Party Size and Old-School Play

Everyone has their preferred number of players, and of course some games are better at different player group sizes.  For myself, I've long since settled on 4-5 PCs as the optimal mix of varied opinions, good energy / impetus, varied PC capabilities and ease of fitting at a table with everyone's gaming junk, and I've found that's held true for pretty much every RPG I've ever played.  That's a sample size of 1, but I've never had the impression that I'm alone in this: in my experience, this is a common table size.

What surprised me, then, when diving headfirst into old-school D&D was how comparatively enormous the standard group size once was.  Looking at the image to the left, that's a pretty beefy party by modern standards.  But it's quite typical for the play environment of the day.  Let's take a look at the various old-school rulesets, starting with OD&D:

It's important to note that this specifically refers to the campaign, rather than the number of people you were expected to cram into your house at one time.  And insofar as this was ever a standard (and note the default assumption of multiple parties and perhaps multiple referees co-existing at once as part of a single campaign), it seems to have been rapidly abandoned, as none of the official D&D products ever published supported groups anywhere close to this maximum.  Less than five years later we have AD&D's rulebooks, which are quite conservative when referring to a player base:

The game is ideally for three or more adult players: one player must serve as
the Dungeon Master, the shaper of the fantasy milieu, the “world” in which all
action will take place.

(That having been said, one commenter observed that the DMG's NPC party generator on p. 175 assumes 1D4+1 characters, "with men-at-arms or henchmen to round the party out to 9".)

Mentzer's Basic has largely the same statement as the DMG, while Holmes is even more minimalistic, stating only that at least two players are required.

So while OD&D goes for an upper boundary, and the other rulebooks of the era go for a lower one, it falls to the official adventures to get a feel for how TSR's designers actually wound up playing the game.  A look at the old-school adventure line-up (through 1982) for D&D and AD&D is quite revealing:

• A1 (Slave Pits of the Undercity): 6-8 players
• B2 (Keep on the Borderlands): 6-9 players
• B4 (The Lost City): 6-10 players
• C2 (The Ghost Tower of Inverness): 5-10 players
  
• D1-2 (Descent into the Depths of the Earth): 7-9 players
• G-1-2-3 (Against the Giants): “The optimum mix … is 9 characters”
• I1 (Dwellers of the Forbidden City): 6-8 players
  
• I2 (Tomb of the Lizard King): 7-9 players
• S2 (White Plume Mountain): 4-10 players
• S3 (Expedition to the Barrier Peaks): 10-15(!) players
  
• S4 (The Lost Caverns of Tsojcanth): 6-8 players
• U1 (The Sinister Secret of Saltmarsh): 5-10 players
• X1 (The Isle of Dread): 6-10 players
• X2 (Castle Amber): 6-10 players
  

This wasn't universal, though it was pretty close.  Not all modules have suggested player sizes, while B1 was for 3-6 players, specifically including NPCs.  L1 was unusual in having a recommended floor of 2 players, but still had a high end of 8; similarly, N1 ranged from 4 to 7 players.  C1 has no specified size, but does note that if the party has fewer than 5 players then the recommended player level for the module increases, so there's a minimum assumption of 6 players there.

A quick look at some of the better adventures published in Dragon in this period wasn't as fruitful, as many don't give suggested party sizes, but 1981's The Temple of Poseidon (Dragon #46, by future TSR employee and co-designer of Star Control Paul Reiche III) was recommended for 7-10 players and that same year's The Chapel of Silence (Dragon #50) was for 6-9 players (6 2nd- to 3rd-level characters, with a possible 3 1st-level characters as backup).  Larry DiTillio (of Masks of Nyarlathotep and Babylon 5 fame) had his adventure Chagmat appear in 1982 (Dragon #63), and it was for 6-8 characters.  As such, you're seeing DMs in this period matching the official TSR environment (or at least, the DMs whose adventures were being singled out as the best by TSR and accepted for publication).

Overall, the pattern is very clear: the minimum player size is almost always 6 players, and the maximum averages out to 9.  This also matches was Gygax wrote on ENWorld in 2004, where he said that for OD&D "it was assumed that a typical adventuring party would have about 8 PCs plus as many hired men-at-arms".

Play Today

What's interesting about this, besides the general historical value, is how the modern play environment has moved away from player groups of this size.  At this point we switch from solid data to anecdata, but I think it's pretty clear that player groups of 6-9 players are no longer the standard.  Organizing and physically accommodating that many people is a headache, and many I think find the play experience created with that many people at the table less than optimal: too many people competing (even if passively / inadvertently) for the spotlight, combat rounds and other around-the-table processes taking too long, too great a chance for cross-talk and so on bogging the game down.

This has an important knock-on effect.  I think part of the reputation that old-school D&D has for lethality comes from this disconnect between the assumptions of designers in the 1970s and early 80s, and the gamers of today.  Yes, old-school starting characters are weak as kittens, and one good hit can take you down.  That is heavily compensated by the fact that old-school groups were not so much parties as squads: many PCs, and backed by all-too-often ignored retainer reinforcements to be even more numerous.

Let's take a look at some of the encounters in Keep on the Borderlands (a 1st-level module):

• 8 kobolds 
• 18 giant rats 
• 6 2-HD lizardmen (limited to one per round emerging) 
• 6 hobgoblins

If you sit down to play Keep with just 4 players with 1st-level PCs, you're going to find it a brutal grind.  Double that number, with a retainer or three for backup, and you're in a much better place; the encounters become ones deserving of respect, but beatable.

Couple this with a major shift in expectations—unlike modern editions, an old-school DM is supposed to be a neutral arbiter, not expected to kill PCs (the common strawman of old-school play) but also not determined to keep them alive—and you get a vastly different play experience.  (It's also worth mentioning the assumptions underlying combat as sport vs. combat as war).

So while old-school D&D is undoubtedly more lethal than the post-TSR versions, I think a good part of its reputation is due to people playing it today with a set of expectations not shared by its designers.

---

In terms of my game Simulacrum, this is why I've based the design on the assumption of 4-5 players.  To enable old-school modules like Keep to be played with this ruleset, I've buffed starting PCs: two hit dice to start, two spells for casters, a higher attack bonus available, and D6 HD for monsters (as OD&D did it).  Rather than deciding to make PCs more capable, I simply hoped to be able to play some of the old-school modules I own that expected 6-9 PCs with the 4-5 players I prefer.  If I expect 4 players to be able to deal with 8 kobolds with roughly the same threat level as Gygax expected, I need stronger PCs out of the gate.  Though it might look like it on the surface, they're not coming away with a lead: they're just keeping up.

Little Bits: Magic Item Charges

I was working on my magic item chapter and, as expected, pretty much copied over the traditional charges structure from ye olde D&D.  A lot of items have charges, and you spend one or sometimes two to make them work.  But after a session or two where my players received their first items with charges I remembered that those rules annoy me.  I find two issues with them:

1) They introduce per-item tracking.  Each item that has charges has to be tracked.

2) I like the general idea of magic being mysterious and unreliable.  However, a fixed number of charges works against this.  You can partially mitigate this as the GM by taking over all charge tracking yourself, which leaves the players never sure as to how many are left but still requires that they be tracked, with the burden shifted to the GM.  It also only adds an element of mystery for the players: the items themselves are still operating on fixed charge amounts.

So what I've done is gone with a roll-based system.  Whenever a charge is fired off from an item, the user rolls two dice.  On double 1s, the item fails to function because it's out of charges.

Which dice you use depends on your inclinations:

1) If you want something that varies in a way approaching canon charge amounts (per the DMG, staves have 20-25 charges, rods 41-50 charges, and wands 81-100 charges), roll 2D4 for staves (1 in 16 chance of failure), 2D6 for rods (1 in 36 chance), and 2D8 (1 in 64 chance) or 2D10 (1 in 100 chance) for wands (whichever you prefer).  For staves, this is a little less than canon, but not by a whole lot.

2) If you want something fast and simple, just use 2D6 for everything.  This boost staves a touch and shorts wands in particular, but I always felt that wands had too many charges.  I favour this approach.

For when an item requires the expenditure of two charges, you can just roll each time for each charge expended.  However, I like to treat this as more demanding on the item.  As such, for the use of two charges, I have the user make a single roll, but replace the two dice with one die that's a step higher.  So if you're using method 1, a staff rolls 1D6, a rod rolls 1D8, and a wand rolls 1D10 (for method two, everything just rolls 1D8).  On a 1, the item fails as normal.  If using three or more charges, re-roll the die once for each charge past the second (or add another die per charge used—same thing); any result of 1 exhausts the item as normal.

If you allow items to be recharged in your campaign, then it becomes a binary process: items either work or they don't, and instead of adding individual charges, recharging reactivates an exhausted item's ability.

I like these charge replacement rules because they remove bookkeeping, which is nice, but they also introduce a point of drama: everyone watches when a charge is fired off to see if that disastrous roll comes up.  It's also readily configurable to your own campaign, as you can see: just alter the dice rolled to get the particular odds you want, if the odds given above don't quite suit you.  And it adds a bit of that fickle, capricious status to magic items that I enjoy.

Tactical Encounter Variety

At the end of my last post I referenced the use of a table that added further structure to encounters.  I've long been fond of What are those wandering monsters up to?, from the late, lamented 1d8 blog, and do use it once in a while.  However, with random encounters being essentially entirely improvised encounters, generated then and there and often several times a session, I found it a bit taxing to constantly roll on that table and come up with a narrative on the spot to match the result.

As such, I went looking for something that generated the encounter variety I was looking for, but in a quicker, easier-to-use fashion.  Not really finding anything I liked, I considered how I was running encounters already.  In the dungeon I tended to follow the environment and setting: that is, the dungeon structure itself suggested the creature motivations and approaches.  In the wilderness things were hazier, as there's less structure to work off of: you're in a forest hex, and you roll up an encounter with forest monsters, and that's often all there is to it (e.g. the bandits are intelligent, but unless you've set up some sort of persistent society or structure for them in the hex, they're just an encounter; the T-Rex just doesn't care at all).

 

What's my motivation?

 

I realized that I had wound up varying them tactically.  Sometimes the party is forewarned, and gets a chance to make some decisions.  Sometimes they can outright ambush the enemy.  Sometimes the enemy returns the favour.  From there, I took the improvisations I had made and codified them.

I don't use the standard D&D surprise mechanics, as I wasn’t thrilled with their seemingly arbitrary nature: doesn’t initiative already measure getting the drop on someone?  Why are a quarter or third of all encounters, all other things being equal, complete shockers resulting in 10 seconds (or more) of pure inaction by at least one side (almost always the players, since carried light negates surprise for its carriers)?  It results in a high lethality, without much in the way of room for player skill to mitigate it.  As such, I've switched to making surprise only applicable during ambush-type situations, such as what could be generated here using these rules.  If you do use full old-school surprise, however, these rules can still be useful, in that an ambush not only means the standard surprise situation is achieved, but also lays down tactical parametres: as a proper ambush rather than just a bad surprise roll, the ambushing side will presumably be taking advantage of the terrain to their best effect: missle weapons at the ready, cover taken, ideal target(s) focused on, etc.

I've found the results to be a very simple way to help keep my encounters fresh.  Even just being able to say "a group of ogres break through the treeline ahead and are already almost on top of you" (a Stumbled-upon result) is a touch of variety that helps.  A Brief Warning result might be spotting those about to be encountered a few seconds before they spot you, or it might be "you hear a lot of noise in the brush directly ahead: something's coming your way fast".

It also feeds into the reaction roll, which I always make afterwards.  For example, if the enemy has managed to pull off an ambush, that's going to shape how I develop the reaction result I roll up immediately afterwards.  A hostile result will resolve as usual of course, but an unfriendly Ambushed result would probably lead to a shakedown attempt if the creatures are intelligent.

This is the table: 

Across the Editions — The Reaction Table

Not what I had in mind when I googled my post topic.

The Reaction Table is one of my favourite elements of old-school D&D.  The idea of not moving straight to combat with a random encounter, but assuming that the creatures have agency and so their own ideas of how things might go down is a wonderful idea that leads to vastly more varied gameplay.

The idea of how the Reaction Table should function has not been consistent across D&D editions; I'm going to take a look at the differences, with the hope of drawing some useful lessons for my own efforts.

 

OD&D 

Despite the above, monsters automatically attack unless they're both intelligent and confronted by an "obviously superior force".  As such, the table doesn't actually come into play all that often.  Otherwise, it's definitely the simplest reaction table, very easy to use, albeit with no advice as to how to adjudicate results.

Reaction Adjustment: None (Charisma has no effect on the roll, even if using the Greyhawk supplement).

 

AD&D 1st Edition 

 

 

 

AD&D moves to percentile, as is often the case with that work: Gygax seems to have preferred a greater granularity with most things by the time he put it together.  It's fairly straightforward, with the exception of the Uncertain results, which require subrolls to see if they are further inclined towards negative / positive (something I'm not fond of: I'd prefer to adjudicate everything with the single roll).  Surprisingly, for an otherwise generally verbose book, there is no further advice or conditionals laid on top, not even OD&D's brief note about the power of bribes, fear, and alignment.

 

Here encounters specifically only apply to intelligent creatures which can be conversed with.

 

Reaction Adjustment: -25% (Charisma 3); +35% (Charisma 18) 

An interesting side effect of the AD&D system is that Paladins, with their minimum Charisma of 17 (+30%), will literally never encounter immediately hostile intelligent creatures if the two can understand one another (and reaction is not predetermined), and have a 35-40% chance of enthusiastic greetings from such.  As such, the crusading, purging holy warrior also tends to leave a trail of magical friendship in his wake.  It also means that it's in the interest of a Paladin (or any other high-Charisma character) to speak as many languages as possible, as the ability only triggers with creatures that can be conversed with.  A high Int score (or some sort of Tongues item) is a godsend to such, rewarding what nominally might be considered dump stats on many characters.

 

EDIT: While I included it in the screenshot, something I passed over without comment in my original posting is the fact that in 1st edition you treat a creature as if it "were a henchman of the character speaking".  Most of the resulting modifiers are nonsensical (though "liege" (i.e. player) modifiers for "surrounded and outnumbered" or "hors de combat" could readily apply, and once in a while "saved life directly or personally" could occur if the player interrupts a battle, but racial preference and alignment are the two big ones that should always come up.  There's also a base adjustment just for having an alignment at all, regardless of behaviour: interestingly enough, Paladins win again, gaining an automatic +15% in all scenarios along with any other character that is lawful good.  This can be more than countered, though, by the alignment factor modifiers, which can impose up to a -35% penalty for polar opposite alignments (e.g. a LG character interacting with a CE NPC).  Overall, you get a fair number of tools to better run reaction than any other edition can (I like the racial preference ones in particular), and the effect is very different from every other edition, though I question how many people are going to run through and apply all these modifiers.

 

AD&D 2nd Edition

 

2nd edition has a reputation for cleaning up some of 1st ed's wonkiness, but in my experience it tends to improve only the clarity of the text.  In terms of mechanical subsystems, it delights in baroque complication almost as much as 1st, and in some cases even more so.  Here we've switched to a 2D10 table (rather than percentile), with the players' base attitude being the primary factor.  Why there needed to be a Threatening vs a Hostile column I don't know, but if players want to be murderhobos (as 2nd ed tends to encourage, since combat became the primary XP source in 2nd unless the DM really embraced the vague story-based XP reward concept or used the now-optional gold for XP rules), then they have a base 36% chance of immediate hostility, the highest chance by far in any system featured here.

One of the most important things the table assumes is that hostility is the "good" result.  That is, hostility is the highest result in the chart, so that any positive chart modifiers lead the players towards it (well, mostly: see Reaction Adjustment below).  This is true even if the players want to be friendly: a friendly approach only reduces the range of possible hostile results.  Similarly, while I don't see the need for a Hostile column separate from a Threatening column, altering the latter so that flight is the highest result would at least make room for a charismatic intimidator.  As it is, there's little difference between the two except that if Threatening you still have a 3% chance of bullying someone into being your friend.

At the same time, the 2nd ed reactions are the most nebulous aside from OD&D.  Even hostile is coded only as "Irritable, hot-tempered, aggressive, or violent": that is to say, there are no behaviours mandated by the results, only attitudes.  However, these rules don't have the same restriction on communication and creature intelligence that the earlier systems do.  As such, it's easy to assume you roll reaction for wolves and the like just as much as you would bandits, although the DMG suggests that "The creatures should act in the manner the DM thinks is most appropriate to the situation" and that you should only use reaction when you don't "have a clue about what the monsters will do."  Considering combat is much more the norm in 2nd ed, I suspect that most campaigns went with DM inclination, that inclination being "attack".

Overall, I can see the value in having a player-directed reaction table, but not in four categories and having Charisma behave in such a monolithic, unintuitive fashion.

Reaction Adjustment: -5 (Charisma 3); +7 (Charisma 18).  It looks at first glance as though the chart tends towards hostility as the "positive" option, so that someone more charismatic is more likely to get into an immediate battle, even if they don't want to (i.e. their initial approach is Friendly).  This is because Reaction Adjustments from Charisma are written so that high Charisma provides a positive modifier, and low Charisma a negative.  However, the text in the PHB on p. 17 (the page before the Charisma chart), though written in natural language, makes it (largely) clear that in this one case the modifiers should be reversed.  In other words, the -5 from having Charisma 3 (for example) would be applied in this case as a positive modifier to the roll (making a hostile result more likely), and a bonus applied as a negative (making a friendly result more likely).  However, other modifier types, such as the morale based ones, are applied in the opposite (standard) fashion (i.e. high = adds, low = subtracts).  It's a baffling implementation, only clarified in the Complete Wizard's Handbook, of all places: errata never tackled this.  (Thanks to commentators below for highlighting this so that I could edit this portion).

Holmes / Moldvay Basic

Clean, simple, and straightforward, this system originated in Holmes.  The use of 2D6 effectively creates a curve leading towards the confused monster result (44.42% of the time), which suggests that most monsters of the world are baffled by the existence of adventurers and means that most of the time the initiative lies with the players.  (The image is from Moldvay; Holmes notes the 6-8 result as "Uncertain, make another offer, roll again").

Holmes and Moldvay, like 2nd ed, don't place the restrictions of OD&D and 1st ed about only applying to certain subtypes of monsters and situations (beyond specific monsters such as zombies).  The entirety of the relevant section is given above: there's no further advice.  Unlike 2nd, however, the reaction roll is given as an equal option to predetermined reactions, not a fallback.

Reaction Adjustment: None in Holmes.  -2 (Charisma 3); +2 (Charisma 18) in Moldvay.  All you need is a Charisma of 13 to avoid all possible immediate attack results (though admittedly these only occur 1 in 36 times anyways).  A Charisma of 18 gets you a 1 in 6 chance of enthusiastic friendship for every encounter.

Mentzer Basic (Red Box)

People are fond of saying that Mentzer and Moldvay are the same, the only difference being formatting, but this is not the case: reaction is one area where they notably differ, as seen above.

I dislike this system as it falls back to a 1st edition sin that Holmes/Moldvay dodged: the reaction re-roll.  It actually triples down on this: rolling a 3-5, for instance, gets a possible attack, but you need to roll again, and if you roll 9-12, you have to roll yet again.  Your possible attack could thus turn into friendship.  I again see the value in varied results, but not via such a clumsy implementation; I'd prefer a reworked version of 2nd edition's master table if I was going for that.  It does have the advantage of giving more guidelines than Holmes/Moldvay, however, including a little advice on negotiations.

Reaction Adjustment: -2 (Charisma 3); +2 (Charisma 18).  The notes given for Holmes/Moldvay in this section apply here as well.  However, unlike Holmes/Moldvay, Charisma adjustments only apply if the players can talk to the monster.  There is also a specific allowance made for character reactions, with the same -2 / +2 range of possible reaction adjustments suggested for this, which is cumulative with any Charisma modifier.  It's not clear if you apply your modifier(s) to every subroll, or just the first roll in the series.

 

Rules Cyclopedia / The Classic Dungeons & Dragons Game

1991's New Easy to Master Dungeons & Dragons Game (TSR 1070) and 1994's The Classic Dungeons & Dragons Game (TSR 1106), the final gasps of old-school D&D base rules implementation, each use the same table.  By changing the instant attack result from 2 to 2-3, immediate hostility moves from about 2.5% of the time to about 8%.  Adding 1 at the start bumps everything else up by 1, which due to the nature of the 2D6 curve makes possible attack/monster growls more likely and uncertain/cautious less likely even if the number of results stays the same, until we get to the "possibly friendly" result, which is now only 10-11 instead of Mentzer's 9-11.  Overall, this table version walks back from the nested subcomplexities of Mentzer, but still relies on re-rolls.  The nature of the 2D6 scale means that a 4-point modifier is huge, making it very likely for you to achieve the result it's pushing you towards.  Still, it is possible you could wind up rolling several times.  Both books state to not roll more than three times, however.  "If the PCs don't do something to get a reaction (talk to it, or attack it, or put it to sleep—something) by the third roll, the monster attacks if the roll was 9 or less (remember to take into account the monster's alignment).  It just leaves if the result was between 10 and 12."

"Monsters" here are specifically defined as anything that isn't an NPC / doesn't have a character class.  As such, the reaction roll is very clearly applicable for most anything in the game.

1991's Rules Cyclopedia also uses the above table, only changing the wording on the first part of result 10-11 to "monster is neutral".  Like Mentzer, it also suggests a -2 / +2 allowance for character reactions, which its more introductory 90s brethren skip.  It oddly decides to be less granular than the intro products, however, if three rolls occur: "If by the third roll the monster hasn't achieved a roll of 10 or better, it will decide to attack or leave."  Interestingly, it is the only ruleset that uses subrolls to rule on when Charisma adjustments apply: they are only used for rolls after the first.  As such, if you get an initial hostile result, no amount of Charisma will save you.  Essentially, the initial reaction is as-is, but if the room for negotiation is there, charm (or the lack thereof) can come to the fore.  I like this a lot, although by being able to stack Charisma adjustments and player behaviour adjustments to these later rolls, it's pretty easy to swing the 2D6 scale in your favour.

Reaction Adjustment: -3 (Charisma 3); +3 (Charisma 18).  Unless using the Cyclopedia, with an 18 Charisma one can avoid any "monster attacks" results, and it's not clear if you apply your modifier to every subroll, or just the first roll in the series.  Also unless using the Cyclopedia, a Charisma of 18 gets you a better than 1 in 4 chance of a friendly result with every encounter.

 

What I Use 

Click for PDF

I wanted results that reflected monster alignment, clear statements of intent about what each reaction means, no nested rolls, and some decent guidelines as to when to use the rolls and how to modify them.  As such, I've gone with a basic 2D6 scale, matching Holmes and Moldvay (and thus maintaining compatibility with products designed for such).  Sometimes I've gone with major changes when implementing a common subsystem, but I really felt in this case that if it isn't broken, why fix it?  The only mechanical change I introduced is a roll penalty for dealing with Evil creatures and an accompanying hostility re-roll when generally dealing with Good ones.

My main "alterations" have been in presenting firm guidelines as to how the system is meant to work: not just when to use it, but what specifically happens when you do.  Holmes and Moldvay were very weak here, lacking the solid advice of Mentzer or even the 1990s efforts.  I wanted a GM to be able to readily able to apply this at the table.

I don't use Charisma modifiers, so the table results stand as they are, which is also a key structural change.  This was because I wanted to completely leave social interactions to roleplaying and random rolls, without even a stat influencing them.  I have nothing specific against Charisma: this was just something I wanted to do here.

There is a note about an "Ambushed" result in the text.  That is because I introduced a basic encounter structure table, specifically for outdoor encounters, meant to adjudicate the general tactical situation so as to provide some encounter variety.  It's possible that the encounter begins with the monsters springing an ambush on the party, and so I wanted to make sure that such a situation was covered.  I'll examine that in a future update.

I must admit to being tempted by the prospect of a cleaned up and more rational 2nd ed approach, and having written this up I am again, but for now I'm sticking with the Holmes/Moldvay interpretation.  I may come back to this, however.