As part of my continuing search, I recently, at long last, acquired a copy of the rules to Gygax & Arneson's first-ever game collaboration, Don't Give Up the Ship (DGUTS below). This is the 2nd Edition (1975), with added optional rules from Mike Carr (designer of Fight in the Skies, referenced in OD&D as inspiration for its detailed aerial combat rules). It's a great read, obviously made with a ton of love and affection to the milieu of fighting ships of the American Revolutionary period; and in very much the same style as a book like Chainmail. In some sense it has about what you'd expect: a Basic Game, optional Advanced Rules, super-detailed Single-Ship Action Rules, more abstracted Fleet Action Rules, rosters of possible ship-to-ship engagements (all historically based), a Bibliography of over a dozen historical texts, etc. The play involves tracking gun weights, wind direction & force (reminiscent of many tables in D&D), sailing points and gunnery with a protractor, and so forth. Boarding is entirely abstracted, with the mechanic based purely on opposed morale checks (and without any men being lost from such actions in the standard rule). Some of the DNA can even be detected in Swords & Spells; like, the overall format of the record sheet, and the percentage-based damage adjustments that likely require a calculator (e.g., the last ship-to-ship optional rule notes that "Various tests made have proved beyond reasonable doubt that U.S. shot was lighter than it should be", and so should do a proportional 7% less damage; whereas French shot was 10% heavier and therefore will do 10% more damage).
It's not immediately clear how this can benefit your D&D game, however. One thing I noticed is that the scale is wisely chosen: 1" = 100 yards, 1 turn = 5 minutes. It recommends ship models close 1: 1200 scale (note this varies from the surface equivalent to scale of 1: 3600, but by less than an order of magnitude). This is quite different from the D&D scale officially 1" = 10 yards, or arguably what should have been 1" = 5 feet to match the size of miniatures in Man-to-Man action. Yet despite this, OD&D recommends the same 1: 1200 scale for ships as DGUTS (Vol-3, p. 30), which is either 3 or 20 times too small depending on how you count that. (As an aside, we can reflect here how much earlier wargaming relied on the player to acquire or build their own materials from other products, as the rules were designed expecting toys like that to be commercially available; this is long before consolidated brands in which wargame rules are part of a company selling their own boxed products.)
The really startling thing (to modern eyes) is none of that, however. First let me observe that DGUTS seems to owe a rather large inspiration and debt to a halfway famous older game, Fletcher Pratt's Naval War Game (published in 1943, but developed and played for more than a decade before that). Pratt's game simulates fighting ships contemporaneous to its play in the era between the World Wars. The intriguing thing that Pratt did with it is to rely on the real-world publication of Jane's Fighting Ships for its ship statistics. I wouldn't know much about Pratt's game it weren't for Jon Peterson's sublime Playing at the World. Peterson writes (p. 280):
Pratt borrowed Jane's method of classifying ships, especially his notation for measuring arms and armor. The thickness of armor and the size of guns are quantified and compounded in an elaborate mathematical formula, to which additional figures are added for amenities like torpedoes or the ability to carry aircraft. This sum is multiplied by the speed of the vessel in knots, and finally the tonnage is added to determine a "value" for the ship. Ship values tend to be large; one example boat given in the rules has a value of 23,034. Guns, when they score a hit with a shell, inflict a certain number of points of damage depending on their size; the weakest 37mm guns might inflict 23 points of damage, the standard 4.7" cannon hits for 244 damage, while the implausibly large 16" cannon deals a whopping 10,550 points of damage. As a ship suffers points of damage, it begins to lose capabilities, including movement speed and the use of its guns. For the convenience of players, a "ship card" typically lists all of these attributes and details exactly which capacities are sacrificed at the various levels of disrepair. When it has taken damage greater than or equal to its value, a ship is sunk.
Gygax & Arneson's (and Carr's) Don't Give Up the Ship uses the same basic idiom, somewhat simplified, for its ship statistics. Two types of damage are tracked: high (sails & masts) and low (deck and hull). The high damage score is simply half the real-world ship's actual tonnage, with these points split proportionally among each individual sail and mast on the actual vessel -- lose a sufficient number of points, and sails/masts are lost, reducing speed appropriately. The low damage is equal to the real-world ship's tonnage -- when damage scored is over 70%, the vessel may possibly sink slowly, while at 100% the vessel sinks automatically; in addition, crewmen are lost proportionally following any low damage scored. Specifically: "crew factors" (CF) are tracked where a ship has one CF for every 21 men on the real-world ship (a seemingly odd conversion rate, but this comes from an estimated 7 men to operate a gun, times 3 guns per fire unit; see p. 17 and below).
Likewise, guns are based on whatever guns the real-world ship was known to have. For example: Ship A in the simple "Training Game" scenario has 12 24-pound guns, 15 12-pound guns, and 3 9-pound guns (denoted 12-24#, 15-12#, 3-9#). Every 3 guns of a given type allows one "fire factor", that is, one d6 roll on the very simple combat table per turn (which can result in either a high or low hit or a miss, depending on range). Damage is simply equal to the poundage of the gun type firing -- for example, one hit from a trio of of 12-pound guns does 12 points of damage. In its way, a breathtakingly elegant mechanic!
The thing that I like about these systems is that by making an explicit connection/formula between real-world entities and game entities, the designers have immediately populated their game world with everything (of the appropriate category) in one fell swoop. Pratt doesn't need to include rosters of ships; he can just direct the reader to Jane's Fighting Ships, or other reference works. The ship rosters in Gygax & Arneson's DGUTS aren't really game statistics, they're just real-life profiles of historical ships (in terms of real tonnage, guns, crew, etc.). The designer doesn't have to spend time laboring over individual game-piece statistics. One's game is enriched by, and stands on the shoulders of, the amount of attention and detail given by military scientists tracking and documenting the things in the real world. If any balancing or revisions are needed, they are only in the game rules themselves (perhaps tweaking the mathematical formulae that simulate ships in the game).
Note how briskly this tacks against the fantasy gaming headwind that attention to simulating concrete, real-world elements cannot have any benefit (link). And how synchronized it is with our long-running observation that real-world solutions almost always give the most elegant in-game mechanics (link). In fact, by making the link explicit and mathematically precise, we instantly profit by the whole universe of whatever real-world thing we have simulated.
Of course, this assumes that there is a pre-existing compendium of measurable information compiled about our topic of inquiry, which we have in abundance for fighting ships -- and much less so for individual people, arms, creatures, or of course fantasy creatures. But I am tempted to think about what other quantitative fields of study exist that we can port semi-immediately, and connect explicitly, into our D&D games to their benefit. Let's be on the lookout for that in the next few weeks.
Can you think of any such field of study that we can mathematically connect to our D&D games in order to enrich them?