Mapping The Entrances to Blackmoor Dungeon

If you have the 1977 print of Blackmoor Dungeon, or even the Zeitgeist games version, it's easy to think that the maps you have are good to go.  But if you try to run a game with these maps, you will soon come to notice stairs that seem to exit above ground without any key or clue as to where.  Even the main stair to the basement on level one is not clearly marked.  In fact, lining up the stairs from level to level just by looking at the maps is far from easy to do.  The only labels the stairs have are "up" or "down" and these don't always seem to be correct.  

A good long while ago, Tavis Allison of Autarch Games created a GIMP file of levels 1-6 of Blackmoor dungeon and began labeling the stairs in order to line up the levels one on top of another.  At that time, I decided to build on what he did by trying to place the stairs and exits that reached the surface onto a map of the castle of Blackmoor.

As simple as that sounds there is one very real problem.  There is no good surface map of Blackmoor castle.  There are two maps of the town of Blackmoor that show the castle, but the scale is tiny.  There is also a planview of the main castle building, but it shows nothing beyond the building itself.

There's another problem.  For reasons I have never been able to figure out, the two maps of Blackmoor town are oriented completely differently.  The original map that Arneson drew was published in the Castle & Crusades Society newsletter issue #13 in July of 1972 and depicts the town south of the castle.  The second, updated map appears in the FFC, and shows the town east of the castle.  Suffice it to say, this second orientation creates some weird alignment problems, and so I ignored it in favor of the original map.  The actual cardinal directions aren't really the issue anyway, because what really matters is the orientation and facing of the castle toward the town.  Here is what that original 1972 castle looked like:

In short, what I did was to create this blow up of the original castle drawing.  Then  I traced out the walls and towers and scaled the traced image to fit the dimensions given in the FFC.  After importing this image into the GiMP files, I noticed right away that the basement on level 1 and the outline of the inner walls of the Castle were obviously meant to match.  So I made a few more adjustments to the castle outline, straightening out some of the lines, and matched it up in what seemed like a logical fit.  The result of that was this:

And so it sat for years.  Just so you know what you are looking at, all those letter/number combos are Tavis' staircase notations on various levels.  The letter is just a label but the number indicates what levels the stairs are on - in retrospect it probably would have been better to start the label with the level the stair starts on and then give a letter tag, but it is what it is.   

Recently, former Arneson associate Jeff Quinn (author of the excelent Blackmoor adventure The Redwood Scar) has been redrawing the maps of Blackmoor dungeon (see here: Map Discussion ).  That prompted me to revist this old project.  I was never very happy with my results and I realized it would make more sense to keep the wall outlines I had made, but use the official castle floorplan instead of the crude sketch from the original map.  That way I could get the scale of the castle building exact and adjust the walls to fit.  Here then is the castle floorplan from the FFC (note that this particular version comes from the Arneson estate with his handwritten notes on it.  The Outline is identical however.)

Using this map allowed me to exactly scale the castle building (170') to the castle basement.  Playing around with that new image resulted this positioning of the castle:

Notice how the side entrance actually lines up on the side of the building and the castle walls line up with or closely parallel the walls of the basement.  There's also a possible second stair inside the castle itself, stair G1234.  I didn't show the outer walls on the above image for a reason.  My original drawing, based on the blowup of Arneson's '72 map, had the walls fairly close.  However, I noticed that on the west (left) side of the castle there was a line of stairs on level 1 that seemed to follow the line of the wall.  So I extended the wall over and discovered a remarkable alignment.

As it turns out, I should have known to extend the castle walls further out anyway.  If you look at Arneson's FFC map of Blackmoor Town and Castle (the one from '77) you can see this same area between the inner and outer wall is larger and is labeled "fair", meaning it is the location of the fair of the elves guarding the castle.  So, after trying scores of combinations and configurations, I'm convinced this is the proper fit or as close as is possible to get.  However, this would mean that the interior of the castle rooms shown in the FFC has to be wrong!

Here it is again:

Notice in this image that the throne room is in the smaller right side wing of the castle, while the tower base on the left is occupied by various rooms.  According to all descriptions, the main stair (Main 1) should go down from the throne room, and that appears to be what the "A" above is meant to represent.  Believe me, I tried dozens of orientations, size variations, mirror images and so forth to get the Main Stair to match with the throne room as shown in this planview.  but it simply won't line up in any way that makes any sense of the basement below and the arrangement of the outer walls and towers.  However, as you can see from the alignment above, the Main stair lines up quite well with the first floor of the Main tower, so it appears to me that the throne room should be in the tower, not the extension to the east where it is shown.  If you ask me, putting the throne room at the bottom of the tower makes more sense anyway.

Now we have one more thing to mention.  Thanks to Dave Megarry's maps for his character HW Dumbo, we know that the stairs marked "B12" and "D12" do indeed exit on the surface, and that the corridor heading directly west from the basement doesn't just end but actually extends to exit on the hillside.  So alright then; using the above map and Tavis's stair labels, here are the surface entrances and their locations around the castle.

Name	Location
A1 Stair	Hillside west of castle
B12 Stair	Tower on NW corner of outer wall
D12	In section of outer wall west
E12	At corner of outer wall west and sw
HY1	At corner of outer wall southwest and inner wall
I12	Inner wall SE near south tower
J1234	Well? near SE corner of outer wall
L12	Hillside just outside SE outer wall
Side	West wall of main Castle
Main	Throne room
Western corridor	Hillside west of Castle

Monster lists and Megarry's Maps

Let's take a closer look at that last section of Dave Megarry's maps - the one for the SE section of Level 7.

What's particularly awesome about these notes on Level 7 is that we can compare them to one of Arneson's original stocking lists.  Although the first 6 levels of Blackmoor dungeon published in the FFC are Arneson's Gen con convention restocking list generated randomly using the OD&D booklets, levels 7-10 are original lists, dating to the first year or so of Blackmoor, prior even to the great Invasion that led to King Funk III taking over the 10th Level with his hoards of minions.

It would not be too surprising for the notes on Megarry's maps to be quite different from the FFC lists.  Arneson may have restocked more than once - though perhaps some levels saw more such changes than others.

Here's Megarry's notes:

Room#	Monster	Treasure
17	10 Ogres	"Gold Dinning"
19	168 Goblins
22	33 Trolls	"riches"

Here's the same rooms in the FFC list, with the information in exact order:

Room#	Gold Pieces	Protection Points	Monster	Magic Treasure
17	3000 GP	150	10 Ogres	-
19	-	300	200 Ghouls (catacombs)	2 Arrows, 2 Gloves
22	-	500	33 Trolls	1/10 (prolly a weakened spell)

As can be readily seen, rooms 17 and 22 are identical, and that in itself is really quite remarkable. but what about those 200 (yikes!) ghouls in room 19? 

Certainly that could be a restocking change, but upon consideration, I think it is more likely to be a mistake.  As most of you will know, the FFC was a lightly edited assemblage of a fairly hodge-podge collection of Arneson's campaign notes and essays.  The original material was both hand written and (poorly) typed, and it is not unusual to spot errors in the FFC print - have a look at this post on the Loch Gloomin material for some examples: Stocking Blackmoor Wilds in 1972  

I think it possible that the editors/typist preparing the stocking list may have misread Goblins as Ghouls.

To explain why I think so,  I need to refer back to Arneson's "Protection Point" method.  In brief, Arneson assigned protection points to a room and then used those points to "buy" the monsters (For a full explanation, please see this post Point Buy Method ).

  The costs of the monsters were those given in the CHAINMAIL^TM  booklet.  Ogres, for example, cost 15 points.  That's why room 17, with 150 protection points, has 10 Ogres in it.  (BTW "points" also functioned as HP and XP, see here On Points )

Arneson's goblins are a little more complicated.  He actually had two types: one that costs 1.5 points, just as in CHAINMAIL^TM, and a tougher version that costs 5 points, perhaps a hobgoblin.  Ghouls however in the FFC lists always cost 10 points, just at they do in CHAINMAIL^TM.  

Notice that the FFC stocking list assigns 300 Protection Points for room 19.  To stock the room with 200 Ghouls would require 2000 Protection Points.  Three hundred points would give you 30 Ghouls.  On the other hand, 300 points would give you 200 standard Goblins.  So the number is right if is really Goblins and not Ghouls.  Further, Megarry's map has 168 Goblins in this room, and it is easily conceivable that any number of things could have led to those 32Goblins being "missing".

So, I think Megarry's map notes have actually helped us find and correct a mistake in the FFC.

The Oldest Dungeon Maps in D&D History

David Megarry, inventor of the Dungeon! boardgame and one of the original players in the Blackmoor campaign that sparked the creation of Dungeons & Dragons, has once again dug deep into his sea chests and pulled out an old Chemistry notebook with a set of maps on graph paper, drawn from sketches made as his character H.W. Dumbo braved the depths of Blackmoor Dungeon. 

As discussed previously (here: Early Blackmoor Characters ), H.W. Dumbo was one of David Megarry's characters in the pre D&D Blackmoor game developed and run by D&D co-creator Dave Arneson.  As Megarry's 5th created character, he probably came to existence in 1972, but certainly within 1 year of that date.  That makes the maps Megarry drew as his character explored Blackmoor dungeon the oldest adventure role playing game maps ever revealed.

These maps were drawn in a bound notebook of graph paper at a time before D&D existed, probably even before Gary Gygax was himself introduced to the game, so naturally, a close study of the images reveals interesting details, especially for Blackmoor fans.

For comparison, I'll include images from the 1977 First Fantasy Campaign booklet by Judges Guild (JG).  These 1977 maps show the dungeon as drawn by Arneson and include a key.  (For a related post see here: # Appearing in Blackmoor Dungeon )  These same maps were, of course, redrawn nearly identically for the Zeitgeist Games (ZG) release of The Dungeons of Castle Blackmoor in 2006, and I'll refer to that publication from time to time also.

Megarry's scans of his notebook are of whole pages and show a lot of blank space.  However, to make the images easier to view, I have cropped together the portions showing the explored areas.  We begin with Level 1.

Here is the corresponding section of the JG map:

First, a bit about notation; Megarry's stairs are little circles.  Frequently he marks these as U.S  or D. S. - up stairs or down, just as on the JG maps, or as "Grand stair", to indicate both up and down.  Passages are also marked D.E. for Dead End.

Notice that D.E. appears where room 9 is marked on the JG map - very likely indicating that H. W. Dumbo failed to find the secret door leading into that room.  We also see in the passage just to the east of where room 9 isn't, an area marked "cave in", just where the Orchian Way stairs should be.  This and other cave ins we will see on Megarry's maps, are probably the result of the earthquake reported in Arneson's newletter the Blackmoor Gazette and Rumormonger released in the fall of 1971.

Curiously, the note "to castle" is written near room 2, and very interestingly, there is a note at the western end of the east west corridor at the bottom of the map which reads "way out; the side of cliff."  On the JG map the same corridor simply appears to end.

The long north/south passage ends in a "Grand stair", indicating both must exit above ground - I'll have more to say about that in a later post - but this does contradict the published JG map where these stairs are labelled down (and probably shouldn't be).  

LEVEL 2

Here's the corresponding JG section:

Perhaps the first thing to jump out is room 4, marked "Torture" on Megarry's map.  In the FFC key, this room is non-descript, but the ZG dungeon key describes this room as the torture chamber of Sir Fang, the vampire.  Further there is a mysterious circle on the JG map, which is described as a pit in the ZG text, but on Megarry's map we see only a half circle in that area marked "pillar".  I wonder if this might have been one of Arneson's elevator shafts, or probably more likely, some kind of a pedestal.

The configuration in room 2 is slightly different but there are the same 8 cells.  These are marked "pens" on the JG map.

Rooms 5 and 6 are slightly bigger on Megarry's map.

Lastly, there is a small note on Megarry's map that says "? 30' above house".  This note refers to a house Megarry had H. W. Dumbo build.  Mr. Megarry told me:

 "It is located on the isthmus between the town and the castle. It is amazing that Arneson let me dig a tunnel in that area. I recall that I dug a tunnel from the northwest shore to the location of the house and then dug down to the 5th level before I connected with the Blackmoor dungeon (I missed other levels by a hair). I earned enough on expeditions to finally build the house, which I put right on top of the down part."


LEVEL 3

And the JG section:
   

There is a slight difference in stair placement in the three pronged passage east of room 2 that could have some significance for map alignment issues.

Room 10 is oriented completely differently.

In the center east section of the map is a single colored in square marked "H. W.'s hole".  That's the location of the tunnel from H. W. Dumbo's house.

At the very bottom of the map, near the area between rooms 10 and 12 and in the east/west corridor below them, is a door in the south wall with a hint of the corridor beyond heading south.  Here is a note reading "Elves ?".

LEVEL 4

Compare to the JG map:

At the North east edge of Megarry's map there is a note that says "Devils Fountain" next to "cave in".  Devil fountains are mentioned in the FFC, where they are said to be marked on the maps with black squares.  Only level 9 of the JG FFC maps have black squares marked on them, and curiously, that map has a note - probably a mistake -saying the black squares are trap doors. 

There's also a note on the nearby stairs that says "damaged stair".

On the eastern edge, the note reads S.P. 50'.  This appears to be the same as the rough hewn passage indicated on the JG map, but at the next hallway just to the west.

Just east of room 16, there is a feature not shown on the JG map that appears to be the fire pit also shown on level 3.  

Just east of that, in the north end of the triangular shaped room, a down stair is marked.  The same stair is marked two squares further north on the JG map, and the significance of this is that the JG map does not align with a stair on level 5, but Megarry's position of this stair lines up perfectly with the published JG map of Level 5.

Once again, on the eastern edge of Megarry's map, we see a note against a colored square marked "Dumbo's hole".  The note in the upper corner reads "50' from house; down in Blackmoor; (4th level)" 

LEVEL 5

Here's the Corresponding are of the JG map:

Megarry's map notes this is "100' down in Blackmoor".  

Room 15 is marked "empty" on Megarry's map, but there is a note there that says "secret passage"

Just east of room 15 is a "down stair" in a hallway with the note "...to 7th level 100' down".

There is also a down stair marked on Megarry's map missing from the JG map - at the end of the hall adjacent (east) of room 4.

A little further east, we see the location where the tunnel breaks through.  The note reads "Hole in roof". "Way up; 20' S; 100' up".  Another little note pointing at a small line in the area says "Spike".  The rubble seems to be marked on both maps.

Lastly, there is a stair marked in a slightly different position in the room just north of room 5.

LEVEL 6

The Level 6 JG map portion:

 The corner notation on Megarry's map has "150' from Dumbo's house"

There's some pretty rich notation on the maps for this level.  The hatched squares on Megarry's map are marked "fire pit" as on the JG map.

The South passage extending from room 1 is marked "slopes down" and the doorway at room 1 is marked "S. P." for secret passage.

Room 14 in the JG map has but one stair marked.  Megarry's map however has two "up" stairs marked on the southeast wall, flanking a semicircle labelled "Throne".

LEVEL 7

Level 7 is the last level contained in Mr. Megarry's maps.  The areas shown are the northwest and southeast corners, so I will present them seperately;

L7 NW

Same portion on JG map:

Interestingly the Megarry map actually shows a small portion of corridor at the top edge not found on the JG map.

The square in room one of the JG map is marked "firepit" on Megarry's map.  

Also note the area indicated by small dotted lines and the word "down" on the JG map at the northwest end of the corridor in the middle of the image.  Megarry has an explanatory note at this location: "Alcove False Bottom". 

The note at the southeast corner reads enigmatically, "To Dwarfs Entrance"

L7 SE

  And this section of the JG map:

The corner notation on Megarry's map reads, "200' down in Blackmoor" 

This section of the map reveals more of the dungeon inhabitants.  The entrance to room 17 is marked with S. P. on Megarry's map and the room itself is noted to have "10 Ogres" and underneath it is somewhat hard to make out, but according to Dave Megarry it reads   "gold dinning room".  

The stair at the west end of the east-west passage extending from room 17 is marked "up stairs 100' up".

Room 19 on Megarry's map is marked "168 Goblins"

Room 22 is drawn much larger on Megarry's map and is noted to have a "secret passage" entrance and "33 Trolls + riches"

Lastly, the unnumbered vertical corridor in the southeast corner is marked "20' wide passage", thus confirming the scale of 1 square = 10 feet.
   
I'll have more on HW Dumbo and Megarry's other Blackmoor characters in future posts, but for the maps, that wraps it up for now.  I believe some additional information and certainly some related content will shortly be available on the Secrets of Blackmoor documentary Facebook page HERE, so keep an eye on that page.

Megarry's Blackmoor Character Sheets IV - Missing Info

This post is about what is and is not on the character sheets, and why.  I'll start with a picture painted in quotations:

"..,Dave played things so close to the vest, when I asked him questions at the "My Awesome Gaming Group" panel, he answered my "is it this or that" question 'yes' because 'some of my players are here'"  Tavis Allison, commenting on the 'blog Beyond the Black Gate Monday, July 25, 2011, "Dave Arneson and Impartiality in the Temple of the Frog"

"Dave did lots of secret rolling and never showed us the results or explained what he was roling for. I am sure that sometimes he rolled dice just to make us nervous..." Greg Svenson Nov 27, 2009, http://odd74.proboards.com/post/40731/thread

" We were not keeping our own records or character sheets as they are called now. Dave had an index card on each of the players (and NPCs) with their attributes, HP, possessions and other useful notes. I only remember seeing Svenny’s character card a couple of times.... he kept our character cards,... We didn’t track our experience points as is done now. Dave simply told us when we had transitioned from one level to another."  Shams Grog.& Blog Q AND A-with-Greg-Svenson

"To be honest, I never had a character sheet. Dave A. had my character (Svenny) on an index card that he kept. He let me use it during a couple of game sessions (maybe 2 or 3) but always collected it again. I never thought to make a copy of it, either."  Greg-Svenson, Post subject: Re: Megarry's Blackmoor Characters Posted: Oct 24, 2016 5:08 pm Comeback Inn forum.

"Don't ask me what you need to hit. Tell me what you're character is doing. I'll tell you what dice to roll" - Dave Arneson

If you are an OD&Der, it is only natural for you to be looking at Megarry's characters for signs of familiar stats, but you will be hard pressed to find them.  A few of the "ability scores" are there, albeit in 2d6 instead of 3d6, but the all familiar hit points, level, saving throws, etc., of the D&D are conspicuously absent.  Is this because these things were not a part of the Blackmoor game?  The answer to some extent depends on the particular stat.  In fact, while there seems to be a lot missing from the sheet, we can be certain nonetheless certain that there were some items of character information, that the players weren't allowed to know. 

As the quotes above show, Arneson prefered what we might call an immersion game, with as little "out of character" talk as possible.

"Just role playing not roll playing." Dave Arneson, Blackmoor at the Piazza, Sep 30, 2008 at 10:09pm ODD74 forum

Perhaps the genesis of this disposition lies in Arneson's contrasting experience of Wesely's nearly rule-less Braunstein games and all the "rules lawyering" that took place in his own Napoleonic games.  Arneson made a number of comments that one of the reason fantasy appealed to him was that no one could argue historical facts with him.  In one interview he said:

"How did it all start in Blackmoor?..... I was also quite tired of my Nappy campaign with all its rigid rules, etc., and was perhaps rebelling against it (in fact I'm sure I was!!).....  (For Blackmoor) rules were actually written down (but closely guarded by the referee and subject to change without notice if things got out of hand)."  Different Worlds magazine, June/July 1979, p6-7.

Given the context, it is pretty easy to see what is going on with Dave Megarry's character sheet.  Arneson apparently preferred to keep important rules details out of the hands of his players, including things that we would now think of as a part of character information.  The information the players had about their character was not the whole picture. 

Hit Points are perhaps the most obvious omission.  We might also expect to find Hit Dice, Level, and Experience Points. 

Despite not being directly referenced on the character sheet, we know these things were a part of a characters makeup.  Hit points and experience points are discussed in this POST, for example.  That post references Greg Svenson notes in the back of his 2nd edition CHAINMAIL.  Svenson had these notes, no doubt, because he was one of the first persons Arneson turned to with help refereeing Blackmoor adventures.  It was "need to know" information.

There's a bit more to be said about experience points, however.  Although it is true there is no XP total listed, Megarry does keep track of the monsters his characters have killed, and in the single case of the Scholaress character, he also writes their point value.  Which tells us at least something about the characters earned experience (more on this in a later post). 

Academically, this is all kind of interesting (to me anyway), but imagine for a moment, if Arneson's approach had become the norm for D&D.  Imagine character sheets where the only statistical information a player has is their ability scores, saving throws, armor class, and maybe their level.  It is the referee, not the players, who track the characters hit points and damage, experience points, and other "rulesy" type information.

So, you don't know what your hit point total is, or how much damage exactly (in points) you can do.  You don't know how close or not you might be to leveling up.  You probably don't know how much of a bonus that magic weapon has either.

Descriptions of action, especially of fights, would naturally be less laced with statistics and mathematics.  Instead of "You take 4 points of damage" it would necessarily come out as "You receive a deep gash and your character is feeling very weak." or some such.

One wonders what direction D&D would have developed in if the D&D player's character information had been limited in the same way it was in Blackmoor.

Megarry's Blackmoor Characters III - A Mystery

The mystery of the circle and squares.

Now here is a fun one.  You'll recall from the first post in this series (in point three) I mentioned two new stat which were added to the character sheets and placed near the character name.  These stats were unlabeled and distinguished from each other by having either a square or a circle drawn around the number.  Here's a close up of a couple:

 

It's hard to see in the pics I copied to the blog so here is a link where you can download better resolution jpegs.

Sheet 1

Sheet 2

So to reiterate, the "circle numbers" came first, at the time the Hercebeiner characters were created.  Then came the square numbers, and Megarry drew circles and squares around them to mark them as different.

What are they?

Before your mind goes racing to all the obvious D&D stuff, take note of the zeros, and keep in mind the ranges found on the 2 sheets; square numbers range from 0-4, whereas circle numbers range from 0-7.

My first thought was that they might perhaps represent levels.  Dave Arneson once said "Because the players in the original campaign could learn different skills and different abilities, virtually everybody who was a fighter also wanted to be able to throw magic.  And it seemed like everybody who was a magic-user also wanted to be able to fight.“  Dave Arneson, Mortality Radio interview, July 9th, 2004.   In FFC material, quite likely to be contemporary with Megarry's characters we read of an assistant to the Ran of Ah Foo who is "Level 7 Warrior and Magic". (1977:19)

However, when I presented the idea to Dave Megarry, he disagreed, "The only character that had some magic type ability was the Scholaress: she had the shape changing amulet and the vial of basilisk blood. I was not a magic user: my vision was more of Aragorn rather than Gandalf, if you get my meaning." Personal Communication, 2016

Mr. Megarry further explained that he knew what the squares were.  "The square numbers were the hit class / armor class of the characters. I am having a problem remembering what the round numbers meant. Arneson added something at the time or made us add the information at the time I did the Herk characters and I filled it in for my previous characters (for consistency, I guess:).  The Herk series is the three characters that were related to each other or part of the household in red.... The armor class dealt with what type of armor you had bought or acquired: no armor = 0; leather armor = 1; chainmail = 2; plate armor = 4. I think these are straight from the Chainmail combat system." Personal Communication, 2016

I suggested to Mr. Megarry that if it was from CHAINMAIL, then the numbers were probably no armor = 0; shield only = 1, leather armor = 2; chainmail = 3; plate armor = 4., and Dave Megarry thought that was probably it.

The puzzling thing about that is that it skips the "with shields" columns, so why use these numbers?  I mean, if you are just counting across on the Man to Man table "3" is leather and shield, not chain.  So it would seem to make more sense to just write "chain" or to number across all the column so as to avoid confusion.

I wondered if Mr. Megarry had gotten the circles and squares backwards, since circle numbers can be found as high as 7 (the Gester) which would be Plate armor, counting across.  Again, however Megarry corrected that notion, "The two characters, Earl's Jester and Hero Trany both have circle 7's and I know they did not have plate armor with shields!" Personal Communication, 2016

In fact, full plate seems to have been a rarity in Blackmoor.  I think it quite likely that the numbers in the squares do indeed refer to the Man to Man armor types in CHAINMAIL just as Megarry claims, but it is simply the case that none of Megarry's characters ever wore armor any more robust than type 4.  He said, "I was not much into shields as they used one arm in combat and was hard to carry. I liked bow and arrow weapons and did not use shields. I did like leather armor as it gave a minimal protection for a relatively cheap cost. The Scholaress did have chainmail as she had money..."  Personal Communication, 2016.  The Scholaress, by the way, has a square number of 4.  For reference here is the CM list, with numbers added:

0	no armor
1	Leather/Padded
2	Shield only
3	Leather + shield
4	Chainmail
5	Chainmail + shield
6	Plate
7	Plate + shield

Perhaps it is worth noting that this is also consistent with John Sniders remarks, "Armor .. I thought it was 1-8, my Boozero character started at 1 if I remember correctly, but then he was shall we say in his cups a tad much so Dave could have adjusted the system to account for this." Personal Communication 2009 

There's also this odd little note in the FFC, "ROBOTS - I roll one Six sided dice for Armor Class," (1977:74)

I also wondered if the circle and square numbers could both be armor class, like ascending vs descending, but the circle and square numbers do not correspond in any way that would allow them to be describing the same sets of armor, therefore, if they both represent a defense number of some kind, and the square numbers are a shorthand for CM armor types, it would seem that the only real option for the circle numbers would be helmets, or possibly shields.

Of course, there is no table in CHAINMAIL for shield types or helmet types - so if the numbers were intended to be a separate armor rating for either shields or helmets, Arneson would have to have created his own tables for them.  That seems a bit fiddly, and there is no evidence for such.  Also, when we look to the FFC, we find three kinds of helmets and 9 kinds of shields (three kinds of small, medium and large).  Since our number ranges in the circle and square figures are 0-4 and 0-7, helmets clearly don't fit the range.

Shields however remain a possibility for the 0-7 range of the circle numbers, but this again is problematic.  Yes, shields are already included on the Man to Man table, but it is conceivable that Arneson developed a shield rule like that found in the Dalluhn draft and in somewhat different form in the 3lbb booklets; namely that a magical shield would only "catch" a fraction (like 1/3rd) of the blows depending on the direction of the attack.  However, even this idea doesn't stand up well to scrutiny.  The numbers can't be a separate shield rating within the Man to Man method because the number combinations on the character sheet don't correspond in any meaningful way.  

Remember also that the circle numbers came first.  What sense does it make, for Arneson to have switched from a single "to hit" target number to determine a hit on a character, to a wholly unknown to hit table based on whether a character has a shield or not?

Well it doesn't.  About the only way it could make sense was if the circle numbers were a shield 2d6 saving throw number.  If a saving throw based on shields came first, then it might make some sense if armor were added in the manner Megarry suggested (no armor = 0; shield only = 1, leather armor = 2; chainmail = 3; plate armor = 4) because then the "with shield" column on the Man to Man table would be redundant.  Then again, type 1 "shield only" would seem to be redundant too.

There are two bits of circumstantial evidence that might support this; namely there is an "X5" shield, otherwise unexplained, in the treasure of the original Temple of the Frog (First Floor, Room 4, p37) and a cryptic comment in the FFC that "...the player received a "Saving Throw" against any hit he received." (1977:2).  Both are rather thin sauce, as the first might be a typo and the second doesn't seem to have any dependence on possessing a shield, but who knows.

Perhaps the only way it could make sense for the circles to represent shields and the squares to represent armor is if the numbers have nothing to do with the Man to Man table, and are instead "ablative" armor, like that found in the original Temple of The Frog and Craig Grasteck's "Rules to the game of Dungeon".  This type of armor will absorb a number of hits until it fails.

Honestly though, there's very little evidence for this type of armor in early balckmoor.  Further the numbers don't really fit well.  If the circles represent shields, then shields are significantly better than body armor according to the numbers.  They also don't come anywhere close to the number of hits the one known example found in the 1975 TotF.  That suit of armor could take 65! hits.  True it was space alien technology, but it is still a tremendous gap between the 2's and 3's we see in the circle and square numbers  and 65.

My last, best guess is that the circle numbers could represent an attack bonus, possibly also applicable to saving throws and such.  If you'll remember a good while back when I discussed Pete Gaylord's character sheet, there was an unexplained "+5" written next to his best weapon, the pole axe.  At that time I pointed to the "Blackmoor Military Manpower Distribution" list (FFC 1977:17), where following each Player Character, their Hit Dice are listed, and many of them include a bonus ranging from 1-5, though only Svenny has +5 and the rest are +1 or 2.  Maybe the circle numbers are this sort of bonus.

Okay, that was a lot of verbiage, so I'll summarize:

Squares: These numbers represent armor class corresponding to the Man to man table in CHAINMAIL.  They range either 0-4 or 0-7.

Circles:  These numbers could be:

* A shield saving throw rating possibly in cojunction with a 0-4 AC range

* A bonus applied to attacks, most likely to  damage rolls

* Something completely different  

l'm afraid I have to leave it at that until more information becomes available, but feel free to speculate in the comments.