Monday, January 3, 2011

Basic D&D: Close Combat Trinity

I'd like to spend some time thinking about the basics of D&D, which is to say, the most fundamental assumptions built into the combat mechanics of classic/original D&D. (This is not to be confused with the B/X style "Red Box" rulesets, which ironically I was just commenting on at another blog a few minutes ago.)

Obviously D&D evolved out of medieval miniatures wargaming (Chainmail). If we boil premodern land combat down to its most basic level, then I would argue that we find three fundamentally atomic types of hand-to-hand combatants (leaving out missile troops for the moment), which I'll call: Swords, Cavalry, and Pikes.

You'll see these three core types in the diagram above. The arrows indicate dominance (which type defeats which other type). In the broadest possible terms, you could track a history in which the ancient Greek phalanx (pikes) was superseded by the flexibility of Roman legionnaires (swords), which were ultimately replaced by medieval European cavalry, which was ultimately bested by Swiss pikes, which began losing battles to Spanish swordsmen in the 16th century (reverse around the graphic), etc.

When I say "Swords" here, I'm ultimately thinking of any shield-and-something combination where you walk up to an opponent and swing a melee weapon in an arc with your arm in order to strike a blow (which could include swords, axes, flails, etc.) Cavalry are charging lancers. Pikes are any long, pointy weapon that you can stick out defensively. Perhaps you could think of this categorization as the "Rider/Pike/Sword" triad (i.e., RPS, so as to emphasize the analogy with rock/ paper/ scissors).

Consider the core D&D combat sequence of movement, followed by an attack (or some other permutation; the important thing is that they're discrete steps within a turn of action). Interestingly, each of the three core combat types interacts with the move/attack sequence in a different way.

Swords (or any handheld weapon that delivers a blow by stroke) work the most elegantly within this system, and they're also the type most emphasized by D&D itself. It does make sense that moving up to an opponent by leg power, and then standing and swinging a weapon with your arms several times, are indeed separate and distinct components of an action. It also makes sense that this type of action takes the spotlight in D&D, with its focus on close combat within the confines of a dungeon, where other longer weapon types are impractical.

As much as it pains me to say it, cavalry and pikes very intrinsically work counter to a separately phased move/attack sequence. The essence of a Cavalry attack is that it happens while in motion -- the horses gallop, the lances held straight ahead and still, the moment of impact, and if successful, the breaking of the enemy line and continuation of the gallop. The attack is fundamentally one with the attacker's motion. Pikes are the inverse of that; primarily the advantage is defensive, again with the pole weapon held out straight ahead and still, but set so as to take advantage of the enemy's movement, as the foe's momentum causes them to be impaled upon the presented points. In other words, the ideal attack is fundamentally one with the opposition's motion. In both of these cases, I would argue, the discrete move/attack sequence breaks down, because it is inherently antithetical to the nature of the attack; the impact is due solely to the forward momentum itself.

Cavalry and pikes, meanwhile, get surprisingly short shrift in the classic D&D rules. The three-page Chainmail section on man-to-man combat (p. 25-27) manages to contain more detail on mounted combat than the entirety of the OD&D and AD&D system, for example. Again, this is not entirely surprising, but notable nonetheless. If you do want to run games with prominent usage of classic cavalry and pike units, then you've got to make certain independent adjudications (filling gaps) about how they work. That's just the nature of the beast, because it wasn't part of D&D's core competency.

In summary:
  1. Historically: Swords beat Pikes beat Cavalry beat Swords.
  2. Sword attack is a separate impulse from movement.
  3. Cavalry attack is intrinsic to the unit's own movement.
  4. Pike attack is intrinsic to the unit's enemy's movement.
  5. Cavalry & pike are only cursorily treated in dungeon-centric D&D.
Hopefully I can continue some of this thought process on a semi-regular basis. Happy New Year!


  1. Would it be pertinent to say also that one of the defining characteristics of pike tactics is the necessity of massed pikemen? (Vis a vis the phalanx.) At least if you're talking about them on the field of battle. A lone pikeman is toast against a lone swordsman AND a lone cavalryman. (Unless the rider or his horse are suicidally stupid.)

    Actually, the only context in which I can imagine a lone pikeman having ANY chance at effectiveness is in close quarter, low maneuverability space like, say, a dungeon corridor.

    So to echo your point (no pun intended) it's interesting that that sort of tactic gets short shrift in D&D. Seems like the 10' pole is even better with a blade at the end of it.

  2. That's part of what I hope to expand a bit on in a future post. Which leads in to the need for extra cohesion to maneuver and present the many pikes in an organized fashion; more catastrophic breakdown if that becomes impossible.

  3. Interesting and well-thought out post, as always.
    Will your "Book of war" follow the R/P/S style of WRG type game (especially DBA)? That would be a plus in my book.
    One thing I'd quibble with, and it is really quibble especially when the focus is on the middle ages, but pikes were also used very offensively in earlier and later periods. The Macedonian pikemen charged at times, and I understand that English Civil War era troops also used the pike to charge at times. But I think your overall point about spear/pike armed troops being more defensive than offensive is of course correct.
    Also -- I like the diagram. I've seen similar ones in military books which add light (bow armed) cavalry, and archers to the diagram, but it gets complicated because some of the arrows go across the center.

  4. Hey Mike -- Yes, I wanted to make sure that BOW plays out this way, and it does (on a cost-ratio basis). Ultimately that's where this series will wind up in a number of weeks. :)

    I totally agree that pikes were used offensively, and also that the place where they really shine and stand out by default is on defense.

  5. Oh, the diagram is excellent--it's something you can show people in the middle of a big fight and they'll understand it.

    What do you think of the trinity used in the old Ancient Art of War video game:

    arrows beat heavy infantry (knights) who beat light infantry ("barbarians") who beat arrows?

    I feel like it makes sense in dungeons but seems kind of tactic-dependent on the battlefield.

  6. Great post.

    Gygax mentions pikes are only good while in formation specifically in the weapon speed section of the DMG, "when weapon speed is a determinant in initiative" section lifted right out of the MtM section of CHAINMAIL.

    Look forward to the next installment.

  7. What blog did you write about B/X?

  8. @Zak: Thanks, I've had a poster of this over my computer to meditate on daily for about 2 months now.

    Re: AAOW triad: it's arguably supportable, although counterargument is if archers lose to barbarians, then why not even more heavily armored knights? (Perhaps case is made that a shield makes the big difference, although in D&D the shield is not represented as being that powerful.)

    For a long time I had archers in a complicated 5-way core analysis, but I recently realized it was good to start with a kernel of pure hand-to-hand types, esp. for the Swiss pikes vs. cavalry issue. In the game I'm developing, archers actually swing the other way: beat light armor but lose to heavy armor (matching D&D, really).

  9. MV: It was actually Zak's. A very tangential point in the comments about where race-as-class first popped up -- but amused me just as I was sitting down to write "don't confuse this with Red Box basic".

  10. Does this mean that pikes are only good in defence?

  11. No, but it means they are best (primary advantage) on defense. Especially as a means of foiling cavalry charge -- using the attackers' forward energy against themselves.

  12. Here's a long response to your post:

    Maybe it can offer you a different perspective on how to incorporate charges and pikes into D&D.

  13. Sorry, I messed up that last comment. Here's a clicky link for your convenience.

  14. Anthony, thanks for the link and thoughtful response. I'm not going in the direction of morale checks myself, but it's interesting to consider.

  15. Does it also mean that pikes are best in rought terrain (because they're standing still anyway), and cavalry is worst (because they rely on speed)?

  16. Just thought of another question: I know in the Middle Ages European armies had pikemen, archers and mounted knights. But did they have D&D fighter types ie heavily-armoured swordsmen on foot? And if not, why not?

  17. Hey anarchist -- My understanding is that both pikes & cavalry suck in the rough (pikes because it interrupts the group-organized facing/maneuvering you need). Swords have the flexibility and advantage in that case.

    And my understanding of army composition is that it evolved slowly over the Middle Ages. Cavalry was the primary thrust through the middle part (1000-1300). Later, use of pikes was found to counter cavalry (1300-1500). Near the end, swords started to come back to counter pikes (early 1500's), but the cycle was interrupted by gunpowder. Military science and tradition evolved very, very slowly in the premodern era.

    (Much of this from C.W.C Oman, The Art of War in the Middle Ages, Cornell 1953).

  18. Unfortunately, this analysis falls down on several points. First, pikes were offensive weapons; Macedonian phalangites worked by either pinning the enemy down or breaking them altogether with an inexorable push forwards, while Swiss pikes were human steamrollers. Second, the idea that "swords beat pikes" is instantly brought to the question by the Battle of Seminara, where Spanish sword-and-target men had their first major encounter with Swiss pikes--and lost. The sword-and-target men were light infantry meant for fighting in rough terrain or fortifications, so they had little chance against pikes in the open. When the Imperials won later at Cerignola, it was with the help of field fortifications that broke up the Swiss formation (thus making them more vulnerable to infiltration by small groups of swordsmen)--and of course loads and loads of firepower.

  19. (Forgot to add that the lesson is that there was a great deal of irreducible complexity in ancient/medieval/Renaissance tactics, and any attempts to simplify the equation--as here--inevitably brings the danger of getting things too simplistic to provide a reasonable simulation of the actual interactions on the battlefield).

    BTW, for anarchist: the real medieval equivalent for the "heavily-armoured swordsman on foot" was the "mounted knight" sans his horse. Dismounted men-at-arms were the prime infantry type throughout the Middle Ages; mind that "men-at-arms" back them largely meant the group of warriors now popularly known as "knights," so the term most clearly didn't refer to common soldiers. That's another wrench thrown into the RPS equation....

  20. l-clausewitz said: "The sword-and-target men were light infantry meant for fighting in rough terrain or fortifications, so they had little chance against pikes in the open."

    Thanks for the comment -- totally agree with that, that sword/shield men are better in rough and pikes in the open. The game I'm developing matches that observation; although on balance, assuming a sufficient frequency of rough terrain where battles take place, then sword/shield has an advantage over pikes in overall cost/benefit analysis (matching Oman and Machiavelli).

  21. I wouldn't rely that much on the authority of Oman and Machiavelli on this subject, though. The evidence we have shows that armies with sword-and-target men only employed relatively small numbers of them in a specialist role--far fewer than pikes, unlike Machiavelli's proposal where he called for a large formation of swordsmen screened by a small company of pike. (Needless to say, his proposed army structure was never adopted.)

    It's also worth noting that "sword" might be too restrictive a term, since rough-terrain specialists in a similar role in other European armies were often armed with polearms (such as Swiss and Landsknecht halberdiers or the Flemish clubmen at Courtrai). Some miniature wargames use "Blades," which I think is a rather more appropriate shorthand.

  22. Very late coming to this post - sorry - but I must point out that your analysis of move and strike as separate actions in swordplay, does not concord with any competent demonstration of swordsmanship.

    Please consider, not only Tai Chi and Ai Ki Do, but also, the many, many cinematic portrayals of renaissance sword fighting.

    For maximum effect, and minimal risk, a sword-stroke always is delivered by a moving target upon a moving target. The guy who runs up, stops, and -- gets skewered by the guy dodging aside -- is not the guy who wins the fight.

    PRESS: the simultaneous actions of charging, cutting, and dodging aside, such that one's opponent stumbles back.

  23. Sigilic -- Thanks for the comment! I would almost entirely agree with what you say above. However, when I used the term "movement" in the blog post, I meant it as straight-line movement only (i.e., covering actual ground as in a scale tabletop game): I'm not talking about "dodging" activity. (Swordplay presumes that, and also likely use of a shield.)

    So I'm thinking that you can distinguish between an attack with lance or pike -- where (a) the only option is to line up the weapon straight-on along the vector of the charge, and (b) you specifically do NOT have any ability to dodge while the blow is being delivered -- versus a sword attack which is at least possible to accomplish while (a) center-of-mass stays in place, and (b) sword-swing comes from a different direction (possibly opposite) than part of the dodging body.

    Perhaps the term "rooted" is appropriate here (sword-strike generally yes, lance-charge basically no).