Sunday, February 3, 2013

How Gaming Got Its Dice

For those of us who grew up with Dungeons & Dragons, it is easy to take the polyhedral dice of gaming for granted. Dice had played an integral role in gaming since Prussian wargamers of the early nineteenth century first developed combat resolution tables. Those games and the many works they influenced, however, relied exclusively on 6-sided dice, apart from a few experimental dead-ends (like Totten's 12-sided teetotum in the late nineteenth century). When modern hobby wargaming culture began in the 1950s, it too stuck with 6-siders: the first Avalon Hill game (Tactics, 1954) requires a "cubit" for combat resolution, and the miniature gamers who contributed to the War Game Digest similarly seemed content to rely on the d6. By 1970, however, polyhedral dice had begun to creep into the wargaming community, as we see in the advertisement above from a 1971 Wargamer's Newsletter. Why do we need those funny dice anyway? What purpose did they serve that an ordinary 6-sider couldn't?

Wargamers constantly reach for greater heights of realism in their simulations, and by the mid-1960s, they increasingly relied on actual military statistics to model their combat. Perhaps the most influential example of this trend is Michael J. Korns's book Modern War in Miniature (1966), a WWII-setting wargame which offers little by way of system other than tables aggregating real-world weapon behavior. Korns reduced these statistics to percentile probabilities: for example, a particular rifle might have a 70% chance to hit a target 200 yards away. But how to resolve those odds during play? There is no intuitive way to extract percentile results from rolling a small number of d6s, but Korns provided an appendix that gave the closest approximation, roughly 5% increments:
Korns's game proved very influential in the American Midwest, and took hold in the two communities where Dungeons & Dragons grew up: Lake Geneva, Wisconsin, and the Twin Cities of Minnesota. For example, Mike Carr, a prominent Twin Cities gamer, incorporated the above table into his 1972 edition of his Fight in the Skies. In Lake Geneva circles, Leon Tucker, himself a professional statistician, advocated strongly for the use of Korns, but was dissatisfied with the means of resolving percentile probability. The need for greater realism demanded more than just approximations of 5% increments. This led to an outpouring of creativity as gamers developed all sorts of improvisational ways of getting percentile numbers: you could draw two cards from a deck with the face cards removed, or pick numbered poker chips out of a hat. Tucker himself proposed an elaborate apparatus involving a graduated tube with a stopper and colored beads as a part of his ongoing collaboration with Gary Gygax and Mike Reese towards a new post-Korns set of modern miniatures rules called Tractics.
Thus, late in 1969, when chatter began in the "Must List" of Wargamer's Newsletter about the commercial availability of 20-sided dice, these implements were presented as a means of generating percentile numbers: the advertisement from the Bristol Wargames Society refers to them as "percentage dice" and doesn't even say how many sides they had. This is an area where the exotic icosahedron excels, as the models sold at the time were numbered 0-9 twice, rather than 1-20. With two throws, one could therefore generate a number from 1-100. Gary Gygax was among the readers of Wargamer's Newsletter at the time, and thus it is unsurprising that he chimed in with a letter in the February 1971 issue saying, "I imagine that sales of 20 sided dice will pick up when Mike Reese starts selling the [Tractics] rules." The d20 gets an early mention in the Tractics rules published in the fall of 1971.

This was not the first time that the use of 20-sided dice for wargaming had been proposed: Lenard Lafoka wrote an article late in 1968 that described the potential applicability of the icosahedron to wargames, but since readers would have no means of procuring one, Lakofka actually provided instructions on how to construct a 20-sided die out of wood or cardboard. The dice discussed in Wargamer's Newsletter in the early 1970s were available only from Japan or Britain, and for Americans ordering from either was a costly and lengthy proposition. Don Lowry, the publisher of Tractics, couldn't rely on an expensive and slow source for supplying the needs of his customers.
News of an American supplier of 20-sided dice began to spread in mid-1972 through wargaming zines like The Courier, as in the notice from Dion Osika above. The dice were also prominently featured in the first issue of the People's Computer Club magazine of October 1972, with an advertisement that showed a spinner and then five "superdice." Intriguingly, the supplier, Creative Publications of California, only sold their 20-sider in a set with four other dice: one of each Platonic solid. These five geometric shapes alone have a special property (that all of their faces, edges and vertices have the same relationship to their center of gravity) which makes them ideal as dice. Both the 6-sider and 20-sider are Platonic solids, and the list is rounded out by the d4, d8 and d12; all of these shapes have origins that go back into pre-history. The applicability of dice other than the d6 and d20 to wargaming was not clear off the bat; although Dave Wesely, in his work revising Totten, had gone on an epic quest to find a twelve-sided teetotum, ultimately his Strategos N rules did not require anything but d6s. Since the Creative Publications dice were comparatively inexpensive, Don Lowry began to stock them as a service to Tractics customers, initially without even a mark-up.
Gary Gygax frequently helped Lowry to promote new products in his mailer Lowrys Guidon, and with the arrival of polyhedral dice came Gygax's seminal June 1973 article "Four & Twenty and What Lies Between." In it, Gygax explains the probabilities that can be resolved with combinations of d4s, d8s and so on, though he concedes that "the most useful are the 20-sided dice." He knew they could be used for more than just generating percentile numbers: you can also "color in one set of numbers on the die, and you can throw for 5% -- perfect for rules which call for random numbers from 1-20." What would you use that for? Coincidentally, Gygax confides in that same article that he was "busy working up chance tables for a fantasy campaign game." That game, of course, was Dungeons & Dragons.

The first edition of Dungeons & Dragons made far less use of polyhedral dice aside from the d20 than later editions; the d4, d8 and d12 make only very rare appearances. Nonetheless, polyhedral dice quickly became a signature feature of D&D. They were moreover an early stumbling block when demand for the game was high: one could easily photocopy rules, but not dice. TSR continued to resell the Creative Publications dice, but at a considerable mark-up: initially $1.75, then $2.50, then $3.00. Many gamers therefore experimented with alternative methods for generating numbers, reminsicent of the Korns table above. Others eventually found the source of the dice and ordered directly from Creative Publications. By the 1980s, TSR had sufficient sales to strike aggressive wholesaler agreements, and it was only then that they augmented their sets of Platonic solid dice with a newcomer: the d10. In the years since, mad scientists like Lou Zocchi have produced all sorts of unusual dice, some more practical than others. Today, we can't imagine polyhedral dice without thinking of gaming, but their association with games is the sort of happy historical accident that frequently accompanies success.




18 comments:

  1. Wonderful post! I remember Lou's very first injection molding machine, one of those high-school shop class machines. Lou didn't empty the feed hopper between different-colored runs of plastic pellets, but sold the resulting 'swirls' for higher prices as they were 'custom' dice! :)

    yours, Chirine

    ReplyDelete
  2. Wow! I have Michael Korns' "Modern War in Miniature" and his later "SUTC" games. They are a fascinating peek into how the RPG hobby might have developed based on WWII wargaming rather than fantasy wargaming. I had no idea Korns' work was so influential in developing percentage-based randomizers.

    ReplyDelete
  3. I cannot speak to how the dice were used in the original Dungeons and Dragons game, but I recall that the d12 had very little utility in Moldvay's basic rules. If I recall correctly, 2d6 were used in determining the results of morale checks for monsters, and 2d6 were also used to determine whether a cleric's attempt to turn undead succeeded. My only memory of ever tossing a d12 was when I was randomly determining treasure for an encounter.

    I recall reading in an old interview that all of the platonic solids were made available from one manufacturer at the same time. The Moldvay boxed that I purchased came with a full set of dice. I have seen pictures that suggest that some of the Holmes boxed sets came with a piece of cardstock with chits that could be cut out and placed in a bag or cup as a replacement for dice. I have no idea whether dice were included with the original games. However, it does seem pretty obvious that nobody felt the need to market the dice individually by giving each at least one unique role, even though the photo above shows that they were sold individually. My guess is that the dice didn't need promotion through manipulation of the rules. They pretty much sold themselves.

    ReplyDelete
    Replies
    1. OD&D did not ship with polyhedral dice in its box, but they were available as a set separately from TSR from the start. It's true that supply problems forced some of the Holmes set to ship with chits rather than dice.

      The photos above don't show polyhedral dice being sold individually: Lowry only sold them in sets, though he sold d6's individually. Pre-D&D, d20s were sometimes sold in lots of two or three without any of the other Platonic solids, but by the time D&D came around you pretty much had to buy sets with one of each Platonic solid. That had changed, of course, by the time Moldvay came into the picture.

      Delete
  4. Great post Jon! What do you make of Wesely's claims that he got a set from a scientific supply store and Arneson's that he bought a D20 in England.

    ReplyDelete
    Replies
    1. Don't leave out Gygax's account that he found them in a school supply catalog. I guess we have three plausible but mutually incompatible claims for who introduced polyhedral dice to gaming. They can't all be right, but they can all be wrong.

      I'm sure there's a kernel of truth in all these stories. When Wesely was working on Strategos, no doubt he would have gone looking for a device that generates numbers from 1-12. But if he found them, where are the chance tables with ranges from 1-12 in "Strategos N" (1970)? If they were available to Arneson during Blackmoor, why does he say in the intro to the "First Fantasy Campaign" that they had "no funny dice back then," and why do we see no "funny dice" ranges in Blackmoor tables from the era? My piece above is about the causal chain that the surviving evidence shows. I think that causal chain is a sufficient explanation for the presence of these dice in D&D and subsequent games.

      Delete
    2. Hey Jon, nice piece, interesting blog. Your book, "Playing..." just arrived to today, can't wait to dive in. Rumor has it the IFW and myself are in it somewhere! ;)

      Delete
    3. Yes, Mr. Speer, quite right, you may read a few words about yourself and the IFW within. Hopefully they will conform to your recollections. Enjoy!

      Delete
    4. Arneson confirmed a draft found in MAR Barker's estate as having been "Probably one of his" - Beyond Here There Be Dragons has extensive percentile mechanics, some d6 mechanics, but no other polyhedrals.

      It is also worth noting that wargaming didn't generally adopt dice other than d20 (as d10/d%) and d6 until the mid 1980's, tho' the odd d12 showed up. Many wargames still stick to d6, d10, d20, and d100... This vastly simplifies the supply chain.

      Delete
    5. I gather you're not aware that I own that "draft" that Arneson referred to here. What you've heard about it is not accurate, in any event. It does contain limited polyhedral usage, though less even than OD&D.

      Delete
  5. > the advertisement from the Bristol Wargames Society refers to them as "percentage dice" and doesn't even say how many sides they had.

    The pic in the advert even shows the correct number sequence on the faces, per http://www.acaeum.com/forum/viewtopic.php?f=2&t=11152&p=177716#p177716 (black die on the left).
    It's almost certain that Arneson didn't pick up those London dice quite as early as he recalled. The Bristol crew were a bit slow on filtering through the d% mechanics to their own role-playing games, too.

    Thankfully Zocchi's early design suggestion for 10 sided percentile dice pairs /didn't/ catch on!

    ReplyDelete
  6. Hey Jon, dove in last night. I would say you have captured the true essence of my concept early on: we play wargames to have fun not "conquer" a region. Your account of our Convention in 1967 is also accurate, to a point. We did not blow out the treasury top cover the $200 hole we were in. My dad, gave us the cash. Quite a chunk for a truck driver with a teacher mother.

    Also, one other quibble. I did not quit the leadership in 1967. It was in '68 when I graduated from high school and entered PA. Military College. There was simply not time to do anything at that point.

    I am still in touch with Scott Duncan, he too lives in Georgia. I have a painting Gary Gygax gave me on my birthday of a Roman Legion being attacked by the barbarian horde and tons of IFW materials including a program from the con and copies of our newsletter you mention. If you are interested in seeing any of these, let me know. You can find me on Facebook quite easily.

    ReplyDelete
    Replies
    1. Thanks so much, I would love to hear more about some of the material you're still holding. I will definitely connect up on Facebook.

      Regarding the treasury and the leadership of the IFW, my account there follows an article Scott wrote up in 1969. I'll follow up about that offline too.

      Delete
  7. The first reference to Len Lakofka has his last name spelled Lafoka. A bit later you have it correct

    ReplyDelete
  8. Nowadays, you can buy all sorts of custom dice sets (or design your own) using 3D printing techniques (see, for instance http://www.shapeways.com/search?q=dice). Not exactly a cheap way to go, but for folks looking to create unique designs (or shapes: http://www.shapeways.com/model/157304/catalan-dice-bundle-1.html?li=productBox-search) it's an interesting option. I've recommended it to a couple of people looking to prototype custom dice sets for games.

    ReplyDelete
  9. It is well worth noting that original D&D was released with rules that required only d6's (and the companion miniatures game, Chainmail, also d6 only), but had an option for the d20 in the Alternate Combat System. No other polyhedrals are referenced at all.

    It was not until the 1975 Supplement 1 that the d4, d8, d12, and d100 uses were noted.

    ReplyDelete
    Replies
    1. That would indeed be well worth noting, if it were true. If you have an original D&D, this is quite simple to verify: look at for example the treasure types (M&T pg. 22), the numerous places where percentile resolutions are required for monsters, spells (like say "Contact Higher Plane") and magical items, the random encounter tables (U&WA pg. 10-11, then U&WA pg. 18-19), the "Number Appearing" lists (M&T pg. 2)and so on. All of these clearly stipulate ranges that require polyhedral dice.

      There are numerous places in OD&D were polyhedral dice are explicitly referenced: e.g., in the "Elves" monster text, "Roll a four-sided die for level of fighting...", or under the "Staff of Wizardry" how for elementals you "roll four-sided die to determine how many...". In the naval rules, "wind direction is determined by rolling an eight-sided die." You "roll two eight-sided dice" to determine how many snakes are created in the "Sticks to Snakes" spell. The "score of a twelve-sided die" determines how many turns it takes for a Confusion spell to take effect.

      All of those quotes are from the first printing of OD&D.

      Delete