Not sure if anyone else has seen them yet, but the preview heroes seem a bit off in terms of conquest values. To review:

Challara : 12 Wounds, 0 Armor, 3 CT, has familiar that does a WG breath attack once per area
Brother Gherinn : 12 Wounds, 2 Armor, 3 CT, Can sacrifice 1 wound to Aim when attacking (before rolling)
Lindel : 16 Wounds, 1 Armor, 5(!) CT, Can swap out one skill for a new random one of the same type once per area

While CT values are still somewhat of a mystery (despite repeated attempts to reverse-engineer the formula with a spreadsheet), these values make little sense given the extant characters. Neither Gherinn nor Lindel have a precedent for their CT values, while Challara is basically just as strong/tough as Astarra only more expenise and slower (the familiar is a neat but tactically insignificant addition. It doesn't add to her per-turn damage output or her survivability unless it's in a position to take out Skeletons and doesn't fail).

I just don't see how Challara and Lindel are supposed to be anything more than a CT burden, while Gherinn is clearly in the top tier of CT 3 heroes in terms of survivability and damage output. Anyone seeing something I'm missing?

Well, we know the following about conquest values:

• Kevin W's custom hero rules suggest he considers the baseline defenses for a 2 conquest hero to be 8/0, for a 3 conquest hero to be 12/1, and for a 4 conquest hero to be 16/2.
• Defenses of 16/1, 12/0, and 8/1 have all been rounded both ways in the past...with mixed results. 12/2 has only ever gone to 4 before, but we only have 3 example heroes (Jaes, One Fist, Nara). And 16/1 going to ct 4 only happened in expansions, not in base game heroes.
• It seems FFG occasionally tries to compensate for offensive ability power with a higher conquest value, even though this is widely considered a bad idea. See: Laurel of Bloodwood, the only hero with 8/1 ct 3. (Note: I consider this practice reasonable for determining rounding, but not for large conquest changes.)

So, if I were to speculate on the reasoning for the conquest values of these heroes, I'd say:

Challara: FFG considers defenses of 12/0 to be a borderline conquest value of 2, and feel justified in rounding up to 3 if the hero has a defensive ability (Ispher, Shiver). Someone considers Challara's ability to be very powerful (I'm not sure I disagree; 58% chance per area to one-shot a breath template of tier 1 monsters in a 4-hero game), so they pulled another Laurel and raised her conquest as a penalty for offensive potential.

In my opinion, 12/0 is better than 8/1, so Challara is probably more justified than Laurel, and could be reasonably balanced, depending on how useful that ability actually turns out to be. Challara will also change wildly in power depending on game size, but since she's got practically the only hero ability that's actually better in smaller parties, I probably won't complain too loudly.

Brother Gherinn: FFG considers defenses of 12/2 to be a weak conquest value of 4 (since 16/2 is "normal" for 4 conquest). Gherinn's ability requires him to sacrifice health; this means he'll be easier to kill (on balance) than a hero with a non-sacrificial ability, so they wanted to err on the low side for the conquest value. Additionally, Jaes and Nara have a high stat total (13 vs. the typical 12, dividing wounds by 4), probably to compensate for being on the weak end of their conquest range; Gherinn has a low stat total (11), probably to compensate for being on the strong end of his. (One Fist's stats only sum to 12, but he's got a pretty strong ability.)

I actually think this hero looks nicely balanced; if sacrificing a health is about the same as spending a fatigue, he's almost exactly at the target price in my editor (358/360 BP).

Lindel: Someone designing this hero thoroughly believes that raising a hero's conquest value in exchange for giving them more power is a good idea. They probably also believe they made Lindel more powerful than they actually did, because for some reason everyone freaks out when they see two dice in every trait as if that were "monstrously overpowered" rather than "about as good as 3 dice in one trait".

Though based on the standards of my hero editor , she's actually frighteningly close to being on budget...if my editor allowed you to get a conquest value that high with those stats. In fact, if you assume her ability is the same cost as "Lucky Skills" (draw twice the listed starting number of skills in each category, then discard down to the listed numbers), I believe she's exactly 360 BP, my recommended limit and the median cost of recreating published heroes.

OK, now I'm wondering if the people who made these heroes looked at my editor. Published FFG heroes range in cost from 282 (Red Scorpion) to 439 (Astarra) in my editor; even ignoring Zyla and looking only at the other 5 heroes from the last expansion, they range from 321 (Shiver) to 399 (Okaluk). But these three heroes could all be within 5 BP of my recommended price limit with plausible estimates for their ability costs. Despite some weird conquest values, that's much more consistent than FFG has managed in the past if my cost scheme is even vaguely accurate.

They used Antistone's formula's and still screwed it up badly!

I'm willing to call it a step in the right direction. Though there may be some ego involved in that opinion.

Consider Zyla: she'd still make a competitive hero if you raised her conquest value to 2. Giving her a conquest of 1 was not only unprecedented, it was also gratuitous overpowering. At least with Lindel, he'd clearly be above-average power with a conquest value of 4, so they had some rational reason to raise him to 5, even if I think they should have given him a different penalty instead.

I'm not super clear on why heroes with minor offensive advantages should get any kind of CT consideration. I know it's easier to survive when monsters die, but most heroes die from being jumped by a timely spawn/trap card and not because they left too many monsters on the board.

The alternative is basically to say that half-way stat lines like 16/1 can't exist AT ALL , because if 12/1 is conquest 3 and 16/2 is conquest 4, then 16/1 can't have any conquest value at all, because it has to be higher than 3 and lower than 4 and only whole numbers are allowed.

The only way you can even let a 16/1 hero through the door is if you're willing to say that he can have a conquest of 3, balanced by some disadvantage in another area (compared to a 12/1 hero), or a conquest value of 4, balanced by some advantage in another area (compared to a 16/2 hero). You don't absolutely have to allow both of those, but if you're going to allow either, I don't see any particularly convincing reason that you couldn't allow both.

And once you've allowed that, you've established some sort of benchmark for trading off between conquest and other abilities, so we have at least a general idea of how a 16/1 hero with a conquest value of 5 (or 2) might want to look, even if such a creation isn't a good idea.

The reasons I think it's a bad idea are:

• I think the true value of a higher or lower conquest value is probably complex, nonlinear, and dependent on a lot of external variables (like the other heroes in the party), so the approximations we use for small changes have a good chance of giving big errors for larger changes.
• Having heroes with widly different defense/value ratios in the same party skews the strategy, which makes the game harder to balance in general, and I don't think Descent is designed to withstand it.
• Having heroes with wildly different defense/value ratios in the same party also makes the game more volatile, and I think it's already too volatile (hence The Enduring Evil).

And yes, the game might well be more balanced overall if we just got rid of 16/1 heroes and their ilk entirely, but we sacrifice balance to get a wider variety of content all the time, and more importantly, we already have a bunch of 16/1 heroes in the game.