Based on the combat traits argument I'm now strongly considering dropping the Shadowdancer altogether and switching back to Fighter once I hit the epic levels. Dodge and Mobility do seem pretty weak at the moment; I haven't even needed to use Haste potions or toss on a Stoneskin against most of the enemies thus far.

Pretty much everything in the official campaigns is pretty weak, frankly.

Remember that Dodge is 1 AC versus your current target only -- good versus a threatening single enemy, bad otherwise. Mobility gives 4 AC when provoking an AoO, which does include something like drinking a potion (or firing a ranged weapon or running past an enemy without having Tumble/Spring Attack)...but it's still a small AC gain overall (best case scenario it's 4 AC on one of two attacks for 2 AC average, "worst" case scenario is 4 AC on one of 4+ attacks for 1 AC average...and that's only if you're provoking an AoO every single round).

He already has 18 dex. Yes, he should invest in STR from now on, but with DEX that high full plate is by no means the only option.

18 DEX and a chain shirt gives AC 8. That and leather and the ring of the rogue gives 8 AC. That and full plate gives 9. Even a breastplate will still give 7 AC. -1 or -2 effective AC is hardly a crippling malus. Now, you indeed probably don't want to wear padded armour and lose a STR belt for a Belt of Agility, but there are plenty of options besides that.

Sure, -2 AC compared to full plate isn't great, but it can be worth it if that suit comes with great properties. Say, the chainmail of speed, bracers of AC +5 and Boots of the Sun Soul can give much better AC than Full Plate +5 and Boots of Speed. At worst that'll cost you 1 AB from missing gauntlets of Ogre Power.

Same for a suit of armour that gives damage resistance, or a strength bonus. Can well be worth it on this build.

And whilst you can certainly switch out armour for stealth, in practice that just means you'll likely use stealth less and makes for a less fun way to play a rogue. Chain shirt only gives 1 AC less than full plate, but it reduces the AC penalty by a whopping 6 points, negating the need for any armour swapping. Plus it looks cooler for this type of character, i.m.o.

When building a character like this from the ground up, it's certainly sensible to go for full plate so you don't need to invest in DEX and can boost STR instead. But once you already have the high DEX, you might as well make use of it.

Is -3 to -8 stealth checks a crippling malus either? Quite frankly you don't need that many ranks in Hide/Move Silently to sneak past the vast majority of enemies (which have Spot/Listen scores of 0ish). The breastplate, for example, would be +3 Stealth for -2 AC.

For anyone not aware of how detection works, Spot/Listen checks only count half skill plus d10 unless in Detect Mode or Keen Sense is present (and neither is true for most creatures). So if the enemy has 10 wisdom and 0 Spot/Listen, you only need 10 Stealth ranks to hide (there are technically some additional hidden modifiers that can work both ways -- trying to sneak directly next to a standing still target is harder than sneaking around a moving target). A level 10 character with 18 Dex will have 17 Stealth ranks by "default" prior to any feats, gear, or armor penalties.

So even if you have Full Plate on with no other positive modifiers you'll still have 9 Stealth ranks which is enough to generally always sneak past something with <10 Wisdom and will still usually be able to sneak past anyone with 0 Spot/Listen. And if you have any gear available that gives Stealth bonuses (and/or you do something crazy like drink a Cat's Grace potion) you can easily boost your Stealth score higher. It only gets easier to make checks from there too. I'd definitely take 2 AC over 3 Hide/Move Silently in a heartbeat. I'd even prefer 1 AC over 8 Hide/Move Silently past a certain level -- either I can sneak past regardless or I'm probably going to get detected regardless (due to high Spot/Listen or True Seeing).

Another issue is that using, say, Ring of the Rogue is *still* losing out on other stats. Either you're losing a ring slot (which could be replaced with a regeneration ring or an elemental resistance ring or a saving throw ring or whatever) or you're losing a cloak slot (since you're wearing a Cloak of Protection rather than wearing a Ring of Protection plus a cloak). Could this be worth it in some modules based on the items available? Sure, but I doubt it generally is.

That's hardly a fair comparison for at least two reasons.

1, you could make that same argument with a Dex of 12. That 12 Dex would become 15 with Boots of the Sun Soul so you still get 7 AC from Chainmail, so same overall AC as starting with 18 Dex.

2, you're assuming very specific items which are not only module specific/dependent but ignoring other similar possible items. In your example you get 5 base armor, 2 dex AC, 5 armor bonus, 5 dodge bonus, and Haste for 21 AC total compared to the full plate +5 and boots of speed netting you 18 AC total in addition to 1 AB/1 damage...so 3 AC versus 1 AB/1 damage. Which is due to Boots of the Sun Soul being quite good and Gauntlets of Ogre Power being bad. Now imagine Bracers of the Blinding Strike (4 armor AC, Haste), Boots of the Sun Soul +5, and any full plate. That's 8 base armor, 1 dex AC, 4 armor bonus, 5 dodge bonus, and Haste for 22 AC total...*plus* whatever bonus you get on the full plate. Full Plate +5 would give you 23 AC total, for example, at a minimum. But you could also use full plate with worse armor bonuses but better other bonuses -- ranging from +2 saving throws to 5 slashing/piercing resist to 20 resistance of an element to Freedom to whatever. And something like Red Dragon Armor would net you that 23 AC (+2 versus the other set-up) plus 20 Fire Resistance.

Depends on your level, really. Like I said, once you hit level 10+ or whatever you can usually be sneaking around in Full Plate versus most enemies if you want. Sooner than that even with higher Dex, Cat's Grace potions, skill boosting feats, and/or stealth gear.

In theory a level 15 enemy should have a much better chance to spot you than a level 1 enemy...but a level 12 (CR 9) Troll Chieftain, for example, still has 0 Spot/Listen. Even Fire/Frost Giants (the male Fire Giant has Taunt while the other three have Spot...guessing that's an error with the male) only have like 7 Spot skill...which is really only 3-4 skill due to that whole "half skill applies while not in detect mode." And those are level 15 creatures -- a PC in Full Plate, 12 Dex, and no stealth/dex boosts could be "guaranteed" (again, sneaking in front of a standing still giant in bright light is harder) to be undetected by level 17.

On the flip side, a default Balor (level 13, CR 15) has 33+ in both Spot and Listen...so talking effectively 16 Spot/Listen. To "guarantee" a non-detection under default circumstances you'd need 26 ranks in Stealth after all modifiers. Is that doable? Of course, but it would take a somewhat serious effort and likely include no armor penalties -- you don't even want that -5 from Chainmail in that case.

So yeah, generally a binary problem. Either you can sneak by stuff with any effort put into stealth or you need a massive effort put into stealth. 60% effort versus 70% effort rarely matters.

True, sneaking in full plate is possible once you reach high enough levels. I never really tried, because.. well, it looks weird, but mechanically you're quite correct that against enemies without high spot/listen the armour check penalty won't matter after a certain point. Still, against enemies that do have a good spot check, the +6 is a big enough difference to be noticeable.

Oh, and another point if you're trying to hide: if any of your items have a light source, you get a huge penalty. -10, I believe. That's worse than full plate. So better leave that Ring of Resistance at home if you want to sneak.

As for the specific item combos: I was naming items I recall finding in the first chapter of HotU, though on reflection I'm not sure the +5 sun soul boots don't come later. Can't recall anything better than ogre power gauntlets there. But anyway, the specific items are not really the point. My point is more that with a build like PalmTree has, the specific properties on the armour are more likely to be useful than 1 or 2 points of base armour value.

You're right that if you lose a ring or bracer slot it's unlikely that the armour's properties will make up for it, but it is still an option if he finds some kind of leather he really wants to use.

True. That's more or less the same point I was making; in general, once you hit a certain level and really powerful magic items show up, it becomes more important to consider other properties besides armour class. (Early on, I find that there's very little that can make up for AC. At low levels you're just too vulnerable otherwise.)

One thing I'm confused about, though. How does Red Dragon armour get you more AC? It's just Full Plate +5 with Fire Resistance, isn't it?

All things being equal, if you don't need the extra stealth, full plate will give the best AC bonus. But it's a small difference at this point. And a chain shirt in particular is a quite competitive alternative.

Is it, though? Look at our Balor friend. 16 Spot/Listen effectively. If he's a "boss" for our hero then maybe our hero is level 11 or so? So 14 skill ranks, throw in another 6 from Dex modifier (18 base and 4 from gear) and we get...20. So if our intrepid hero rolls an 5 (on a d20) on his Hide and the Balor rolls a 9 (on a d10)...the Balor sees his next meal. Or the same happens on the Move Silently/Spot opposed checks. If either of those ever happens *once* the Balor will detect the player. Which isn't even bringing in modifiers like the Balor getting a +5 bonus to both checks if it's standing still (which can be potentially countered by the player to some degree (like being 15+ meters away and behind the Balor) but then you're not just freely sneaking around).

So, sure, the Full Plate wearer will *probably* get detected first but both are likely to get detected pretty quickly. You could eliminate that detection chance by bringing in stealth boosting feats/items, but in such a case it might be possible to eliminate it for both characters depending on the items available.

Penalty at night only, yes. Of course, you could be using a ring with bonus saving throws that *doesn't* have light or could be using something like a Ring of Regeneration and a (Lesser) Ring of Power.

Can't you find monk gloves with +5 AB and 1d6 sonic or something? My point is that "Ogre Power" gloves only go up to +2 while something like Bracers of Dexterity go up to +5, which is stupid in the first place and more importantly means you should be trying to use gloves/bracers that don't suck in comparisons (unless we're comparing all +2 items). Regarding your last sentence, that is *always* true. Full Plate seems objectively better than Half Plate, right? Well, not if the designer decides he thinks Half Plate is totally rad and makes some with 20 resistance to all damage and +6 AC while the best Full Plate has 10 resistance to all damage and 4 AC. Because reasons.

Hell, in the Aielund Saga my 13 Dex (needed for Weapon Master) strength based Fighter type was using Padded Armor at one point...because it gave Permahaste.

I'd disagree. For the vast majority of low level enemies, for example, I'd trade several AC for 5 Slashing/Bludgeoning/Piercing resistance in a heartbeat. Or even something like 5/+2 damage reduction. At low levels you tend to have few enough options that a powerful item (for the level) can surpass many other considerations. At higher levels the rest of your gear can often "make up the slack" of missing properties on armor...while making up the "missing" AC is difficult to impossible.

Bracers of the Blinding Strike have +4 AC and Haste. So if you use Full Plate with +4 or less AC, you get +4 AC bonus (but presumably get other properties). But if you use Red Dragon Armor, then that's +5 AC so you gain an AC over just using Bracers of the Blinding Strike.

1 AC typically means you can take 10-15% more damage (or rather, you'll avoid more blows to live 10-15% longer) and in some cases can be vastly more (if enemies can only hit you on an 18+, then gaining 1 AC is effectively +50% effective HP versus physical attacks). I wouldn't call that small at all. I mean, if you were doing 30ish damage per hit with a weapon and someone offered to add 1d6 bonus damage to it, would you say "Nah, going from 30 to 33.5 damage is a small difference at this point?"

If so, then maybe we need to define exactly what qualifies as "small" -- to me that means like 1-2% or something.

Indeed you should, but I find that in many modules there are none to be found unless you're a DEXer. Or indeed a monk. But unless you're unarmed, those +5AB gloves won't do you any good. In practice, I recall many of my characters sticking with Gauntlets of Ogre Power well into their late teens and early epic levels when every other item slot had been filled with better equipment, even in Hordes of the Underdark. (Bracers of Blinding Strike aren't much use there, because you can just stick haste on your weapon.) A quirk of the stock item design in NWN.

Ah, right. I didn't get that you were still talking about the BoBS. Thanks for clearing up.

Really? I admit I never made the full calculation, but I was under the impression that 1 AC would typically provide a much, much smaller bonus than 10-15% once you reach high levels. If you're right, then yeah, AC will remain paramount.

Earlier though, you said "The AB of most enemies in the official campaigns is so low you could probably afford to skip the Natural Armor amulet." So I'm a bit confused here. A natural armour amulet gives a much bigger boost than upgrading from a chain shirt to full plate does. (And yes, I have been talking about the specific example of PalmTree's character in the OC the whole time, here.)

But anyway, let's try and calculate how much AC helps in more generic situations because I'm rather curious now.

We're talking a level 13 characters, so most enemies will have 2-3 attacks, more if they dual-wield or have haste or something. In the OC, you'll likely have mostly a mix of +2 and +3 gear at that point, though some better items are available, so let's assume +3 average. It's possible to find boots of speed somewhere around this point, but let's assume they're not in play yet. Some other single player modules give much more powerful gear, of course, but let's discount those because it would just complicate things.

A fighter/rogue wearing full plate, using a large shield and wearing all armour items including Boots of the Sun Soul +3, will get a total of 10 (base) + 11 (Plate +3) + 1 (Dex) + 5 (shield +3) + 9 (Boots, amulet, deflection AC items) + 3 (Tumble) = 39 AC. The same character wearing a chain shirt will have 38.

Say he's fighting a bunch of Fire Giants. (CR 12) These have an 23/18/13 attack sequence, but because of the size modifier this effectively becomes 22/17/12. This means the giant will hit the plate guy on a 17/20/20, and the chain guy on a 16/20/20. In other words: the plate guy gets hit an average of 0.3 times a round, the chain guy an average of 0.35%.

The giant inflicts an average of 29 damage a hit. (Assuming no fire or slashing resistance.) So the plate guy suffers an average of 8.7 damage a round. The chain guy suffers 10.15. I have to admit that this is quite a significant increase in damage suffered.

If we suppose our fighter/rogue has some 110 hitpoints, he survives for 12.6 rounds in plate, and only 10.8 rounds in chain. That's only 85% as long, or indeed a 15% decrease. So in this example, you are exactly right and I have to admit every little bit makes a big difference.

Of course, both fighters might decide they're too vulnerable and activate their Expertise feat. Then, no matter what the fighter is wearing, the giant will hit on 20/20/20 for an average of 4.35 damage a hit and 25.3 rounds expected survival time. So in this example, it indeed makes no difference whatsoever what you're wearing.

Now, for completion's sake, let's bring on a nastier foe. The fighter-rogue is up against an Old Green Dragon. Attack sequence, accounting for size, 32/27/22/17. Conveniently enough that's almost the same as the fire giant, except with one extra attack at +32. This means the plate guy suffers 0.8 hit average and the chain guy 0.9.

At an average damage of 17.67 a hit, that's 14.1 damage a round suffered for the plate guy and 15.9 damage for the chain guy. That means the dragon needs 7.8 rounds to kill the plate guy, and 6.9 rounds for the chain guy. In this case, the chain guy survives 88% as long as the plate guy. The difference is smaller, but still pretty significant.

So in conclusion: a single point of armour class does matter more than I thought it did, even when it's only the first attack or two that hit on anything other than a 20. It only ceases to matter when your AC gets so high the enemy only hits on 20s regardless, but that's unlikely to be true for all enemies even in the OC. I concede the point.

Yes, I'm aware -- my point was that 5 AB gloves were "on par" with +5 items in general. And comparing +5 items to +2 items isn't exactly a fair comparison. It's like saying "Well, the Stonefire Greataxe is +2 with 1d6 fire damage while the Harbringer Kin is +3 with 1d6 fire damage, therefore Greatswords are better than Greataxes."

Don't feel bad, it's a very common mistake to make. People think "If I have a lot of AB/AC, then adding more doesn't really matter, right?" Problem is that it's a constant d20. 10 AB vs 20 AC, 30 AB vs 40 AC, 60 AC versus 70 AC, and 100 AB vs 110 AC are all the same chance to hit. You're better off having 30 AC versus a 20 AB enemy than 55 AC vs a 50 AB enemy.

It actually doesn't, though -- Barkskin potions will still give 3 Natural Armor AC (so 1-2 AC lost max) and you're also gaining 3-4 Strength from Daelen's Amulet. Even if you ignore the Barkskin potions, though, there's still always a tradeoff between offense and defense. If I only take 2% of my HP in damage each round on average, then even doubling my defense to take 1% of my HP in damage each round doesn't really help (because I take so little in the first place) and making it so I don't take forever to kill stuff is more helpful. On the flip side, going from taking 30% to 15% of my HP in damage each round would be a massive difference.

Another note: gaining 5 AC as the chain rogue allowed him to survive 134% longer (25.3 rounds compared to 10.8 rounds). Versus the dragon he wouldn't gain quite such an insane amount, but this helps illustrate why I hate Expertise so much. If a player is expected to be hitting an enemy, oh, 75% of the time so he doesn't feel bored and the enemy is designed to be hitting a player, oh, 25% of the time so the player doesn't get crushed...then activating Expertise would change that to player hitting 50% of the time and enemy hitting 5% of the time. So the player is basically choosing between 50% more damage (don't use expertise) or 400% more survivability (with expertise).

This is also a huge problems with how shields work by default -- a player with a 2H or dual-wielding goes from "losing" 3 AC at most (tower shield) to potentially losing 8+ AC (8 AC loss would be Tower Shield +5).

The same general idea applies to AB too. This is why I'm annoyed when people make comments like "Oh, it's just 1-2 AB" -- yeah, that's like 10-30% damage in most cases. And it boils down to that d20 that never changes which causes nightmares for balancing at higher levels -- say a 2H level 40 fighter has 18 Strength modifier and a 1H + Shield level 40 rogue has 17 Dex modifier. That rogue will have 4 more AC from Tumble, 8 more AC from Dex, and at least 8 more AC (something like a +8 Tower Shield would be 11 more AC) from a Tower Shield...that's 20 AC right there. And that's assuming equivalent gear too, and not the rogue being able to wear something like Boots of the Sun Soul while the fighter can't. Also assuming no Expertise and not factoring in Epic Dodge.

Even if you make an enemy that can only hit the rogue on an 18+ (for the first attack with 4 attacks, no haste)...the rogue will take 0.15/0.05/0.05/0.05 = 0.3 hits per round while the fighter will take 0.95/0.9/.65/0.4 = 2.9 hits per round. Or about 10 times the damage.

Yes, it seems my intuition was very off here.

Not entirely: at higher levels, the difference is smaller in some cases. If you're attacked by something with 1 attack per round that hits on a 19 and you get an extra point of AC, the damage you suffer goes down by 50%. Instead of 0.1 hit per round, 0.05.
If something with 3 attacks hits you on a 19 with the first attack and you get an extra AC, the damage suffered goes down by only 25%. From 0.2 hits per round to 0.15. Only... a 25% reduction is still pretty huge, so it remains very important.

I furthermore thought that the effect of AC would go down once you face enemies that can hit you more easily. At lower levels, this is actually true. Going from being hit on a 9 to being hit on an 10 only reduces the damage suffered by 8.4% (0.60 hits a round vs 0.55) But when facing something with 3 attacks, the damage goes down by 16% (0.95 hits average vs 0.80) because the AC now also helps against subsequent attacks.

Same dynamic applies to AB, of course. It's easy to think "I don't need 1 more AB, I'm hitting anyway." Yeah, with your first attack. Third and fourth, not so much.

Let's just say I'll reconsider the value of the Weapon Focus feat now.

I suspect a part of what obfuscates things is that high-level characters tend to have many layers of defence. Both passive, like concealment and damage reduction, and active, as in stuff that stops the enemy from attacking you. (Disabling and debuffing spells, feats like Knockdown, or even good use of terrain to minimise the numbers of enemies attacking you at a time.)

If enemies don't even get to make attack rolls, it becomes that much harder to see the effect of your AC. (i.e. what you said about trading off offence vs. defence.)

Of course that only adds to the balancing nightmare. There's such a huge difference between a "recommended" greatsword fighter and a customised fighter/bard/RDD with a shield that every fight will either be a pushover for the second or absolute suicide for the first. Modules usually err on the side of the former, because otherwise tons of players are excluded. But I don't envy them the effort it takes to try and balance this.

Well...

1, you're adding in number of attacks per round which I did not. With the same number of attacks per round my statement holds true -- one attack at 2 AB versus 12 AC is the same as one attack at 12 AB versus 22 AC.

2, one attack per round skews it in *both* directions. If the enemy has one attack at 29 AB versus my 30 AC, then going to 31 AC is barely a 5% decrease in damage taken (0.95 -> 0.9). On the flip side, if that enemy has 4 attacks then we're looking at 0.95/0.7/0.45/0.2 = 2.3 vs 0.9/0.65/0.4/0.15 -> 2.1, or nearly a 9% decrease.

I'd point out that even 8.4% is pretty huge -- I doubt people would turn away a piece of gear that said "take 8.4% less physical damage" because it's too small of a difference or something.

Yes, (Epic) Weapon Focus and feats like Epic Prowess are extremely valuable. That said, even if you're hitting 80% of the time on your first (and only) attack gaining 1 AB is still over a 5% damage increase -- and it's a multiplicative effect with damage per hit.

This is one reason why I generally hate modules which don't have reasonable weapon options and/or try to make people swap weapon types constantly especially in Epic levels -- losing 3 AB, 6 damage, and Improved Critical makes it very hard to do so without actually being worse off.

"Hmm, well, I have 25 AB with 26 damage per hit and 17-20/x2 for criticals against that enemy with 30 AC. That's 0.8/0.55/0.3/0.05 = 1.7 hpr which is 44.2 dpr and factor in criticals (20% increase) to get 53.04 dpr overall. But crap, that enemy is 50% immune to my weapon's damage type because reasons so that's only 26.52 dpr. I guess I'll need to swap to that featless weapon which leaves me with 22 AB, 20 damage, and 19-20/x2. Let's see...0.65/0.4/0.15/0.05 = 1.25 hpr which is 25 dpr and factor in criticals (10% increase) to get 27.5 dpr overall. So swapping weapons will gain me...3.7% increased damage. THIS FEELS LIKE A FUN AND REWARDING MECHANIC."

I really don't like the extremely specific weapon feats. I'd much prefer either no weapon specific feats or least make them weapon categories (like 1H vs 2H vs ranged or something).

Not only that, but most modules aren't even halfway tuned reasonably to the point where having 30% more damage and/or 30% more HP even matters -- you could do most/all modules without ever using rings or a helm, for example (just leave three equipment slots completely blank, no problem). And if you have effectively infinite Heal potions of top of that, it usually doesn't make a difference whether you drink a full heal every 30 seconds versus every 42 seconds...despite the fact that means you're taking 40% more damage (or have like 29% less health).

I actually had some significant complaining about my Siege of the Heavens module initially (and I feared a repeat with A Peremptory Summons but was pleasantly surprised -- maybe I scared away the problem people) -- because despite literally having a potion with infinite uses of full heal (I dislike pretense -- I find it hilarious when modules shower players with hundreds of thousands of gold and have default Heal potions available for a mere 2000 gold...but try to act like the players don't have effectively infinite healing), you actually needed a half-way reasonable character. If you tried to make a Wizard with 8 Con it would be extremely, extremely difficult (potentially impossible) not to die. If you made a super low DPS character some fights would take forever and/or be much harder (no, your Intelligence based fighter is not a special snowflake).

You might notice a discrepancy there -- I deliberately designed/tuned the module so that you didn't need to optimize your damage but your defenses do have to be reasonable. Part of that is due to defenses being more "standard" -- I can assume every Fighter is going to have 14 base Con, 400 base HP from class levels, wearing full plate, etc. They might have more defenses, of course (from feats/stats/multiclassing/etc) but they won't have less. Harder to assume what people will do for offense (and, as we've seen, a Fighter that takes Epic Weapon Focus will do massively more damage compared to one that doesn't). IIRC I think my bosses had 60 AC normally and a "reasonable" 40 Fighter would have 30 (BAB) + 18 (Str) + 6 (weapon AB) + 4 (feats) = 58 AB for about 0.95/0.7/0.45/0.2/0.95 = 3.25 hpr while one without EWF would get 0.8/0.55/0.3/0.05/0.8 = 2.5 hpr. The former is flat out doing 30% more than the latter (technically slightly more due to how confirming criticals works). Now imagine the latter didn't max out his strength via starting stats/leveling (maybe put more into Con or something), didn't take all 7 Great Strength feats, and/or had some other issues. We could easily get two level 40 Fighters using the same weapons who have a 50%+ damage discrepancy.

So I made the modules with the principle of you'd eventually win if you could survive...to a reasonable extent. For example, one boss summons adds who try to buff him and you need to kill those adds. The difference between those adds buffing the boss for 12 seconds vs 24 seconds is survivable. The difference between those adds buffing the boss for 12 seconds vs 60 seconds is not. The "penalty" for poor offense characters is some fights taking forever...but you can still win.

Oh, and to make something explicitly clear -- I designed/balanced the module around reasonable pure class characters (40 Fighter, 40 Rogue, 40 Cleric, 40 Druid, 40 Sorcerer, etc). Yes, that means a 12 Fighter/3 Rogue/25 WM can come in and wreck face compared to that 40 Fighter. But nothing I can do about that without overhauling the combat system (which I didn't want to do for that specific module) or leaving non-powergamed builds out.

Indeed. With this issue specifically there is, thankfully, a relatively easy way to address the shield issue -- don't have shield AC scale up with levels. It's AC, it already scales. The downside is that it's less fun from a player perspective to use the same shield for 40 levels or something...but I'm not sure what can reasonably done about that without massive changes. I also have larger weapons get more damage bonuses. For example, you might see something like this...

Greatsword: 1d12 fire damage
Longsword: 1d8 fire damage
Short Sword: 1d6 fire damage
Dagger: 1d4 fire damage

Those two changes alone (along with some buffs to dual-wielding -- like flat out adding 2 AB to Dual-wielding in general (so no AB loss) and not penalizing for a "non-light" weapon in the offhand (so you could dual-wield long swords without penalty) helps make 2H vs 1H/Shield vs Dual-Wield far, far more balanced.

By default it is, frankly, stupid to not use a 1H/Shield as a Strength fighter. And if the module is so easy that it doesn't matter then...the module is so easy things don't mater in general, do whatever you want.

Never played those modules. But I see they're pretty short, so I should be able to check them out one of these days.

Weapon enchantments help equalise characters. Baldur's Gate 2 kinda did this, or at least it said it did, though in practice they didn't really follow through.

Another thing that helps here is party play, or having a lot of enemies using non-physical attacks. Even if you have a henchman or two, a two-handed weapon becomes much more valuable simply because that shield is dead weight unless you're actually being attacked. If your henchman is getting his skull bashed in, doing 1.5 times STR damage suddenly becomes very attractive. Both because there is no downside and because killing the enemy quickly lowers the amount of damage that enemy can do.

Same goes, of course, for fighting wizards and the like.

I always thought dual wielding sucked unless you're a rogue, and now I've done the AC calculations I can't see that -2 AB ever being worth it, so your changes sound like a good idea.

Why weaken enchantments on light weapons though, if there's no penalty to dual wielding heavy one-handed weapons? DEX builds may have good defences, but their damage output sucks already and fights that last forever don't tend to be much fun.

Why not just include a bunch of shields with non-AC enchantments? Spells cast, skill bonuses, stat boosts, whatever. Should be easy enough to create a sense of progression, and you can balance things out so the bonuses on your shields and one-handed weapons combined are about as good as those on a 2-handed weapon.

Of course, Clerics can still cast Magic Vestment and boost their AC into the stratosphere in low-magic settings. You can fix that with a hak or something, but spells are a whole other story of balance-issues.

I was just referring to what you did with weapon enchantments, i.e. giving more damage to big weapons makes them more competitive. I didn't mean in general, of course.

(Big damage enchantments do bolster dual wielding over a two-handed weapon.)

Well, 28% isn't to be sniffed at. And RDDs gain even more. But yeah, in general it's still a bad trade-off. Especially because 95% of enemies in 99% of modules do use direct physical attacks.

Re shields:

Yeah, the gnomes using longswords two-handed does mess things up. Honestly, I'd be tempted to just consider them a casualty of expedience. Very few people play Strength-based gnome or halfling fighters/clerics using two-handed weapons anyway.

Same for people keeping the off-hand empty. That's always going to be sub-optimal. Characters who do that generally are going to be less front-line oriented anyway. (Except monks, admittedly.)

Other than that, giving two-handed weapons the same kind of bonuses as two items held in a single hand doesn't seem too impossible.

The way I understand it, your issue is this: A shield's +2 or 3 AC is always going to help a character immensely, but people don't want to use the same shield their whole career, nor do they want a non-magical shield in epic levels.

I can think of a couple of potential solutions:

1] Give the shields "cosmetic" properties. A shield that casts Barkskin and Identify won't unbalance anything even at lowish levels, (It's basically free potions) but finding it still gives a player a new toy. A shield that casts Combust on people who hit you is pretty cool, but at high levels it won't make a big difference. Bonus spell slots are great for clerics or druids, but at high level they typically are more convenience than a raw power boost if you have infinite rests.

2] Give more substantial stat boosts, but tie them to the item type. I.E. give all shields some kind of progressive CON bonus, but no CON bonuses on other items. Give STR bonuses only on weapons, with the bigger bonuses only appearing only on two-handed ones. Maybe DEX bonuses on light weapons. This will make shield characters more defensive and dual-wielders and greatsword-wielders more offensive, which makes sense. It would encourage STR based dual-wielders to use medium weapons, which isn't a bad thing, I think, with your tweaks to dual-wielding.

Downside is the +12 cap. With items alone that's not too big a difference, but spellcasters will already be getting +5-+7 from empowered/maximised animal spells, and that doesn't leave much wriggle room. Of course, you can just ignore the spellcasters.

3] Give shields various damage resistance/immunity properties instead. That's what you did in your mod, I saw. No reason these can't be progressive. Even if you don't want to give big damage immunities/resistances, you can still add protection against more types of damage as you go along. Not very exciting, though.

4] Give shields negative properties. If all shields come with a STR penalty, deciding to use them becomes more of a trade-off. You could also penalise saves, or even give negative damage immunity against certain attacks.

This also covers the people wanting to use a single weapon. This would then give some more wriggling room to give them positive stats to compensate.

Of course, people can just unequip shields when they're not being attacked, so in a sense this would make them a little like the expertise feat. Still would make them less of a no-brainer though.

Unfortunately, it often is. The vast majority of NWN content is tuned around "Not dying" and demands/expects very little DPS. How many modules/worlds have you seen where you could walk around with hundreds of healing potions -- and often even hundreds of full Heal potions? As long as you manage to not die in a given burst window (and/or manage to outheal the incoming damage overall) you can eventually wear the enemy down.

I'm tempted too. But given the type of game NWN is and the audience it generally attracts, I am extremely hesitant to disrupt that option. It's clearly non-optimal but people do a lot of non-optimal things.

More or less, yes. The exact AC amount is of less concern too -- could argue, for example, that a Tower Shield +1 (4 AC) is the perfect balance between the offense of a 2H or DW and the defense of a shield. But even if that's true, then you're still using a Tower Shield +1 for all 40 levels.

As you noticed, I actually decided to give shields a small immunity bonus instead -- so that A, the AC gap remains smaller while still providing roughly the physical defense I desired and B, the shield offers some protection against non-physical attacks too (which can also include enchanted weapons).

The Barkskin/Identify worries me as I don't want the player to feel obliged to hold on to that shield they got at level 4 for another 30 levels either.

You'd have to define infinite rests -- do you mean unrestricted resting instead? Rest anywhere you want at any time? This is also a "Difficulty matters" kind of deal -- on easy stuff, sure, the spell slots make no difference. On hard stuff the spell slots make a big difference.

You're also forgetting buff potions, class buffs (like Barbarian Rage), and the fact the spellcasters can buff the melee.

The amount I gave at 40 was basically the max I thought they should have, which doesn't leave much room to scale from 1 to 40. And by default you can't even work in increments of less than 5%.

The adding protection against more types is tricky too -- it might be possible to do something like "physical only" to "physical and elemental" to "physical and elemental and exotic." But I'd want to avoid having resistance to Fire but not Cold or Slashing but not Piercing.

Except the player using an "Immunity: ability drain" item and could ignore that penalty. Something like negative saves would be possible but then are you talking about making the penalty worse and worse as the shield gains more AC or something?