Slawbug

The poison attack isn't that powerful. The poisoning is comparable to a Venom Baton, and the attack itself is comparable to a regular Vlish attack -- except it doesn't slow.

The real problem is that Terror Vlish cost three times as much as regular Vlish. They are a few levels higher, but they also have a reduced attack bonus, so they aren't really any more powerful.

The one thing they have going for them, IMHO, is that their attack does poison-elemental damage -- it's not just a physical attack that also poisons. This is useful against a few enemies, like mages and gazers, that resist most forms of attack, but not poison.

There seems to be a rough consensus that Khyryk is the best part of G3.

His dialog options are great. They also are very different from the standard iloveserviles vs. ihateserviles options. They aren't even on the same continuum. Include more of those, by all means!

Hurray!

Honestly, the dialogue choices weren't much worse in G3 than in previous games. And I liked them in G1. It's just that longtime Geneforge fans found themselves answering the same questions for the kazillionth time.

Hurray!

The formulas for damage and hit rate for ANY type of attack made by any character are very simple.

Base hit rate is determined by type of attack and has mostly stayed the same throughout G1-3. Basic rule of thumb: 60% physical, 60% fire, 70% magic/acid, 80% ice. Then add 5% per point of attack bonus. Enemy ability to dodge can affect this, I believe reducing it by 5% per point of Dexterity -- not sure.

Base damage is attack specific, with an additional amount per level of attack bonus. In G3, Firebolt has a base of 1-4 and gets an additional 1-3 per level of attack bonus. The enemy's armor or resistance is then applied to the total amount rolled.

To get your attack bonus:

PC Melee attacks:

Bonus = Strength + Melee Weapons + Level of equipped weapon

PC Missile attacks:

Bonus = Dexterity + Missile Weapons + Level of item used

PC Magic attacks:

Bonus = Spellcraft + Battle Magic + Spell Skill

Creation attacks (melee OR missile):

Bonus = Strength + Appropriate attack Skill

Attack skill is a hidden stat, but it is generally in line with creation power level. Not always, though -- Fyoras have a better melee attack skill than Plated Bugs do.

Note that the descriptions for many of the stats involved are inaccurate in both the manual and the game. Strength and Dexterity come to mind.

Also, note that the attractive thing about missile weapons is that they tend to have higher base levels compared to spell skill. They won't be scaled back anytime soon, though. They are -- at best -- mildly more powerful than magic, but they have the huge restriction of being limited to what items you collect and save. Reapers are nice, but even they are typically weaker than a Gazer.

So, after being inspired to start up another G3 game, I was poking around the scripts, looking at how the game balance has changed. I didn't realize just how widespread the changes have been. It's not just melee weapons that were weakened. Here's a list of damage per level across games:

G1    G2    G31-6   1-4   1-3   Firebolt, Fiery Wand (Fyora)1-6   1-6   1-6   (Drayk)1-7   1-5   1-4   Searer (Artila/Roamer)1-5   1-6   1-5   Ice Spray, Icy Crystal (Cryoa/Cryodrayk)1-8   1-4   1-5   Searing Orbs, Spray Crystal (Drakon)1-8   1-6   1-6   Diamond Spray, Swarm Crystal, Jeweled Wand (Ur-Drakon)1-5   1-5   1-5   (Vlish)      1-12  1-10  Kill (Gazer)      1-12  1-10  Aura of Flames (Eyebeast)1-8   1-8   1-4   Regular melee attack1-3   1-5   1-4   Vlish melee attack            1-5   Alwan's melee attack1-5   1-6   1-3   Thorn Baton1-6   1-6   1-3   Venom Baton1-8   1-5   1-3   Acid Baton1-10  1-7   1-3   Submission Baton1-14  1-10  1-12  Reaper Baton1-6   1-5   1-4   Javelin1-16  1-12  1-8   Discipline Wand1-5   1-4   1-4   Terror Wand1-5   1-4   1-4   Ensnaring Crystal, Null Wand

This brings up a few interesting points. For one thing, whatever effect the weakening of melee has had on guardians, it's completely nerfed most battle creations. If you look at their stats, Clawbugs and Battle Alphas are pretty much complete crap next to comparable Magic and Fire creations, despite being more expensive -- and that was much less true when physical attacks hit harder. Thahds get a HP bonus that's noticeable early and Rotghroths get bonuses to strength and quick action, though they still don't impress me.

This also reveals just how Vlish became so overpowered. Originally, all their attacks were noticably weaker than those of a Fyora of comparable level; that was to make up for the powerful slowing effect. Now their melee attack is on par with a Clawbug's, while their missile attack is stronger than a Roamer's, and as good as Ice Spray with a more consistent crippling effect. Drayks have also gotten somewhat better.

Reapers are spectacularly good in G3. I knew that, I just didn't realize how good.

They have neither the sprawling, messy power of a shaper, nor the elegant efficiency of an agent.

Of course, if this poll had happened after G2, guardians would have gotten lots of votes, thanks to being broken.

I agree with most of what's been said, but I want to point out that the first two games were a lot better on this front.

The Awakened were crazy in G2, sure, but not in G1. G1 also had some respectable Sholai.

In G2, there was a non-aligned option. Although it was never explicitly outlined, it had its own ending, and I found playing it quite satisfying.

Also, I think it's interesting that Jeff has only made ONE game in the past decade that hasn't made you pick sides, and that was arguably Jeff's least-plotful game -- Avernum 4. (The lack of picking sides in the other Avernums was predetermined based on Exile.) Nethergate and G1-3 make you pick, as does one of the mainline scenarios in BoE/BoA.

There are still a bunch of Exile series pages out there. Do a google search.

*nod*

Normally, I would never advocate such a tactic. However, it is actually pretty useful as a singleton on Torment. As a singleton, coming up with the gold to train is no problem whatsoever. And as a singleton on Torment, every skill point you get is critical -- there are basically none to spare.

*nod*

Normally, I would never advocate such a tactic. However, it is actually pretty useful as a singleton on Torment. As a singleton, coming up with the gold to train is no problem whatsoever. And as a singleton on Torment, every skill point you get is critical -- there are basically none to spare.

Magma, the ONLY differences between the classes are the cost to train in different skills, the base skills you start with, and slightly different formulas for HP, energy, and essence.

A guardian with 8 battle magic and an agent with 8 battle magic will cast exactly the same strength battle magic spells. But it will take way more skill points to get the guardian there. Therefore, magic-reliant guardians are worse than magic-reliant agents. Using the editor ignores almost all distinctions between classes. Do you understand why your comments about a character you created by breaking the rules of the game are not relevant in a discussion about characters that follow the rules of the game?

If not, allow me to quote myself:

This time, it really is me you're thinking of.

Agents aren't actually better at melee skills. (They are better at missile skills, which is silly, but hey.) They are nearly as good at melee. What makes them better for a melee build is the truly horrific number of skill points a guardian has to plunk down in order to access the good pump spells, which are key for any good melee fighter.

If you are an aggressive min-maxer, custom is still the way to go; but even then it's a close call. Certainly, few of the preset templates are actually bad.

If you are an aggressive min-maxer, custom is still the way to go; but even then it's a close call. Certainly, few of the preset templates are actually bad.

A guardian who focuses on magic is significantly worse than a melee agent in a number of ways. It's just not a good build.

Unfortunately, I'm also reconsidering the shaping agent. I did some more math, and things aren't as pretty as I'd hoped. I forgot that G2 and G3 still implement 10- and 20- caps on shaping skill effects. This means that, given reasonable skill point investment, a shaper will use 34 skill points to reach 10 in magic shaping, versus an agent using 70 skill points. The problem is that Create Vlish typically becomes available before you've accrued much more than 70 skill points, yet without points into Intelligence, an agent can basically make one vlish at that point. The agent can put more points into Intelligence, but subsequent vlish lose out on levels gained through experience. Meanwhile, the comparable shaper is done on this front and gets ~20 skill points while the agent is catching up, making up for the shaper's more critical lags in magic skills. So the shaper gets extra essence forever and slightly better vlish (averaging 1-3 levels better, depending on how many you make), whereas the agent gets slightly better stats in a few random places.

And I had such a good name, too -- Sss-Vlish.

Vlish and others have posted a number of interesting character builds since the release of G3.

Shaper - Deadweight

Shaper - blessing/dazing magic user

Shaper - Missile

Guardian - Missile

Guardian - Melee

Agent - Melee

Agent - all out mental magic

Agent - all out battle magic

Have I missed any? (Noobs, this is not an invitation to post "I made a guardian with skills X and Y" or whatever. Please.) What else is out there?

THE SHAPING AGENT

One type of build that hasn't gotten so much investigation is guardians and agents relying on shaping. I was looking into the mechanics of shaped creature stats today, and I think shapers have less of a monopoly there than is commonly assumed. (Certainly less than I've always assumed.) This is because a cheap, weak creation, when levelled up appropriate, is largely indistinguishable from better creations.

The first hurdle is essence: guardians and agents get 75% the essence of shapers, plus shapers get a bonus to Intelligence. This cramps shaping early on, but if you aren't a shaper, you can just use usual guardian or agent tactics to survive early. Later, this may make it impractical to deploy Eyebeasts or other overpriced creations, but there's nothing stopping you from using, say, a bunch of Vlish.

The other hurdle is creation strength. Agents especially won't be pumping their shaping stats too high. This, too, can be compensated for by relying on weak creations, making them early, and allowing them to level up with you. They won't be as strong as pimped out shaper creations, but they'll be perfectly usable.

This is, admittedly, a more strategically demanding build than a pure shaper is. There are potential advantages, though. Let's look at an agent using vlish. The agent will be 3-5 skill levels behind the shaper in Magic Shaping, for any given point investment. That skill, however, is only relevant when you first make the creations, so you can push it early and then ignore it. OTOH, the agent will be 3-4 levels ahead in missile ability (missiles + dex) for the same investment. The agent will also be 2-3 levels ahead in mental magic and 1-2 levels ahead in blessing magic, plus 1-2 points ahead in spellcraft. There are some other differences on both sides (quick action, battle magic, healing craft, etc) but those can be worked around.

The main reason to go through all that is that (unlike previous games) mental magic requires dedicated attention for it to be effective throughout the game. With Daze and other spells at full effectiveness, you end up with a team of creations that basically use agent tactics to survive. The vlish can incapacitate several creatures easily; the agent, with some points in Quick Action, can pre-emptively daze anything that's fast or close enough to threaten the safety of the vlish, as well as using missile and wands in tough spots. Your creations will be weaker, but you end up with more tactical options.

My plan is to complete the first island spending as few skill points as possible -- probably using items to help it along. As soon as I get to Harmony, make a beeline for San Ru and pick up Create Vlish. Dump the skill points into some appropriate combination of Intelligence and Magic Shaping, and make a squadron of vlish. Fill them out a little over the next few levels, then start putting points where they belong, probably starting with Mental Magic. After a little while, life is good.

I'm probably overlooking something, of course.

Oh man. That's great. Nice discovery.

Do you know if that works on all versions?

There goes another eminently useful post.

I believe she also had a gender change along the way. She was a he in G1, or something like that.

Playing out of character is already an issue. If you party is all sliths or all nephils, there are frequent situations where the dialog makes no sense and you just have to pretend you're humans. There are workarounds, as (exhaustively) demonstrated in Kel's scenario Bahssikava, but I'm not sure the effort would be worth it, for a full-sized game that focuses on humans.

Troglos, I think, could actually work; they can communicate with humans, they are the same size, have similar bodily needs, and are of comparable capabilities. But they wouldn't add anything to the game, and it would be odd.