Constitution is not as rewarding as it should be
 

When I was creating my PC Chanter, I noted that Constitution was marked as a recommended attribute. But now, I'm getting the notion that Con does not make a huge difference - especially for my Chanter. It appears that maxing Con does not create good tanks. The difference between a con of 10 and 14 is a measly 12% increase in health/endurance. If I increased my con to 18, I would get a 22% increase. That's not much (I'm not really sure, so I'd appreciate others' opinions).

As for my chanter, he has 16 in con, but I have started to think it would be better to reduce my con to 10 (and lose the 18%) and instead, increase other attributes. Int and Res are no-brainers, but perhaps I could benefit from increasing my Per (and thus, increasing deflection and reflex), or Mig and consequently, my damage output which is low.
I'm not really sure if Dex is beneficial enough (for a Chanter), but then again, I might be wrong.

If you want to use the Chanter as a straight up tank (not the best idea), you need to get him to survive melee combat first of all, so I'd go with maxed Per and Res and a shield. You'd need to accept that other sources of damage will take him out soon though. If you just use him as a tank + perm buffs (phrases) though, why not also dump Int and put that into Con?

edit: my Chante is a second rank character (3 tanks, 3 glass cannons). I gave her max Con and max Res and go with a shield. Sort of a good compromise, I'd say.

The results of an attack is relative, checking the linear increase does not fit.
It is better to check how the increase of con protects from all the available attacks.
A four point increase might mean near impunity against 90% attacks etc

Front line units are malleable units that will only last as long as the buff they are applied fit the situation. In this regard, con plays a specific part.

Chanters are more units to control the space behind the front line. They make good body guards to protect a spell caster.
They can also supplement a falling front line by being able to move in position instead of a front line unit.
When a front line unit falters, a chanter can move into action to allow the faltering unit to disengage, rest out of engagement range (potion might help) then using a spell, the chanter might buy time to disengage himself etc

I might as well do that with Mig, since they increase Fortitude by the same amount?

My Per can be increased a little. Now I wonder, would it be better to set Con to 14 and spend the 2 remaining points on Per, or on Mig, or set my Con to 16 and forget about them.

Since might is also involved in computing fortitude…

Assessing the efficiency of this or that would require to list all the attack value of all the enemies and to check what an increase in con produces against those attacks.

One point more in this or that might mean significantly improved performance against 20 per cent of enemies etc

After completion of my run on hard and building a tank in a wrong way (very high resolve, high con and low might), it appears that optimisation is not really the focus of the game.

The gameplay is flexible enough to allow to make the most of many builds.

You finished the game already? I haven't started it yet. Just waiting for a stable patch.

Constitution is great on tanks and especially monks. Because they have access to skill the more damage they take.

Yes, for monks and barbarians Con is great, especially for monks that use wounds as a resource. And both of those classes have a relatively large health pool, therefore even a slight increase in Con means a noticeable increase in health/endurance. For a human chanter, the difference between con10 and con 16 would be Health 144 and Health 170 which is not that good. For mages, priests and ciphers the difference is even less noticeable.

blackcanopus, I think that extra health is useful, especially if you plan on using a Chanter as a front-line fighter. It's not a huge difference, no, but I can see it keeping you alive for a few more hits than if you didn't have that Con invested. :thumbsup:

It has enough material to beget multiple playthroughs.

The optimization process is hard to refine as it turns out a matter of playstyle.

Imo, the whole thing has many cracks, many exploits, that does not matter though because the objective is more about providing a tool set to support multiple ways of playing a same character.

Chanters might be played differently, and this must be associated with the appropriate micro.

Putting points here and not there is only optimal with a certain play style in mind and the associated micro to support it and the proper addition of skills.

The game is not only about running numbers, it is also about using properly.

Thank you for your opinions. I think I will settle with con 16 for now. Sadly, I don't have enough time for replays.

There was a thread on the obsidian forums discussing how, all things, considered, if there was a dump stat it would be constitution. I'd have to agree, though I'd never go below 8-10 con.

Well, it depends on whether you like to keep someone in the middle of the fighting or in the back. For anyone in the back, it's surely a dump stat. However, in the middle it only takes a few stray blows to smash any character with low Con to bits.

I generally prefer to keep everyone but Fighters, Monks, Barbarians and Paladins as far away from the fighting as possible, as I've had too many "accidents", but I still keep 1h + shield weapon setup ready for teleporting enemies (shades etc).

Con is only meaningful if your class has a decent starting health pool. During character creation, as you click on each class, it shows what some of your starting stats are for that class, including the health/endurance, along with a description like "high", "low", "very high", etc.

If you're making a frontline class like Fighter or Paladin or Monk, then your base Health/Endurance are high, and a Con's percentage increase can greatly help. But if you're making a back-line class like a Wizard or Cipher, then your base Health/Endurance are crap, and a percentage increase of crap will be just slightly higher crap.

For my Cipher, I did an experiment and left her at the starting 10 Con, and she started with 49 Endurance. I then re-made her with an 8 Con, for a 6% decrease in Endurance/Health, and she instead started with 47 Endurance. Trading 2 precious stat points for 2 measly Endurance was a no-brainer, so I went for an 8 Con. Yes, this difference will grow over time as she levels, but the Health/Endurance gain per level is also low for this class, so I think this is still the right call (even with her going as a dual-wielding melee fighter…there's far more to be gained from an extra 2 points in Int or Dex or Might).

PoE Constitution works opposite to how IE games had it. There low base HP classes got a biggest boost from high Constitution (16) (up to 50% boost for wizards and sorcerers) while melee classes got the lowest boost (33% boost for Con 18 for Barbarians; 40% boost for Fighters at Con 18).

But even in IE games, Constitution was just one part of making a GOOD tank.
Good tank has good Life, it is hard to hit and it is resistant to spells/has good saves.

Even in IE games if you got 18 or 19 Con but ran around it Leather armor with no vs spells gear you died fast.

Definitely. Base class has the biggest impact on Health. Three percent per point is just meh. I think they ought to overhaul some attributes.

Well, I'm fine with it. I don't really fret about dump stats - though they shouldn't be too blatant.

I'm actually more against enforcing rigid balance through immersion-breaking systems, like Might adding to Healing and stuff like that.

I prefer systems that are based on how the real world works - for the most part - and then rationally adjusted according to fantasy and game necessities.

Reason for liking that more, is that I don't actually have to study mechanics in-depth to get a reasonable feel for what to do. That's because we're all familiar with reality.

That said, not a lot of systems do that too well. Except my own, naturally ;)

IMPNHEO*, I think they went too far trying to be clever in Pillars of Eternity - and most of it seems to be about correcting dislikes in D&D - instead of simply doing their own thing from scratch. They could probably reduce the mechanics complexity a good 25-50 percent without losing anything of significance in gameplay richness.

(In-my-probably-not-humble-enough-opinion)

I do feel they somewhat over-engineered combat systems. I am looking at you, Josh Sawyer.

Personally, I suspect the problem is that they felt they had to stick close to D&D - to appease fans. They just couldn't resist the temptation to change a lot of little details. But that's a bad foundation, in my opinion.

I have a feeling the system would be less opaque and much more playable if they'd simply invented their own from scratch.

Only thing they really kept from D&D is 6 attributes and general feel of the classes and some spells feel similar to their IE games counterpart. Most everything else was new. Just because it reminds you of D&D and IE games does not make it bad.