I read that note as pertaining only to the Attack Matrix for Fighters etc., clarifying how NPCs of various races do or don't use the table, so standard MM elves, dwarves (both hill and mountain), and gnomes all attack
as 1st-level fighters (THAC0 20) regardless of their actual HD and
regardless of whether they are considered equivalent to 1st-level fighters in some other respect. I.e. they might still be considered as "unclassed" individuals. Whereas, unclassed humans and halflings attack using the "0" column.
I was thinking of this passage from
Swords & Spells (1976), p 6:
NOTE: Scale figures representing human/humanoid (and highly intelligent) creatures of 1st level or above or with 1+1 hit dice are always considered as having elite guard status. For example, the following types of troops are classified as elite guard status:
Veterans (1st level fighters)
Elves
Hobgoblins
Gnolls
This ties in with the Loyalty Base Modifiers given on DMG, p 36. Under "Training Or Status Level" are the following:
trained regulars ... +10%
elite, sub-officers, minor officials/expert hirelings ... +20%
guards, officers, or major officials/henchmen ... +30%
My thinking is the distinction between hirelings and henchmen is relevant. I.e. henchmen are classed individuals while hirelings are not. Thus, 1+1 HD MM demi-humans, like elves and mountain dwarves (but not hill dwarves), not only attack as 1st-level fighters on the attack matrix, but they also have a status/training classification that is the equivalent of that of a classed character of at least 1st level. Because MM hill dwarves lack that status due to their inferior HD is why I use the term "level 0" to refer to them, by which I mean "unclassed".
I'm not sure there's really a 'rule', but overall it is pretty messy, actually. So, in Chainmail there were basically grades of forces, light foot, heavy foot, armored foot, and then light, medium, and heavy horse. Furthermore peasant/levy is a sort of subcategory of light foot (even worse). In D&D this got translated into hit dice, with 'peasants' being basically 1d6-1 hit points, and then 'normal soldiers' (veterans) having 1d6, and thus the 'heavy' types had to be 1d6+1. Fighting Men then got a d8 for hit points instead of d6+1 for whatever reason. The Monster Manual is an unreliable guide to ANYTHING because it was written BEFORE 1e AD&D, and thus we don't even know what a 'hit die' meant, it was defined later!
So, is an AD&D 1e goblin basically a 0 level figure? Yeah, kinda. In the case of a dwarf we have no way of knowing, but AS ORIGINALLY ENVISAGED a 'peasant' has a 1d6 hit points, maybe with a -1 but we don't know for sure as D&D doesn't have ordinary people stats. At some point during the development of the Monster Manual hit dice were redefined to a 1d8 (presumably this marks the point in Gary's development of the PHB rules where Fighters got a 1d10 hit die). It is widely accepted that at this point 'zero level humans' (IE peasants) have a 1d6 still, but that's kind of an assumption AFAIK. I don't think the DMG or MM actually define 'zero level' as a thing. There's a section called 'typical inhabitants' on p88 of the DMG which lists hit points for various sorts of 'people', but mentions nothing about race, just that they are 'non-classed'. The combat tables have a column under the fighter matrix for level 0, and here is the only place where '0 level' is defined (in a footnote which is where your 'humans and halflings only' comes from). This same footnote explains that dwarves, elves, and gnomes are 'never below first level' but what this means, especially in terms of the 'ordinary inhabitants' rules, is never explained.
I have a copy of Chainmail, and have read most of it. I've never owned or read Swords & Spells. And I've never fully tried to comprehend the translation of these various elements - HD, and associated combatant/training status - from system to system.
I do know that AD&D presents multiple options/frameworks, without explaining how to choose between them. Many of the "Men" in the Monster Manual have 1d6 hp, which is statistically close to Goblins' 1d8-1 (average 3.5 vs 3.635). Halflings also have 1d6 hp. Presumably these are the 0-level characters the DM refers to: but whether they attack on the 0-level column or the less than 1-1 HD column doesn't matter, as those two columns are the same. But "Men" or Halflings vs Goblins have a hard time of it: while Goblins have nearly identical average hp, they have a +1 to attack (due to attacking on the 1-1 HD column, which is the same as the 1st level fighter column).
(In Moldvay Basic Halflings, like Goblins, have 1-1 HD. And in Expert the "Men" have 1 HD. I remember being struck at the time, when moving from B/X to AD&D, that the "Men" were debuffed in the latter system.)
Then there's the p 88 table in the DMG, which gives sex-based hp. An active but not labouring male peasant has 1d4+1 hp, which has the same average as 1d6. A male labourer has 1d6+1 (the same average as 1 HD) but gets stuck on the 0-level attack chart.
Then there is the DMG information about mercenary soldiers and man-at-arms. According to p 30 of the DMG, a man-at-arms has 4-7 (1d4 +3) hp, which is the same average as 1d10; but attacks as 0-level rather than as a 1st level fighter. A 1st level sergeant, on the other hand, has the same average hp though a wider spread, and a better attack bonus. The chart on p 100 of the DMG notes that a man-at-arms has +1 to STR and +3 to CON, although whether that means they get CON modifiers to their hp is left quite ambiguous. But it does say that they have a minimum of 4 hp, though what die is supposed to be rolled is left unstated on that particular page. (For completeness, the p 100 chart gives labourers a +1 to +3 to STR.)
It's also not really clear why men-at-arms are around 1.5 times as resilient as bandits, brigands, buccaneers and pirates, but have the same chance to hit. Nor why they are similarly more robust than Goblins, but have a lesser chance to hit. Nor why Goblins have the same chance to hit as sergeants, although they are slightly easier to cut down with a single blow.
There are other points of arbitrariness as well. The monster attack chart in the DMG says that "Any plus [to HD] above +3 equals another hit die, ie 6+6 equals 7 hit dice". There is a similar, though not identical, note beneath the saving throw table that says that, for monsters,
Further die levels are added for each increment of four additional points. Therefore, for the purpose of determining saving throw levels, 1+1 though 1+4 hit dice becomes 2, 1+5 through 1+8 becomes 3, 2+1 through 2+4 also becomes 3, 2+5 through 2+8 becomes 4, etc.
So a 6+6 HD Troll is considered 7 HD for attack purposes (making no difference, as there is a single column for 6-7+) and 8 HD/levels for saving throw purposes. A 7+7 HD Type V demon is considered 8 HD for attack purposes (stepping up one column), and 9 HD/levels for saving throw purposes. And a 8+8 Type VI demon is considered 9 HD for attack purposes (making no difference, as there is a single column for 8-9+) and 10 HD/levels for saving throw purposes.
Having the Troll as either 7 or 8 HD, the Type V demon as 9 HD, and the Type V demon as 10 HD would make almost no difference to the game, but eliminate the confusion of pluses to hit dice, and make those table footnotes redundant. The 1+1 category seems rather arbitrary (and Swords & Spells apparently eliminates it, grouping Hobgoblins with Gnolls if I've read
@Hriston's post correctly). And things like black, brown and cave bears as 3+3, 5+5 and 6+6 rather than just 4, 6 and 7 or 8 are arbitrary. Or making a basilisk 6+1 HD - what is the point of that extra +1 except to create confusion as to which saving throw numbers to use?
To be honest, it's all a bit of a mess. It's hard to take seriously that any of this minutiae is a "simulation" of anything: the differences in hp and to hit numbers of these various peasants, man-at-arms, bandits, sergeants, Goblins, etc all seem rather arbitrary, and to reflect various pathways out of Chainmail and into the AD&D framework, none of which was every fully thought out and some of which (like the p 88 "Typical Inhabitants" chart) reflect a growing concern with "simulationism" that has not been integrated into other elements of the system.
If you mean "there's lots of weird, inconsistent rules which make it hard to state anything categorically about AD&D" then I'd agree, but I don't think it invalidates the wider point. YMMV.
And I don't disagree with this either. I'm pointing out that some people like a predictable, shared mechanical model for their PCs and NPCs. And that such a model is frequently - albeit inconsistently - employed in AD&D.
Honestly, I'm having trouble understanding why stating the case for NPCs often using the same mechanisms as PCs in AD&D is receiving such pushback. It seems fairly self-evident to me.
Well, see above for the basis for my view.
In the end, all we have is that NPCs and monsters, like PCs, have attack numbers and saving throws determined on a chart, and have a certain number of HD (of some or other size) rolled to determine hit points. There is some loose correlation between attack numbers, saving throws and HD but it is not precise and in some cases is not even straightforward to work out for that particular NPC or creature.
My point is that if we have rules, they should consistently be linked to fictional elements. If PC elves have dark vision of 60 feet, then NPC elves probably should too, and it should work in the same way.
You are pointing to a rule, here, that is in very close contact with the fiction:
dark vision to 60 feet is an in-fiction property of Elves. This is quite different from things like HD, hp, saving throw numbers, classes, levels etc.
But even in this sort of case, there is precedent for non-uniformity. For instance, while Drow have the in-fiction property of being resistant to magic, PC Drow lose that ability (with a bit of patter to explain why - they are no longer exposed to the "strange radiations" of their under-earth homelands). If we wanted to, some similar sort of story could be told to explain why PC Elves don't have dark vision as typical Elves do.
(EDIT:
@AbdulAlhazred beat me to this point.)
In my current D&D game, I basically assume that most powerful people basically have class levels. It helps me to gauge on the fly what their capabilities might be even if I had not made full datasheets for them. If I know that the eldri shaman is a sixth level shepherd druid, or that the leader of the Corpsecrafter's Guild is a 13th level necromancer then I already have rough ballpark for their powers. It is not exact, and small deviations are fine by me, but I find these sort of rough guideline helpful and they also set limits to me as a GM. I can't just suddenly decide that this friendly NPC can totally cast a helpful spell the PCs need, because I have already established that they're of a level that has no access to it.
On this, I agree with
@Bedrockgames: it produces weird or constraining outcomes, such as no one being able to work miracles unless they are also a mighty warrior.
This relates back to
@AbdulAlhazred's comment about whether or not the pirate MU can wield a cutlass: AD&D most strikingly, but many other versions of D&D too, simply don't tell us which parts of the PC build rules represent in-fiction "lifepath-y" stuff (eg
no one can wield the magic of raising the dead, before they have first learned the simpler art of curing light wounds) and which are meta-game constraints intended to secure game balance, variation and niche-protection among PC options, etc.
