2021-09-13

d6 Usage in OD&D

Red six-sided die

The d20, of course, is closely associated with the D&D game. But it's easy to forget that the early classic editions used the d20 in fairly narrow circumstances: just attacks and saves, in fact. Everything else about running an adventure was done with d6's -- it wasn't until the 3E version of the game that all of these other functions were replaced with the "core mechanic" of d20's for everything. 

There's something about that I really like, in fact. (And I've written about it several times.) The granularity of the game is usually at the d6 level, except when death is on the line (an attack or saving throw), at which point the detail "zooms in" to the d20 level. That's a bit like a movie slowing down when a character is in mortal peril, simulating the cognitive effect from increased adrenaline. When everything else is d6's, they probably don't show up to clutter the character sheets. There's a small enough list of modifiers that the DM can just remember them all. And when things go off-book -- like they should -- it's easier to correctly estimate the real-world chance of success out of 6 pips than out of 20 points. 

Here's an encyclopedic run-down of the use of the d6 for adventuring function in Original D&D. The majority of these cases appear in little brown book Vol-3, the DM's guide analog:

  • Damage from falling into pit: Occurs on 1-2 on d6 (suggested example). (p. 5)
  • Random dungeon stocking: Monsters appear 1-2 on d6. (p. 7)
  • Random dungeon stocking: Treasures appear 1-3 on d6 with monsters, or 1 on d6 without. (p. 7)
  • Searching for secret passages: Success 1-2 on d6 generally, or 1-4 on d6 for elves. Or elves can possibly find one 1-2 on d6 just by walking by. (p. 9)
  • Opening doors: Success 1-2 on d6, or 1 on d6 for smaller characters. (p. 9)
  • Spiking doors open: Success 5-6 on d6. (p. 9)
  • Traps activating on trigger: Occur 1-2 on d6. (p. 9)
  • Listening for sounds: Success 1 on d6 for humans, or 1-2 for elves, dwarves, and hobbits. (p. 9)
  • Surprise: Occurs 1-2 on d6 unless some signal prevents it. (p. 9)
  • Wandering monsters in dungeon: Occur on a 6 on d6, rolled each turn. (p. 10)
  • Monsters continuing pursuit: Occurs 1-2 on d6 when party passes a corner, door or stairs; or 1 on d6 when party passes through a secret door. (p. 12)
  • Castle occupants turning out: Occurs 1-3 on d6 within the castle hex, 1-2 at 1 hex distance, 1 at 2 hex distance. (p. 15)
  • Becoming lost in wilderness: Occurs on either 1, 1-2, or 1-3 on d6, depending on terrain type. (p. 18)
  • Wandering monsters in wilderness: Occur on 4-5, 5-6, or 6 on d6, depending on terrain type. (p. 18)
  • Castle inhabitants pursuing party: Occurs 1-3 on d6 if hostile, or 1 on d6 if neutral towards party. (p. 19)
  • Damage from fall off ship rigging: Occurs at one low pip on d6 for every ten feet fallen (more detail below). (p. 31)
  • Ship crew in melee obeying other commands: Occurs 1-4 on d6. (p. 32)

So let's take stock of what we have there for a "core d6 mechanic" sensibility. We've found 17 cases in OD&D Vol-3. We note that in 14 of the cases success is indicated by a low roll (82%), whereas in only 3 cases is success indicated by a high roll (18%).

Clearly, the fundamental instinct of the writer with these d6 mechanics is for the low roll to indicate success. That doesn't mean low is "good" exactly -- consider a trap being sprung or a party being surprised, for example. But generally some new-thing-of-note pops up with a low d6 roll -- a change to the status quo.

Of course, since most of these mechanics have a base success of 1-2 on d6, the inverse is in the majority, and we might say that's thereby the status quo by definition. 

The falling-off-ships-rigging case is interesting (Vol-3, p. 31), because it highlights that either side of the random occurrence could have certainly been phrased as the event of interest. In this case, the rules text is phrased in terms of taking damage from a low roll ("one chance out of six for every level fallen that damage will be sustained"), whereas the associated example is phrased in terms of saving with a high roll ("i.e. a fall from 40 feet will require a 5 or 6 to save"). 

And in the list above I'm not even counting cases from other books, like the end of OD&D Vol-2 (the monsters & treasure book), in which a roll of "1" on d6 bumps an individual gem in a batch up to the next-higher price level. Whereas, if you were in a different headspace, you'd likely think that a high value would indicate, well, a high value.

So where do the 3 outliers come from? To be clear, those are: (1) spiked doors failing, (2) wandering monsters in the dungeon, and (3) wandering monsters in the wilderness. Let's ignore the first of those for now. But the two wandering monster cases have a clear source -- that's exactly the mechanic in the earlier Outdoor Survival board game for the chance of a daily encounter (where the rule is itself optional). This is in contrast to all the other mechanics in Outdoor Survival, which are notably roll-low-on-d6 to break status quo -- e.g.: getting lost, finding food, or finding water (as allowed in some scenarios; finding food or water occurs 1-2 on d6 when permitted). Note that the lost/food/water mechanics are printed on the Scenario cards there, whereas the optional Encounters rule is in the separate rules pamphlet -- so they weren't synchronized with any core mechanic. And that's exactly why these rules appear in the same form in OD&D, since they were just wholesale lifted from that source & tweaked a bit. Examples below:

Outdoor Survival, Scenario 3: Search rules

Outdoor Survival, Rules of Play: Optional Encounters rule

So if I were going to get my rules-design steam press, and iron out the wrinkles in this particular system -- for both Original D&D and Outdoor Survival, because the system is at least conjoined if not identical at their root -- what I'd do is swap around the wandering-monster rolls and make them appear on low results on a d6, e.g., a "1" on d6 for encounters in the dungeon. Same for spikes failing, too. Taking the opposite tack and saying you're going to flip all the d6 rolls around so a high result is success entails a lot more editorial fixup-work (e.g., as Menzter tried to do with opening doors in BECMI; and as I've also stumbled towards doing in the past myself).

In conclusion, there's also a number of things that are attractive about what I might call an "accuracy" roll-low core mechanic. Principally, it's that announcing a target number is simultaneously communicating the probability of success. (As opposed to a high-roll mechanic, where the conversion between the two requires subtraction and then an off-by-one adjustment.) I assume that's why the writer of OD&D fell into this habit; you don't even need to mentally distinguish which way you're thinking about it as you furiously pound out the rules text on your typewriter. Additionally, to my mind, the die-roll then has the feel of communicating the amount of "error" in your task attempt, which is a statistically robust concept; as opposed to (I guess) "goodness", implied by a roll-high mechanic. 

So there are days when I wistfully daydream of a D&D tradition in which all the mechanics were always roll-low by default, instead of the legacy we have. Imagine celebrating being "Number one!" on an attack roll with as much gusto as we now do a "Natural twenty!". (Although I suppose it might not be immediately as clear that an exotic die type was in play.) It would also synch up with the old roll-under-ability mechanic, which at one point seemed natural and obvious (rather than convert to a modifier, and now have many people ask, "why do we record ability scores anyway?").

Do you agree with the suggested roll-low tweak to wandering monsters in OD&D? Did I miss any notable d6 mechanics in the DM's rules for OD&D?

51 comments:

  1. I appreciate the round-up of old D6 rules!

    You can keep it consistent without changing a thing in the three aberrant cases of D6 rolls and say you are testing an outcome with a low roll if you just declare you're rolling for the alternative. Spiked doors get closed anyway on a 1-4. Players *avoid* wandering monsters on low rolls. Now you have a universal "roll low on D6" system with the D6.

    Part of the inconsistency may be in the unanswered question of who rolls for each thing (Ref or players?) and who, if anybody, is "winning" by these rolls (again, Ref/game world/dungeon or players?). The implicit orientation of the game's story, towards either Ref or players or somewhere in between, has always been ambiguous and is one of the main things people debate in these games (usually without realizing that's what's going on).

    It's not at all clear that the generality of D&D players follow the procedures about "who rolls?" that were originally envisaged and not stated. After all, OD&D Men and Magic makes it explicitly clear (p. 10) that *the Referee rolls for player character ability scores* and hands them out to the players. No "old-school" player follows this rule as stated in the book, as far as I know, despite chest-pounding oath-taking about "3D6 in order."

    ReplyDelete
    Replies
    1. Great point on the "who rolls?" question. Definitely been an evolution of more stuff moving over to the players' side.

      Honestly I've been thinking a lot lately about trying out the "DM rolls abilities" rule -- in fact, in an online setting with old-school body count, I'm thinking that might nicely accelerate the new-character creation (have a stack of prepared ability'd PCs at the ready).

      Delete
  2. Honestly, I'd completely forgotten encountering a monster on a 6. I've rolled Wilderness encounters as <= the terrain cost to enter the hex for so long...

    ReplyDelete
    Replies
    1. Could you expand on what you mean by the terrain cost. It would be great to be able to simplify my Hex Crawling rules even more.

      Delete
  3. OD&D came out of a D6-based wargame tradition. Yes, Chainmail, but most dice-based games...including Don't Give Up the Ship, Fight in the Skies, etc. and the games from which THOSE evolved...were ALSO d6 based. And these had large influences on OD&D systems.

    In many ways, the inclusion of non-sixer polyhedrons in OD&D was more "gimmick" than a proactive attempt to change/influence probability scales.

    When I wrote my "fantasy heart breaker," Five Ancient Kingdoms, one thing I did was to restructure the OD&D rules to ONLY use d6s. It's certainly possible to do, though I added a few wrinkles (like Rule Zero) to create additional probabilities.

    ReplyDelete
  4. Playing OD&D over the last few years, I've definitely drifted into using a d6 mechanic for most challenges. I also use the universal -1/0/+1 mechanic from S&W, where the modifier changes the RANGE of success, i.e. 1 in 6 becomes 2 in 6 with a +1. Then those little pluses and minuses really feel significant. But I suspect that one of the reasons for doing this isn't just keeping the game to d20/d6 for everything (I also use d6 weapon damage), but as you described above, there is something really intuitive and easy to "eyeball" about d6 chances. Your description of using the d6 as the "zoom out" view and the d20 as the "zoom in" view is brilliant. Also, I find that relying on the d6 roll really helps naturally avoid the urge to add more granular circumstantial rules and modifiers, and players don't even seem to miss them.

    ReplyDelete
    Replies
    1. Great points, 100% agreed. And yes, the fact that a +1 bonus is literally adding to the target/chance number is another sweet spot for it.

      Delete
  5. While I appreciate the consistency and elegancy of an universal roll-low-on-d6 approach, I have two reservations against it. First is simple human nature; most people I've played with struggle a bit with rolling-low-is-good, because it's simple human instinct to want to get high numbers. While of course one can get used to this, it's one more obstacle to start play. Secondly, in my opinion, modifiers are much more counter-intuitive when applied to roll-low. In roll-high, you can simply have a +2 bonus or -3 penalty and adjust easily; having a -2 bonus or +3 penalty can introduce errors both in application and difficulty of reading spells/powers.

    ReplyDelete
    Replies
    1. I think it's significant that with the possible exception of open doors, all of the rolls are GM facing and many of them are secret and would rarely if ever have a per character modifier.

      Delete
    2. With roll-low, modifiers can be simply added to the chance in d6 for success. If the base chance to open doors is 2 in 6, with a +1 modifier that becomes a 3 in 6. Seems easy.

      Delete
    3. ^ I agree with Jon here, in my view that's actually an advantage (add to the target/chance number). Note Anathemata further up said the same thing previously.

      Delete
    4. Q: Do you think it's easier to digest a percentile system for a roll-low mechanic? Those TSR systems I learned shortly after D&D seemed equally comfortable to me -- I think I'm recalling that correctly.

      Yet I actually had another math professor over as a guest once and when I said that was the mechanic (in Star Frontiers) he paused and had to think for about 5 minutes whether that was legitimate. Which was very surprising to me personally.

      Delete
    5. If people struggle with rolling low, consider reframing it as a blackjack system. You want to roll as high as possible without going over the TN. Then they are looking for a high number, just not too high.

      Delete
  6. I've used roll encounters on 1, 1-2, or 1-3 for so long I sometimes forget the real rule is roll high. One of the few core rules I changed in my mind long ago.

    And regarding using d20 for attacks, that was only with the optional combat rules, the Chainmail rules used 2d6, as did Arneson's original combat rules (again, apparently, and in a different fashion, as a roll under).

    IIRC, Arneson's original rules may have also required a 2d6 saving throw under the character's ability score (which were 2 to 10 not 3 to 18). So the saves being d20 high and the alternate d20 high combat rules were apparently a Gygaxian innovation, I suppose.

    Why he changed that would be interesting to know. Was that a development from Wooden Ships & Iron Men?

    ReplyDelete
    Replies
    1. Argh. I meant Don't Give Up The Ship...

      Delete
    2. Right: brief review of DTGUTS shows it's all low-roll mechanics, with either 1d6 (cannon fire, melee) or 2d6 (morale saving throws).

      It does remind me that all the weather tables (wind in DGUTS; precipitation in Chainmail) were roll-high for more severity, carried through into D&D. Which is intuitive, but throws a wrinkle into my thesis here, honestly.

      Delete
    3. Interesting. How do the morale rules compare to Chainmail and BX? Is there continuity?

      Saw some nites on the cannon fire rules. Dead simple, really. Wish Wizards would release it in PDF, collector's prices are way beyond my meager purse...

      Delete
    4. Interesting question -- Comparing them, there's different terminology and mechanics running in opposite directions. It has what's called a "Morale multiplier", but that's actually for melee combat proficiency (contrast with morale multipliers in CM Post-Melee Morale). Then it has a 2d6 "morale saving throw" by troop type, but it's roll-low (contrast with CM Instability Due to Excess Casualties, which is roll-high).

      I just have a PDF of DGUTS myself; I have no idea where I got it now!

      Delete
  7. I think that it's interesting that one of the things that Gygax did in AD&D to differentiate the game was to take care to alter the die used from the d6 to other polyhedrons. The most notable instances are damage done—though that was introduced in the supplements to D&D—and the surprise roll, where circumstances can alter which die is being used. For instance, when trying to surprise a Ranger, use a d8 instead. This leads back to a little exchange in your recent interview with Michael Curtis, sort of, bringing up how the dice-step mechanic in DCC has a clear evolution from that mechanic.

    Of course, the other big change that AD&D made was an increased reliance on d% rolls, and surprise is another place those found a use, in the instance of confusion about which type of die to use.

    ReplyDelete
  8. Introducing the Target6 system :)

    ReplyDelete
    Replies
    1. I was honestly thinking that in prior versions of OED!

      Delete
  9. Roll-low on d20 would also work well with descending AC; the target's AC would be the number or lower the attacker must roll in order to hit.

    ReplyDelete
    Replies
    1. Thing that always bugged me was "+1 ring of protection" changed AC from 5 to 4, for example.

      Delete
  10. Which wargames used the d20 before D&D? What influenced Gygax to adopt the d20 for D&D? There seems to have been a general progression among Gygax and other designers towards increasing the size of the die rolled, from d6 (Chainmail), to d20 (D&D), and then to d100 (RuneQuest, and later with Gygax himself in Dangerous Journeys and Lejendary Adventures).

    ReplyDelete
    Replies
    1. If I recall from Jon Peterson, prior to D&D pairs of d20's were being used mostly for percentile systems. E.g., grab real-world casualty charts in percentages, roll your d%, design finished.

      There was some famous early wargame I was reading a while back that gave real-world percentages for various things and literally never stated a core mechanic by which to resolve them -- players just had to make that up on their own. (Can't recall name just now.)

      Delete
    2. ::Invoke Homer Simpson::
      Mmmm... Dangerous Journeys... Droool...
      ::Dismiss Homer Simpson::

      Loved DJ Mythus...

      Delete
    3. I also recall seeing a document in the past (preserved in PDF, but originating in historical wargaming) that had a long list of ways to roll percentages using one or more 6-sided dice. For example, something with a 25% chance might be "8 and 9 on 2d6."

      Delete
    4. Agreed, that was likely in multiple wargame appendices, and even an early Dragon article at one point.

      Delete
  11. There's a third benefit to the roll-low-something-happens design -- you can easily alter the odds just by changing the die size.

    So, a 1 in 6 chance becomes a 1 in 8 chance, and you still only have to see a "1" to know something happened.

    ReplyDelete
    Replies
    1. Oooooooh, that's a nifty point! Definitely outside the box for my brain.

      Delete
    2. Yes! I remember reading Stefan Poag's Mines of Khunmar and noticing that he had some secret doors that were extra-difficult to find (using a d12 rather than a d6, as I recall).

      When I was running adapted ICE Middle-earth modules using Swords & Wizardry, I did the same thing and expanded the range of difficulty. I discussed the (very simple) mechanical details here. https://snw.smolderingwizard.com/viewtopic.php?t=175

      Delete
  12. With time I came down to this simple system: when interesting things are susceptible to happen, roll 1d6.
    1-2:bad things happen to players
    3-4: things remain unchanged (for good or bad)
    5-6: things go better for the players.
    Circumstances, player action or high/low stats can add or remove one pip to the roll.

    ReplyDelete
  13. Do you use forcing every door in the dungeon in your odnd/oed group?

    ReplyDelete
    Replies
    1. I actually do. I've largely boiled it down to "can you burst it on the first kick and possibly catch anyone inside by surprise?". So often times it's roll the first attempt and hand-wave later attempts.

      Delete
    2. I view it similarly: open door allows surprise, and thereafter it's just a measurement of time (because I use turns for wandering encounters instead of table time).

      Delete
    3. The other thing is that I believe in the books it says that you can add your bonuses together to open doors, and if you use 3 people then none of them can act. I haven't modelled it but it seems clear that you always use two and sometimes use three to open a door (assuming each person gets 1-2/6, so 1-4, then 1-6, presumably with a back row firing through the doorway and a roll for surprise).

      Delete
    4. I assume for OED you would use a base 1-2/6 and then let everyone else contributing to a door push via their strength modifier only? I also assume you would reflect reality with 1 additional person tops. Anyways, please let me know. I'm a perma armchair DM with like 3 sessions of experience so I'm curous how it works.

      Delete
    5. The book rule in O/AD&D is that multiple openers all make separate die rolls (most obvious in 1E DMG example of play). Personally, I only allow one opener per standard door -- I totally can't visualize how 3 people work on a dungeon door at once. Usually there's at least one PC with +2 strength so open 4 in 6. If PCs get a rope or ram or something then I'd add bonuses together, like you say.

      https://youtu.be/JkZ79cUy8UQ

      Delete
  14. One last thing, for now: When using target 20, do you tell the players what the target's AC is?

    ReplyDelete
    Replies
    1. Most of the time, I don't. Personally I want that (a) unknown, and (b) not burdening player's math. Some other DMs play it differently and that's fine. One argument is usually in OD&D AC is obviously implied by visible armor worn.

      Delete
  15. I've done random encounter as 1 on a d6 for years. But maybe it's a little disingenuous of me to say that because I key stuff to the other number results, too 😬

    ReplyDelete
  16. There is but one extra step to remove the d20 completelly and adapt Saving Throws and To hit rolls to the superior d6 method

    ReplyDelete
  17. I already do everything as "roll low" in my games. The best possible roll on any attack, save, or check is calling acing the roll. :)

    ReplyDelete
    Replies
    1. Hey, that's very neat, thanks for sharing that! Love the "ace" terminology, that fits. And some interesting interpolations there, nice work.

      I don't suppose you ever consider looking at the Dragon #80 article by Lakofka/Gygax, "the 5% principle", where they had expanded tables so everything always got modified 1 point at a time?

      Delete
    2. No, I didn't even know about it, but thanks for pointing it out! Now I'm probably going to spend the weekend scrutinizing and obsessing over that article, so thanks for that. :Þ

      Delete
    3. You're welcome/my apologies for that. :-D

      Delete