monster ability score question -- players of rick's untitled campaign stay out :-)

[rrwoods]

The "improving monsters" section of the SRD (http://www.d20srd.org/srd/improvingMonsters.htm) mentions three ways in which a DM can determine the ability scores of a monster (or other NPC, I assume). Those ways are average array (10s/11s across the board), nonelite array (13/12/11/10/9/8), and elite array (15/14/13/12/10/8), each as modified by racial bonuses/penalties. From what I understand from that (and from how monsters as characters typically work), this means that the given ability scores for a monster entry reflect racial modifiers; that is, their racial modifiers are essentially how far you need to go to get from 10 or 11 to the listed score. Of course this lets you customize a monster by specializing it (nonelite array plus racial) or making it better (elite array plus racial).

If I'm misunderstanding anything in the above paragraph, tell me, because my question is based on those assumptions.

Magic of Incarnum has this nifty "Lost" template (page 183) that I want to use. The ability score modifiers are Str +4, Con +4, Int -6. The first sample Lost they give is based on a human commoner 1 has listed ability scores of 14/10/14/2/8/8, suggesting initially assigned scores of 10/10/10/8/8/8 -- a "below average" array. My main question is: Is this a typical case of a stat block simply being inconsistent with the rules, or is there's something special (possibly about statting commoners in particular) that isn't present in what I've read?

As a side question, Lost is a CR +1 template, and their sample lost is CR 1/2. I assume, then, that a "normal" human commoner 1 is CR 1/4 (or whatever 1/2 - 1 works out to in fractional-CR-land)? I have my doubts, since their other sample lost is a nymph barbarian 3 which is somehow CR 9 (7 + 3 + 1 = ... 9?). Either way I'm not too too worries about it since I know the CR/EL system is pretty messed up anyway, and I'll be taking my PCs' abilities into account for sure.

[ExLibrisMortis]

Commoners don't have worse stats or anything, so I have no idea why they used those stats. They're all even, too, which is unusual - WotC stat blocks usually have three odd and three even. I guess they forgot to assign three threes. You can't go below 3 int as a human, either.

As for the barbarian, it might be that they added the barbarian levels as non-associated class levels, for 1 1/2 CR, rounded down to 1.

[rrwoods]

OK, good to know I'm not crazy then.

I hadn't heard of the "non-associated" class levels rule, but it makes some amount of sense.

[FocusWolf413]

Don't pay too much attention to the given stat blocks. They aren't going to do anything for you. WotC is notoriously bad at providing consistent rules when it comes to monsters and stuff. There's holes, inconsistencies, and things that just outright defy all sense or logic. If you want to build something with a given template, don't worry about the stats too much. Just have fun with it. Feel free to roll 7d6 for monster stats, or to send players up against things above their CR.

[Chronos]

Associated and non-associated class levels (http://www.d20srd.org/srd/improvingMonsters.htm#associatedClassLevels)
Basically, if a class is one that plays to a monster's particular strengths, then it's an associated class, and each level adds +1 to the CR, and if it doesn't play to its strengths, it's non-associated, and only adds +1/2. Nymphs aren't really much good at melee combat, so barbarian is nonassociated for them. An ogre, though, is right in line with barbarian, so it's associated for them. Bard, sorcerer, and marshal would probably all be associated for a nymph, because they leverage her super-high charisma, and druid is also associated, because they already have druid casting and it stacks.