What's true about aerial combat in Advanced Dungeons & Dragons is that:
Nobody probably uses it as intended
It is designed to either avoid combat or drive it to the ground
If this doesn't sound particularly appealing, I understand; but remember, this is AD&D, where combat seldom lasts more than three rounds anyway. The brutality of the aerial combat system is such that your players will likely try to evade or land and seek cover; which one might say is the sane thing for humans and demi-humans to do, seeing as how they are not native to the firmament and have little business trying to contend there with Air Elementals, Efreet, Griffons and the like.
Aerial Combat comprises not-quite 3 full pages of the 1979 Dungeon Master's Guide. It starts over half-way through page 50 and ends exactly mid-page 53.
Here are the essentials.
Short version: No one should attempt aerial combat; but hey, if you are riding a griffon you are probably OK until it gets charmed or paralyzed.
Long version starts here with Maneuverability. The gist is that all fliers can climb 1 hex in elevation per 3 hexes traveled [or] dive 1 hex in elevation per 1 hex traveled. Maneuverability Class A is excluded from this and can ascend or dive vertically if they so choose.
For simplicity sake, the way to run this system during an actual game is as follows:
Get a BIG sheet of hex paper with MANY hexes.
Make each hex = 30'
Whoever WINS initiative decides whether to move FIRST or SECOND.
Each piece is then moved in three hex increments, taking turns until all movement has been expended.
Pieces can move up to their full MV rate but, unless they can hover (MC A & B only) they must move AT LEAST 3 hexes.
If at ANY point during movement range allows for attack, resolve attacks simultaneously at that time and then continue with movement.
If two pieces ever occupy adjacent hexes, allow for physical blows.
Note that you will not find steps 1----7 in the DMG. Those are my steps for bringing process to the rules as written. Also note that this is NOT a standard combat round and is almost a different sort of animal entirely. Initiative has nothing to do with who attacks first. It has to do with who is forced to begin moving their pieces first, thus showing their hand. Aerial combat is 95% about two things: Maneuverability Class and Speed. Whoever moves first has a slight disadvantage if they are trying to evade in that the foe can react. Whoever moves first has an advantage if they are nearly within range and are desiring to attack. This is why I allow the initiative winner to decide who moves first.
I'm going to show you how this actually plays out below, but before you get all surprised and bored at the same time, brace yourself for the terrible truth: physical blows rarely take place unless there is a great disparity in speed between the two combatants [or] both potential combatants actually WANT to strike physical blows (which as you may have realized by reading the sidebar above is very dangerous for both participants----especially if one or more of them have membranous wings). Even missiles and breath weapons are unlikely to get an opportunity for use if one of the combatants is actively avoiding combat (I.e. fleeing) and has superior mobility.
Check out this SUPER boring progression:
Notice above that the Dragon and Couatl are assumed to be at the same relative elevation and the Dragon neither climbs nor dives, therefore "+0". Yes, the Couatl could simply turn ethereal at this point but it decides to "play" with the dragon. Dragon MV = 30". Couatl MV = 18". Ergo the Dragon can move up to 10 hexes this "round" while the Couatl can move a maximum of 6.
To simulate both fliers moving at the same time, I enforce alternating token movement based on the initiative winner's choice of "who goes first". Note above that the couatl chose to climb +1 hex while moving forward 3. Because it is MC:A, it could have chosen to ascend +2 while moving forward only 1 hex. Or it could have ascended vertically 3 hexes. It could have also dove etc.
Note above that neither flier has used any of their allotted turn maneuverability...yet.
The next bout of movement is very simple. The couatl chooses to keep flying in a straight line and ascends another hex, placing it at a relative altitude of +2. The couatl has now expended ALL of its available movement (speed 18" = 6 hexes) so the dragon is allowed to finish out the remainder of its move (30" = 10 hexes). The dragon also climbs to relative +2 altitude and banks right 30° (which is the maximum it can turn this round).
At this point, the dragon and couatl are at same the altitude but about 270 feet apart so no attacks are exchanged.
Roll for initiative!
Note above that after two sets of alternating moves the couatl has expended ALL available movement by ascending vertically like a rocket and is now at +8 altitude. IF they were at same altitude, the dragon (climbing as hard as it can) would be within 90' breath weapon range (I don't know why a gold dragon and a couatl are fighting----feel free to narrate that part yourself). But alas, the dragon is only +4 (couatl +8 | dragon +4 = dragon is 120' below the couatl because, as we recall, each hex is 30').
If we finish out the round and stack another round on top of it, we can see that the dragon can bank another 30° and continue climbing to +7 (one of the notations is hidden under the couatl's token) while the couatl ascends vertically to +14.
The dragon is grumpy but smart enough to know that this is futile, and flies off at maximum speed. We can sum up what happened above like so:
Clearly Dragon vs Couatl is the worst sort of match-up but it illustrates HOW to plot AD&D's aerial system in a way that won't turn your brain to jelly.
I believe there is a very sound argument that goes like this:
If all participants are airborne, do not roll dice for initiative. Give it instead to the flier with the HIGHEST MV SPEED.
This method would always allow the fastest flier to decide who moves first over the course of the round and might help decouple the aerial combat round from a normal round, setting it apart in all participants' minds as a different kind of system.
I suppose you could concoct a hybrid method as well wherein each side rolls a die that is modified by either MC or MV to determine the winner (who then decides the order of movement).
I myself stick to the d6 even for aerial initiative with the caveat that MC A always wins. To adhere to procedure, ties in aerial initiative are always broken with a dice-off.
Remember: initiative serves NO purpose in aerial combat other than to determine order of movement. If at ANY POINT during movement, range allows for attack, resolve qualifying attacks simultaneously and then continue with movement.
Thus, if while moving pieces on the hex paper, there is EVER a moment when the pieces are positioned in such a way that arrows or wands or breath weapons or even flasks of burning oil COULD REACH the enemy, STOP and exchange missile fire simultaneously before continuing with the move.
Dropping stones or missiles on targets passing directly BELOW is certainly possible, but I would apply missile fire penalties as normal depending on range: -2 or -5 as listed.
What about physical blows, Anthony?
If, at any time, aerial combatants are in adjacent hexes, I allow for blows to be struck. I do this the same as ranged. That is, both combatants roll to hit simultaneously, deal damage and then the movement of flight carries them onward in their trajectory.
Adjacent hexes, Anthony? Don't you think 30' hexes are kind of far....?
No. That's SUPER close. The combatants (traveling at speed) could be ANYWHERE inside those adjacent hexes----even touching----especially given the size of many fliers. One small adjustment as the fliers pass each other is all they need to attempt a to-hit roll. If they miss, there you go. They pass one another in a very close call.
Now, if the fliers enter the exact same hex, you could conduct business the same way and roll to-hit [or] you could compose a House Rule that checks for unintentional full speed collision. A House Rule for unintended collision MIGHT look something like this:
You would add the chances of the two fliers together. So when a dragon enters the same hex as a couatl there is a 12% chance of unintentional collision (10 + 2) and the blame would rest mostly with the dragon. Whereas two air elementals battling in the same hex would have only a 4% chance of such an error. Two dragons vying in the same hex would be the worst case scenario: 20% chance of crashing into one another. How much damage collision does to each participant is likely a matter of size but you could make a House Rule chart like so:
Cross reference for both crash victims and voila, you have sound mechanical support for Gygax's assertion on DMG p51 for various large creatures acting as aerial battering rams. If the Impact is intentional, you could double the instigator's base chance. So a sprite trying to avoid collision retains a 4% chance, while a Dragon intentionally using its mass to ram into the sprite would double it's collision chance ( 10 x 2 = 20 ) for a 24% chance of impact. On collision, the sprite would take 1d20 dmg while the Dragon would take none. Don't forget the Crushing Blow for any hit in the 1d20 range! Obviously this is not by the book and is offered up as a way of extrapolating Gygax's own notions by relying on extant systems.
What about spell casting, Anthony?
No. Just no. Rules are clear. You CANNOT cast spells WHILE MOVING.
But Anthony, what about my 16th level mage on a flying carpet. He can just stand there. Technically he's not moving.
What do you mean "technically" he's not moving?! He's MOVING. And he has a face full of wind! His arms are getting buffeted about! He must hold on, or balance himself. Remember my last post? Say Yes to the player? Forget that. NO. No, no, no.
BUT, if you want to be kind, like I ALWAYS am. You could say that spells of 1 segment casting time or less MIGHT be fired by flying casters with a chance of Spell Failure thus: "Roll your caster's level or less on a d12 or the spell fails!" I picked a d12 here because it correlates exactly with the benchmark for magic resistance (11th level) which assumes that casters above 11th level are exceptionally powerful and skilled. Ergo, 12th level casters could fire off 1 segment spells at passing aerial targets with 0% chance of failure.
In this wise, you inject a possibility for 1 segment spells to fire at moments of opportunity (when range allows) during the movement of tokens during the round. Note that such spells would still need to be declared prior to initiative and if range never allowed them to launch, they would be wasted. Further, any hit upon the caster prior to casting would still spoil the spell that round even if it happened BEFORE the 1 segment casting had begun. Keep in mind that spells are always spoiled if the caster is hit BEFORE the spell is cast during the round declared.
You say: "Woo-Hoo! I'm ready to do aerial combat!"
Ok, but here is the salient bit. When your PCs are flying through the mountains on their flying carpet and you roll a dragon, you don't actually need to get out the hex paper UNLESS the PCs WANT to fight it. Because at similar speeds (and carpets go min 24") even one MC class difference means evasion is almost certain. Thus, you are free to query the party.
"Hey Party. Do you want to fight this dragon that's approaching?"
If they choose to evade, you simply say that after executing a few turns and climbs, the dragon seems to grow disinterested and is left far behind.
Thankfully, Players will usually have this choice because:
Broom of Flying: MC: C @ 30" speed
Carpet: similar to broom (though tbh I give Carpets MC: B because I believe carpets should be able to hover).
Fly Spell: MC: B @ 12" speed
Wings of Flying: MC: C @ 32-ish
When you look at Speed and MC together, Player Character options are all on par or significantly better than most flying monsters.
Remember that speed is a huge part of the equation.
If neither flier is MC A (player characters are never MC A) and if the MCs of the potential combatants are similar...AND if one of the fliers is slower, break out the hex paper. Shorter version: if the aggressor is much faster and MCs are close, there is a decent chance of aerial combat.
Look at this travesty of aerial warfare. Much thanks to Black Blade for the hex paper though, which allows us to appreciate Gold Dragon vs Pit Fiend! And yes, Allan, I realize now that I should be using the OTHER side with the SMALLER hexes for more real estate.
In the combat map above, the dark dots represent end of round. Therefore:
Round 1: Pit Fiend climbed and banked as hard as it could in 5 hexes (60°) but since you need to fly 3 hexes to gain 1 altitude, it only manages +1. Meanwhile, the dragon climbs to +3 and only banks near round's end to adjust course for the Fiend.
Round 2: Pit Fiend goes into a dive (1 for 1 hex) and drops to relative altitude -4. The dragon remains level in its first leg, then goes into a dive and banks 30°. Its tremendous speed allows it to end the round in the hex adjacent and directly behind the Pit Fiend at identical altitude (-4). As the dragon comes within range, I would of course allow it to breathe AND at end of round, when they meet, I would allow for the dragon to reach out and attack the Fiend from behind.
So, as you can see, even though the Pit Fiend is one MC better than the dragon, the dragon is able to intercept by virtue of having double the Fiend's MV speed.
I've already covered aerial melee and the way I would do it. You can follow Gygax's advice in DMG for the breakdown on each monster OR you can just roll with whatever makes sense...which is, in fact, what Gygax is telling you to do by giving you sample examples.
I myself would allow for missles, wands, breath weapons, etc. in the lead up to contact and then simultaneous exchange of melee attacks at the moment of contact with the Dragon gaining +2 to hit from behind AND x2 dmg from the dive.
Of course, the Pit Fiend could also Teleport out prior to engagement or after the exchange assuming it survived. The Fiend's membranous wings mean he's likely to be forced down or even plummet, so Teleport is really his best choice.
Note that I would apply none of the penalties for aerial combat collated above to either of these creatures as they are natural fliers. Penalties for missiles and melee are for Player Characters bold enough to tackle denizens that call the firmament home.
A battle with an Air Elemental whilst standing on a Flying Carpet is almost certain to end badly. The elemental will track directly to the carpet and, I would rule, attempt to shred it. (I'd likely fiat a Cloth vs Acid save b/c I think the Elemental can shred cloth with about the same power. Applying a +2 (because the carpet is magical) means the carpet would need a 10 or higher to avoid destruction).
Thus, once again, Player Characters with an ounce of sanity will try to land immediately or possibly even jump off if they have access to feather fall and the like.
The short of it, is that aerial combat is, well... short indeed and does not fit into the standard combat round. It is rather about movement and position and, as I run it, simultaneous exchanges of missiles, magic or blows when range allows.
I strongly recommend using the relative altitude system of +/-- n, as it is simple and can be tracked directly on the game surface.
If the combat is taking place very high, you can of course impose a hard ceiling. Perhaps no fliers can exceed +20 for example if the fliers are already in the stratosphere.
Likewise, if you have floating islands you can notate their tops (and thickness) with a simple range (+2 to --2 for example) while pinnacles of rock rising from the ground might be solid at --4 and lower. This allows you to very simply make abstraction of elevation for quick calculations while availing yourself of real numbers (each increment represents 30 real feet) for purposes of range and falling damage.
Using a scale of 1 hex to 30' is the key to this level of simplicity and I would strongly advise against detracting from it. Nevertheless, to each their own.
Obviously, in large open dungeon spaces, such as Geir Loe Cyn-crul or the Underdark, you may have maps on a grid set to 10' increments that preclude easily switching to 30' hex paper. In these cases, I generally make due and the aerial combat is more akin to flapping leaps or bounds that give the fliers mobility advantage.
I pay no heed whatsoever to Gygax's admonition about acceleration. MC: A accelerates to full speed in 1 segment; whilst MC: E requires 4 rounds to go from zero to full. He likewise mentions braking distance for fliers that can hover. This is a fun mental image and you can use it (I suppose) to help you ballpark whether evasion is likely, but IMO it is added baggage that I doubt Gary ever thought of again...let alone actually used.
Sometimes, I feel guilty putting words in his mouth or making assumptions, but I was able to play a game with Ernie at Gary Con IX so I think I have a vague inkling of how those guys ran.
My guess is that I adhere FAR more rigorously to "holy writ" than they ever saw reason to. I'm intentionally doing something different.
My whole shtick is finding a way to adhere to the books while also keeping the pace fast and simple.
Still, poor magic-users. The rules hate them. Even with my lenient method of counting casting time from beginning of round, they get interrupted quite frequently. Which is why I propose (for aerial combat) the 1 segment spell allowance. Allowing them to squeeze off only the simplest of spells during flight gives them a modicum of self-reliance outside of magical devices. In AD&D, fighters are KING by default. In my opinion, you have to throw casters a bone now and again, sometimes through encounter design; sometimes with fiat or house rulings.
That's about it. Like many other aspects of AD&D, aerial combat is a bit loose and fast, open to in-the-moment interpretation. I think the outline provided for breaking up movement and resolving combat along its path (as well as suggestions for the role initiative can play in this process) go a long way toward making aerial combat feasible while adhering to the text.
and happy gaming.