Resolving a combat round.

[rapanui]

There's nothing more basic to a combat system than resolving the combat round. That is, take input from the players regarding the actions of their characters, figure out what the NPCs would be doing and figure out how the round turned out.

Most systems out there use something like this:

Start Round
Initiative Loop
{
Movement
Actions
}
Back to Initiative Loop for next character
End Round


The problems with this are apparent to anyone that has thought about it for half a second. If a round is 6 seconds, that means one character gets to take his full 6 seconds worth of actions before the next character on the initiative queue gets to do anything.

That means a character can kill another character and not get hit back, even though theoretically both creatures were acting simultaneously.

Furthermore, movement and position is incredibly poorly resolved. Suppose you have two warrior types at 60 feet from each other. Fighter 1 gets initiative, charges and takes a swing. Fighter 2 takes a full attack action (this is obviously something akin to D&D 3e).

Now, that's all well and good, but what if Fighter 2 had wanted to charge too? He can't, even though it isn't too hard to imagine that they both simultaneously charged forward and clashed somewhere in the middle of their starting points (maybe a bit more towards F2's starting point, since he was a bit slower in initiative).

OK, now that I've elaborated what I perceive to be a problem, I have a question:

Is there any way to resolve this problem caused by the level of abstraction in the rules, and not turn the RPG fight into a difficult to run wargame-style simulation?

I've though about it some, and I have few possible solutions, but none of them satisfy me very much. So, suggestions would be appreciated.

[Murtak]

I can't see a simple solution to this. The obvious approaches are:

Split up rounds into many segments and let everyone take an action before the first player has taken his second action. If you break this sort of system down far enough it handles situations like your simultaneous charge nicely. Of course it also bogs down each turn immensely.

Or you can have everyone announce their actions at the start of their turn and then set up special rules for all the special cases you want covered. This keeps the turn relatively short, but creates a crapton of extra rules.

You could also go for some sort of hybrid approach where everyone announces their actions for the turn and then you resolve the actions normally except for when you have conflicting actions (like to simultaneous charge). In those cases you slow down to split-actions for those characters only. That gives you a sort of rules-light system that does not bog down much. Unfortunately it also has the potential to be very very confusing. What constitutes a conflicting action? Do non-conflicting actions happen before or after the conflicting actions and why? What happens when some sort of contigency effect or readied action retroactively turns a non-conflicting action into a conflicting one?

I could see the last approach working if you designed a game with this system in mind, but even in that case it would restrict what kind of abilities you can put into the system. Retroactively changing a system would be right out I think. The former two approaches won't work at all I would wager.

[Aycarus]

Probably the fact that this problem is so clearly exasperated is due to the granularity of the initiative system; ie. there are at least 25 divisions between the first person's 6 seconds and the last person's. Conflicts do occur in D&D, but they're rare because it requires two characters to have the same initiative out of the 25 possible options. In most of these situations the DM either makes something up or he simply goes with the character with the highest dexterity.

If you want to make things seem more simultaneous, what I'd propose is to cut down on the granularity a bit and reduce it to 5 or 6 possible times. That way it's likely that two characters will have conflicting actions but not for the entire group, and so you don't have to deal with writing down *everybody's* actions before a combat round. When there are conflicts, hear from the 2 or 3 people who are acting at that point and decide how things turn out.

[rapanui]

OK, at least I'm glad to see that I'm not the only one who sees an issue.

Murtak, your proposed solutions are very in line with my previous thinking on the subject. I outright rejected the second option you mentioned, because it would be too hard to think of every situation in which simultaneous actions need to be resolved.

The third option you mention is actually something I had thought of, but couldn't actually build into a coherent ruleset (either I'm not smart enough, or it's just a tough problem). Besides, if someone specializes in a kind of combat that tends naturally to necesitate constant slowing down to figure out what exactly is happening, then that person's actions monopolize the game. Not fun.

So, we're stuck with option #1. Break down the round. And that's where Aycarus's comment proves interesting: how high is it necessary to increase combat resolution? Is it necessary to know a character's position to the inch? To the foot? To the nearest 5 feet?

The more specific you get, the more 'realism' you have, and the more emphasis on tactics, but the more bogged down the combat round gets, particularly with more than 2 characters.

I guess it boils down to where on the wargame-roleplaying continuum I want my new system to fall.

Any more ideas would be appreciated, but good stuff so far.

(Oh, and Aycarus, perhaps I presume too much, but I believe the word you're looking for in your first sentence is 'exacerbated', as in "to make things worse")

[Josh_Kablack]

Honestly, in practice "everyone annonounces, then everyone resolves in order" iniative systems nearly double the execution time of a round. And in 3.x D&D that is more unacceptable to me than the apparent lack of realism in caused by compacting continous actions into discrete points.

My only suggestion would be to limit the available initiative positions to a fixed number (something like +25 to -5) and then dividing the combat round into that number (in this case 30) of segments or sub-rounds. When a character's initiative comes up they may BEGIN taking actions, but each action takes a number of segments to execute, and there iniative pin gets reset to that number of segments after the current one.

For example a Full Round Action now takes 30 segments, a Move Action takes 15 (and could occur with 1/15th of the move taking place each segment), a Standard Action takes 15, and each attack in a full attack takes 30/number_of_attacks.

This sounds complicated, but in practice I've seen it work better than any "all declare, then act in initiative order"

[Modesitt]

Is there any way to resolve this problem caused by the level of abstraction in the rules, and not turn the RPG fight into a difficult to run wargame-style simulation?

Maybe. There's some options available to muck around with initiative and the combat round for more realism without over complicating things. Before I get into that, I'm going to point out that a lot of the simultaneity in D&D comes in form of attacks of opportunity and readied actions. You can get a lot of what you're asking for with liberal use of those.

Steal a copy of Spycraft d20 from someone and read it. It's Good Stuff. Here are the things that apply to you and fixing up the init system:

1. They have removed the distinction between move-equivalent and standard actions. You just get two half-actions. This would make subdividing the round up easier on you if you opted for that, as Murtak suggested higher up.

2. This has little to do with what you asked, but you may end up implementing it. Every class in Spycraft gets an Initiative Bonus, which is just like BAB or armor class bonuses under d20 modern. I'd suggest it be implemented in any system where initiative order plays a large role in combat success.

----

Other ideas -

Don't do what D&D does and start from the top, start from the bottom. Start on the lowest init count and work your way up. A higher-init character may interrupt a lower-init characters action and take his own at any time. This means combat is less "We go down the little list one at a time" than "We go up a list...But it can change at any time!" You might add some extra restrictions, such as you can only take a half action when interrupting or something like that.

Or think M:tG. Combat involve both characters simultaneously dealing damage or getting a chance to do so. You could either have it be "If you attack him, he also attacks you" or you could do something like Shadowrun 3Es melee system and make it an opposed test.

To expand on the above, maybe you could only counter-attack a ranged attacker if you have a ranged weapon, only able to counter-attack a melee attacker if you also wield a melee weapon, you could only counter-attack against a spell if you also wielded a spell, etc.

[Murtak]

So, we're stuck with option #1. Break down the round. And that's where Aycarus's comment proves interesting: how high is it necessary to increase combat resolution? Is it necessary to know a character's position to the inch? To the foot? To the nearest 5 feet?

To solve your example problem (simultaneous charge) you would need to break movement down to 5 foot increments. Well, infinitely small really, but I am going to pretend that no creature has a facing of less than 5x5 feet. Of course at that point we are already deep in "too clumsy for fast gameplay" territory.



Modesitt:
While the reactive system you propose is not inherently worse than say, DnD, I can't see it helping Rapanui's main objective (making combat simultaneous). Sure, higher initative characters can interrupt slower characters at any time, but that merely changes the action order from A->B to B->A.

Opposed checks, while solving the original issue, introduce a new problem: It is no longer beneficial to initiate combat and thus initiative becomes meaningless. Additionally depending on how the opposed check works fighting back against a superior opponent can become detrimental.

[rapanui]

OK, I now have an idea for a system. It uses both reverse action declaration with respect to initiative, and breaks the round down into 12 smaller subsections (yes, this system resolves to the nearest 5 feet). I need to streamline it so that it doesn't bog down though.

[Modesitt]

Murtak: Yes, the action interrupt wouldn't help much unless combined with a broken down combat round. Example, you can only take a half-action when you interrupt and your second half action later in the round.

Your post is on the money about the disadvantages of an opposed test, except for the part about it not being advantageous to initiate combat.

Shadowrun uses an opposed melee test system, so I'll use it as an example. In Shadowrun, there are three levels of melee combatant. People with little or none melee combat skill, people with some skill, and adepts. People with some skill beat up on people with little or none and sparred against each other. Dedicated adepts raped everyone in melee except each other.

There was no mechanical advantage to initiating combat except that ties went to you. On the surface, there looked to be little advantage to a melee-centric character having a high initiative(In Shadowrun a high initiative gets you extra actions). The thing is, only an idiot would initiate melee combat with a blatantly superior opponent. You needed a high initiative in order to force people into melee combat with you.

Rapanui: Is there some reason you picked the number 12 and not 20? Or 40? If I was intent on breaking the game down more, I'd probably first try to implement something like M:tG's rounds. You could have a Movement Round(Or two, maybe one before attacks and one after), an Attack Round, etc. Although if you're willing to have people declare their actions five feet at a time, I don't think that's going to be quite enough.

[PhoneLobster]

To put it simply I've used a system that went a bit like this.

Order of initiative is resolved.

Everything is considered to happen simulataneously.

Slowest characters declare (and resolve) their actions first. (yes it sounds crazy).

A resolved action, even if it kills you, does not prevent a simultaneous action of another character from being resolved.

However. An action CAN cause other potential actions to become valid (moving for instance placing you in range of melee attacks of characters who were faster than you and thus read your intended move).

Thus the advantaged gained by higher initiative was an increase in options for your single action as you knew what slower characters were doing, and even the outcomes of their actions.

The only major difficulty I encountered at the time (it was a simple system that saw little actual use) was dealing with characters tackling and holding other characters. To which I decided a hold only invalidated movement actions if it existed at the comencement of a combat round and that in order to tackle or intercept a character you simply had to move with them.

I imagine similar actions such as summoning a brick wall dividing a field could be similarly resolved (the wall doesn't prevent characters moving in that same round but does from the start of the next). This fits in nicely with the speed characters advantage as characters slower than the wall summoning only end up on the right side by chance (having resolved their actions before knowing about the wall appearing) and faster characters have the option of using their action that round to leap onto the correct side by choice.

Oddly in order to use this simultaneous system no two characters actually can have equal initiative or "go" at the same time.

My next system will probably use something similar.

[RandomCasualty]

I do share Josh's opinion that "declare first" combat systems are very slow and add a lot of complexity to the game.

Adding segments can also get unbearably slow as well. Basically it amounts to spreading one round's actions over several smaller rounds. If you do this enough, you'll eventually get a total simulation of real time combat, however only a computer is actually capable of doing all that with enough speed to make it workable.

I really can't find any really workable solution to this problem. I think it's probably best to worry about balancing the actual mechanics as opposed to trying to change the combat round structure itself. For instance, it's probably easier to balance full attack with charge than to try to work out some way for two people to simultaneously charge.

Basically you'd probably have to create a "movement phase" where everyone moves. The high initiative moves 5', the next highest moves 5' and so on. Then the highest moves another 5', the next highest moves another 5', etc. As one might imagine, this would be painfully slow, but it's really all you can do if you want to try to set up a charge simulation where both parties can charge each other, as opposed to one guy standing still while the faster one goes 30' and attacks.

Mostly it's a speed tradeoff, and really when you're talking about combat systems, having something that resolves quickly is very important.

[PhoneLobster]

People are too hung up on this two simultaneous charges thing.

Most sane time management systems handle time in a vague and fuzzy kind of way that is more relative to resolution of game actions rather than any real measurement of any kind.

So why the hell do you need to know the exact position of any character down to a teeny five foot square at any exact slice in this fuzzy not sure when it is time measurement system? Even at the beginning or end of rounds.

I suggest that instead of a finer granularity solving simultaneous movement tracking a more chunky granularity solves it much more nicely and in a way that can be resolved quickly and smoothly.

If square size (or something similar) were more like 50 foot or something like that then simultaneous charges would be easy to adjudicate.

[rapanui]

There are many reasons tactical realism is desirable, but it depends on what kind of game you want to play. The system I'm designing does resemble a wargame more than a traditional RPG.

In essence, I'm trying to translate something like "Freedom Force" (a great little PC game) to tabletop gaming. Only Fantasy and not comic book superheroes.

And yes, the system I've been testing recently runs a tad slow for my liking even with only 4 creatures on the board. It's a tough little puzzle, but I'll keep working at it until I find a decent workaround.

[PhoneLobster]

Tactical Realism and square size are not intimately connected.

Big squares help the simultaneous charge beusiness because the answer to the questions "where am I when I start charging?" "Where am I while I am charging" and "where am I when I finish charging" are much more likely to be the same answer or at least span fewer squares.

But what do you pay in "tactical realism" in exchange for a big square?

Character relevant positioning? Well most moves seem to be "I move near a target, I move away from a threat or I turn and run like hell" and big squares can handle that fine.

Weapon range and reach rules? Large squares don't have a big impact on most missile weaponry range granularity unless they are REALLY large, so whats being effected? "Reach" weapon type rules for pseudo ranged attacks that reach less than the range of a large square but more than the range of a standard melee weapon. Personally I for one don't care. Melee range == same square. Its easy and fast and good, I don't want to see Melee range == somewhere between 0 to 4 squares (or more) depending on size and weapon type as under 3.x after all what have the 3.x reach rules ever done for us? Anyway I'm sure you could come up with some OTHER reach mechanic not involving squares and positioning that could give peoples extra long dagger on a stick attacks that special little "flavour" of their own.

Taking cover behind stuff? Under a small square based system cover is still highly relative to various positions of characters and obstacles and often has little to do with actual squares or game measurement units of any kind. So if you can be in one of the 7.35 five foot squares that give you cover from the guy shooting at you from a particular direction then you can also be in the part or section of a 50 foot square that provides you the same cover in the same circumstance.

In summary. Big squares != lack of tactical realism. Think about it, its a good idea. Honest.

[Murtak]

But what do you pay in "tactical realism" in exchange for a big square?

You pay with your minimum areas going up in size. Off the top of my head that affects area attacks (duh) - both magical and mundane, the maximum amount of characters being able to attack a single target and the effectiveness of bodayguards.

[PhoneLobster]

The minimum area for an attack is one character, not one square. For those area attacks that can't handle the scale of everything in X squares then they just have to switch scales to X characters in Y squares.

[Murtak]

Let me explain. When you increase your square size your game gets more fuzzy as far as exactly locating creatures is concerned. For some combat maneuvres this does not matter at all. But any action that depends on locations - such as fireball, whirlwind attack, blocking movement lines or a simple attack action gets more fuzzy as well.

In the case of area attacks increasing your basic square from 5 to 50 feet means your system can not handle attacks with a range of, say, 30 feet.
In the case of a bodyguard trying to interpose himself between a character and some random monster it means you either create special bodyguard rules or you simply won't be able to physically protect characters.
And in the case of multiple creatures ganging up on someone it means you can now have more creatures gang up (again, unless you create special rules).

I am sure there are some issues I missed. I am also pretty sure it is possible to work out a decent system with bigger base squares (side note: why squares? Hex bases are soo much better ...). But it is not simply an improvement to do so - it's a tradeoff. Speaking for myself, setting the base square size to "one normal character" seems both intuitive and a decent compromise of precision and ease of handling.

[PhoneLobster]

Which burning hands spell is more fuzzy?

One that targets an arbitrary small area (which is not so easily convertable to the small scale five foot squares in 3.x) which may or may not have 1 or more targets in the area.

Or one that targets a fixed maximum number of targets in a single square.

The same for whirlwind attack, fire ball the lot.

Fuzzy here is a pretty arbitrary term. I mean look at d20, small squares, but then ranges measured in cones and arcs circles and spheres. THAT is fuzzy. Spells that are supposed to be of similar power and effectiveness using vastly differing areas that could effect anywhere between 0 and umpteen targets. That is fuzzy.

Seriously. Big squares are not fuzzy. They don't even prevent you from having a specific position within a square. The functional difference between "in square 9-R behind the dumpster diagonally sprawled across a dozen squares" and "behind the dumpster in square 1-A" is simply that one is quicker than the other.

And yes indeed, why squares? But whatever, why not have a completely grid free system based on points, radius and ranges. There are plenty of other ways of doing things there, I'd go with squares (or better yet abstract extra "fuzzy" areas of no fixed shape), but you know, whatever.

[Murtak]

None of these spells is more fuzzy, since your version of the spell is not even an area attack anymore. And yes, mapping a circle to squares is fuzzy. Mapping a circle to even bigger squares is even more fuzzy. That was my whole point to start with.

And yes, area effects work better with more targets in the area. Again, that is the whole point of attacks targeting an area (as opposed to a single target or a number of targets).

By the way, when you let characters take up specific and discernable locations within a square, then how the heck can you claim that is your square size? That is the whole point of having a square size to begin with - it determines the locations of those in it. Squares are the smallest used measuring units. And if you can move around within a square then they are, by definition, not your smallest measuring unit.

[User3]

Phonelobster, it sounds like what you're talking about with Burning Hands is a spell which has a certain number of targets and a fixed range. There are already spells like that.

Is also seems to me that is would be simpler to talk about 'spheres of influence,' as opposed to "squares."

If all you want is an abstact system where characters are either in or out of the range of an effect, that is...

I mean, if you draw the squares on the ground my character could be 3 inches away from yours and still be out of his "square" and thus out of reach.

[User3]

I missed the last bit of your post. Boy am I dumb

[PhoneLobster]

And if you can move around within a square then they are, by definition, not your smallest measuring unit.

Thats what I'm trying to say.

The smallest measuring unit as I would propose it is not the square but the character, and to some degree the state that they are in.

d20 for instance already does this in a variety of ways, cover being the prime example, characters take cover as a state they enter relative to other characters and NOT relative to squares.

You can do the same in big squares.

As for saying "gee there are already spells which affect X targets in Y squares and that is NOT an area of effect..."

Yes there are spells that work like that in d20 (magic missile for instance). But the mechanic could (and I believe SHOULD be used to determine area of effect results. I also believe it is a superior more uniform mechanic and one more measurable and reliable for the effectiveness of a power when seeking to balance it against other abilities.

Of course in the end area of effect as currently used, and it appears as supported by you works in EXACTLY the same way. It has a fixed number of squares for potential targets (its area) and a fixed maximum number of targets (the most that can fit in those squares) the ONLY difference is the area is much more fiddly and the potential variance in number of tagets is much more unpredictable and unballanced.

What you are basically telling me is that bigger squares means that I'm forced to use the same mechanic you already use anyway in a manner that is BOTH quicker to administer and has a less wildly variable outcome based on circumstance and therefore using larger squares is less tactically accurate.

That is just plain incorrect.

Is also seems to me that is would be simpler to talk about 'spheres of influence,' as opposed to "squares."

(I note you noted that you missed some vague mention of this, but anyway)

What I'm considering using next time round for this isn't really spheres of influence, because they can overlap in strange ways. I WOULD rather have a situation with people in the same square 40 feet away effecting each other but differing squares 2 feet away cannot. We do it already with ranges and reach and such in d20 its not a big deal.

But I would like to make it seem less obvious by using a more vaguely defined non square based set up. Most RPG "boards" or maps or whatever look like a chess board, all squares and tiles etc... I think I might go for something more like the board from diplomacy. So for instance a room with a chasm spanned by a small rickety bridge might have 4 areas, this side of the chasm, the far side of the chasm, the chasm and the bridge. It seems a VERY functional way of dividing a combat area like that in the ways that actually matter.

Also note the influence of an interest in simultaneous resolution there, diplomacy after all is one of the most infamous simultaneous resolution games out there.

[Murtak]

The smallest measuring unit as I would propose it is not the square but the character, and to some degree the state that they are in.

d20 for instance already does this in a variety of ways, cover being the prime example, characters take cover as a state they enter relative to other characters and NOT relative to squares.

Actually cover is a function of squares. You draw a line from your square to your target's square and if there is anything solid in between the target has cover.

You can do the same in big squares.

Sure you can. However you proposed adjucating cover by using sublocations like "behind the dumpster in square 1-A". That is fundamentally opposed to the basic idea of your location being determined by the square you are in.

As for saying "gee there are already spells which affect X targets in Y squares and that is NOT an area of effect..."

Yes there are spells that work like that in d20 (magic missile for instance). But the mechanic could (and I believe SHOULD be used to determine area of effect results. I also believe it is a superior more uniform mechanic and one more measurable and reliable for the effectiveness of a power when seeking to balance it against other abilities.

So you think when the ceiling collapses in a cave it should only bury a select number of targets and not effect others at all, even though they are in the same location?

What you are basically telling me is that bigger squares means that I'm forced to use the same mechanic you already use anyway in a manner that is BOTH quicker to administer and has a less wildly variable outcome based on circumstance and therefore using larger squares is less tactically accurate.

That is just plain incorrect.

Huh? I don't get it. What are you trying to say here?

But I would like to make it seem less obvious by using a more vaguely defined non square based set up. Most RPG "boards" or maps or whatever look like a chess board, all squares and tiles etc... I think I might go for something more like the board from diplomacy. So for instance a room with a chasm spanned by a small rickety bridge might have 4 areas, this side of the chasm, the far side of the chasm, the chasm and the bridge. It seems a VERY functional way of dividing a combat area like that in the ways that actually matter.

Ok, from what you wrote you actually want not big squares but infinitely small squares to determine character location (read: exact locations). Then on top of that you want to layer dynamic locations - which basically amounts to collapsing arbitrary areas into a single location for combat purposes.

One some levels I like that idea. It works great for area attacks for example. No more "can I fit x and y creatures into my fireball?" - you just fireball a single area. It also neatly solves the classic DnD problem of the moving fighter only getting a single attack while the fighter with the lower initiative gets a full attack.

I also see problems though. Firstly, exact locations - those don't work too well unless you have an exact map and to-scale minitures for everyone involved in the fight. That is not a big deal though as the rest of what you propose works just fine with, say, 5-foot squares.

Your combat area system also has all the problems associated with big squares however. You can not have area effects not measured in whole combat areas (which may be fine, depending on how big you make them). You will have very deadly combat as everyone within one area can and will gang up on a single opponent. And you make guarding someone from physical attacks near impossible.

As to whether the benfits of such a system outweigh the downsides, well, that is a matter of personal preference.



Edit: me spellz guud

[PhoneLobster]

Ok, from what you wrote you actually want not big squares but infinitely small squares to determine character location (read: exact locations).

Utterly wrong. I'll explain clearly for you on a new thread as this isn't enough about simultaneous stuff no more.

[Murtak]

I think I may have a somewhat workable solution to the original issue. Two problems were to be solved:
- The DnD style of handling your entire turn in one go leads to weird situations.
- Actions which should be really simultaneous don't work at all if one action invalidates the other (such as the simultaneous charge).
An additional constraint was that the solution should not bog down gameplay too much.


Ok, the solution to the first issue is simple and obvious. Divide up the turn into smaller actions and have characters take turns taking these actions. If you take care not to go overboard this also does not bog down combat too much.

Example:
At the start of combat (or at the start of each turn) initiative is determined - rolling, from a fixed stat, whatever. You start with the highest initive and go down. Each action a character takes uses up a certain amount of initiative and the character initiative for the turn is lowered appropriately. Keep going until all characters have an initiative of zero.

So we have a fight between Joe fighter and Bob warrior. Joe has an initiative of 10, while Bob only has 7. Joe starts and uses 2 points of initiative on moving towards Bob, bringing his total down to 8. He still has a higher initiative so he gets to ake another action, moving again and ending up next to Bob (and bringing his initiative down to 6). Now Bob gets to act. He uses 5 points of initiative to take an attack against Joe. Then Joe uses 5 Points to smack Bob back. Since none of the two has anything useful they want to do with their leftover initiative the round ends.



Now, the trick is to not break down the turn too much here, but I think it is possible to work out something that still allows for fast-paced combat, while getting rid of the standard DnD problem described above. That leaves the problem of handling simultaneous actions. I propose handling these as reactions. That means, under certain circumstances (the trigger) it is possible to take actions while or directly before someone else is acting. This action the substracts it's cost from your initiative total as normal.

Example:
In our above example Joe does not want to get hit by Bob, so when attacked he elects to actively dodge Bob. He lowers his initiative by 3 and adds some fixed value to his AC in return. Since that lowers his initative to 3 he does not get to attack back this turn (since attacks cost 5 points).



Hmm, now that I think about it this sounds exactly like the Feng-Shui system. Yet another reason to convince my friends to play that game some time.