AD&D 1st Edition Surprise Revisited

The guidelines and examples for Surprise in the DUNGEON MASTERS GUIDE are very basic. They only really explain in detail the results when two sides with typical surprise chances meet. They really don't stress enough the facts of how surprise might be run during actual game play when players are taking precautions (like having an advance scout, not using light, moving silently due to spell or ability, etc.) to avoid being detected. In fact, I'm not really sure how Moving Silently is supposed to be treated, since in most circumstances, it would be used by an assassin or thief to sneak up on a victim and surprise them. So then, is the thief penalized by having to make two rolls in order to surprise? Is the Move Silently check merely to see if they gain an extra 1 in 6 chance to surprise? Why then would an elf or halfling gain a 4 in 6 chance to surprise simply by moving silently? Should the Move Silently check be used in place of the surprise roll? None of these questions are answered anywhere. If the surprise roll is basically meant as a detection roll, then shouldn't a thief of 1st level have a better chance than a fighter or cleric of the same level in gaining surprise?

Normal chances to surprise are 2 in 6. That roughly translates to 33-1/3% chance of surprising an opponent. A look at the Thief Function Table in the PLAYERS HANDBOOK reveals that Move Silently is below 33% until they reach 4th level. So Moving Silently in and of itself does not make sense as a "replacement" for surprise chance. So what benefit does it provide? A lowly human thief in leather armor should probably have the same chance as an elf or halfling in non-metallic armor for surprising an opponent. After all, their very class functions include this stealth. But the rules as written do not support this. Additionally, it appears on the chart that Moving Silently for an elven thief is only 5% better than a human, and for a halfling thief is 10% better. So, the percentage does not really equate to surprise at all, but rather an increased chance of surprise.

In fact, looking at the PLAYERS HANDBOOK description of the Move Silently ability, it even states that:

"Moving Silently can be attempted each time the thief moves. It can be used to approach an area where some creature is expected, thus increasing chances for surprise (q.v.), or to approach to back stab, or simply done to pass some guard or watchman. Failure (a dice score in excess of the adjusted base chance) means that movement was not silent (see SURPRISE). Success means movement was silent."

The DMG defines silent and invisible movement as an increase to 4 in 6 chances of surprise. Now, looking at some of the monsters in the MONSTER MANUAL (again, a book designed at the end of the Original D&D era), one can see that most ambush predators only have a 3 in 6 chance of surprise, and most invisible opponents have a 4 in 6 or 5 in 6 chance of surprise. Does this mean that elves and halflings are unseen while moving? Does this mean that the lowly human thief has only a 3 in 6 chance to surprise? The Hide in Shadows ability strictly disallows the use of this ability while moving, so one can never Move Silently while Hiding in Shadows. So why do the elves and halflings who aren't even thieves get the 4 in 6 chance to surprise? Again, this harkens back to the days of Original D&D when elves were the masters of camouflage in the forest, able to move without sound and be considered invisible when in foliage. Halflings also had a similar ability in natural surroundings, mainly due to their size and natural stealth. So how does this translate to AD&D? Well, Gygax obviously allowed this traditionally outdoor ability to be used indoors, thus increasing chances for surprise for elves and halflings. But there is a caveat here. The ability first off does not apply to half-elves, only full-blooded elves. It stipulates that the elf or halfling must be alone and not in metal armor (or well in advance (90'+) of a party not consisting entirely of elves and/or halflings) in order to gain the 4 in 6 chance for surprise. If a door has to be opened to proceed, the chance drops to normal 2 in 6. Again, what if the character is an elven or halfling thief? Should this further increase their chance of surprise to 5 in 6? After all, the elf or halfling is being granted this extra ability due to moving silently. The rules are silent on this...

It is my belief that normal Move Silently successes merely increase the chance to surprise to the same as an ambush predator that also moves silently (like some monsters, especially the bugbear, whose 3 in 6 chance to surprise is directly stated as being due to silent movement). Elves and halflings somehow are able to move so stealthily that they have increased chances of surprise as if they were silent and invisible (4 in 6). Thus, any elf not wearing metal armor, is essentially better at sneaking around than a non-elven thief when it comes to surprising in combat situations only. It does not, however, mean that an elf can simply sneak past a guard or tip-toe near a dragon, since they have no Move Silently ability as a thief does. However, if the surprise roll simply means it is a detection roll, then it could mean that an elf or halfling might have the ability to do so even though they don't possess the class ability (but they must be alone and not in metal armor). If you are a halfling or elf, and not wearing metal armor, you are giving up protection for the increase in stealth, so it is somewhat balanced. Should an elven magic-user be sneaking up on a monster for surprise? Probably not. But most leather-wearing elves and halflings would likely be thieves anyway. In any case, it doesn't work unless they are alone, and ambushing a creature alone in a dungeon is a very risky proposition!

Okay, so if moving silently normally grants a character 3 in 6 to surprise, and moving silently and invisibly grants 4 in 6 to surprise, how does one achieve 5 in 6 to surprise? A few basic monsters actually have this chance, including huge spiders, the giant lynx, giant owls, xorn, and invisible stalkers. What does it represent? Well, the huge spiders represent a class of hunting spiders that ambush by leaping or lurking, such as wolf spiders or trapdoor spiders. Giant owls are said to glide to the attack on silent wings, so the attack is coming unexpectedly from above I suppose. Xorn can emerge from the solid rock to attack, so there really is no hint that anything is even there. Invisible stalkers are naturally invisible but also fly, so there is nothing to even warn the target before they are attacked. And the giant lynx? Who knows, that's a weird one. Maybe Gygax had a soft spot for this creature since there is nothing else to distinguish it from any other great cat (which typically have 3 in 6 or 4 in 6 chances of surprise). I don't see any reason to give a giant lynx that kind of advantage, but how often does one encounter this creature anyway?

Right off the bat, you can see that there are better chances for monsters to surprise characters than for characters to surprise monsters. Also, there is no automatic surprise for any creatures. However, this is where we get into that gray area in the rules. Some monsters are stated as being 90-95% undetectable before they attack (shadows in dim lighting, mimics, piercers, etc.). Their entire existence and mode of attack is dependent upon surprise. Does this percentage represent their ability to surprise opponents? It is a detection roll, after all, and what is a surprise roll if not a detection roll? A piercer is 95% likely to attack by surprise. How many segments of surprise does 95% represent? Since the roll is not being made on a d6, can one even use the segments rule of surprise?

An article was written in the May 1988 issue of DRAGON Magazine ("Surprise!" by Leigh L. Krehmeyer) that tried to resolve this conflict by equating non-standard surprise percentages to a number of segments depending on the die roll. I suggest that this may be unnecessary. As I stated in my original surprise article, I believe that surprise can only result in two conditions - surprise or complete surprise. These terms are used in the books themselves (although they are never defined), and I later found their definition in Original D&D (so they are a carry-over from those rules). You can still use the segment rule just fine for normal surprise chances (2 in 6), but if one side has a higher chance to surprise, you run into some issues. Say a hunting spider (5 in 6) surprises a party who had no chance to surprise the spider (using light, armored in metal armor, not really trying to be quiet, etc.). If the roll to surprise was a 5, then according to the DMG, those poor characters will now suffer 5 rounds worth of attacks as the spider mercilessly decimates (and likely poisons) them. Is this fair? Not from the players' perspective. On the other hand, per the rules, an elven archer dressed in leather armor alone in the woods comes across a camp of orcs. He can surprise them 4 in 6. Per the rules, if the surprise roll was a 4, that's 4 segments of attacks with a bow that attacks at 3x the normal rate of fire! Basically, that elf just emptied his quiver before those orcs even knew what was happening! A surprise attack is basically a flank attack, since the attack is coming at you from an undetermined/unexpected direction, so no shield or Dex bonuses for those orcs (if the DM even has this information). Most elves have higher than normal Dexterity, have +1 to hit with bows, and this elf is likely a higher level than the orcs, so his chance to hit them is pretty good. Should a lone spider be able to decimate a party, or a lone elf decimate a dozen orcs? The rules seem to support this.

Where this really gets crazy is when almost ALL the monsters encountered have non-standard chances of surprise. Consider the first time the module D1 was run at GENCON. The rules were still new, the DMG I think was not available yet, and the players were going on the word of the DMs when it came to running combat. Suddenly they were faced with drow, jermlaine, and svirfneblin, all sorts of ambush creatures in the underworld, and most of the PCs used in the tournament adventure were elves! Right there, that's a nightmare for a DM to determine all the variables involved with the surprise roll. The solution - use a second roll to determine the number of segments of a surprise. Keeping it simple allows the combat to run smoothly and still gives the benefit of surprise without an overwhelming advantage to those creatures with increased chance of surprise.

Of course, if you would prefer running surprise by-the-book, you are more than welcome to try. It's not as nightmarish at it would seem at first. Let's take the spider example above and use the scenario from T1: The Village of Hommlet as an example. In this scenario, an abandoned tower houses a huge spider lurking in the shadows. The characters enter through a broken door on the north face of an 18' diameter round tower. The setup states that it is a wolf spider (so it leaps up to 3" (30')  to attack) and that it is lurking on a portion of the ruined floor of the collapsed upper part. So, if it gained 5 segments of surprise, it would likely leap to the attack on the 1st segment (treated as a charge attack at +2 on an unsuspecting target, so no shield or Dex bonuses to AC), if it failed to hit (not likely) it would thus have 4 more bite attacks on the target before the party could even do anything. But remember that Dex adjustments affect surprise for the individual, so if anyone in the party had an 18 Dex, that person would only be surprised for 2 segments. What happens to the other party members during the other 3 segments? Does that mean that the spider gets only 3 segments of complete surprise on the party, then has to roll for initiative with the higher Dex combatant while still conducting combat in surprise segments? This is what it seems like to me. That means that the spider and high Dex party member roll initiative for the last 3 segments of surprise, have to continue moving at segment speed, and can otherwise act as if attacking during a surprise segment. If the character were a thief, they could move to the spider (1 segment), then attack twice (2 segments); if they were an illusionist or magic-user, they could cast a 1 segment spell (1 segment), move (1 segment), then stab the spider with a dagger (1 segment); if they were a fighter they could move and attack as a thief, above; and if they were a cleric or druid they could move and attack, but spellcasting would likely be impossible since segment speed must be used, and most cleric/druid spells take 3+ segments to cast. If the high Dex character was the target of the spider's attack, then he would suffer 2 segments of attacks from the spider (one leaping charge attack, plus one normal attack), followed by 3 segments of initiative between the two combatants. In any case, the character involved in the combat above is likely level 1-3, and surviving 5 segments of poison bites is not very likely. What is more likely is that the character would be attacked successfully on the initial leap/charge attack, fail his poison save, then be dragged by the spider off into the shadows to be fed upon. The outcome would be death for that character. And considering that this is only the second encounter in the module as written, that player would be forced to spend the next few hours hating the game or rolling up another character that the DM has to somehow work into the game.

If you decide not to use the segment rule as written, you also have to modify the Reaction adjustment for Dexterity rule, otherwise characters with 17-18 Dexterity scores are NEVER surprised. If this sounds fair to you, then by all means use this. Otherwise, I would suggest that you only negate one segment of surprise if they have any Reaction bonus of +1 or higher. Thus, if the creature gains surprise, their reflexes counter it. If the creature gains complete surprise, however, then they only mitigate surprise for one segment and are able then to act before their slower companions.

Surprise can be devastating by-the-book with archers at the ready. Assume a typical adventuring party of 3 fighters, a cleric, a magic-user, and maybe two thieves enters the underworld and is ambushed by a band of 8 drow fighters of 3rd level all armed with hand crossbows hidden on an upper ledge. Assuming that there are no rangers or monks in the party, they are at the mercy of the normal surprise chances for the drow (which is 75%). Since the light sources of the party are torches, the drow are not affected by the party's light and cannot be surprised since they will see this light approach from a long distance away. Hand crossbows have a fire rate of 1, so if they gain surprise, they can each fire 3 bolts each segment (DMG, p. 62, 1st column, 3rd paragraph). So how many segments of surprise are gained?

Well, you can do this one of two ways. By the book there is no explanation; you could convert the percentage rolled to segments by dividing that number by roughly 17%. Thus if the roll was 73%, the party is surprised for 73/17 = 4.29 segments, rounded to the next nearest whole number, so 4 segments of surprise. Assuming that the crossbows are loaded and each drow has enough bolts, that would result in 4 segments x 8 drow x 3 shots/drow = 96 attacks before the party even has a chance to react. Remember that this does not count reaction adjustments for the characters, and attacks should be made as if the characters are unaware, thus they are treated as flank attacks unless some drow get rear attacks at +2 (this means no shield or Dex adjustments). Add in the -4 saves versus poison for the drow darts and you can see how devastating surprise rules are by-the-book.

It has been suggested by some on the Dragonsfoot forums that the 3x fire rate is based on the round, not the segment. So an arrow that has a fire rate of 2 normally, would have a fire rate of 6 during surprise. This would result in an arrow being shot every other segment with two shots coming back-to-back at the middle of the round (X o X o X X o X o X). So you can then determine that if you have 4 segments of surprise, an archer should get a shot on the 1st and 3rd segments. Therefore, they are NOT getting 3 shots each segment, but rather an increased fire rate that allows possibly more shots to be fired faster. Note that spells (and magic item usage) that are based on segments still require that amount of time to use! So if you have a 3 segment casting time for a spell or 2 segment activation time for a wand, you still need that time to cast or use the item regardless of whether you have surprise segments or normal rounds. In the case of the hand crossbows, 3x fire rate would result in no more than three shots per round, or roughly one shot every 3 segments of surprise. Thus they would have 1 shot on the 1st segment and another on the 4th segment. This is much more in line with the intention of the rule. Given that fire rates in the PHB range from 1/2 to 3, one can make a chart based on how many and when they go during surprise.

Fire Rate   3x Fire Rate Round (X = shot, o = delay or reload)
1/2            (X o o o o o X o o o)(o o o X o o o o o o)
1               (X o o o X o o o X o)
2               (X o X o XX o X o X)
3               (X X X X X X o X X X)

According to the above, if you are using a heavy crossbow (fire rate 1/2) you would get, at most, one shot off before requiring to reload, even at 3x rate. A troglodyte hurling a javelin (fire rate 1) would get only 1 shot off unless they somehow got 5 segments of surprise (which I don't believe is possible since they can only surprise 4 in 6 at most). An elven archer with 4 in 6 surprise and fire rate of 2 for his bow would at most get 2 shots off. And lastly, a jermlaine throwing darts fire rate 3) with 5 in 6 chance for surprise, would be able to throw a missile every segment of surprise gained! In this way, only melee attacks are treated as if it were a round. This makes sense since a round of time for melee is supposed to consist of dodging, riposting, and deflected attacks; if the surprised person cannot defend themselves, then all those thrusts and slashes are actually uncontested and have a chance of hitting and scoring damage.

Your other option is to simply use the method I've suggested in other posts. Each surprise roll has two components: the check to see if surprise exists, and then if so how many segments it lasts. If the first roll indicates surprise, the second roll of d6 means that 1-3 = one segment, and 4-6 = two segments. You might allow stealthy opponents to cause +1 or even +2 segments of surprise (or change the ranges so that 1-3 = 1 segment, 4-5 = 2 segments, and 6 = 3 segments), and if this is done Reaction adjustments should be used as normal. Otherwise, if a character gains a Reaction adjustment from high Dexterity, simply reduce the second roll by 1. In this way, complete surprise (2 segments) still has a chance of affecting a high Dexterity character and does not make them immune to surprise. Using the drow example above, the DM could roll for surprise chance - let's say he rolls 53%, thus indicating surprise. Then a roll of 2 on d6 indicates that there is 1 segment of surprise. High Dex characters would be able to negate this surprise individually. If the roll indicated there were 2 segments of surprise, even high Dex characters would be surprised. Using this method, allow the 3x fire rate per segment, as it will be far less devastating. Thus, if the drow gained 1 segment, they could fire freely on those surprised at 3 shots per segments, but only get 1 shot on those who are not surprised. If the drow scored 2 segments (complete surprise) then they get off 6 shots each (3 shots per segment for 2 segments) before the party can react. This could still lead to a bad ambush situation, but much better than being surprised for 4 segments by-the-book.

Again, interpretation of the rules and how the DM wishes to view Surprise is important and should be worked out ahead of time so that the players know the mechanics of surprise before the condition occurs. Although I favor by-the-book rules in most circumstances, this is because I have yet to use non-standard surprise creatures in my adventures. That is about to change with the characters in my AD&D campaign about to engage multiple drow in the dungeons of G3. Hopefully, the experience won't end my campaign!