Spells Through The Ages – Duration

Here's another entry in the top-level Spells Through the Ages investigation, with one issue in particular for OD&D that I'll focus on:

Some initial observations: Where no specific duration is listed, the spell in question is either by concentration, instantaneous, indefinite, permanent, special in some complex way, or simply with duration not given (undefined). Most of the Chainmail spells seem to presume ending when the caster wills it or someone else dispels it. Of the 6 with specific durations, half of those were in some sense carried over to OD&D. Again we see the usual transition from 1E to 2E, in which almost everything was directly copy-and-pasted. But what you do see here, which is unusual, is an almost complete and total makeover (in both numbers and units) between OD&D and AD&D 1E. Why is that?

OD&D confuses its own time units. The Vol-1 spells are all entirely in units of "turns". (With one single exception: infravision has a duration of 1 day.) The spells from Chainmail mostly carry forward with similar durations, still in "turns". However, according to OD&D Vol-3, the meaning of the word "turn" has ostensibly changed. Whereas in Chainmail "turn" means a single combat sequence of 1 minute, in OD&D it supposedly indicates a 10 minute unit that many of us are familiar with. Instead, the 1 minute combat sequence is now supposed to be called a "round" (Vol-3, p. 8).

But that's Vol-3; Vol-1 (the one containing the spell list) doesn't display any knowledge of that definition anywhere. None of the spells are given durations in terms of "rounds" (nor "hours" or anything else). The first usage of more varied units comes in Supplement-I; there you have some new spells with durations in hours (web, wizard lock) or a week (suggestion).

And something even more intriguing: the various monster summoning spells in Sup-I are given durations in units of "melee turns", which is not even a defined term anywhere in OD&D. Somehow the writer in question clearly threw away, or forgot, or was not aware of, or was writing prior to, the distinction of "round" as it appears in Vol-3, p. 8.

So consider what happens with the switch to AD&D. The great majority (72%) of spells with duration given in "turns" in OD&D changed to "rounds" in AD&D 1st Edition. (There are 36 such spells: 26 change to "rounds", 10 remain as "turns".) This one switch -- really, a belated realization that the term "turn" has changed -- primarily constitutes the highlighting of the entire "AD&D1" column in the table above.

Case study: I'll use the haste spell about which I wrote previously. In Chainmail, it lasted for 3 turns (i.e., 3 minutes or combat-cycles). In OD&D, it also lasts for 3 turns (but now that means 30 minutes). In AD&D 1E, it's 3+level rounds (i.e., back to 3+ minutes/combat-cycles), and this unit is carried forward uniformly to 2E, 3E, and 3.5E. So OD&D is the only edition where haste -- and combat-oriented spells like it -- last for whole or numerous combats. (And in my last game, the spontaneous player opinion was that haste would be very powerful even if it did indeed last just 3 rounds.) However: Note that when the Holmes/BXCMI line branched off, it retained the OD&D duration measurements, such that haste was always 30 minutes throughout that series.

Question: Was OD&D's use of "turn" for spell durations fundamentally a mistake? I believe it was, as evidenced by the intent of the editions which came both before and after it. In fact, my main hypothesis is that Vol-1 (with its spell list) was written prior to the terminology change in Vol-3 (with its 1 turn = 10 minutes), resulting in big unanticipated consequences.

As one last piece of evidence, it's worth pointing out that several of the spells we normally consider to be instantaneous -- fireball, ice storm, and dispel magic -- were given a duration of "1 turn" in OD&D, which is totally nonsensical if that should mean 10 minutes, but somewhat more defensible if it was meant to indicate 1 individual combat cycle.

So here's a new poll -- If you had to interpret the OD&D Vol-1 spell durations one uniform way, do you think that "turn" should be understood to be 1 minute (as in Chainmail) or 10 minutes (as in AD&D)? Or something else? (See poll results here.)

Final notes: The troubling issue of what "turn" meant in Vol-1 spell durations is something I entirely sidestepped when writing Original Edition Delta: Book of Spells. I used more-or-less the same terminology throughout, and figured the reader could apply their same standard interpretation.

As far as duration numbers go, in OD&D durations were fixed, most commonly 3, 6, or 12 turns. In AD&D, durations became a function of caster level, usually 1, 2, or 5 times the level (in rounds). This is another place where I think OD&D has a leg up on AD&D, through both of the lenses that I usually consider: (1) As matter of verisimilitude, if a spell is a fixed "formula", it's hard to see how caster identity would alter their effect; and (2) as a matter of gameplay, I look dimly on having to do numerous multiplications every single time a spell is cast.

Also, I stopped the table above prior to the 3E of the game, partly for space considerations, and partly because 3E put all the spells in a radically different order, such that it's an enormous pain to search through for this purpose. Suffice to say that a lot of changes occurred, with the "turn" unit being disposed of, and many spells given in terms of rounds (now 6 seconds), minutes, tens of minutes, and many other units.


  1. I use 1 turn = 10 minutes for spell durations, but that's me reading the B/X D&D definition back into OD&D.

  2. You're actually simplifying the OD&D terminology quite a bit here.

    The section in question is titled THE MOVE/TURN IN THE UNDERWORLD on page 8 where it appears. But on the table of contents it's listed as "The Move Turn in the Underworld". Compare the use of the phrase "move turn" here with "melee turn" from Supplement 1 & 3 is perhaps suggestive of Gygaxian usage at the time.

    But not exactly, because the actual section on page 8 of Volume 3 includes the following:

    "Movement (distances given in Vol. 1) is in segments of approximately ten minutes."

    "Thus it takes ten minutes to move about two moves--"

    "Two moves constitute a turn, except in flight/pursuit situations where the moves/turn will be doubled (and no mapping allowed)."

    "Melee is fast and furious. There are ten rounds of combat per turn."

    So we have:

    melee turns (1 minute?)
    move turns (10 minutes?)
    turns (10 minutes?)
    segments (10 minutes)
    moves (one-half turn or one-quarter turn while running)
    rounds (1 minute)

    But also take a look at page 18 of Supplement 1: You've got a giant snake doing 2-8 points of damage per turn. That certainly can't mean a 10 minute turn, can it? It must mean a round/melee turn, right?

    And on page 1 of Supplement 1, monks are given the ability to stun opponents for 3 to 12 turns. That probably means rounds, too, right?

    But on page 2, monks are allowed to simulate death for 1d6 x level "for the number of full turns". Should we interpret "full turn" as a "move turn" from Volume 3? Context seems to imply that it must mean something different than the use of "turn" on the previous page.

    But how should we interpret the use of the term "turn" in the assassin class' poison ability, then? (Chance per turn of opponents recognizing the poison and turning hostile.)

    Or what about this use of the term "turn" on page 8 of the same:

    "Decapitation of limb (sic) will cause one damage point per limb lost per turn left untreated."

    That's gotta mean that "turn" is being used to mean "round" here, right? Particularly when we read for snakes (page 11): "Destruction of the head will result in death but the body will writhe for 2-16
    turns." He surely didn't mean that the snake would keep writhing for at least 20 minutes and up to 2.5 hours, right?

    It's therefore tempting to read into this a usage of "turn vs. full turn" used by Arneson in the same sense that "round vs. turn" or "melee turn vs. move turn" was being used by Gygax.

    And if this is actually the case, it would certainly go a long way towards explaining why this terminology is so hopelessly confused in OD&D: You had two different authors using the term "turn" with completely inverted meanings.

    Perhaps the central crux can be found when we hit the description of the giant squid: "Squids tire easily; there is a 50% chance that they will withdraw after three rounds of melee, with a 5% increase
    each turn thereafter."

    Arneson displays knowledge of the term "round", but does his use of the term "turn" in immediate juxtaposition indicate that he's using the terms interchangeably? To my eyes it seems almost certain, but perhaps others disagree.

    OTOH, there's a fair degree of evidence that they just used the term "turn" to mean whatever the hell they felt like it should mean. For example, on pg. 43 of Blackmoor:

    "There is a 1 in 6 chance that when entering passages marked with a dashed line or when crossing one of the bridges that 1-3 trolls will be encountered. Any fighting will bring an additional 1-3 trolls every turn the fight exists."

    It's surely not possible that Arneson meant that 1d3 new trolls should show up every single round, right?