Wednesday, February 7, 2024

Mass Combat in AD&D 1st Edition, a New Approach

I've been thinking about "domain-level" play in AD&D 1E lately, going through old Dragon magazine articles on the topic and picking out the best ideas. For those interested in that broader topic, I'd suggest that two issues are especially relevant to that matter: 125 and 108 (the latter because one of the articles in 125 made use of ideas found in "The Role of Nature", an article in that issue, but poor editing scrambled them somewhat). Issue 125 has articles titled "Meanwhile, Back at the Fief" and "Armies from the Ground Up" that cover some of the most important concepts relevant to domain-level play. However, that digression is not what this article is about. Mostly, this article is me thinking through an idea and playing with it to see if it's useful.

Here, I want to consider how to use AD&D to play out mass combat without going far outside of the rules in the Dungeon Masters Guide. I realize that sounds like a tall order, but bear with me.

A couple of attempts have been made over the years to make mass combat wargames for AD&D, most notably Battlesystem. In its first edition, the author created a somewhat abstracted method of discovering how many general "hit dice" of damage a unit, stand, or figure dealt in a combat turn. The second edition abstracted things further, reducing a set of attacks to a roll on one of the various polyhedral dice. Those are very convenient, but aren't mathematically stringent. This is, of course, not really a concern for most players and DMs, but other options do exist.

Recently, Daniel Collins of Delta's D&D Hotspot used large-scale simulations of the sort only possible in the modern computing era to develop a set of miniatures rules called Original Edition Delta: Book of War. This is mathematically stringent (Collins is a college lecturer in mathematics in real life), but is mostly set up for Delta's particular house rules of original D&D. On the other hand, it bears some resemblance to 2nd edition Battlesystem, for those who like that game, and those house rules are sufficiently similar to other D&D editions as to be close enough for most purposes.

That said, I wanted to just find a way to get through all of the hundreds of rolls required in a large-scale AD&D combat as quickly as possible. That led me to an article published in Dragon magazine issue 113 titled "One Roll, to Go". This article used statistics and probability to reduce 5, 10, or 20 rolls of a d20 to one simple percent roll, with the roll on the d% cross-referenced to a binomial table to determine how many of the rolls would be successful. Using this table with an average result for damage (count each d4 of damage as 2.5 points, each d6 as 3.5, each d8 as 4.5, and so on), as well as for hit points for each member of a unit (a "normal man", with 1d6+1 hit points but a minimum of 4, the normal amount for a soldier would have 5.5 hit points, for example, since that is equivalent to 1d4+3). The main issue is working out how many soldiers can attack at any given time, and spreading those attacks out across a unit randomly as per the DMG.

My suggestion would be to use the figure scales from Battlesystem 1st edition, where a unit with so many hit dice per individual is at a 10:1 scale, with more hit dice being at 5:1, 2:1, or individual figures depending on the exact hit dice amount. The front rank of such figures would be the ones able to attack, with ranks further back being able to bring longer weapons to bear. So, a 10:1 figure might bring 5 individuals to bear in melee, or all 10 if the second rank is equipped with spears (for example), a 5:1 would bring 2.5 (I'd allow it to round up, but just counting 5 individuals for every two figures is easy enough, too), while 2:1 and 1:1 figures would allow all of the individuals so represented to attack normally when in contact with the enemy.

There's a lot still to work out, of course. As melees go on, the ranks of each unit start to interpenetrate in melee circumstances, for example, maybe bringing one more rank to bear each melee round until reaching the back rank of one or the other unit. But beyond that, morale rules already exist in AD&D, so that is handled. I'm well aware that this system would not be as quick-playing as those more abstracted and streamlined systems, but my intention isn't to make it into that, only to offer an option for just using AD&D with minimal additions. I also fully realize that most people won't want to use this method. But it's something to consider, anyway (chances are, I'd just use Battlesystem, 1st edition, myself; it doesn't beef up individual heroes in the same way 2nd edition does, and the results are close enough to AD&D results to keep me happy).

Perhaps this method is best used for mid-scale battles, where there are maybe 50-100 goblins fighting the PCs and their retinue of henchmen and maybe a couple dozen mercenary soldiers or whatever. That is, battles that are barely worth breaking out Battlesystem or Book of War for. After all, 5 or 10 figures of goblins on the battlefield might not be worth the time, but on the other hand that many are very unwieldy when every one has to be rolled for individually.


  1. The first edition of Battlesystem was mathematically stringent. The main resolution chart was based on a binominal distribution of what happened with 30 attacks were made (10 men attacking for three rounds).

    When you plot the results, they follow a bell curve.

    1. Thank you for saying so, I'd hoped that was the case, but I was very unsure of the situation there, since the designer's notes articles I'd seen were a little obscure on the details.

    2. While I was able to reverse engineer the math, it was also obvious that it was tweaked in the interest of playability.

      As I mentioned in the next comment. The odds shift slightly more than one whole integer (# of successful hits). Resulting in a two integer shift at the end of each +15% change in the odds. So there is a slight inaccuracy, but still way better than any other D&D mass combat made before or since.

      It is so good I was able to adapt it easily to D&D 3e and D&D 5e.

      My other battlesystem posts.

  2. As a result, any D&D mass combat that doesn't use a bell curve to resolve how much damage is dealt by X number of combatants is not mathematically consistent with how D&D works.

    1. Yes, though there are different ways to get a bell curve. Battlesystem 2E and Book of War get there by rolling multiple dice, one per figure in a unit. The binomial table article I referenced gets there by digitally interpolating the results on a d% chart.

    2. While you are rolling multiple dice for both BS 2E and Delta which does produce a bell curve. However, it is still inaccurate for what a figure would do if you rolling multiple d20s across how many rounds a mass combat turn comprises of.

      BS 1e avoids this fallacy, whether it is just one figure or many. In addition, I reverse engineered in part the table and found out that there is an important reason why 1 mass combat turn equals three AD&D combat rounds.

      When you calculate 30 trials across the range of d20 possibilities (0% to 100% in 5% increments) each 5% jump also shifts the peak of the binominal bell curve by little more than one whole integer.

      While not an even integer, the result looks amazingly like the d8 column of the BS 1E resolution chart.
      Changing the distribution to using 10, 20, or 40 trials produces a very different looking chart.