Tuesday, September 29, 2015

How Much Does a Dragon Eat?

Obligatory Joke: As Much as it Wants!

But seriously... the other day, my daughter and I were talking about dragons. She'd been re-reading Wizard of the Coast's "Practical Guide to Dragons" series, and, well... let's just say that while the art is great and the writing is good, the books could be used as poster children for the "Writers Cannot Do Math" trope.

They thankfully don't even try to give amounts that dragons eat. So, when she asked about how much that might be, my physics/computer science background came out, and I started doing some estimating. I started with the knowledge that a working cart horse needs 30 to 45 pounds of food a day. Estimating the horse's weight at 1000 to 1500 pounds, and looking up the Calories / pound of various foods, I estimated that an animal of the same weight eating meat would need about 1/6 as much food... for 5 pounds of food per day per 1000 pounds of body weight, of 10 pounds/day/ton.

We then decided that since dragons fly and breathe fire, they'd likely need more energy. So we decided to somewhat arbitrarily double the amount, to 20 pounds/day/ton. Given a 7-ton dragon (the weight of a large tyrannosaur or elephant), that comes to 140 pounds/day, or 980/week... that is, about a sheep a day, or a cow a week.

Of course, that's presuming the animals being eaten are 100% meat, which isn't too likely. So let's make it two sheep a day, or two cows a week. (Some of which will likely be wastage.)

However, predators tend to eat less regularly, eating all they can at a meal, then not eating for a considerable length of time. While some types of reptiles can eat their own body weight or more, most animals cannot. If we assume a dragon can consume up to 1/4 its own weight and still fly, which seems reasonable (and about matches how much a lion will consume at a meal in the wild), then our 7 ton dragon will eat about 1.75 tons at a sitting, or 3500 pounds - which is about 3.5 cattle.

That gives us a dragon who kills and (mostly) eats four cattle about every two weeks. This doesn't seem unreasonable, and would certainly make a big impact on local farmers, especially when you consider that each time the dragon feeds, it's likely to panic the herd of cattle it's feeding from, causing additional injuries (and possibly deaths), knocking down fences, and so forth.

Later on, I looked up how much predatory birds eat, thinking that would be a reasonable model for flying dragons. The relevant formula there is:

Maintenance Metabolic Rate (kcal/day) = 1.5 (78 (weight in kg 0.75))

(found at The Modern Apprentice, a falconry site)

This is adjusted for bird activity, with multipliers from 0.7 for an inactive bird, to 1.3 for a highly active bird. To two significant digits, we get 83,000 kcal / day. Now, here, we need to note that food industry Calories are actually kilocalories. Using beef's roughly 850 kcal / pound, we find that this is 98 pounds of beef / day, a bit less than the 140 pounds / day we arrived at above. However, digging into some other sources, carnivore digestion is only about 90% efficient... and, of course, we have the fact that a cow isn't pure beef. Between those, we can probably keep the above figure of 20 pounds/day/ton, and not be too far off.

(I also remembered later that Gregory Paul's Predatory Dinosaurs of the World, which I am a happy owner of a copy of, has an appendix on predatory dinosaur energetics. There, I found a similar formula, of 140 (weight in kg 0.75)).

The one thing we might want to do is use the weight 0.75 scaling factor from above, when figuring larger weights, at least. So, for a huge dragon, say weighing the estimated 90 tons of the largest titanosaurs, multiply by ( 90 / 7)  0.75 = 6.8... for a creature that eats around 25 cattle every two weeks!

Of course, if dragons hibernate, then this changes significantly. Bears depress their metabolic rate to about 25% of normal when hibernating - using that model, our huge dragon might each 25 cattle, then return two months later.

Another alternative would be to use cold-blooded animals' metabolic rates, which are about 1/10 that of warm-blooded animals... so our dragons would eat about once every 5 months. If they also hibernate, that could be every couple of years!

Monday, September 28, 2015

Toward a History of RPG Mechanics, Part 6: Tunnels & Trolls Magic and Miscellany

Last time, we started off looking at the second-created RPG, Tunnels & Trolls, with a bit of delving into the character creation and combat system. Now, we turn to magic.

Magic-Users and Spells

Having already seen how Ken St. Andre chose to make attributes more meaningful in T&T than they were in OD&D, it should come as no surprise that he continued this with the magic system. OD&D's magic system is based purely on class and level - while we're told that Intelligence is the primary attribute for magic-users, the magic rules of pre-Greyhawk OD&D do not actually use the character's Intelligence in any way. Instead, what spells they can possibly cast depends on the character's level, and how many spells they can cast does too. Spells are also assigned a level, but spell levels do not have a 1-to-1 relationship with character levels.

In T&T, spells are still organized into spell levels, but what determines whether you can cast spells of a given level is your Intelligence (IQ, as T&T abbreviates it) and Dexterity (DX). 1st-level spells are not given an IQ requirement, but higher level ones require an IQ of (8 + 2 x spell level). All spells require DX of (7 + spell level). The highest level of spells given in the 1st edition of T&T is 17th, requiring IQ 42 and DX 24!

As in D&D, however, magic-users do not automatically know spells simply because they're theoretically capable of casting them. They do start off knowing all 1st-level spells, but gaining access to higher level spells requires learning them. T&T, however, posits a magic-users' guild, which teaches spells. Thus, characters can spend money to learn new spells, with set amounts being given for each level (in contrast to D&D, where the acquisition of new spells for magic-users was left essentially to GM fiat in placing spell books as treasure, and/or deciding what spells NPC magic-users would teach and for what price).

Strength (ST) was also a considerable factor for magic-users. Instead of using the D&D 'Vancian' system, with a set number of spells of each level that a magic-user could memorize and cast, T&T instead gave each spell a cost to cast, measured in ST points (taking the fairly common fantasy idea that casting spells fatigues the caster). For 1st level spells, these costs ranged from 1 to 10 ST (excepting the Detect Magic 'spell', which was described as an inherent power that cost no ST to use). Spell costs tended to increase at higher levels - the lowest cost for 3rd and 4th level spells was 7 ST, the lowest for 5th level spells was 15, and the 17th level spell "Born Again", which could allow the magic-user to reincarnate another PC or even themself cost a whopping 208 ST! Spent ST would recover at 1 point / turn (10 minutes of in-game time).

There were, however, two ways to lower the cost of a spell. The first was by gaining levels: a magic-user of higher level than the spell they were casting had the ST cost reduced by 1 for each level they had over the spell's level. Thus, a 3rd level magic-user casting a 1st-level spell had the cost reduced by 2 points. This could not, however, reduce the cost below 1 ST - Detect Magic was the only "free spell", no matter how good a magic-user you were.

The second was through use of a magic staff. One could either be bought or made, and either type reduced the cost of casting any spell by the magic-user's level (again, to a minimum of 1). Self-made staves would "burn out" once the magic-user had used it to cast spells with a total ST cost of 2 x IQ, but bought ones would last forever. Of course, any player of a magic-user would make it a point to buy a permanent staff as soon as possible. Thus, in practice, a magic-user would almost never pay 'full price' to cast a spell.

T&T also designated several spells as being able to be cast at higher levels. Doing so increased the cost of the spell - adding one level to it doubled the cost, adding two levels tripled it, and so forth. However, it also doubled the potency of the spell for each additional level it was cast with, giving an exponential increase in spell power. When cast in such a way, however, the new level of the spell would be what was compared with the magic-user's level when calculating any discount on the ST cost. (The new level would also be used for the IQ and DX requirements to cast the spell - thus, you couldn't "boost" a spell up to a level you weren't qualified to cast!)

For an example, let's use the basic attack spell of T&T: "Take That, You Fiend". This 1st-level spell allows a magic-user to use their IQ as a weapon in combat, and costs 6 ST to cast. A 3rd-level magic-user casting it in its basic form, using a staff, would pay only 1 ST for the casting (6, -3 for the staff, -2 for being 2 levels higher than the spell). Casting it at 2nd level would cost 10 ST (6 doubles to 12 for boosting a level, -3 for staff, -1 for being one level higher than the new spell level), and casting at 3rd level would cost 15 ST (6 triples to 18, -3 for the staff)... but would add 4x the caster's IQ to the party's combat total! Since the minimum IQ to cast a 3rd-level spell is 14, this would be at least 56 points.

Interestingly enough, the rules of 1st edition T&T do not limit a magic-user to casting spells of their own level or below - hypothetically, a 1st-level magic-user lucky enough to have IQ 18, DX 12, and ST 14 could cast the 5th-level "Dear God?" spell, if they spent the 2000 gold pieces to learn it!

A total of 69 spells are presented in the original T&T. These are heavily weighted toward low-level spells - while they go up to 17th level, more than half are 4th level and below. However, 27 of the spells presented can be boosted in level. Many of them are similar in effect to D&D spells, but they often vary significantly in level from their D&D counterparts. There are also quite a few that do not correspond to any D&D spell, including several spells meant to boost the effectiveness of other party members by enchanting their weapons and armor. These include some effects that D&D would add to its spell lists later on - Wall of Thorns, Ghostly Going (an astral projection spell), Summoning (to call up demons), and even Take That You Fiend (which is effectively D&D's later-added magic missile).

Several D&D spells don't have equivalents in the original T&T. These are mostly spells that would be useful outdoors, such as move earth, hallucinatory terrain, and massmorphPass-wall is also given no equivalent, which, together with the others, may betray a focus on the 'dungeon' environment greater than that imagined by Gygax. Several of the more Biblically-oriented clerical spells are left out as well, such as turn sticks to snakes, part water, and insect plague.

But while most of the individual spells are similar to D&D ones, the system as a whole is definitely very different. T&T's magic system would be an inspiration to the writers of many other early fantasy RPGs, much more so than D&D's, as we will see as this series continues.

Odds & Ends

It's of note that the original version of T&T gives nothing to warriors to give them any sort of advantage over rogues. Magic-users are limited to one-die weapons (if they wish to retain combat adds), disadvantaging them significantly in combat; rogues start off not automatically knowing any spells, and can only learn new ones if a magic-user who knows it is willing to use the Teach spell to teach it to them (they can't learn spells by reading them, and can't create new spells). They also don't get the ST cost discount for casting lower level spells that magic-users do (it's unclear whether they can use a staff in 1st edition).

This is likely the motivation behind limiting rogues to 7th level. Interestingly, 1st edition T&T does allow characters to switch classes, with the limits that a character who began as a warrior can never become a magic-user, and that anyone who switches classes loses all previous experience points and must start over at 1st level. They apparently also lose all the ability score advancements they had gained, since the rules comment that they might as well just roll up a new character. Rogues, then, get a special benefit here, since at 7th level they can choose to become a 5th level warrior or 3rd level magic-user. Whether they kept ability advancements isn't stated, but it seems likely that they did. Indeed, the rules suggest that if you have a character whose Luck is their highest ability (thus, making them good at surviving saving rolls, if nothing else), that you start them as a rogue. This would make good sense if you could earn XP as a rogue and advance the attributes you need for the class you plan to eventually take.

The rest of what's in 1st edition T&T is a bit of a grab-bag. There's a page of adventuring equipment, rules for weapons that are made of things other than steel (iron, bronze, copper, or stone), various GM advice on creating and stocking dungeons and running the game, and rules for berserkers. Let's look at that last:

PCs with IQ 8 or less always go berserk when they are eligible to. Those with IQ 16 or greater never do, unless forced to by magic. Characters using missile weapons do not go berserk. A character has an opportunity to go berserk when they get doubles on their combat dice, or get a 6 if they only have one combat die.

While berserk, the character gets no combat adds or subtracts (making it very appealing for characters with low combat attributes!). They still roll their normal weapon dice, and if two or more of their dice come up with the same number, they re-roll those dice and add them to the previous total, continuing to do so as long as they still have matches. Each matching group is treated separately, however, so a character with 5 combat dice who rolled 1, 3, 3, 5, 5 would roll the pair of 3's and see if they came up with another pair until they failed to do so, then repeat with the pair of 5's.

Those with only one combat die simply get an automatic six. A character with only two dice who does not roll a pair lowers their higher die to match their lower one to make a pair, then continues from there.

Each round that a character is berserk, they lose 2 ST. If their ST goes to 5 or less, they become too weak to fight, and stop berserking. Otherwise, they can be stopped either by someone with a Charisma of 15 or better choosing to calm them down, or by making a saving roll for them to recognize their friends and stop; this saving roll can only be tried once. Lastly, any magic-user spell that knocks them out also ends their berserking (so, a Sleep spell can be useful to stop them!).

An exhausted Berserker (reduced to ST 5 or less) cannot fight, but can still be in a fight - that is, can still take damage if the other side wins a round. Lost ST from berserking recovered the same as that used for casting spells - at 1 point / turn.

Man-like monsters could also go berserk, but had very different rules for it:

They would go berserk on a reaction roll of 2 (worst possible) or if cornered alone and 'facing certain death'. Further, when half or more a group of monsters was slain, the GM was to roll 2d6 for each remaining monster: on a 10 or higher, that monster would go berserk; otherwise, it would try to run away.

Berserk monsters got to roll their normal dice and use 1/2 their MR as adds. Further, they would get to re-roll 1s on their monster dice and add. Their combat score would be taken separately from that of other, non-berserk monsters instead of being added in. Thus, berserk monsters do not coordinate with others!

They would lose 5 MR each round, but would never become exhausted - the rules say "they continue snapping feebly at your boots until they have gasped their last gasp."

Summing Up

The authors of Runequest put this dedication in the second edition:
This book is dedicated to Dave Arneson and Gary Gygax, who first opened Pandora's box, and to Ken St. Andre, who found it could be opened again.
Tunnels & Trolls wasn't really needed to show that - plenty of others were working on other RPGs at the time. However, it did expand the field of RPG mechanics in some important ways. The first was the emphasis on using attributes to do things, instead of class and level. Outside of its use to lower the casting cost of spells, level in T&T was nothing more than a track of when and how much you'd get to increase attributes. This in turn pointed the way toward the idea of level-less and class-less systems.

Second, it was the first deliberately simplified RPG, a reaction to the confusing mass that D&D was, with 112 pages split into 3 books (156 if you include Chainmail). T&T managed to put its basic rules into 31 pages, with 10 pages in the "Elaborations" (excluding the magic rules, which I'm counting as part of the basic rules). The smaller size made T&T easier to learn, and also meant that finding things was not as painful as in OD&D, where, for example, to find all the abilities that an elven player character had required looking in all three books (the basic description of the elf PC race in Men & Magic, the special abilities of elves mentioned in Monsters & Treasure, and the rules for finding secret door in The Underworld & Wilderness Adventures).

The use of attributes directly as "hit points" and "mana pool" was a simple and seemingly obvious jump that many RPGs would borrow. T&T also introduced the idea of exploding rolls, with the "doubles add and roll again" rule for saving rolls, permitting any character to have a chance, however slim it might be, of succeeding at any saving roll.

Third, as Dale Newton pointed out in his comment on the last column, T&T was leaning in a narrativist direction before there was such a term. Gygax spent a good deal of effort on realism - we can see that in the weapons vs armor matrices of Chainmail, which would be brought into the D&D "alternative combat system" in Greyhawk, and make another appearance in the first edition of AD&D. Ken St. Andre chose to model T&T primarily after 'sword & sorcery' fiction, but also brought in elements of superheroes and of children's fantasy (the Wall of Thorns spell actually states "as in Sleeping Beauty"). While D&D also it emphasized the use of common sense and going with 'what makes a good story', the size and detail of the rules presented tended to drown those admonitions out somewhat. T&T's smaller size lent proportionally greater weight to the admonitions in it.

Also, Gygax presented the D&D rules largely as a fait accompli - as if they'd sprung fully-formed from his and Arneson's heads. In contrast, T&T was written with a tone more as if Ken was talking to you, with phrases like "if anyone ever gets more than a million experience points, let that person decide what the higher levels are up from there", examples mentioned of what other people had done in their games (with their names!), Ken addressing the reader with "I" when describing how he did certain things, but finishing with the suggestion to try it a few different ways and see which you liked, and the crediting of certain rules to other creators.

The net result is that T&T's rules invited those using them to tweak and change them, in a way more clear than D&D did. By the time I joined the hobby, in the early 1980s, Gary Gygax was regularly making admonitions in Dragon magazine that the AD&D rules should not be changed or tinkered with, but T&T never lost its open, somewhat 'gonzo' feel.

Next time, we'll be stepping on to the first RPG to step outside the fantasy genre - Blume & Gygax's Boot Hill!

Monday, September 21, 2015

Toward a History of RPG Mechanics, Part 5: Tunnels & Trolls Character Creation and Combat

Dungeons & Dragons was, of course, the first published RPG. However, when it was first released to a wide audience in 1974, it was quite expensive (thanks to the magic of inflation, those original three little books cost the equivalent of $50 in today's money - and dice weren't included, bringing it up to $65!), and was not well-distributed.

So it was that a young man by the name of Ken St. Andre heard about D&D late in 1974, and thought the concept sounded interesting, but was unable to find a copy to buy, or even peruse in his hometown of Phoenix... until one evening in April of 1975, when he visited the Flying Buffalo game store, and found that someone from out of town who was there that evening had a copy with him. Ken got to spend about an hour looking over the books, and, to quote him, said, "What a great concept! What a lousy execution! I’ll write something that me and my friends can play."
Over the course of the next week, he did exactly that, and then played his new game with friends the following weekend. It was a hit, and over the course of the next two months, more and more of his friends and their friends wanted to play... and many wanted copies of the rules.

Ken had done fanzines before, and decided to publish his game as a proper book. They'd been simply calling it "Dungeons & Dragons", but he knew they couldn't publish it with that name, so a discussion was held, and the name Tunnels & Trolls settled on. Ken prevailed on a friend to draw some illustrations, took the results down to the university copy shop, and had 100 copies made and stapled together for $60, planning to sell them to his friends at $1 a copy to recoup his costs. He also decided to spend $10 to register copyright for it.

He managed to sell 50 copies of that original print run to his friends, and persuaded Rick Loomis of Flying Buffalo to try to sell the rest at a convention. Those copies sold out, Rick became interested in helping to publish the game, and T&T was off to a running start!

But we're about mechanics here. So, how did T&T differ from D&D?

Ken was a gamer, but he wasn't a miniatures gamer - thus, T&T had no "inches of movement" - indeed, it had no means of measuring movement at all, the presumption being that the action would simply be envisioned in the heads of the GM and players. It did encourage making maps on graph paper, but this was simply to aid the GM in keeping track of what was where in the dungeon, not to be used as a playing surface.

Characters had a very similar set of attributes to those from D&D, with the only difference being that Wisdom was removed as being too similar in concept to Intelligence, and Luck substituted in its stead. Just as in D&D, the attributes were generated by rolling 3d6 six times. Everything was done with six-sided dice, because Ken knew of no way to obtain polyhedral dice... but had plenty of six-sided dice from games that used those. As noted in part three of this series, OD&D made very little use of attributes. Tunnels & Trolls, however, made much fuller use of them, which we'll describe below.

Classes, levels, and races were features of T&T, but they were quite different from those in OD&D. To start with, the fantasy works that Ken was familiar with featured religion only in minor ways - and thus, he dumped the D&D cleric class (which went quite naturally with removing Wisdom as an attribute). The fighting-man and magic-user were retained (although "fighting-man" became "warrior"), and a third class was added: the rogue. This, however, was not a thief-type. Rather, it was a character who combined the ability to fight and do magic, along the lines of the Grey Mouser or Cugel the Clever. (The name going with them being "rogue magic-users" - people with the necessary talent, but who weren't members of the magic-users' guild.)

Levels worked very differently as well. In D&D, a character's level determined their combat ability and hit points. T&T, however, based both of these things on attributes. When a character gained a level in T&T, they could increase an attribute, with Strength or Constitution rising by the number of the new level (e.g., a character advancing to level 3 could choose to add 3 to their Strength or Constitution). This could be split, with each rising by half the level. Luck could be raised by double the new level, or one of the other attributes (Intelligence, Charisma, or Dexterity) could be raised by half the new level.

In addition, all three classes used the same experience table. Rogues were limited to 8th level, and could then choose to become either a full warrior or magic-user, becoming 5th level if they chose warrior, or 3rd if they chose magic-user.

The term "race" was avoided - T&T instead referred to "type", or, in later editions, "kindred". Types gave attribute adjustments, but via multipliers rather than adding or subtracting from them. Thus, a hobbit in 1st edition T&T has 1/2 their rolled Strength, but twice their rolled Constitution, and 3/2 their rolled Dexterity. The list of types was different as well, including D&D's dwarves, elves, and hobbits, but adding leprechauns and fairies. Multipliers were also given for giants, trolls, ogres, half-orcs, orcs, goblins, and gremlins, for the benefit of game masters who might wish to personalize their monsters a bit.

Available money was generated randomly, and used to buy equipment. Carrying ability was based on Strength. Weapons in the optional "Elaborations" (i.e., expanded rules, which were in the same book, just a different section) had minimum Strength and/or Dexterity requirements to use. Armor had no minimum attribute requirements, but did have weight, so low-Strength characters could not wear the heaviest armors.


As in D&D, combat was assumed to be a major activity in the game. A character's combat ability was based on their attributes, however: Strength, Dexterity, and Luck over 12 gave "combat adds", while those same stats gave minuses for values below 9. Thus, a character with Strength 14, Dexterity 10, and Luck 8 would have "combat adds" totaling +1.

These "adds" were added to the "dice and adds" that came from the character's weapon. A dagger gave 1 die (1d6, that is), a standard sword 2 dice, a flamberge 3 dice + 3 adds, and a poleaxe a massive 6 dice. Magic-users were limited to using only 1 die weapons, and did not receive combat adds for high attributes... but could still get penalties for low ones. This was meant to reflect that they had no training with weapons. They could wear armor.

Combat resolution was by each side rolling their dice and adding any adds they had to generate a total, which was referred to as their "hit points". The side with a higher total won the round, and the difference between their total and that of their opponents was done to the opponents as damage. This damage was to be divided between the members of the losing side as evenly as possible, with the note that magic-users should take the least if damage could not be divided evenly. Damage came off of Constitution, and when that was reduced to zero, the character was dead. Armor increased Constitution, absorbing damage until it was destroyed.

Before we get into further details of combat, a few observations. First off, the phrase "the best defense is a good offense" fits this system very well. This makes quite a bit of sense - a longer weapon helps keep a foe at bay, and also offers a greater area with which to defend. Attacking a competent sword-wielder with a dagger in real life is quite likely to get you killed without significantly hurting your opponent.

Second, the system of having players (and monsters!) add their combat rolls together makes a good deal of sense as well. Essentially, it abstracts the idea that party members are going to be assisting each other in combat. A poor combatant might not be a threat to the monster at all on their own - but by teaming up with someone else, they're creating a distraction, and they might be able to slip behind the foe and get a jab in while it's occupied with the fighter who's the "real threat".

This system abstracts melee to a much greater degree than D&D. There, each player rolls "to hit" separately, then rolls for damage if they did hit. Some gain more than one attack, and these attacks are resolved separately. Who dealt damage and to whom is definitely known. On one level, this all makes logical sense. On another level, however, it breaks down in conjunction with the details of the D&D combat system. Specific numbers of attacks per round, the use of defender facing in determining hit probability, parrying rules, the use of weapon length and speed, and using miniatures on a map all contributed to the feeling that D&D combat was supposed to be detailed and realistic... but the abstracted nature of hit points, armor that lowered the chance to hit rather than absorbing damage, and the one-minute melee round all argued for an abstract system.

Tunnels & Trolls' melee system, however, doesn't even allow attributing damage done to a particular combatant when there's more than one on each side. (Missiles and magic do, however, as we'll see.)

At the same time, though, the text of Tunnels & Trolls encourages the GM to adjust things as needed based on logical considerations. To quote:
"Each combat is a unique and individual experience that must really be played by ear. It is impossible, when you think about it, to have all 14 people in the party fighting three 30' high giants while in a passageway only 10' wide and 10' high. In the first place the giants couldn't get into the passage. Likewise if the tunnel is only wide enough for a single-file passage, you can't have all 14 hew into that cave bear that just came around the corner."
A couple of pages later, we have:
"[When] the group has enough time to prepare a plan, the caller will decide who is fighting what. He may say we're all attacking all the monsters [...] or he may specify that Furd is fighting the cave troll with his slingshot while Jiriel and I, Yardoom stand off the blue-fanged trolls."
So, T&T allows for such tactics as fighting in a narrow corridor so your stronger combatants can keep enemies off the weaker ones, forming a wall to prevent a monster from reaching someone, and so forth. When these things happen, the PCs and monsters would form multiple groups in the same combat, each group with its own opponent group. This allowed the players some level of control over who took how much damage.

For missile attacks, things worked very differently. If the target was a PC, they would make a Saving Roll to avoid being hit. If this failed, the GM would roll the monster's normal attack, but the PC would use their Luck as their defense, taking hits equal to (monster's attack total - PC's Luck).

If the target was a monster, the Dexterity and Luck of the PC would be added together, and the monster would roll its "monster dice" (which we'll discuss later on). If the monster's roll was lower, then it would be hit. The PC would then roll their attack dice for the weapon being used and add their combat adds, but adds (or minuses) from Dexterity would be doubled. The monster's roll of their monster dice would be applied as defense, so that (PC's attack total - monster dice total) would be the damage done.

This is an uncharacteristically complicated procedure for T&T, and would change in future editions.

We've just mentioned Saving Rolls and Monster Dice, so let's talk about those things quickly.

Saving Rolls

Where D&D's "saving throws" were mostly focused around avoiding the effects of magic, T&T's "saving rolls" were not. The rules state that they "are used chiefly to try and save a character from some form of mechanical (as opposed to magical) trap, for dodging missile weapons, for immunity to poisons, etc."

They were based on Luck, with the player rolling 2 dice and adding their Luck, trying to get a total of 15 + (5 x dungeon level) - i.e., 20 for the 1st level of a dungeon, 25 for the second, and so on. This however, was expressed a bit differently, as needing a roll of "20-luck number", "25-luck number", and so on. T&T continues to express saving rolls in this way even in the most recent edition (Deluxe Tunnels & Trolls), which somewhat puzzles this writer, since most people find addition easier to do than subtraction.

Doubles on a saving roll allowed the player to roll again and add, continuing to do so as long as doubles continued to be rolled. This gave even low-Luck characters a chance of making difficult saving rolls. However, on the opposite end, a roll below 5 was always a failure. In modern versions, this has been changed to "a roll of 3 is always failure" - a moment's thought will show that these are equivalent, because of the way doubles are treated.

Unlike D&D, saving rolls were not normally allowed against magic - whether in the form of spells or magical items.

Although not seen in this edition, in future versions, it would be mentioned that saving rolls could be based on other things than Luck - so, for example, a character with better Dexterity than Luck might use Dexterity to avoid a trap, Intelligence might be tested to try to spot a trap, and so on. While the idea of saving roll levels would remain, it would be decoupled from the dungeon level, making it simply an expression of difficulty.


As mentioned above, there were optional rules for treating the more humanoid sorts of monsters like player characters. However, in general, T&T used (and still uses) a very simplified system for monsters, called Monster Rating.

In this system, Monster Rating (MR) is the only statistic that most monsters have. In combat, a monster in 1st edition T&T gets a number of attack dice dependent on monster rating. They also receive combat adds dependent on MR, equal to MR/2 the first round of a combat, and MR/4 each round thereafter.

MR also doubles as the monster's damage capacity. The rules in 1st edition do not specify whether the monster's combat dice and adds fall as it takes damage. My understanding is that GMs interpreted this in several ways, from "yes, they do", to "no, they don't", and some using a middle ground of "adds fall, but dice stay the same". The last is the current rule in Deluxe Tunnels & Trolls.

Since monsters had only a Monster Rating and no Luck score, saving rolls worked differently for them as well, with MR used in place of luck, and difficulty being 50 + (50 x dungeon level). Unlike PCs, monsters could automatically make saving rolls. Of course, monsters with player-style attributes simply used their Luck.

Because the MR system is so simple, the original T&T rulebook had no monster section at all - the GM could simply make up any monster, assign it an MR, and be done. The general T&T rule of "adjust when it makes sense" applied, of course - as the quotes about combat mention above, a 30' tall giant shouldn't be allowed to enter a 10' tall passage. A dragon's breath weapon would simply be treated as a missile weapon attack. Any additional abilities a GM might want to give a monster could be represented via saving rolls (e.g., a poisonous snake with a low MR, but whose bite requires a saving roll to avoid death), or, in rare instances, by allowing the creature to cast one of the system's spells.

There was a suggestion that "puny monsters, like rats or dogs" should have ratings below 30, that a troll would be between 26 and 40, that a balrog might have MR 250, and a dragon MR 500. Lastly, it was specified that invisible monsters should have their MR doubled.

MR continues to be the standard method of handling monsters in T&T to this day. On the positive side, this means that T&T, unlike other systems, has never needed extensive bestiaries. The MR system also has the bonus that it makes balancing fights very easy on the GM - in general, an MR that gives dice and adds comparable to those of the party as a whole will give a balanced fight. (Of course, old-school games didn't worry about balance in the same way that modern ones do; rather, parties were expected to flee from obviously superior monsters, and to find ways to try to even the odds against others.)

On the negative side, in the hands of a new or unskilled GM, this system tends to cause all monsters to feel the same - it's up to the GM's narrative ability to make fighting different types of monsters feel different. Further, if players choose to interact with a monster in a way other than fighting it, it's entirely up to the GM how that will work. Saving rolls can be used for a general mechanism, but early versions of T&T don't even suggest that.

Next time, we'll discuss the magic system of 1st edition T&T, and talk about how the system compares vis-a-vis the original Dungeons & Dragons.

Monday, September 14, 2015

Toward a History of RPG Mechanics, Part Four: Combat in Original D&D

A Brief Detour into Race

I realized after I started writing this that I forgot to mention race as a part of character construction in OD&D last time! So, here's the brief on how that works:

It's a common assumption that race in OD&D is like race in the later Basic D&D - that each race is effectively its own class. This isn't true, although it's close to it in pre-Greyhawk OD&D. There are three non-human races available as player characters here: dwarf, elf, and hobbit (which would become 'halfling' in later printings and editions).

Dwarves only have one class option: fighting-man. Dwarves can only progress to 6th level, but they gain four benefits: +4 to their effective level to save against magic of any type; access to additional abilities of a +3 Magic War Hammer; the ability to note slanting passages, traps, shifting walls, and new construction underground (there is no roll or other mechanic given for this); and the ability to speak Gnome, Kobold, and Goblin.

Hobbits likewise can only be fighting-men. They are limited to 4th level, get the same +4 to their level for saves as dwarves, and "have deadly accuracy with missiles as detailed in Chainmail". The specific rules there are that they "can fire a stone as far as an archer shoots" (but without specifying what type of bow is meant) and that two hobbits shooting missiles count as three shooters on the combat tables.

Elves are the most unusual of the three. They can act as both fighting-men and magic-users... but rather than being treated as in later D&D, an elf's player chooses at the start of each adventure which class the elf will be in this adventure. It is specified that they can use magic armor "and still act as magic-users" (presumably meaning they can cast spells in armor). It's also stated that they "may use both weaponry and spells", but it isn't clarified whether this means at the same time (that is, whether they can use any weapon while acting as a magic-user).

In compensation for this flexibility, elves are limited to 4th level as fighting-men, and 8th as wizards. Experience gained as each is kept separately, so an original D&D elf could be, for example, a 2nd level fighter and a 7th level wizard. They "are more able to note secret and hidden doors", locating them on a 4 in 6 chance instead of the 2 in 6 of other races. They speak Orc, Hobgoblin, Gnoll, and Elvish, "and the other usual tongues". They also "gain the advantages noted in the Chainmail rules when fighting certain fantastic creatures". Those rules give elves using magical weapons the ability to "add an extra die in normal combat" (two dice against orcs, three against goblins), and to attack several sorts of fantastic creatures. Effectively, an elf with a magic weapon becomes something between a Chainmail "hero" and "super hero" in combat ability. Of course, for those who didn't have Chainmail and were using the alternative combat system, these special abilities effectively didn't exist.

Elves possibly have other special abilities in OD&D. In the creature description for elves, it's stated that "elves have the ability of moving silently and are nearly invisible in their grey-green cloaks." Whether this is meant to also apply to PC elves is unknown. The fact that it isn't in the elf description is meaningless, since OD&D places many rules that pertain to races or classes outside their own description. It's also stated there that elves get +1 to damage when armed with magical weapons, and that elves on foot may split-move and fire (i.e., move half their movement, fire missile weapons, then complete their movement).


We've discussed the Chainmail combat system (which was supposed to be the 'normal' combat system for Original D&D) fairly extensively in Part Two of this series. But, as noted there, it was the "alternative" system presented in the OD&D books themselves that became the commonly used system. So, let's take a look at that in more detail, and, in particular, to how its differences from the Chainmail system would shape things.

The alternative system closely resembles the one familiar to modern D&D players: the attacking character's class and level are used to pick the column used for their attacks. This is cross-referenced with the defender's armor class to determine a number. The attacker then rolls a 20-sided die and, if the result is greater than or equal to the number from the table, hits.

OD&D at this point does not yet differentiate weapon types vs. armor for this system in any way - the weapon used doesn't affect either the chance to hit, or the damage done on a successful hit. All successful hits do 1d6 hit points of damage instead.

The choice to use 1d20 instead of the 2d6 of the Chainmail system here has a huge impact. The potential spread is much greater, with 20 possible results instead of 11 (2 to 12): this means that die modifiers are much less significant - a +1 in Chainmail is a big deal, but not much of one in D&D's "alternative system". The probability curse of 2d6 makes this even more significant in many cases. For example, a sword against chain & shield in Chainmail hits on 9+. This is a 10/36 chance, or about 27.8%. If that sword is +1, though, allowing a hit on 8+, then it will succeed 15/36 of the time, or about 41.7% - a 13.9% increase in the chance to hit! Using 1d20, these would be about the same as the difference between needing a 16+ to hit and only needing a 13+ to hit - making that +1 equivalent to a +3 on 1d20!

There's a second difference in how the systems are worked, though. Let's illustrate it with an example:

When using the Chainmail combat system with OD&D, a 3rd level fighting-man attacks as 3 men. This literally means that the player rolls to attack three times. The rules do not make it clear whether this means it's possible for them to do 3d6 damage to single defender, or if only one hit 'counts'. Regardless, though, in our above example with sword vs chain & shield, a 3rd level fighter now has a 62.3% chance to hit. While how this is meant to apply against a single foe is unknown, the FAQ for D&D presented in The Strategic Review for Summer 1975 shows that it does allow attacking multiple foes who count as 'ordinary troops'. In that example, a 4th-level fighting man fights ten orcs, and is allowed to attack four of them in each round of combat.

(Actually, it's more complicated. The rules state that a 3rd level fighting-man attacks as "3 Men or Hero - 1". However, it doesn't indicate how to make the determination about which to use. Does the player get a choice between rolling 3 attacks or rolling 4 with a -1 modifier? Does the -1 modifier apply to all four attacks, or just one of the four? Is the "Hero - 1" part only supposed to be used when the "Fantasy Combat Table" is used (i.e., against monsters, other heroes, etc)? Is there any significance to the fact that some entries are ordered differently, so that a 5th level fighting-man's attack strength is listed as "Hero + 1 or 5 Men"? The rules as written don't address any of these questions. This may have contributed to the quick adoption of the "alternative combat system" as the official one, since none of these questions apply with it!)

The alternative system, however, instead increases the chance to hit on the single die roll... but it does so in jumps, rather than the smooth climb later versions of xD&D would adopt. So, here, a 3rd level fighting-man attacks the same as a 1st level one, needing a 15+ on 1d20 to hit an opponent with chain & shield. At 4th level, this would improve to needing a 13+. This significantly changes the impact of level advancement: as noted above, the 3rd level fighting-man using the Chainmail combat system hits with sword vs chain & shield 62.3% of the time; using the alternative system, however, only 40% of attacks are hits in the same situation.

On the other hand, though, someone using a dagger against plate & shield with the Chainmail tables needs a 12+ to hit... which is a 1/36 chance, or 2.8%. With the alternative tables, a 17+ hits on 1d20, so the attacker has a 20% chance of hitting - a vast improvement! Since the initial chance is so low, multiple rolls don't help a lot, and our 3rd level fighting man only has an 8.1% chance to hit, even rolling three times.

On the other hand, however, the initiative rules of Chainmail allow multiple attacks for fast weapons. So, if the foe in plate & shield is wielding a two-handed sword, the dagger wielder gets three attack attempts a round... so our third-level fighting-man against an ordinary foe will get to roll nine times in this situation! The chance that at least one of these hits is... only 22.4%. There is, however, the possibility that all three attacks will hit... although that possibility is about 1 in 2,000.

There are more questions that arise, though. For example, the FAQ makes it clear that a fighting-man can attack multiple foes that count as normal soldiers in Chainmail in one round. But what about, say, a 2nd level fighting man vs a 3rd level one?  Do they both count as normal soldiers, since neither can yet attack as a Hero? Or would they only each get one attack against each other normally?

These questions are, again, avoided when using the alternative combat system. That system, however, removes all differences between weapons... and the main difference between a fighting-man's combat ability and that of a cleric at levels 1 to 3 was what weapons they could use. (The same was true of magic-users when attacking at those levels. Their lower hit points and lack of armor were only important on defense - on offense, a 3rd level magic-user with a dagger was as good as a 3rd level fighting-man with a sword!)

Likely because of these considerations, in Greyhawk both weapon vs armor adjustments for the "alternative combat system" and differing damage dice for weapons would be introduced. Greyhawk would also see the introduction of higher magical weapon and armor bonuses, going up to +5. It would seem likely that experience with the alternative system showed how little significance a +1 bonus had there relative to its worth in Chainmail, prompting this development.


Of course, weapon attacks are only part of combat. As also mentioned in Part Two, while Chainmail had a detailed and relatively sophisticated initiative system, OD&D gave no rules at all about who attacked first or the order of movement. We turn again to the FAQ presented in The Strategic Review. The actual rule as given in the FAQ is:
"Initiative is always checked. Surprise naturally allows first attack in many cases. Initiative thereafter is simply a matter of rolling two dice (assuming that is the number of combatants) with the higher score gaining first attack that round. Dice scores are adjusted for dexterity and so on."
Yes, that's all of it. We're left to fill in the reasonable assumptions that the GM decides when surprise allows first attack, that if there are more than two combatants, they each roll and go in order, from highest to lowest, and to make up our rules regarding ties.

The example given then proceeds to somewhat violate these assumptions, by facing a lone 4th level fighting-man with 10 orcs, and having the orcs all roll as one. This would become the standard rule in AD&D1 - that each side rolls initiative as a group. Dexterity adjustments to initiative were eliminated for melee there, but retained for missile fire. (Of course, Dexterity adjustments in OD&D pre-Greyhawk are only a -1 or +1.)

All the considerations of the relative length and speed of weapons that Chainmail had are gone here. Who moves first isn't explicitly addressed; in Chainmail, the question of order of movement and order of attack are related, but separate, but it's impossible to tell whether this is intended here as well.

It's quite likely that Gygax intended that game masters and players should simply work out such considerations using reason and common sense. Chainmail had been intended for miniature wargamers, who were mostly college-aged and older. However, the combination of D&D's subject matter and the lower requirements for preparation (not needing one to buy and paint miniatures) made it appeal to a younger group than Chainmail had, and it would later become apparent that high schoolers and middle schoolers were a large part of the contingent adopting D&D. Later versions of D&D would accordingly give more guidance to players and game masters.


Magic-users and clerics both cast spells as part of their major abilities. The individual spells are sketchily described, and each is essentially its own special rule, a tradition which D&D has kept up. For most spells, there is no roll to hit or to cast; instead, the target makes a saving throw, trying to roll higher than a threshold set by class and level (and affected by race, as mentioned above) on 1d20. Depending on the spell, a successful saving throw either means no effect from the spell, or a reduced effect (generally half damage for damage-causing spells).

The saving throw needed also depends on what is being saved against, with these categories:

death ray or poison
all wands, including polymorph or paralyzation
stone (presumably meaning petrifaction, rather than thrown rocks)
dragon breath
staves & spells

Progression on the table is in the same bands as for combat ability in the alternative combat system: groups of three levels for fighting-men, four for clerics, and five for magic-users. Maximum ability is reached at 13th level for fighters ("fighter" is used on this table as an abbreviation for fighting-man) and clerics, and 16th for magic-users.

Wrapping Up

This finishes our coverage of the major systems of OD&D. Some notes and observations:

While Chainmail is referred to several times, it was quite possible to run the game without it... and most people did. D&D vastly outsold Chainmail, and while there were supplements for Chainmail announced in the first issue of The Strategic Review, none of those were ever actually produced, so far as I know. D&D's massively greater popularity caused the fledgling TSR to essentially abandon the game.

Chainmail combat was much more sophisticated... but that may have been its own undoing, along with its original design for mass battle. To properly continue the Chainmail combat system forward would have required creating a new row of the Fantasy Combat Table for each new monster type added to the system. Each would also need to be assigned an armor type, and possibly an adjustment to that armor type, and they would also need weapon classes for initiative. In contrast, the much simpler alternative combat system did not require nearly so much detail - it worked with numeric armor classes, and used monster hit dice to determine how well they could strike in combat.

It seems likely that the switch to the "alternative combat system" as the standard one was due to both this, and the far greater popularity of D&D. Indeed, two years later, TSR would produce a new mass combat system for D&D - the supplement Swords & Spells. The difficulty of expanding Chainmail to encompass the new monsters, spells, and so forth in the D&D supplements was probably a large factor in this.

As has been noted so far, OD&D had nothing even remotely approaching a "universal mechanic". What mechanics it did have were largely ad-hoc constructions, each made up for the specific situation in question. Attributes had little use beyond granting experience point benefits and minor adjustments. This had the positive property that there was little or no pressure to have high attributes. The major breakpoints were at 9 (below which some attributes gave a -1 penalty) and 13 (at which and above a +1 bonus was granted by some attributes). A 15 or higher in a prime attribute gave experience bonuses, but other attributes contributed as well... since a fighting-man could add 1/2 of Intelligence and 1/3 of Wisdom, an "average" character with 10 in each attribute had an effective 18 Strength for experience purposes. It was a rare character who didn't get an XP bonus in OD&D!

Greyhawk would change this in major ways, with new tables that gave much higher bonuses from high attributes, and most especially with the introduction of exceptional strength, which would allow those fighting-men characters lucky enough to have an 18 Strength to roll for greater abilities. The introduction of new character classes and subclasses with high minimum attribute requirements would affect this as well, and so would the introduction of psionic abilities in Eldritch Wizardry, with the upshot that a character who didn't have at least a 15 in one attribute was at a major disadvantage. Each roll of 3d6 has about a 10% chance of giving a result of 15 or higher, though, so the chance that a character generated by the original method would have one (or more) was only 44.2% - thus, a bit better than half of all characters came to be seen as "undesirable". This led to the use of alternative generation systems, such as the well-known 4d6 drop the lowest, and to allowing players to arrange their ability scores in a desired order instead of simply taking what the dice gave them.

But that was all in the future at this point in D&D, and we'll talk more of such things in future columns. For now, this will end our coverage of original D&D, and in the next column, we'll be moving on to talk about the second fantasy RPG to be created: Tunnels & Trolls!

Sunday, September 6, 2015

Toward a History of RPG Mechanics, Part Three: OD&D Characters

The key difference between early RPGs and the miniature wargames they diverged from is a simple one: the idea that each player controls not an army, squad, or other group, but a single character. Of course, this means that instead of generating or selecting an army, players must generate or select characters to play. Characters are defined largely in terms of their attributes, class, race, and level, each of which we will cover in turn.


D&D, both the original game and all subsequent versions, has six 'attributes' that express the raw physical and mental qualities of characters. These are, in the order the original game gives them: Strength, Intelligence, Wisdom, Constitution, Dexterity, and Charisma. Attributes and class are inter-related, with three of the attributes being the prime requisite of a class.

The attributes are assigned by rolling 3d6 for each one, in order. OD&D gives little explanation of its attributes, and makes very little mechanical use of them. Let's go through them and what they're used for in the OD&D rules quickly:

Strength is the prime requisite for fighting-men: high Strength gives them experience bonuses, and low gives penalties. It is also of assistance to clerics, who get to add 1/3 their Strength to their Wisdom when determining their experience bonus or penalty. Its description says it "will also aid in opening traps and so on." However, no actual mechanics are either given or suggested for this. In fact, beyond giving fighting-men an experience bonus, Strength has no mechanical uses whatsoever described in OD&D - it doesn't help with attacks, doesn't increase damage, and doesn't even influence the maximum weight a character can carry. The first OD&D supplement, Greyhawk, would rectify all this, but it seems a curious omission.

Intelligence is the prime requisite of magic-users. Fighting-men and clerics both get to add 1/2 their Intelligence to their own prime requisites to determine experience adjustments. It also allows additional languages: characters speak two languages to start with (the 'common tongue' and their alignment language), but can add another for each point of Intelligence over 10 they have. This, however, was all the mechanics that were associated with it - it did not at this point influence what spells a magic-user could learn. This, too, would be added in Greyhawk.

Wisdom is the prime requisite for clerics, but assists both fighting-men (adding 1/3 its value to their Strength for experience purposes) and magic-users (who get to add 1/2 its value). Here too, there were no other actual mechanics associated.

Constitution gave a bonus or penalty to hit points (+1 or -1 per hit die), and gave a percentage chance of "surviving adversity".

Dexterity could give a +1 or -1 penalty to missile attacks.

Charisma actually had more mechanics associated with it than any other attribute: it determined the maximum number of hirelings a character could have, and modified the "loyalty base" of NPCs relative to the PCs they were hired by or were henchmen of. This in turn modified morale rolls, and Charisma also modified reaction checks.

Unlike later versions of D&D, the original game gives no guidance on how to use the attribute numbers to help determine the success or failure of actions. This led to individual gamemasters each creating their own methods to figure such things... and, following the example of the mechanics given in D&D, often completely ignoring the attributes in those mechanics.

The simple expedient of rolling a d20 and trying to roll under the attribute seems to have not been much used. Indeed, the first issue of The Dragon (June 1976) has an article on "How to Use Non-Prime-Requisite Character Attributes", which presented a system which by modern standards is almost comical in its complexity: the GM is to roll 1d100 + attribute, use that to consult a table, roll a d4, d6, d8, d10, or d12 based on that, and multiply the attribute by the result of that die... then use that result as the percent chance of success.


As we've seen, the three classes of OD&D are fighting-man, magic-user, and cleric. Although the thief or rogue would come to be considered one of the core classes of the game, it was officially added to the game in Greyhawk.

The fighting-man and magic-user hold obvious and well-known places in fantasy literature. The cleric, however, doesn't really seem to correspond to anything in the most popular and well-known fantasy literature - there's no cleric-like characters in The Lord of the Rings, nor in Leiber's Fafhrd and Grey Mouser stories, nor in Howard's Conan stories.

According to players in Dave Arneson's early games, the cleric was originally the 'priest', and the main inspiration for it was Van Helsing in the Hammer films - one player wished to play a vampire hunter, and his character was given abilities to wield holy powers, and to do some healing.

Gygax changed this to the cleric, and here we see his apparent love of medieval lore. The term 'cleric' itself originates in Ecclesiastical Greek klerikos "of or belonging to the clergy". It was Latinized as clericus, and as such, was used throughout the medieval period to refer to clergy. He also drew on the medieval Knights Templar and Knights Hospitaller, plus Victorian stories that indicated that medieval clergy who wished to participate in battle would use clubs or maces as weapons, to get around a prohibition against priests shedding blood. Mixed into this, we get clerical magic drawn from medieval saint stories and miracles described in the Bible - so D&D clerics produced food and water, healed the injured, cured disease, raised the dead, turned sticks to snakes, and so forth.

Thus, we get the OD&D cleric: a holy warrior with saint-like abilities, but limited to using only certain weapons. Their magic was treated essentially the same as that of magic-users, but with a different list of spells available to them, and with clerics receiving the bonus that they did not have to find and learn new spells - upon gaining a new spell level, they immediately gained access to the clerical spells of that level.

Magic and Magic-Users

Chainmail already had a magic system, for its 'wizards', consisting of ten spells. These were: phantasmal forces, darkness, wizard light, detection, concealment, conjure elemental, move terrain, protection from evil, levitate, slowness, haste, polymorph, confusion, hallucinatory terrain, cloudkill, and anti-magic shell. Each spell had a 'complexity' from 1 to 6, and less powerful wizards could not cast those of higher complexity.

In addition to these, wizards could throw either fire balls or lightning bolts (which one had to be selected before play). Wizards were also impervious to normal missile fire, and could counter the magic of other wizards (which became the D&D spells protection from normal missiles and dispel magic). They came in five different degrees of power, from a Seer who could only use one spell, to a full Wizard who could use six or seven spells. However, exactly how this worked is very unclear in Chainmail - there's nothing which says whether a Seer gets one spell that can be used over and over, or whether they can simply only cast one spell from the above spell list in the course of a game, but can freely choose which one at the time of casting. (The fire ball and lightning bolt aren't listed as being spells per se in Chainmail, and thus, can be used repeatedly.)

All this made Chainmail wizards terrifyingly powerful on the battlefield... but that was all right, because Chainmail balanced fantasy forces by means of a point system. A full wizard was worth 100 points, the same as a dragon, and more than any other type. Heroes were only worth 20 points, making a wizard as valuable as five of them!

D&D, however, had to balance things between individuals... and so, while many of the elements of Chainmail's wizards were retained, others were dropped. A Chainmail wizard was quite powerful in melee, being counted as two Armored Foot or two Medium Horse (if mounted), and capable of using any sort of magical weapon, including magic swords. On the Fantasy Combat Table, they are hard to kill, with Giants and Super-Heroes (defined to be exceptionally powerful heroes, such as Conan) being the only types who are more likely to kill a wizard than to be killed by one.

All this went by the wayside for the D&D magic-user. They were vastly demoted in combat ability, being the weakest of the three classes in melee, having the fewest hit points, and being unable to use most weapons (whether magical or not) or any armor. The spell 'complexity' of Chainmail became spell levels, and were used to control how powerful of a spell a magic-user could cast, dependent on their own level (although not in a simple one-to-one relation).

Most importantly, though, the magic-user was limited in the number of times a day they could use each spell. For this, Gygax borrowed a concept from Jack Vance's Dying Earth novels, wherein spells were extremely complex and alien formulae, which a magician had to devote considerable time to memorizing... and which, when used, would sort of burn themselves out of the magician's memory, so they would be forgotten and have to be laboriously re-memorized.

However, this fact was only mentioned once in the rulebooks (in the first paragraph on page 19 of Men & Magic), and seems to have been very easy to miss. It was either missed or ignored enough that Gygax regularly felt the need to call attention to it, doing so in an FAQ about D&D rules presented in the second issue of The Strategic Review (Summer 1975), and again less than a year later in the article "The D&D Magic System" in the April 1976 issue (TSR had switched from quarterly to bi-monthly along the way). Thus, in campaigns where the Vancian system was not in use, the magic-user frequently outstripped other characters in power.

It should be noted too that in original D&D, the starting spell selection for magic-users had no direct combat spells. Such spells as magic missile, burning hands, and even shield did not exist yet. The most useful spells that starting magic-users had for combat were protection from evil, sleep, and charm personPhantasmal forces and invisibility were in the list of second level spells, but magic-users didn't gain spells that could cause hit point damage until the could cast the third level spells fireball and lightning bolt... when they themselves were fifth level. Lower level magic-users wouldn't gain officially sanctioned damaging spells until the release of Greyhawk.

... and that's it for this time. We haven't really touched on the fighting-man class yet, but it makes more sense to save that for Part Four, where we'll dive into OD&D Combat!


Gygax, Gary and Arneson, David, Dungeons & Dragons. 1st edition, Tactical Studies Rules, 1974.
Gygax, Gary and Perren, Jeff, Chainmail, 3rd edition, Tactical Studies Rules, 1975.