Open up to the Vol-2 Monster Reference Table and you'll see a column for "Number Appearing*", with entries such as 30-300 men/orcs, 40-400 goblins/kobolds, 20-200 hobgoblins/gnolls, etc. Then there's a footnote that says:
*Referee's option: Increase or decrease according to party concerned (used primarily only for out-door encounters). [Vol-2, p. 4]So, these numbers are primarily not for dungeon encounters. Where to find encounter numbers for that purpose? You might consider in the DM's booklet under "Distribution of Monsters and Treasure":
Roll the die for every room or space not already allocated. A roll of a 1 or 2 indicates that there is some monster there. The monster(s) can be selected by use of the Monster Determination & Level of Monster Matrix which is given later in this booklet. The number of monsters is best determined by the level being considered and the kind of monster inhabiting the room or space. The Monster Table from Volume II can be most helpful here. [Vol-3, p. 7]Thus, we are to determine monster numbers "by the level being considered and the kind of monster", which is about as vague as you can get. If you follow that latter reference back to the table in Vol-2, of course, you wind up again looking at the "used primarily only for out-door encounters" dictum. So let's try pursuing the forward reference to the "Monster Determination" table, after which it says this:
Number of Wandering Monsters Appearing: If the level beneath the surface roughly corresponds with the level of the monster then the number of monsters will be based on a single creature, modified by type (that is Orcs and the like will be in groups) and the number of adventurers in the party. A party of from 1-3 would draw the basic number of monsters, 4-6 would bring about twice as many, and so on. The referee is advised to exercise his discretion in regard to exact determinations, for the number of variables is too great to make a hard and fast rule... [Vol-3, p. 11-12]This paragraph goes off on a wild tangent after this point, saying, "There can be places where 300 Hobgoblins dwell," and then ruminating about how many can fit abreast in a corridor at once ("Allow perhaps 3 in a ten foot wide passage.") But do we get any specific recommendations? No, we don't. (This is actually the kind of writing that happened a lot more in 2E, and which I would complain bitterly about.) Let's say we've generated some Orcs, since they're mentioned above. We know they will not be a single creature, but rather in a "group". What is the group size? The world may never know. (Other than "greater than one", which provides a rather large amount of wiggle room.).
This is a place where the BXCMI line might have some advantage, because it explicitly includes separate figures for dungeon-versus-wilderness settings in its "Number Appearing" entries. Other lines tend to include specific numbers only as part of their dungeon encounter tables: see Holmes or the AD&D DMG. (The AD&D Monster Manual, as usual, just copies forward the numbers from the Vol-2 Monster Reference Table, with the equivalent warning note, "It is not generally recommended for use in establishing the population of dungeon levels." [MM p. 5])
Let's consider those encounter tables a little bit more. In the lines that do have specific numbers in their tables, non-Gygax products tend have very small base numbers. If we look at the 1st-level tables (for example) in places like Holmes, Mike Carr's module B1, or the Monster & Treasure Assortment, then you tend to see things like 1-4 Bandits, 2-5 Orcs, or 2-8 Goblins. But if you look at products like the AD&D DMG or Gygax's module B2, then you're looking at 5-15 Bandits, 7-12 Orcs, or a Goblin lair with a minimum of 6-10 fighters per room. (As a side issue, I'll also point out that the DMG tables are highly "naturalized" in that they're heavily weighted against interesting or exotic monsters -- chances are 10-15% each for men, rats, beetles, and shriekers; compare to just 5% for goblins, or 2% for skeletons or zombies. See DMG p. 175.)
That seems like a pretty big swing between different sources: Something like double, triple, or quadruple basic numbers when you switch from any non-Gygax source to the Old Man himself. No wonder it seems like such a hard road if you switch play action from module B1 to B2!
So, what do you do for stocking 1st-level dungeons (for example) if you're playing with just the OD&D booklets? Do you tend more towards 1d4 Orcs, or 1d6+6 Orcs per group?