The genesis of the Dungeons & Dragons cleric class is relatively well known. Inspired by the character of Van Helsing and his analogues as portrayed in the Hammer Horror films of the preceding decades, the cleric was created to combat a particularly troublesome vampire player character known as “Sir Fang”. To this somewhat narrow archetype were appended the trappings of a pseudo-medieval warrior priest typology, and the infamous restriction against the use of edged weapons (later “edged and/or pointed weapons which draw blood”). This latter clause was derived from a nineteenth century visual interpretation of the depiction of Bishop Odo in the Bayeux Tapestry, which was current in academic circles up until the late twentieth century, and remains embedded in the popular consciousness, even amongst those who really ought to know better. Gygax showed some trepidation in this assertion by the time of the Player’s Handbook, when he stated that the cleric has “a certain resemblance to religious orders of knighthood of medieval times” (p. 20), who he must have known were under no such compulsion. Regardless, the most formidable abilities of the class, casting spells and turning away evil spirits, have little to do with the military orders.
Whilst spell casting was already well established in the developing milieu, the ability to turn away undead, as well as lesser demons and devils, was a new addition. It fairly clearly has its roots in the conventional Hammer Horror scene where a character attempts to keep a vampire away by holding up a cross, with varying degrees of success (perhaps most amusing of these is an instance in which a character played by Peter Cushing destroys a vampire with the shadow of a burning windmill). These sort of scenes no doubt owe their currency to the myriad superstitions concerning the warding off of evil with magical amulets or sacred objects, and the reputed power of holy men to themselves drive away evil spirits. However, even in the original version of Dungeons & Dragons the vampire is singled out as averse to garlic, mirrors, and the sight of the cross, over and above any power of the cleric, if “presented strongly” (Monsters & Treasures, p. 10). Indeed, to turn away a vampire ordinarily a cleric must be at least sixth level (and thus equivalent in fighting ability to a hero) and roll a nine or more on two six-sided dice, a probability of only ten in thirty-six or just less than twenty-eight percent.
It is interesting to note that, just as the Chain Mail man-to-man combat system and its two six-sided dice gave way to the alternative combat system and its twenty sided die, Advanced Dungeons & Dragons did the same with regard to the turning ability. This was not the case with B/X or BECMI, both of which retained the shortened bell curve approach. Frank Mentzer, primary editor and designer of the later, has mentioned in the past that this also mirrored the B/X and BECMI morale rules, which used two six-sided dice, and that this was similar to how he envisioned a turning attempt, which is to say as a morale test for the undead. Of course, morale was not described in the original Dungeons & Dragons game, presumably the game master was expected to borrow from the byzantine Chain Mail version, though many must have used the “reaction test” as a stand in, again using two six-sided dice. Nonetheless, the morale rules for Advanced Dungeons & Dragons used percentiles, with modifiers mainly in five-percent increments, which obviously would work well with a twenty-sided die. Oddly, the second edition of the advanced game compromises between the two methodologies by using two ten-sided dice for morale, but not to generate percentiles.
Once the numbers for the original chart are converted to percentiles (8.340%, 27.78%, 58.33%) it quickly becomes evident that Gygax rounded them off (0.10, 0.30, 0.60), inserted an additional increment between the largest step (0.10, 0.30, 0.45, 0.60) then shifted the probability down one for the latter three numbers (0.10, 0.25, 0.40, 0.55), paralleling what he did for armour class, before converting the probabilities to target numbers on a twenty-sided die (19, 16, 13, 10). He then extended the range downwards by increments of fifteen percent, and made “20” the top of the range (20, 19, 16, 13, 10, 7, 4). For whatever reason, levels 4-7 (hero to superhero −1) omit the 19 between 16 and 20, increasing the overall effectiveness of the cleric from what might be expected of the pattern between levels 1-3. Unsurprisingly, second edition standardised the table to follow the initial pattern, resulting in a corresponding decrease in effectiveness. The expansion in level range and decrease of one step between “D” and “D+” (or D*) somewhat flattened out the curve, along with the switch from “1-12 affected” to “2-12 affected”, and “7-12 destroyed” to “2-12 destroyed and 2-8 turned”, respectively.
Although the original version of turn undead specified the number to be turned as 2-12, it did not indicate how frequently the ability could be used, its range, area of effect, or for how long it was effective once employed, amongst other things. B/X somewhat clarified things by allowing turn undead to be used as frequently as desired, but reduced its effectiveness to 2-12 hit dice, albeit with a minimum of one creature affected. Advanced Dungeons & Dragons took the opposite approach, restricting use to more or less once an encounter (under limited conditions it could be used in consecutive rounds against different types) and specifying the duration as 3-12 rounds, with previously affected undead “being subject to further turning by the cleric” (DMG, p. 76). The second edition returned to vaguer language, and duration seems to be as long as “he continues to maintain his turning”. Whilst that works well for turned undead, it makes destruction results extremely effective; for instance, to a tenth level cleric a pack of 2-16 wights is potentially little to no threat, but 2-12 wraiths remain foes to be feared! Happily, it does at least specify that cornered undead will fight back, “breaking” the turning effect.
For my own campaigns none of the approaches above have entirely sufficed, even less so when chaotic or evil clerics are stirred into the mix, not to mention the unlooked for vulnerability of paladins. Perhaps a better way to handle it would to be to treat it like a spell, an approach that has certainly been postulated elsewhere. However, that is not really any different than restricting it to a once per day effect, which would avoid the complication of introducing an additional slot. As long as an encounter can be defined, there is no reason not to keep it at once per encounter. The variable number of affected undead is a little unpalatable as an all or nothing affair, and the obvious solution is to roll one turning attempt against each target up to twelve, which would result in a more average spread of results, but “T” and “D” effects would always be the maximum. One way to counter that would be to spread out the probabilities for one half of the matrix on a 1:1 basis and use a higher ratio for the other, such as 1:2 or the 1:3 of the original scale. As things stand turn undead is somewhere between a saving throw based fear spell and a percentile morale test, neither fair nor foul. For those interested, comparative charts can be downloaded here.