Need Help Designing a Combat System

[wkwkwkwk1]

Hey there, everyone!

First off, please note that this is a rather long post. I'll try to be brief, but it's not easy. If you can't, or don't want to read a long post, skip to the last few paragraphs.

So, as a way of flexing my creative muscles, I'm designing a tabletop RPG. Set on the aftermath of a zombie apocalypse, it's meant to be realistic (as in, the best decision in-game would be the best decision outside the game; also, combat should be about as deadly as one would expect), but not overly simulationist (tends to result in failure).

I've been able to overcome most problems the design has thrown at me, except for this tiny little problem: combat rounds and time management. I will now talk about the different methods (that I know of) for doing this, and the advantages and disadvantages I find for each of them.

Roud Length

Second-long rounds: Used by GURPS, this system sees each second as a round. This is actually a modified "Action by Action" system in disguise, because, generally speaking, each character is allowed one action per round (except for longer actions, naturally).

Pros:
- More granular than longer rounds
- Allows everyone to act more often (in terms of in-game time) than in longer rounds, where large segments are resolved at the same time

Cons:
- Is it really more granular? Attacking with a knife takes as much time as attacking with a sledgehammer. Some weapons, after being used, have to be "readied" (1-turn action) to be used again, but this strikes me as extremely binary.


Rounds longer than one second: We all know this system; if nothing else, then from D&D.

Pros:
- The longer time frame allows for the differentiation between slow attacks and fast attacks, and for a difference in speed between the combatants.

Cons:
- Gives you the impression of standing still while the enemy acts. Characters can't react to the environment around them. Can obviously be justified, but it is unsatisfactory from a player perspective.

Action-long rounds: One action per combatant per turn. Attacking and scaling a wall takes the same "time". The only narrativistic system.

Pros:
- For my current purposes, not enough

Cons:
- No granularity to speak of.

Actions within a round

"Set" of actions: The system D&D uses. In this case, each character can take, per turn, the following actions: 1*Standard; 1*Move; 1*Swift/Immediate; N*Free. Or a single Full-Round action.

Pros:
- Simplicity. No player overshadows the other in terms of actions per round. Unless you're a Wizard :smalltongue:

Cons:
- Not enough granularity. No setting apart of fast and slow attacks. Very abstracted.

Action Points: Each player has a given number of points per round, that they spend to perform actions. There are actually two possible systems for modelling the speeds of different characters. Either the number of APs increases with the increase of speed, or the AP cost for each action decreases with the increase of speed.

Pros:
- Much more granularity.

Cons:
- Takes longer to resolve each round

The problems with the "AP Cost Decrease" system: Let's say each character has 12 AP per round to spend. It's a good number, divisible by 2, 3, 4 and 6. The problem is that the reduction of AP cost yelds massively different results:
- 7-12 AP: 1/round
- 5-6: 2/round
- 4: 3/round
- 3: 4/round
- 2: 6/round
- 1: 12/round

It gets more noticeable as the cost nears zero, much like a hyperbolic funcion.

And that's it. I find that 6-second long rounds strike a nice balance between too short and too long. Interms of actions within a round, I I'll be using the Action Points system offers much more granularity and many more tactical options. But now, a few other problems pop up.

The problem with long rounds still stands: each character's turn takes place independently, as if each character were moving on their own agenda, allowing someone to chase them while they stood still, for example. There are many rationalizations behind this, but none remove the actual problem. If someone wants to react to another's actions, they'll have to predict them a full round ahead.

My current solution for this is to let the players spend only a fraction of their APs each time they act, with a lower limit of one action. This has the problem of increasing book-keeping and fiddlyness by quite a bit. I'm thinking of maybe tracking it with beans or something.

Another possible solution would be to reduce the round length to 3 seconds / 6 AP, and allow the players to "borrow" or AP from the next round, or "save" AP, with an upper limit of half their AP per round. The problem this raises is that characters with a +X to "Speed" in a 12-AP round now either become much more powerful, effectively doubling their bonus, or apply only half their bonus per 6-AP round, which makes it even more fiddly with odd numbers.

Now, either way, this has the problem of comparing time-frames: when does each action take place? The best less bad solution I've come up with is having some sort of comparison table with several same-sized column split into a number of lines equal to the possible AP totals, such as this (where the middle column is for 12 AP; each blue line is a second):

http://i.imgur.com/6qltPkM.png

Problem: Not terribly intuitive. In my opinion.


So, what I ask to you Playgrounders is:
- Do you know of any other system besides the ones I mentioned above that would fit this sort of game?
- What do you think of the system I've developed? What would you change?

[manny2510]

It may be easier to keep movement separate from action economy, or have actions halve movement each use in a turn, becoming less mobile as you do things but separating it from action points so players can say " I will attack attack and shift back a bit" since in a zombie setting the actions of fight and flight should hurt each other.

[wkwkwkwk1]

It may be easier to keep movement separate from action economy, or have actions halve movement each use in a turn, becoming less mobile as you do things but separating it from action points so players can say " I will attack attack and shift back a bit" since in a zombie setting the actions of fight and flight should hurt each other.

Hence my choice of using an Action Point system: want to commit to a full attack? Good luck getting out of there alive? Best choice is to keep attacking and retreating, or just lunging forward to attack. Therefore, both options are available during the same time frame, and both hurt each other.

Is that what you meant? :smallsmile:

[Mr.Moron]

What about one of those initiative/time consumption system type systems?


You have a printed track of initiative values and participants are represented by tokens or miniatures on the track. At the start of an encounter you use some stat to give people their starting positions.

So say we have 4 parcipants: Bob, Steve, Mary and Joe they'd start at their initiative values

10 (Bob) <-
9
8 (Mary)
7 (Steve)
6
5
4 (Joe)
3
2
1
0

Starting at the top, bob is up first. Bob attempts to start up the truck up his truck, with a time value of (7) he moves down to initiative 3.

10
9
8 (Mary) <-
7 (Steve)
6
5
4 (Joe)
3 (Bob)
2
1
0

Mary is up next, and takes a penalty to her observation roll to make quick look around, with at ime value of (3) moving her down to 5.

10
9
8
7 (Steve) <-
6
5 (Mary)
4 (Joe)
3 (Bob)
2
1
0

Steve is up next and reloads his gun, with time value of 2 putting him at 5

10
9
8
7
6
5 (Mary) (Steve) <-
4 (Joe)
3 (Bob)
2
1
0

Mary is up again, since her and steve are at the same value but she was there "First" she runs for truck with a time value of (6), putting her back up at the top with 10 wrapping around. Steve does the same, but since he's quicker on his feet this only a 4 value action for him putting him at 1.

10 (Mary)
9
8
7
6
5
4 (Joe) <-
3 (Bob)
2
1 (Steve)
0


Joe is up, and readies an action to shoot any zombies that pop out. When a zombie pops out he'll take the action and his time value will move relative to the initiative count that happens at, it won't matter in this example but the shooting on his gun is 3

10 (Mary)
9
8
7
6
5
4 (Joe){waiting}
3 (Bob) <-
2
1 (Steve)
0

Bob peels out in the truck, rushing to get to joe. with a time of 6


10 (Mary)
9
8 (Bob)
7
6
5
4 (Joe){waiting}
3
2
1 (Steve) <-
0

Finally, steve is up. Like joe, he readies a shot. We then end the round and between time "0" time "10" resolve any affects our systems care about with set durations, advance any timers on effects by the length of our round and so on. We then enter a new round where the initiative now looks like this

10 (Mary)
9
8 (Bob)
7
6
5
4 (Joe){waiting}
3
2
1 (Steve) {Waiting}
0<-

with mary up first for the next round.

This flows pretty intuitively, and keeps the time-per-turn relatively low and the book keeping is not that intensive. The con is it requires the use of a special track, players with long actions can be left sitting on their thumbs for a while and you need to assign a lot of time values to everything - which can make action lists/character sheets very cumbersome.

This can be mitigated by having a list of set times with values that all actions choose from
for example:
Short: 2
Moderate: 5
Long: 9

also note the "10"-length track here was chosen arbitrarily. You can make it as long or as short as you need in order to make it flow well. ~20 seems like a good value off the top of my head.

[Just to Browse]

Action Points lead to analysis paralysis. From a pragmatic standpoint, you want a game that people will actually sit down and play, which means using rules that are nice to new players and people running 4 hours of sleep. For this, I heavily recommend set actions. Recycling the action system from another game (word-for-word) lets you sneak in more complicated rules if your prospective players are familiar with the parent ruleset, but I strongly suggest against making an AP system from scratch.

Round length should be tied to game type. Humans doing human things should probably have rounds that are about as long as 1 camera shot during a chaotic fight in your average action TV show. 3-6 seconds shouldn't cause you any problems. 10+ seconds is better in an over-the-top anime like Naruto / FMA, where 1 second is better for games that cater to people who use this kind of thread (http://www.giantitp.com/forums/showthread.php?493127-Got-a-Real-World-Weapon-Armor-or-Tactics-Question-Mk-XXI).

[wkwkwkwk1]

What about one of those initiative/time consumption system type systems?
*snip*

I've given it thought before. Look closer, and you'll notice that this is actually a modified "action points" system, the big difference being that the number of AP changes for each player from round to round. :smallsmile:


Action Points lead to analysis paralysis. From a pragmatic standpoint, you want a game that people will actually sit down and play, which means using rules that are nice to new players and people running 4 hours of sleep. For this, I heavily recommend set actions. Recycling the action system from another game (word-for-word) lets you sneak in more complicated rules if your prospective players are familiar with the parent ruleset, but I strongly suggest against making an AP system from scratch.

It does lead to analysis paralysis, that's mostly the reason I posted this thread. :smallbiggrin:

Set actions = lack of realism :smallfrown:

As I said before, there's no differentiation between actions' times, and this makes for a big lack of realism (what's the point of choosing a knife instead of a sledgehammer?). Besides, this makes it even harder to solve this problem I detailed in the OP:


The problem with long rounds still stands: each character's turn takes place independently, as if each character were moving on their own agenda, allowing someone to chase them while they stood still, for example. There are many rationalizations behind this, but none remove the actual problem. If someone wants to react to another's actions, they'll have to predict them a full round ahead.

Thanks for your feedback! :smallsmile:

[Mr.Moron]

I've given it thought before. Look closer, and you'll notice that this is actually a modified "action points" system, the big difference being that the number of AP changes for each player from round to round. :smallsmile:

There are some critical differences:

-You make decisions on each action independently. You don't have a pile of points you have to divvy up during your turn. You do something and that determines the next time you get to act. The potential for "Analysis Paralysis" is lower, at least on an individual level. This also makes things feel snappier and more responsive.
-Removes all ambiguity as to "When" actions happen. There isn't any need to muss about with timings, things happen as they happen.
-Low time per turn. Unlike with set actions or APs you don't wait for someone to resolve their "Full Attack" or plan out complex actions. In other words the spotlight changes changes hands more.

[Just to Browse]

Set actions = lack of realism :smallfrown:

As I said before, there's no differentiation between actions' times, and this makes for a big lack of realism (what's the point of choosing a knife instead of a sledgehammer?). Besides, this makes it even harder to solve this problem I detailed in the OP:

As I'm sure you understand, realism is already taking a hit because you're playing a game at a table instead of slaying zombies IRL. Using an initiative system with discrete actions taken in order is unrealistic too, but some things you just can't compromise on.

The combat rules of your game are bound by soft constraints (wanting to pass this off as realistic), but the hard constraints are more important (players should be able & willing to play it). Your differences between knife & sledgehammer should come from somewhere other than that particular part of the initiative system, because doing otherwise means likely losing some non-trivial portion of your audience.

Good luck!

[wkwkwkwk1]

There are some critical differences:

-You make decisions on each action independently. You don't have a pile of points you have to divvy up during your turn. You do something and that determines the next time you get to act. The potential for "Analysis Paralysis" is lower, at least on an individual level. This also makes things feel snappier and more responsive.
-Removes all ambiguity as to "When" actions happen. There isn't any need to muss about with timings, things happen as they happen.
-Low time per turn. Unlike with set actions or APs you don't wait for someone to resolve their "Full Attack" or plan out complex actions. In other words the spotlight changes changes hands more.

All very good points indeed. I'm starting to give this some serious thought. I just need to solve this little problem: how do I handle different combatant speeds, without falling into the problems of decreasing cost?


As I'm sure you understand, realism is already taking a hit because you're playing a game at a table instead of slaying zombies IRL. Using an initiative system with discrete actions taken in order is unrealistic too, but some things you just can't compromise on.

The combat rules of your game are bound by soft constraints (wanting to pass this off as realistic), but the hard constraints are more important (players should be able & willing to play it). Your differences between knife & sledgehammer should come from somewhere other than that particular part of the initiative system, because doing otherwise means likely losing some non-trivial portion of your audience.

Good luck!

As for realism, well, what I'm shooting for is not utter simulationism (this isn't Dwarf Fortress, after all). It's that players can expect actions to feel natural and have a natural impact in the game world. If you stab someone, they bleed. If you hit someone with a stick, bones break. Injuries cause pain, which is tempered by adrenaline and willpower. Fighting makes you tired. Someone wielding a knife could attack three times where a guy with a sledge would attack only once, but if you get hit by that sledge, good luck with that concussion, and so on.

As for the audience, well, as I explained in the OP, I'm doing this for myself, maybe a couple of friends, maybe someone here in GitP :smallbiggrin:

Thanks! :smallsmile:

[Just to Browse]

I know you're designing the game for yourself, but I assume you want to play it with other people at some point in the future. I build things exclusively for me + 3 people I game with, and that's already enough constraint to force simple design.

As an aside, you may want to look at knife fighting and "sledgehammer fighting" (likely not by that name), especially if you're going for realism. 3 attacks with a knife per attack with a hammer is an idea inspired from action scenes on TV screens rather than a brawl on the street. You'll probably get in 1 from the knife per 1 from the sledgehammer, and the fight will probably be over after the first guy lands a good hit, because that's the nature of 2 squishy bodies using lethal weapons against each other. "Realism" versus "feels right when I think about it" are not necessarily the same thing.

[Mr.Moron]

All very good points indeed. I'm starting to give this some serious thought. I just need to solve this little problem: how do I handle different combatant speeds, without falling into the problems of decreasing cost?

"Movement is a moderate (5) action, when you take the movement action you advance a distance equal to your speed value"

[Zman]

What about one of those initiative/time consumption system type systems?


You have a printed track of initiative values and participants are represented by tokens or miniatures on the track. At the start of an encounter you use some stat to give people their starting positions.

So say we have 4 parcipants: Bob, Steve, Mary and Joe they'd start at their initiative values

10 (Bob) <-
9
8 (Mary)
7 (Steve)
6
5
4 (Joe)
3
2
1
0

Starting at the top, bob is up first. Bob attempts to start up the truck up his truck, with a time value of (7) he moves down to initiative 3.

10
9
8 (Mary) <-
7 (Steve)
6
5
4 (Joe)
3 (Bob)
2
1
0

Mary is up next, and takes a penalty to her observation roll to make quick look around, with at ime value of (3) moving her down to 5.

10
9
8
7 (Steve) <-
6
5 (Mary)
4 (Joe)
3 (Bob)
2
1
0

Steve is up next and reloads his gun, with time value of 2 putting him at 5

10
9
8
7
6
5 (Mary) (Steve) <-
4 (Joe)
3 (Bob)
2
1
0

Mary is up again, since her and steve are at the same value but she was there "First" she runs for truck with a time value of (6), putting her back up at the top with 10 wrapping around. Steve does the same, but since he's quicker on his feet this only a 4 value action for him putting him at 1.

10 (Mary)
9
8
7
6
5
4 (Joe) <-
3 (Bob)
2
1 (Steve)
0


Joe is up, and readies an action to shoot any zombies that pop out. When a zombie pops out he'll take the action and his time value will move relative to the initiative count that happens at, it won't matter in this example but the shooting on his gun is 3

10 (Mary)
9
8
7
6
5
4 (Joe){waiting}
3 (Bob) <-
2
1 (Steve)
0

Bob peels out in the truck, rushing to get to joe. with a time of 6


10 (Mary)
9
8 (Bob)
7
6
5
4 (Joe){waiting}
3
2
1 (Steve) <-
0

Finally, steve is up. Like joe, he readies a shot. We then end the round and between time "0" time "10" resolve any affects our systems care about with set durations, advance any timers on effects by the length of our round and so on. We then enter a new round where the initiative now looks like this

10 (Mary)
9
8 (Bob)
7
6
5
4 (Joe){waiting}
3
2
1 (Steve) {Waiting}
0<-

with mary up first for the next round.

This flows pretty intuitively, and keeps the time-per-turn relatively low and the book keeping is not that intensive. The con is it requires the use of a special track, players with long actions can be left sitting on their thumbs for a while and you need to assign a lot of time values to everything - which can make action lists/character sheets very cumbersome.

This can be mitigated by having a list of set times with values that all actions choose from
for example:
Short: 2
Moderate: 5
Long: 9

also note the "10"-length track here was chosen arbitrarily. You can make it as long or as short as you need in order to make it flow well. ~20 seems like a good value off the top of my head.

I second this system and was using a variation myself for a game I was using. I included a default "reaction time" as a small penalty for changing actions i.e. The time it takes to decide to do something hing, so even something that would have a very minimal time cost may have a bit longer. For example, drawing a pistol takes only a fraction of a second, but the decision to do so and execution takes significantly longer. Another example, taking off running takes longer than con I using to run. Firing at a different target takes much longer Han continuing to fire at your current target.

[wkwkwkwk1]

I know you're designing the game for yourself, but I assume you want to play it with other people at some point in the future. I build things exclusively for me + 3 people I game with, and that's already enough constraint to force simple design.

As an aside, you may want to look at knife fighting and "sledgehammer fighting" (likely not by that name), especially if you're going for realism. 3 attacks with a knife per attack with a hammer is an idea inspired from action scenes on TV screens rather than a brawl on the street. You'll probably get in 1 from the knife per 1 from the sledgehammer, and the fight will probably be over after the first guy lands a good hit, because that's the nature of 2 squishy bodies using lethal weapons against each other. "Realism" versus "feels right when I think about it" are not necessarily the same thing.

Might be, might not be. I want the game to be playable, yes, though not necessarily appealing to the general public :smallsmile:

Well, I don't really watch action movies, so no, it isn't :smallwink: What I do know, however, is that melee fighting is almost outright impossible to simulate. It's painfully obvious just from watching kendo fights and reading medieval fight reports.

That said, I'm designing this mostly from the standpoint of someone who's used knives and cleavers in the kitchen and sledgehammers and pickaxes to take down brick and concrete walls, and know that a knife is way, way lighter, and not nearly as damaging.

In a real knife fight, if having the element of surprise, an attacker could stab or slash a target multiple times per second (gives me an idea, actually), but let's say you're fighting zombies: you want to get in a solid hit, probably through the eye socket, the temple or from under the chin. That's going to take a little more coordination and time to aim and follow through.

As for the lethality, well, good luck getting the fight over so soon if both combatants are in survival mode. Unless you hit a critical spot, you might well have a man coming at you with 20+ bleeding holes.

Or, they might drop from a shallow slash, either because you hit a critical spot (say, the femoral artery, the trachea, or the neck vessels) or from shock.

Thanks again! :smallsmile:


"Movement is a moderate (5) action, when you take the movement action you advance a distance equal to your speed value"

Ok, that's not what I meant, let me rephrase :smallbiggrin:

I meant speed in terms of agility. Some people are naturally faster at doing everything, while others are just slower. Now, in that system, maybe the Agility stat could modify the "initiative roll". I should have thought of this earlier. I'm liking this system :smallsmile:


I second this system and was using a variation myself for a game I was using. I included a default "reaction time" as a small penalty for changing actions i.e. The time it takes to decide to do something hing, so even something that would have a very minimal time cost may have a bit longer. For example, drawing a pistol takes only a fraction of a second, but the decision to do so and execution takes significantly longer. Another example, taking off running takes longer than con I using to run. Firing at a different target takes much longer Han continuing to fire at your current target.

So much yes. I had actually thought of this, but it was for a different game, so thanks for bringing it up :smallbiggrin:

[Just to Browse]

Humans usually aren't berserkers, and a knife fight will not lead to "shallow wounds" most of the time. People running at you with 20+ wounds is the stuff of movies and rare anecdotes, because most people don't operate well under blood loss. An blow to a gut or torso with a knife means the sledgehammer is no longer useful and your knife-wielder can follow up with whatever number of slashes necessary in order to stop sledgehammer-wielder from getting up.

This is just something you should watch for. It's cool to have a system where using a knife means you fight fast & weak while using a hammer means you fight strong & slow, but if your end goal is "realism" then you're headed in the opposite direction. So far it seems like you're willing to lose some realism there but aren't willing to lose realism for your action system, and that's confusing to me, but it's your game so have at it.

[Djinn_in_Tonic]

You could do a second-by-second round timer where characters get X actions of Y type. This is, however, just off the top of my head. As an example, however:

A round is a 6 second increment. There are ten ticks per round.

A player may move up to 10 feet per tick, to a maximum of their movement speed (or 5 feet and starting an attack as a single tick). Making a single attack takes two ticks (one to prepare, one to hit). There is a penalty to moving away from an attack in progress -- perhaps a chance of it landing immediately.

Each tick happens in initiative order, with players going around determining their actions for that tick. Player can therefore intercept movements, ready actions, counter longer actions (spells, for example), take defensive actions, etc.

So in close combat:

Player A goes first, taking a 5 foot step next to Player B and initiating an attack. Player B goes next, and must decide whether to attack himself, retreat (and possibly be struck), or take a defensive action. He opts to defend, gaining a bonus against Player A's incoming attack. Player A's next tick happens, dealing damage to Player B. Player B then initiates the attack...

And so on. Obviously this is SUPER rough, but I might have to explore it later.

[wkwkwkwk1]

You could do a second-by-second round timer where characters get X actions of Y type. This is, however, just off the top of my head. As an example, however:

A round is a 6 second increment. There are ten ticks per round.

A player may move up to 10 feet per tick, to a maximum of their movement speed (or 5 feet and starting an attack as a single tick). Making a single attack takes two ticks (one to prepare, one to hit). There is a penalty to moving away from an attack in progress -- perhaps a chance of it landing immediately.

Each tick happens in initiative order, with players going around determining their actions for that tick. Player can therefore intercept movements, ready actions, counter longer actions (spells, for example), take defensive actions, etc.

So in close combat:

Player A goes first, taking a 5 foot step next to Player B and initiating an attack. Player B goes next, and must decide whether to attack himself, retreat (and possibly be struck), or take a defensive action. He opts to defend, gaining a bonus against Player A's incoming attack. Player A's next tick happens, dealing damage to Player B. Player B then initiates the attack...

And so on. Obviously this is SUPER rough, but I might have to explore it later.

This seems very interesting, yes. :smallsmile:

Just a few things, though:

1. 10 ft. per half second? That's... 5 meters/second? I don't know, seems too fast :smalleek: :smalltongue:

2. Anyway, isn't this really the same as this?:


Second-long rounds: Used by GURPS, this system sees each second as a round. This is actually a modified "Action by Action" system in disguise, because, generally speaking, each character is allowed one action per round (except for longer actions, naturally).

Pros:
- More granular than longer rounds
- Allows everyone to act more often (in terms of in-game time) than in longer rounds, where large segments are resolved at the same time

Cons:
- Is it really more granular? Attacking with a knife takes as much time as attacking with a sledgehammer. Some weapons, after being used, have to be "readied" (1-turn action) to be used again, but this strikes me as extremely binary

[Djinn_in_Tonic]

1. 10 ft. per half second? That's... 5 meters/second? I don't know, seems too fast :smalleek: :smalltongue:

Adjust times as needed. :smalltongue:


2. Anyway, isn't this really the same as this?:

Not quite. An important distinction is that the system I proposed is broken up into larger rounds. In the system you suggested a series of four actions could be attack / attack / attack / attack, and everyone takes an action on every round.

In my proposal, a character might have, say 2 attacks, 40 feet of movement, and an miscellaneous action per round, but 10 ticks to take those actions on. They can use some or all of this as they see fit: perhaps they don't move for the first few ticks to see what the opponent does and react accordingly. Perhaps they save the attacks to defend in the middle of a two-tick attack. There are more permutations because you can track actions and react accordingly: if an opponent uses up their attacks at the start of the "round," you know their remaining actions will NOT be attacking, and can plan accordingly.