related but different bonuses (such as direct buffs to AOE size, which INT improves independently) would still stack here.

Finally, power levels multiply EVERYTHING that they affect, so those are awesome. In a nutshell, each power level of abilities gives +5% to most figures that scale (e.g. damage, healing, duration, etc.), 0.25 to penetration, and a +1 to accuracy. By nature this means that some abilities are impacted in more ways than others, as a heal with no over-time effect wouldn't benefit from a percent increase to duration. Note that power level scaling happens from the ability base power level, so at power level 5, a level 1 ability would have a +4 power level bonus but an ability you just unlocked would not be impacted. Static numbers are usually not impacted, such as abilities that directly impact attribute scores.
So what's the most fun build?
Of all the builds I've tried, my favorite by far has been a Seer, with the Cipher (Ascendent) and ranger (Ghostheart, boar) classes. This may seem a bit funny because ciphers and rangers both tend to be "meh" on tier lists, so I'm going to explain why this multiclass is both very fun and very strong.

Ranger for real?

Yes, ranger for real. Rangers get a lot of shade because most of their kit is straightforward and they don't provide much oomph. For example, fighters and rogues are common multiclass options because their kits provide easy synergies and quite a bit of damage and/or tankiness. However, rangers' pets aren't really all that great and require heavy feat investment in order to make them worth much, so if you want a class that lets you do damage and tank, there are lots of martial halves of a multiclass that will pull that off better.

However, there is one thing that rangers get that other classes don't, and that's a ridiculous amount of STACKABLE accuracy (something like +30 by the end of the game). So the ranger half of the multiclass is quite simple; get all of the accuracy abilities and Driving Flight, throw in the gunner trait for the Red Hand arquebus, slap on Maia's gunner armor (or whatever floats your boat), take Navi the pet, eat some koiki fruit for recovery time reduction, and you're ready for anything. We take the ghostheart boar so 1) we have a little extra tankiness, 2) reviving or recalling the pet is basically instant, 3) we don't require a feat investment that can be reserved for those juicy cipher spells, and 4) if the pet dies it's not an instant disaster due to Bonded Grief.

OK so what does the build actually do?

Basically, the way the build works is that you let your ranger do stupid levels of dual-target damage (with driving flight) to get focus, then use that focus for whatever the combat requires. The reason this build is so fun is that it allows for unprecedented flexibility compared to other builds. Most builds eventually fall into "use the same ability order every fight because you don't really have that many options." Ciphers are unique because they are full spell casters that are not limited to 2 spells per power level, so they can use whatever spells they want. With the Ascendent subclass, this becomes even stronger because you max out focus in just a few attacks with the Red Hand, and then you can spam whatever abilities your team needs. And considering the cipher tree is jam packed with good options, you can literally just pick a bunch of conditionally useful abilities and then let them loose whenever the need arises. I personally like to get at least one damage and at least one battlefield control ability that targets each defense. For example, at tier 1 we have a control ability that targets fortitude and one that targets will (blind and charm, respectively), and an aoe damage ability that targets reflex. All three of these options are great.

I do need to make a shout out to Disintegration, which as far as I can tell, does like 10x the damage of any other damage spell, and it's SPAMMABLE with Ascendent. There are, however, two things to beware. The first is that Disintegration will be reflected by abilities that reflect spells, which means your character will die rather than falling unconscious if you try to cast it on one of these enemies, even if there are no injuries on the character. So READ THE BUFFS before you fire away. The second thing is that this ability targets Fortitude, so it's always going to land on casters and the like but is harder to land on most other enemies. However, since the ranger gets so much extra accuracy, it will work on front liners too.

The other specific ability I want to call out is Mind Wave. At first glance it's like "ok it's basically the level 1 shock," but upon further consideration it's actually a massive-AOE-high-damage-hard-interrupt. While some enemies are immune to prone, many times this ability can just pause the enemy's turn which can let your priest get off that emergency heal that was half a second away. Given that you're an aAcendant, it's also spammable as well... So yeah that's fun.

What should I wear?

Be a man, wear what you want. Unless you're a woman, in which case... wear what you want anyway!

Regarding gear, you have options here, which is part of the fun, but any weapon with a reload capability is viable. As mentioned elsewhere in this guide, this is valuable because it means you can interrupt your recovery to cast a spell. This means you can cast immediately after getting max focus, you can use the best spell for the scenario as soon as you need to, and if you're not watching your bar closely (which happens to me a lot) then you aren't punished for missing your cast window. The Red Hand is probably the best individual weapon here due to the double tap + knockback (for defense) + gunner passive + Maia's armor + driving flight + high damage + high pen + reloadable weapon, but I swapped out a few times throughout the game just for fun. With this build, you're firing twice every 2-3 seconds with Arquebus level damage that hits two enemies... it's nuts and lots of damage, focus gen, and fun.

Armor is also flexible. Even without Maia's armor you still fire quite fast, so you can swap out for what you want, but I would avoid heavy armor unless you are going double pet from Berath's Blessings to drop your armor recovery significantly and even then probably only if you were going front line. Medium armor is usually a sweet spot, since we mainly just want to avoid overpen from the enemy (where they have double your armor value in penetration). Some of the better light armors are still good, but I wasn't a fan of the cloth armors.

Regarding other gear, the gunner tricorn is good as is the cipher willpower accuracy helmet. Other than that, go with whatever floats your boat.

In the later mid-game when you get access to the accuracy/defense ability from the cipher tree, you suddenly become an off-tank, so you can swap to melee weapons if you want (either full time or conditionally). I like two handed high accuracy weapons personally, but there's flexibility here. Just remember that shield penalties to accuracy apply to spells as well so we wouldn't want a medium/large shield in the build.

This build actually works so well that I find myself simultaneously running into and not caring about action economy issues, because I have so many options in combat and they all work quite well.

Summary

To close out, this build comes online at level 1 with strong CC and damage options, stays strong in the mid game with high flexibility, and caps out the end game with abilities that are individually fight-changing. So is it the most fun after all? Let me know what you think, and throw me a 5 star rating if you find it as fun as I do so other people can enjoy it too!

Thanks for reading! Let me know if I missed anything worth throwing in there.