PDA

View Full Version : What is necessary in a combat system?



Vitruviansquid
2017-03-25, 09:26 PM
I have been banging my head against the wall now for a long time trying to think about what I want the combat system to look like in a system I have been writing. Then I came to the realization that I've been having troubles not because I can't find mechanics to make the results I want, but because I don't really know what I want.

So, assuming that combat is a major deal in the game in question, what does a good combat system need to have, in your eyes? These can be as abstract and theoretical or as concrete as you want.

For me, I think the following are non-negotiable:

1. There must be meaningfully different options that support different play preferences.
2. There must be time or turns enough that a player can see something wrong with his choices and then rethink his tactics in order to win.
3. You must be able to do things that are cool and special and make the player feel powerful.
4. It must not be clear who will win a combat before the first die is cast.

Malfarian
2017-03-25, 09:40 PM
In essence I think

Combat needs timing, so not everything happens at once.
Fairness, not easiness, but people need to think it's fair.
Uncertainty, many great games end up with a min/max action computation and that's really hard to prevent. So if initiative order varied a little turn to turn that could achieve that.
Deadliness - that is combat should NEVER be routine, always a risk.

BayardSPSR
2017-03-25, 09:55 PM
I have been banging my head against the wall now for a long time trying to think about what I want the combat system to look like in a system I have been writing. Then I came to the realization that I've been having troubles not because I can't find mechanics to make the results I want, but because I don't really know what I want.

So, assuming that combat is a major deal in the game in question, what does a good combat system need to have, in your eyes? These can be as abstract and theoretical or as concrete as you want.

I've been thinking combat systems lately not exclusively in RPGs, but in tabletop games at large, after having a mediocre time with a well-liked new boardgame. What I've realized I dislike is combat systems that either don't have a clear purpose behind them (and seem to exist just so that a game will have a combat system) or that make combat an end, rather than a means (where winning fights is a heavily incentivized by game mechanics, rather than combat being a way of resolving conflict). The latter point is something I've disliked in specific RPGs in the past (experience points, I'm looking at you), because it tends to lead to players looking for a motivation for a given course of action, rather than looking for the best course of action for a given motivation.

Working back from my dislikes, what I want in an RPG combat system is a) that it actually has a reason to exist separate from non-combat action; and b) that the RPG isn't just built around winning fights, even though it may include combat as a fleshed-out means of achieving goals. This is more of a big-picture design preference than something that could be translated into mechanical changes, though.

That said, while those are my preferences, I can't claim that they're necessary in a combat system. After all, our hobby started as a combat system in a void, onto which all the things that compose an RPG were tacked on...

RazorChain
2017-03-25, 10:26 PM
I have been banging my head against the wall now for a long time trying to think about what I want the combat system to look like in a system I have been writing. Then I came to the realization that I've been having troubles not because I can't find mechanics to make the results I want, but because I don't really know what I want.

So, assuming that combat is a major deal in the game in question, what does a good combat system need to have, in your eyes? These can be as abstract and theoretical or as concrete as you want.

For me, I think the following are non-negotiable:

1. There must be meaningfully different options that support different play preferences.
2. There must be time or turns enough that a player can see something wrong with his choices and then rethink his tactics in order to win.
3. You must be able to do things that are cool and special and make the player feel powerful.
4. It must not be clear who will win a combat before the first die is cast.


Wll for nr 4 I would rather say that you should be able to estimate weather YOU can WIN or not. When you are surrounded and outnumbered the system should give you some indication that you are in over your head. I sometimes get DnD players into my more gritty, realistic games and often they don't get the notion of retreating, surrendering or try to bluff, sneak etc. It usually ends in a splat! "Whaddya mean...those are only 20 goblins...how on earth did we die!" "My character collapses from couple of crossbow bolt through the spleen and kidney?!?! Unfair....the city guards should only be lvl 1 noobs!!!"

I would rather say that combat should always represent danger


For me I like to have options, both in weapons and tactics and a realistic feel where you aren't just trading blows until someone loses his ablative HP shield and then suddenly falls down dead.
I like either mechanism that use opposed rolls Attack vs. Defense or you roll against target number to hit and target number to defend. And I like armor to subtract damage rather than be included in how hard you are to hit.





I've been thinking combat systems lately not exclusively in RPGs, but in tabletop games at large, after having a mediocre time with a well-liked new boardgame. What I've realized I dislike is combat systems that either don't have a clear purpose behind them (and seem to exist just so that a game will have a combat system) or that make combat an end, rather than a means (where winning fights is a heavily incentivized by game mechanics, rather than combat being a way of resolving conflict). The latter point is something I've disliked in specific RPGs in the past (experience points, I'm looking at you), because it tends to lead to players looking for a motivation for a given course of action, rather than looking for the best course of action for a given motivation.

Working back from my dislikes, what I want in an RPG combat system is a) that it actually has a reason to exist separate from non-combat action; and b) that the RPG isn't just built around winning fights, even though it may include combat as a fleshed-out means of achieving goals. This is more of a big-picture design preference than something that could be translated into mechanical changes, though.

That said, while those are my preferences, I can't claim that they're necessary in a combat system. After all, our hobby started as a combat system in a void, onto which all the things that compose an RPG were tacked on...


We've come a long way from the void where combat system has tacked on an RPG, but I really agree with you on most points, but there are a lot of choices out there where the point of the game is not combat. Yes DnD may be the most popular system and incentivizes combat but most systems don't. For me combat is only one possible conflict resolution as my games aren't centered on the combat minigame.

Vitruviansquid
2017-03-25, 10:28 PM
Fairness, not easiness, but people need to think it's fair.

What does this mean?

I've toyed with the idea of prescribing difficulty level to challenges in the past, and I know at least in 4e, Dungeons and Dragons has some DM hints about how to budget a combat encounter so that it works against some number of players at some level. But every time I've tried this, I just thought it was pointless because it was always the GM's job to make balanced encounters (or unbalanced encounters when he decides they should be).

Koo Rehtorb
2017-03-25, 11:56 PM
Interesting question. I suppose, in the end, I can't actually think of anything intrinsic to a combat system other than that it involves combat in some way, shape, or form. It really entirely depends on the game.

NichG
2017-03-26, 12:02 AM
One important aspect IMO is for the combat system to create a larger space of outcomes and variations than success/fail. For example, take all the different variations of what can happen if the party wins: you can win but use few resources, win but use many resources, win but suffer a temporary disadvantage, win but suffer a long-term setback, win but the enemies escaped, win but a party member died, win but failed at a secondary objective, win and managed to achieve some additional objective in the process, ...

Because combat is finely resolved and rather definite, it creates a sort of map of outcomes where these various possibilities don't have to be specifically worked into the encounter itself, but rather emerge organically from aspects of the system. That allows these kinds of gradations to be as much in the hands of the players and how they choose to approach the combat as in the hands of the DM. That in turn makes combat (or finely resolved systems in general) potentially a lot more expressive.

So with that in mind, I'd try to avoid compartmentalizing combat too much - resist the urge to have everything reset to status quo for the victors. Similarly, the more hooks you can provide for creating secondary objectives or tying in-combat behavior to longer term advantages/disadvantages/progress/outcomes/etc, the more gameplay you can squeeze out of the system.

Martin Greywolf
2017-03-26, 11:56 AM
There are degrees here.

What a combat system needs to function is one thing, but there additional requirements that you need to make combat system actually good. I assume we're talking about the latter.

1) Mechanical core

A combat system must adhere to mechanical core of the entire system. This is where DnD fails, and fails hard - rules for fighting are completely mechanically different from literally anything else. You use different dice, roll difficulties are calculated differently etc etc. This creates a problem where you're asking players and DMs to essentially learn two games instead of just one. Avoid it, avoid it like the plague.

2) Simplicity

A good combat system shouldn't make you spend 5 minutes recalculating things in your head if someone casts a buff spell. Again, DnD fails here, even for simple stuff, you need to go through lists of bonuses, and while this only requires you to do addition and substraction mathematically, it takes your focus away from doing cool stuff.

There is a balancing act here, of course, a too simple a system will not reflect an in-game reality as well as a more complicated one (assuming competent designers). Which one you end up going with depends on preferences of your group, but you'll never go into, say, World of Tanks-style modelling of projectile penetration.

3) Verisimilitude

A system needs to both reflect an in-game reality and be consistent with the tone of the world. FATE is a pretty awful system to run something like gritty Conan or Attack on Titan style game in, because it lacks the ease of death for your PCs. Conversely, a gritty, down to earth system will not mesh well with a superhero game. Most systems can be adjusted slightly to one side or another if you give it time, but it's a triicky thing to do most of the time.

As for the in-game reality, this is where you go from pure crunch to more fluffy territory. In a wuxia game, parrying an arrow with a sword is perfectly normal, and dodging bullets in One Piece is something that even low level folks can do (as demonstrated by Straw Hats during the beginning arcs, you don't need Haki, just fast legs). This is sometimes easy to do (all right, add your parry skill to defense vs arrows), but can be problematic at times (just try and add wuxia-length jumps to DnD and watch your move speed become irrelevant... though everyone leaping everywhere is pretty wuxia).

Honestly, the more varied set of options you can easily adapt the system to, the better. This is where FATE falls short, actually, since it is designed around creating movie/book paced adventures, it can't really do anything else very well.

4) Depth

Not complexity. Depth. The two are different.

What you need to have is a set of meaningful choices you can make. That said, as little of these should be based in the mechanics by themselves as possible, and should be rooted in verisimilitude. You shouldn't be deciding whether your gun has more bonus damage than your bow, you should be deciding whether you want to keep shooting from afar while not hitting as often, or rushing towards someone, catching an arrow or two, but comprehensively spanking them after you get there.

That is to say, only courses of action that should be less effective mechanically should be those that are less effective for that character in that world. This, in the end, means that if I decide to go archer, dual-wielder or sword and boarder in a world where all of these are a thing, I should be equally effective in the long run. Sure, I may be less ideal in specific situations, but over the course of several adventures, all the classes should be able to contribute equally.

Sometimes, having characters that aren't great fighters but can contribute in other areas does work, but not that often. You need a system where amount of outside of combat rolling equals amount of outside of combat rolling, and where fights aren't long (forcing skillmonkeys to sit and do very little for long stretches of time).

This is where FATE Accelerated truly falls apart, as you have no reason but to use the top skill (well, approach) once fighting starts.

5) Accounting for the enviroment

If I want to make a fight take place atop a moving train, I should have a set of tools that easily allows me to do so. These will usually be situational modifiers, but you need to think about incorporating them in your system from the get go, not just slap them on as an afterthought. This is where FATE shines, as any description can be mechanically meaningful very easily.

6) Modularity

If I want to create a new class/skill/stunt/perk, I should be able to do so without much trouble. This is, more often than not, a failing of books presenting a system - your thought process when assigning numbers to things should be clearly outlined, making the lives of the rest of us easier when we decide we want to have a knight that is accurate to a specific period and therefore needs dance or needs to be illiterate.

Conclusion

One final thought is that if you want your combat system to be good, these six points are all equally important. Sometimes you need to sacrifice a bot of one to get a lot of the other (verisimilitude and simplicity come to mind), but you can't mostly toss out one of these. Actually making a system like that is pretty damn difficult. Which is why people pay other people to do it for them.

Cluedrew
2017-03-26, 12:34 PM
The only thing I think is truly necessary is that it must be able to resolve combat. That is it. To make it good what you need entirely depends on what you want it to be good at, which includes the context of the rest of the system. For instance looking at your list of non-negotiable items I am going to work backwards to guess that you are looking for a combat system that is meant to be a relative centerpiece of the system. It is tactical, using player skill, and most characters will be combatants so there must opinions to differentiate between them. Furthermore, although the thread of loss should be present, things will mostly be slanted towards the players.

Does that sound right? You said that you are not sure what you want, is that it?

Quertus
2017-03-26, 01:01 PM
For me, I think the following are non-negotiable:

1. There must be meaningfully different options that support different play preferences.
2. There must be time or turns enough that a player can see something wrong with his choices and then rethink his tactics in order to win.
3. You must be able to do things that are cool and special and make the player feel powerful.
4. It must not be clear who will win a combat before the first die is cast.

This is a really neat list. I most strongly agree with #2; or, rather, most strongly prefer games where #2 is true. This is in part because I prefer role-playing over metagaming, and, to put it in D&D terms, don't want character survival to be based on whether the player has read and memorized the monster manual.

#4, otoh, is a bit of a tricky one. At the extreme, do you really want to be in doubt as to whether Superman can successfully pull a weed without dying? Or whether a housecat can kill a deity?

Looking at war games and Magic the Gathering, if you really think about it, you can determine which army or which deck will probably win any given matchup. But there is enough randomness involved that it probably could go either way, with enough luck. At least until you get into the "Superman vs a weed" level of mismatch.

If that is what you want, then I'd say look at how some good war games / CCGs handle things for inspiration.


6) Modularity

If I want to create a new class/skill/stunt/perk, I should be able to do so without much trouble. This is, more often than not, a failing of books presenting a system - your thought process when assigning numbers to things should be clearly outlined, making the lives of the rest of us easier when we decide we want to have a knight that is accurate to a specific period and therefore needs dance or needs to be illiterate.

This is where most systems - combat or otherwise - fail for me. I have a concept, and it simply can't be represented in the system. If you can solve this, kudos!

Skorj
2017-03-26, 03:45 PM
Lots of great stuff.

Martin Greywolf said most of what I came to say, but let me add a bit at a less-detailed level. D&Ds various systems through the years all have right concept.


You do want to break combat down into rounds, for playability and verisimilitude and, well, it's how most game are played from card games on up.
You do want to have numerical skills of some sort related to combat (I think early-3x D&D errs in making combat roll too different from skill rolls).
You also want situational abilities that give bonuses - and that should be the focus, and where any complexity lives. "I'm better with an axe," or, "I'm better when I'm behind my target".


Where D&D is weak is in letting you as GM easily "wing it" with new skills or oddball situations. It you focus the combat narrative on the environment and tactics (but I repeat myself), it's all about the circumstances specific to this combat. Can you attack from hiding? Can you take cover from ranged attackers? That's where the game is "people playing together", instead of an awkward MMO. You don't need a detailed map-and-miniatures system for that stuff, if you don't want it, but you do need to describe well the place where the combat happens, and have abilities that key off the details.

I think what most grognards like myself have concluded is: RPG combat needs to be different in feel from a wargame or an MMO. Focus on the colorful stuff that players can come up with, the stuff that makes for good stories later. But do in a system with numbers to make the odds objective, not just "rule of cool". So, what's necessary is enough structure that the GM can fairly assign a numerical bonus or penalty to whatever colorful tactic the players think up next.

Steampunkette
2017-03-26, 04:03 PM
A core balancing metric.

For most online games it's damage magnitudes or a permutation thereof.

Once you've got an idea of the core balancing metric, stretch it across your mind like a string and hang other mechanics on it like Christmas lights.

You want players to take 12 damage magnitudes before they die, but want they player able to defeat 6 enemies before dropping? Balance the enemy's health mags against the player's damage mags (modified by mitigation) and vice versa.

Easy Peasy.

jok
2017-03-27, 05:09 AM
I have been banging my head against the wall now for a long time trying to think about what I want the combat system to look like in a system I have been writing. Then I came to the realization that I've been having troubles not because I can't find mechanics to make the results I want, but because I don't really know what I want.

So, assuming that combat is a major deal in the game in question, what does a good combat system need to have, in your eyes? These can be as abstract and theoretical or as concrete as you want.


A perfect combat system would be:

Easy to learn
Has many options
Fast to resolve
Meaningful decissions at every step
Allow for theory crafting
Does not require theory crafting
Reward tactical thinking players
Allow for hack and slashers
Is fair and detailed
Has room for improvisations

How to achieve that I don't know. Many points are mutual exclusive. At least at first glance. A real one in a century genious is needed...

Frozen_Feet
2017-03-27, 07:03 AM
I have been banging my head against the wall now for a long time trying to think about what I want the combat system to look like in a system I have been writing. Then I came to the realization that I've been having troubles not because I can't find mechanics to make the results I want, but because I don't really know what I want.

So, assuming that combat is a major deal in the game in question, what does a good combat system need to have, in your eyes? These can be as abstract and theoretical or as concrete as you want.

For me, I think the following are non-negotiable:

1. There must be meaningfully different options that support different play preferences.
2. There must be time or turns enough that a player can see something wrong with his choices and then rethink his tactics in order to win.
3. You must be able to do things that are cool and special and make the player feel powerful.
4. It must not be clear who will win a combat before the first die is cast.

4 has no place in a systems design.

Why? Because a system which can create all of uncertain, transparently unwinnable and transparently unloseable as well as their non-transparent versions is plainly more powerful and useful than a system which can only create uncertain and non-transparent ones.

Remember: you are not designing a game. You are designing a system which can be applied across games, with different scenarios.

If you have difficulty understanding the difference, consider Chess. You are probably familiar with the vanilla version where the sides of the board are symmetric. That's Chess-as-a-game. But there also specific Chess puzzles which can start from any board configuration, even ones impossible to reach during a normal game. Changing the position of the pieces allows for scenarios which are asymmetric but winnable, as well as unwinnable or unloseable scenarios, even though all the pieces still move normally. That is Chess-as-a-system.

If your system can create transparent scenarios, you will also have easier time wlth 2: you, as a scenario designer, can design an unwinnable or unloseable scenario, and then your players can, via example, learn what to built towards or what to avoid. (This also allows for informed surrender and self-sacrifice later down the line.)

Though overall, the same applies to 2 as to 4: give players time via scenario design, not via systems design. Your system should be able to create both scenarios where there is a chance to adapt, and scenarios where there is not. It doesn't matter if you dislike the latter type, a system able to do both is, again, plainly more powerful and useful.

Martin Greywolf
2017-03-27, 08:19 AM
4 has no place in a systems design.

Why? Because a system which can create all of uncertain, transparently unwinnable and transparently unloseable as well as their non-transparent versions is plainly more powerful and useful than a system which can only create uncertain and non-transparent ones.

Remember: you are not designing a game. You are designing a system which can be applied across games, with different scenarios.

If you have difficulty understanding the difference, consider Chess. You are probably familiar with the vanilla version where the sides of the board are symmetric. That's Chess-as-a-game. But there also specific Chess puzzles which can start from any board configuration, even ones impossible to reach during a normal game. Changing the position of the pieces allows for scenarios which are asymmetric but winnable, as well as unwinnable or unloseable scenarios, even though all the pieces still move normally. That is Chess-as-a-system.

If your system can create transparent scenarios, you will also have easier time wlth 2: you, as a scenario designer, can design an unwinnable or unloseable scenario, and then your players can, via example, learn what to built towards or what to avoid. (This also allows for informed surrender and self-sacrifice later down the line.)

Though overall, the same applies to 2 as to 4: give players time via scenario design, not via systems design. Your system should be able to create both scenarios where there is a chance to adapt, and scenarios where there is not. It doesn't matter if you dislike the latter type, a system able to do both is, again, plainly more powerful and useful.

This is very much a difference in opinion. There are some systems where a lvl 1 farmer can one-hit a dragon (usually involving exploding dice - that is, you roll a top number, you roll again and add up), there are others that make things like these impossible from the get go. This alone doesn't make a system good or bad, and merely ties into the tone the system will give your game, and both approaches have their pitfalls:

The exploding dice has a problem with instakills being too easy to get - the most infamous is the optional instakill on triple 20 house rule for DnD 3.

The no instakill possible systems run into a problem of a single PC being able to pretty much hack apart an army if the numbers are high enough, which may or may not sit well with your game tone.

As for your Chess examples, well, no. Chess is a game with a rigidly determined starting positions, if you change those, you're no longer playing chess - you are playing something very similar to chess, but not chess. The starting positions did change over the time to make the game faster and more exciting, along with movement of pieces, incidentally, so medieval chess is by no means modern chess, and so on and so forth.

It's a bit like E6 DnD based on 3.5 - it's a game very like DnD 3.5, but not truly DnD 3.5. Is it still a usable game system? Yes. Better than 3.5? I'd say it is, but that's a matter of opinion. But it's still its own thing.

Lastly, almost any system period is capable of handling nigh-unwinnable scenarios, just drive up the numbers, whether numbers in a stat block, or numbers of opponents. Sure, exploding dice give you a chance of OHKO on that dragon, but if the odds are one in a trillion, you better think again.

Not one in a million, tho, if that are the odds, the dragon's dead meat.

CharonsHelper
2017-03-27, 08:35 AM
The no instakill possible systems run into a problem of a single PC being able to pretty much hack apart an army if the numbers are high enough, which may or may not sit well with your game tone.


That's an entirely separate problem. Plenty of systems don't make that viable without using insta-kills as a patch to fix it.

Frozen_Feet
2017-03-27, 09:23 AM
This is very much a difference in opinion. There are some systems where a lvl 1 farmer can one-hit a dragon (usually involving exploding dice - that is, you roll a top number, you roll again and add up), there are others that make things like these impossible from the get go. This alone doesn't make a system good or bad, and merely ties into the tone the system will give your game, and both approaches have their pitfalls:

The exploding dice has a problem with instakills being too easy to get - the most infamous is the optional instakill on triple 20 house rule for DnD 3.

The no instakill possible systems run into a problem of a single PC being able to pretty much hack apart an army if the numbers are high enough, which may or may not sit well with your game tone.

You border on talking about nothing I was talking about. Because I wasn't speaking about instakills nor exploding dice, those are individual mechanical decisions much lower down the pole of game design than choosing how broad number of scenarios your system is meant to handle.

Again: a system which can create all of transparently unwinnable, unloseable & uncertain scenarios as well as their non-transparent versions is plainly more powerful & useful than one which can do only the latter. You can achieve both systems with exploding dice, instakills etc., or without them. Discussing those individual mechanics is hence completely beyond the point I was making.


As for your Chess examples, well, no. Chess is a game with a rigidly determined starting positions, if you change those, you're no longer playing chess - you are playing something very similar to chess, but not chess. The starting positions did change over the time to make the game faster and more exciting, along with movement of pieces, incidentally, so medieval chess is by no means modern chess, and so on and so forth.


You either didn't read or didn't understand "you are not designing a game. You are designing a system which can be applied across games".

calam
2017-03-27, 10:36 AM
For me, I think the following are non-negotiable:

1. There must be meaningfully different options that support different play preferences.
2. There must be time or turns enough that a player can see something wrong with his choices and then rethink his tactics in order to win.
3. You must be able to do things that are cool and special and make the player feel powerful.
4. It must not be clear who will win a combat before the first die is cast.

On 1 I think you have to go further and not only have different play options for different play preferences but also make sure having different viable options on 1 character can be easily done. Otherwise if they aren't viable you fall for the D&D problem where the fighter can technically do many things like bull rush but each have to be specialized in for it to be effective, leading to characters that can dungeoncrash bull rush and full attack or trip and full attack making them have about 3 options: do the thing you're specialized in, do the thing everyone can do or do one of many useless things.

On 4 This seems to be more for the realm of GMs since there's a narrative need for both enemies you almost certainly need to run from and enemies that you can easily defeat. There are mechanics like having rolls that always succeed or always fail (like the natural 1 and natural 20 rolls) which promotes that minimum level of swingyness though.

I agree with rules 2 and 3 will add a few I find are important:

1)If a character improves they should play differently and not have the majority of improvements be "+1 to hitting" or "+5 to strength score" with improvements that just give you a bonus to rolls being the worst offenders. Instead there should be abilities that actually change how you play.

2)You should be able to easily customize all enemies. It's always annoying when playing a game that everyone is used to the players have to try not to know the attacks of various creatures. Having a couple potential options for abilities for each creature can do wonders such as instead of an example creature having 3 abilities it can have 3 of the 5 abilities from a list.

3)Every action in combat should be uncertain. A player should always have some worry that a blow won't hit or hope that an incredibly lethal attack can miss.

4) The game should minimize stupid deaths. There should be ways put in to know certain abilities before hand and ignorance for one round shouldn't lead to death in all but the most exceptional cases (except if the game in question is exceptionally gritty)

Quertus
2017-03-27, 10:39 AM
Where D&D is weak is in letting you as GM easily "wing it" with new skills or oddball situations. It you focus the combat narrative on the environment and tactics (but I repeat myself), it's all about the circumstances specific to this combat. Can you attack from hiding? Can you take cover from ranged attackers? That's where the game is "people playing together", instead of an awkward MMO. You don't need a detailed map-and-miniatures system for that stuff, if you don't want it, but you do need to describe well the place where the combat happens, and have abilities that key off the details.

I dunno, this pretty well describes my early D&D experience. :smalltongue:


I think what most grognards like myself have concluded is: RPG combat needs to be different in feel from a wargame or an MMO. Focus on the colorful stuff that players can come up with, the stuff that makes for good stories later. But do in a system with numbers to make the odds objective, not just "rule of cool". So, what's necessary is enough structure that the GM can fairly assign a numerical bonus or penalty to whatever colorful tactic the players think up next.

Hmmm... And here I am, a grognard who was just advocating looking to war games as examples of how to do combat right.

Don't get me wrong, I can certainly see how your preferred style of combat can be fun, but, to me, that's just extra bells and whistles - the core mechanics need to be good in the first place, or else I just view it as a waste of time. You can't shine a sneaker.

Do you know any systems that do a good job with both fluff and crunch in combat, to point to as an example? And that might be fun to play a "civilian" in, as well?


A perfect combat system would be:

Easy to learn
Has many options
Fast to resolve
Meaningful decissions at every step
Allow for theory crafting
Does not require theory crafting
Reward tactical thinking players
Allow for hack and slashers
Is fair and detailed
Has room for improvisations

How to achieve that I don't know. Many points are mutual exclusive. At least at first glance. A real one in a century genious is needed...

Good call. I often forget to explicitly point out that "allowing but not requiring" tactics, creativity, improv, etc, is a preferable state.

erikun
2017-03-27, 11:40 AM
Well, this is not exactly something that I had thought about before. But if you want a list of features which I think would be necessary, it would be something like this.


Capable of attacking a specific target. If a sniper is attacking, there needs to be some method of attacking the sniper directly in order to stop them.
Capable of avoiding an attack from a specific target. If the sniper is attacking, there needs to be some method of taking cover so that their attack won't hit.
Capable of preventing a character from attacking a specific target. If somebody is charging to tackle the sniper, there needs to be a way to stand between them and keep them away from the sniper.
Capable of resolving a likely action. There needs to be some way of resolving tackling a person to the ground. There needs to be some way to resolve grabbing someone from behind. It doesn't need to be mechanically complex (roll d20 to see if successful works) but it needs some method of resolution.
Capable of moving to a specific location. If I am in room A and need to move to room B (to attack something else, to pull a level, to escape) then there needs to be some method of allowing that to happen.
Capable of preventing movement to a specific location. There needs to be some method of preventing someone from moving into room B, such as blocking a doorway.
Having a clear indication of the results, at the end of a conflict. No "first side to score five victories wins the combat" sort of situation. It should be clear which side lost, why they lost, and a good idea how everyone ends up after the conflict.


Please note that not all options need to be available in every situation. If the sniper is attacking from a ledge, then there might not be a way to get up to them. If the characters are being attacked in a rowboat, then there might not be anywhere to move to. Sometimes, there just isn't cover. However, the system does need to be capable of resolving the action when the action is relevant in the game. Otherwise, you run into some pretty big problems just trying to resolve basic actions with the system.

And again, you don't need specific mechanics for each particular situation. Making an attack roll and then a strength check to replicate a grapple works just fine: you just want SOME method of resolving "I try to grab the person and hold them down" when it comes up in a game. So you don't need detailed movement rules, or detailed wound damage rules, for it to work. You just want to avoid situations where you can't try to simply throw a rug over an opponent's head (D&D4e) or where you don't have a clear idea what happened to everyone once combat is finished (some older FATE versions).

Frozen_Feet
2017-03-27, 02:29 PM
Quertus: if referencing war games makes you a grognard, what does that mean of me and my chess references? :smalltongue:

Erikun: a lot of those boil down go something a lot of RPGs have tried to do away with:

Positioning.

I wouldn't say keeping track of positions is necessary, and it does have the difficulty of needing some form of markers to keep it straight (even if those are just lines on a paper), but at the same time, having actual space and actual distances between represent the battlefield is one of the easiest ways to increase depth and expand movespace of a game.

If you don't do positions, you will have to come up with a lot of alternative mechanics to represent movement, terrain, line of sight etc.

Psikerlord
2017-03-27, 05:59 PM
I have come around to the view that the most important aspect - game wise - is a party retreat mechanic. Removes the need for "balanced" encounters all the time.

Cluedrew
2017-03-27, 06:09 PM
To Psikerlord: I'm not sure if I agree with that, but that's clever. Beside balance it brings a retreat onto the table as an official option, gives a more natural out besides fight to the death (both sides in theory) and gives not-combatants a bit of power (just to survive) on the battle field. In other words, I like it.

NichG
2017-03-27, 07:08 PM
A variant on that idea I used recently was that every combat has an engage duration and a specified default outcome which specifies what happens if e.g. both sides just sit out the duration with some conditions on how that default can changed (when its not obvious stuff like 'both sides survive, unless they're killed'). After that duration, combat ends regardless of the state of the participants, and outcomes are assessed. Character statistics and abilities can determine or manipulate that duration.

It was a bit hard to get used to DMing, since it means that you have to avoid 'here is a room, there are monsters, fight until one side is dead' type situations, and conceptually it can be a bit difficult to always justify why the battle would end (depending on the setting, this could be a matter of endurance, morale, sanity, outside interference, the sun going down, battles taking place in a collapsing abstract demiplane that allows the nascent godhood of the PCs to be tapped into, etc).

But the upside is that retreat and things like that are not just part of the system, they're assumed defaults that have to be actively avoided. So you actually get things like deciding whether eliminating an enemy or achieving an objective is more important, and the opponents also have that kind of decision to make. 'We just have to hold for 3 rounds!' - 'I know, but then the bad guy will escape and we'll have to deal with him later, lets attack!' and things like that emerge a bit more naturally as a result.

Psikerlord
2017-03-27, 09:27 PM
To Psikerlord: I'm not sure if I agree with that, but that's clever. Beside balance it brings a retreat onto the table as an official option, gives a more natural out besides fight to the death (both sides in theory) and gives not-combatants a bit of power (just to survive) on the battle field. In other words, I like it.

Low Fantasy Gaming has a Party Retreat mechanic which you might take a look at (free PDF: https://lowfantasygaming.com/ ).

Works well from what I've seen. As a GM I feel able to throw whatever makes sense at a party, rather than feeling bound by CR tables and so on. Helps the world feel more alive in a way. As a player I have the confidence that if we bite off something more than we can chew, there is a formal mechanic to allow for escape (generally with good odds, albeit at a cost).

Psikerlord
2017-03-27, 09:31 PM
A variant on that idea I used recently was that every combat has an engage duration and a specified default outcome which specifies what happens if e.g. both sides just sit out the duration with some conditions on how that default can changed (when its not obvious stuff like 'both sides survive, unless they're killed'). After that duration, combat ends regardless of the state of the participants, and outcomes are assessed. Character statistics and abilities can determine or manipulate that duration.

It was a bit hard to get used to DMing, since it means that you have to avoid 'here is a room, there are monsters, fight until one side is dead' type situations, and conceptually it can be a bit difficult to always justify why the battle would end (depending on the setting, this could be a matter of endurance, morale, sanity, outside interference, the sun going down, battles taking place in a collapsing abstract demiplane that allows the nascent godhood of the PCs to be tapped into, etc).

But the upside is that retreat and things like that are not just part of the system, they're assumed defaults that have to be actively avoided. So you actually get things like deciding whether eliminating an enemy or achieving an objective is more important, and the opponents also have that kind of decision to make. 'We just have to hold for 3 rounds!' - 'I know, but then the bad guy will escape and we'll have to deal with him later, lets attack!' and things like that emerge a bit more naturally as a result.

I have certainly seen this sort of thing done for set piece encounters - like a magic ritual is happening, after xxx rounds this occurs, that sort of thing. But I have never thought of extending it generally to all combat and putting a timer on things. Definitely interesting. I must ponder further!

Pauly
2017-03-27, 11:57 PM
There are three basic steps in a combat system.

1) Do I hit the target?
Contrary to the assumptions built into most roleplaying games, analysis of combat data shows that that the skill of the target is much more important than the skill of the attacker. That's why modern wargames are moving to the defender's skill level as the primary determinant of hit/miss.

Following this you have a number of secondary factors that can influence the to hit chance. Generally speaking simple models that take into account the most important secondary factors only are better at modelling real life outcomes than highly sophisticated models that try to account for every possible detail. Target size and anything that obstructs the attack (cover) are factors that are probably minimum factors.
(to hit roll)

2) Is it a damaging hit?
A hit doesn't mean a damaging hit. It can be anything from a scratch or a hole in the clothing through to full front to back penetration. Armor is probably the most obvious and basic consideration here, but also total target mass to vital spots should also be factored in.
(to save roll)

In addition there are Psychological effects, PCs are the heroes in the story so they generally should be exempt from this. But for NPCs taking hits (not damage) will affect their psychology. i.e. the fear of dying. There is a chance that on taking hits the NPC will adopt defensive postures, and also a chance of fleeing. Again studies have shown that the chance of flight will depend on accumulated casualties in the unit and/or surprise.
(motivation roll)

3) The effect of the damage.
Very few things keep performing at 100% efficiency until they suddenly fall over dead. But for simplicity there are two main types of damage: -
- The first type of damage is something that are like flesh wounds, enough of them over time will reduce your ability to fight to zero. This type of damage will also reduce endurance and things such as your ability to run, your ability to do acrobatic feats, or the ability to perform fine detail tasks. This type of damage generally needs rest and repair to recuperate.
- The other type of damage is critical damage that stops a system (biological or mechanical) from functioning.
Critical damage can be repaired within combat, but it isn't necessary to model in exact detail what the actual damage is. Your sword arm has stopped working is what is important, not whether it's because the fingers of your hand have been crushed, or your AC joint has been ruptured, or your arm has been hacked off at the elbow. That stuff, if you want to model it, is important after the combat is over.
(damage effect roll)

jok
2017-03-28, 01:04 AM
In my opinion RPGs don't realise the current technology enough. Everyone has a powerful computer
In their pockets nowadays...

A detailed and complex combat system has the problem that it is rules heavy and requires a lot of fiddling with little modifiers. An app should take care of this. I imagine it like this:
Instead of little numbers the GM and the players just add predefined tags to the actions they are trying to to do. For example in a modern setting a shootout against some thugs in the streets. The GM sets general encounter tags like #lightrain #moonlessnight #flickeringillumination #mediumwind.
Positioning and movement is still handled traditionally either theatre of mind or on a grid. On the players turn he decides to run behind a car and start spraying the thugs with his Uzi and after selecting his weapon from his digital char sheet he puts clicks the required tags #run15 #halfcover #drawuzi #fullauto #thug1 #thug2 #thug3 #distance20. The app now calculates the target number/dice pool and takes all into account and shows the players what to roll or rolls automatically. Then it calculates the effect of the hits and the damage done. This can be really detailed depending on ammo, amor, caliber, hit location...

Basically give players and DMs tags to throw at things. Calculate crazy tables and interdependencies under the hood.

Added bonus for the company: Since everyone needs the app to play the company might come up with subscriptions or another buissness model that allows it to profit from the game rules more then once.

erikun
2017-03-28, 09:51 AM
Erikun: a lot of those boil down go something a lot of RPGs have tried to do away with:

Positioning.
Fate Core basically covers all the points I made, and it's positioning system is just "Living Room, Hallway, Stairs, Basement". Heck, AD&D covered most of those points as well, and it generally resolved combat by putting letters on a hand-drawn map and with "I step between the orc and the wizard."

Ironically, D&D3e failed a couple of those points precisely because of its positioning system, and because characters could just run around other characters with little penalty.


And sure, there are systems where combat is basically just rolling the Combat die and determining who won. Sometimes fights don't matter, and the game isn't going to focus on that. But I wouldn't call that much of a combat system, much like how I wouldn't consider making a Diplomacy roll in D&D3e to be much of a social encounter system.

Quertus
2017-03-28, 09:58 AM
In my opinion RPGs don't realise the current technology enough. Everyone has a powerful computer
In their pockets nowadays...

A detailed and complex combat system has the problem that it is rules heavy and requires a lot of fiddling with little modifiers. An app should take care of this. I imagine it like this:
Instead of little numbers the GM and the players just add predefined tags to the actions they are trying to to do. For example in a modern setting a shootout against some thugs in the streets. The GM sets general encounter tags like #lightrain #moonlessnight #flickeringillumination #mediumwind.
Positioning and movement is still handled traditionally either theatre of mind or on a grid. On the players turn he decides to run behind a car and start spraying the thugs with his Uzi and after selecting his weapon from his digital char sheet he puts clicks the required tags #run15 #halfcover #drawuzi #fullauto #thug1 #thug2 #thug3 #distance20. The app now calculates the target number/dice pool and takes all into account and shows the players what to roll or rolls automatically. Then it calculates the effect of the hits and the damage done. This can be really detailed depending on ammo, amor, caliber, hit location...

Basically give players and DMs tags to throw at things. Calculate crazy tables and interdependencies under the hood.

Added bonus for the company: Since everyone needs the app to play the company might come up with subscriptions or another buissness model that allows it to profit from the game rules more then once.

Well, not everyone has such a device (I say as I type this from my phone), but, so long as the GM could do the work for the players, or it technically could be done by hand, it seems fine. Heck, I've calculated modifiers like that plenty of times. Darn lazy gamers and their lack of elbow grease!

However, the amount of time it would take to select or type all these modifiers on a phone? Certainly typing them, or selecting them from a dropdown list, would take far too long. Hmmm... Multiple pages, with lists of environmental and character-specific modifiers, that work like toggle buttons to activate? That might work.

You don't want to slow down gameplay by interrupting the app with an ad. Maybe a full-screed ad when you start the app, plus a small running ad along one edge?

ExLibrisMortis
2017-03-28, 11:05 AM
I don't like it when a combat system uses the wrong words. For example, don't call a mechanic "mordhau" when you mean "coup de grace". I know it's very superficial, but there it is: a combat system must have proper names for things.


If you refine this and take it a step further, you get verisimilitude: the relations between your in-game concepts must be (approximately) the same as the relations between the real-world concepts they are supposedly mapping to. That is, if a pike hedge repels horsemen in combat (and I believe it does), you want a "pike hege" to "repel" "horsemen", whatever those things are in your game.


You can make a simple model of combat, as if making a giant rock-paper-scissors game: pike beats horse beats shot beats pike, but X, Y, Z. Horse is more vulnerable to rough terrain, shot requires more supplies, pike is unhappy when flanked. Of course, this works just as well with magic and ultratech in the mix, but you'll have to define the "real world" you are pretending to map to with your mechanics.

Frozen_Feet
2017-03-28, 01:28 PM
@erikun: can't speak for Fate, but AD&D has a fairly sophisticated positioning system which uses either inches (for miniatures) or in-world distance units (feet etc.). That it can be played on paper doesn't make it less so; you can play Go, or Chess, with pen and paper just as well.

3.x. is equally sophisticated in some regards, but curiously failed to codify some common maneuvers like aforementioned interposing yourself between an attack and a target, so it isn't readily apparent how it can be done. (It can be done, but IIRC you have to look up rules for prepared actions, rules for cover, rules for obstructed space, rules for attacks of opportunity and rules for individual combat maneuvers and then put them together yourself.)

CharonsHelper
2017-03-28, 02:50 PM
3.x. is equally sophisticated in some regards, but curiously failed to codify some common maneuvers like aforementioned interposing yourself between an attack and a target, so it isn't readily apparent how it can be done.

I thought that that's basically what AOOs (combined with single attacks when you move) were for - though you can't do it as perfectly as you can in some systems.

Cluedrew
2017-03-28, 03:59 PM
For example, don't call a mechanic "mordhau" when you mean "coup de grace".Question that I don't actually know the answer to but is also makes a point. What does "coup de grace" mean? I've heard it used enough that I know what you are referring to, but the words are kind of nonsense. If I had to guess: "attack on helpless opponent" by context and "graceful overthrow" by the words.

ExLibrisMortis
2017-03-28, 05:16 PM
Question that I don't actually know the answer to but is also makes a point. What does "coup de grace" mean? I've heard it used enough that I know what you are referring to, but the words are kind of nonsense. If I had to guess: "attack on helpless opponent" by context and "graceful overthrow" by the words.
I would translate it as "gallant strike", as in, it's gallant to spare your opponent the horrible fate of slowly bleeding to death from a gut wound, so you give a mortally wounded opponent (or anyone, really, but let's be knightly about it) a quick death, which is the coup de grace.


(Dammit, it's strike of mercy, not gallant strike. So close!)

Cluedrew
2017-03-28, 05:32 PM
First, thanks. Second, that may be thematic, but it still is not a direct statement of what it is. We have just learned what it means. I think renaming things for greater thematic coherence is a good idea. Can be a good idea. You can take it too far as well or just do it badly.

ExLibrisMortis
2017-03-28, 05:43 PM
First, thanks. Second, that may be thematic, but it still is not a direct statement of what it is. We have just learned what it means. I think renaming things for greater thematic coherence is a good idea. Can be a good idea. You can take it too far as well or just do it badly.
Umm, I'm not sure what you mean. If it's about what I said six posts up, it may be relevant to know that 'mordhau' is not the German equivalent of 'coup de grace'. It's an armoured fencing technique that involves bludgeoning your opponent with the pommel of your sword. My point is that you shouldn't rename things to random senseless cool words, not that you can't use different words to get different flavours.

Cluedrew
2017-03-28, 05:57 PM
What we have here is a breakdown in communication. Mostly on my end. I'll confess I know very little of fencing and misunderstood the point. That would probably fall under "bad thematic renaming". Of course for all I knew when I first heard the term, coup de grace could have been a type of croissant.

Max_Killjoy
2017-03-28, 06:13 PM
Capable of attacking a specific target. If a sniper is attacking, there needs to be some method of attacking the sniper directly in order to stop them.



If no one has a weapon capable of effectively engaging at the sniper's range, why should the system automatically allow the targets to return fire?

Quertus
2017-03-28, 06:57 PM
I don't like it when a combat system uses the wrong words. For example, don't call a mechanic "mordhau" when you mean "coup de grace". I know it's very superficial, but there it is: a combat system must have proper names for things.


If you refine this and take it a step further, you get verisimilitude: the relations between your in-game concepts must be (approximately) the same as the relations between the real-world concepts they are supposedly mapping to. That is, if a pike hedge repels horsemen in combat (and I believe it does), you want a "pike hege" to "repel" "horsemen", whatever those things are in your game.


You can make a simple model of combat, as if making a giant rock-paper-scissors game: pike beats horse beats shot beats pike, but X, Y, Z. Horse is more vulnerable to rough terrain, shot requires more supplies, pike is unhappy when flanked. Of course, this works just as well with magic and ultratech in the mix, but you'll have to define the "real world" you are pretending to map to with your mechanics.

Although I thoroughly enjoyed learning about "mordhau" and "coup de grace", I'm gonna touch on the other half of this post.

As I write this, someone is in the other room is playing a video game which has this rock scissors paper style of interaction, where pikemen deal +50% damage to cavalry, but only take 50% damage from cavalry, or something like that. And it feels very... artificial. The units aren't designed to be better by virtue of how they function (pikes have longer reach or whatever), but because of seemingly arbitrarily inflated numbers. Sure, if it's historic units, and you understand the underlying reasons, it can be ok, I guess, but I have no basis to understand why, say, a Hydralisk might deal extra damage vs, say, an AT-AT. I want their relative effectiveness to be encoded in the way they operate, not tacked on +-X% damage modifiers.

But maybe that's just me. Maybe most people are fine with the way it's often done, and it doesn't cause them any, for lack of a better phrase, cognitive dissonance.

Then there's another thing which might be just me. A lot of systems implement rock scissors paper by having concepts like, "immunity to fire". My personal preference, in video games, was a video game that had the monsters be kind enough to take a single point of damage when they were "immune".

This means that themed characters, like a Fire Mage, suddenly don't get to play, when they encounter something that is immune to their shtick.

Personally, I hate everyone else sitting out while the net runner / face / heavy / whatever is doing their thing. No matter which of the three roles I'm in - GM, featured player, or inactive player - it just isn't fun for me when a group game... isn't.

But I like "immunity to fire", because it discourages focused builds. I like the 2e D&D toolkit fighter, with his vast array of weapons, picking the right one for the job, so much more than the 3e min-maxed single combat style fighter. I like the "all the spells" wizard so much more than the focused fire mage / enchanter / whatever. I like having rock, paper, and scissors, and being able to choose the tool that's best for the job at hand.

Pauly
2017-03-28, 08:32 PM
In my opinion RPGs don't realise the current technology enough. Everyone has a powerful computer
In their pockets nowadays...

A detailed and complex combat system has the problem that it is rules heavy and requires a lot of fiddling with little modifiers. An app should take care of this. I imagine it like this:
Instead of little numbers the GM and the players just add predefined tags to the actions they are trying to to do. For example in a modern setting a shootout against some thugs in the streets. The GM sets general encounter tags like #lightrain #moonlessnight #flickeringillumination #mediumwind.
Positioning and movement is still handled traditionally either theatre of mind or on a grid. On the players turn he decides to run behind a car and start spraying the thugs with his Uzi and after selecting his weapon from his digital char sheet he puts clicks the required tags #run15 #halfcover #drawuzi #fullauto #thug1 #thug2 #thug3 #distance20. The app now calculates the target number/dice pool and takes all into account and shows the players what to roll or rolls automatically. Then it calculates the effect of the hits and the damage done. This can be really detailed depending on ammo, amor, caliber, hit location...

Basically give players and DMs tags to throw at things. Calculate crazy tables and interdependencies under the hood.

Added bonus for the company: Since everyone needs the app to play the company might come up with subscriptions or another buissness model that allows it to profit from the game rules more then once.

The problem with this approach is that inputs end up being more important than outputs. i.e. The designer spends so much time trying get the right number for each variable that the output (hit/miss the target; damage/not damage the target) that the overall outcome is wrong.

This type of simulation only really works with long range gunnery duels between targets with limited ability to hide -WW2 naval combat is the only example I can think of. Although space ship combat is also possibly another one.

If you want to simulate outcomes the better models use a target number based on the situation, and then apply a small number of modifiers, no more than 5. Usually 3 modifiers are enough.
For example in a d20 environment. You might say:
Target number is X.
Easy situation - no modifier
Moderate difficulty +2 to the target number
Difficult: +4 to the target number
Extreme difficulty: +8 to the target number.

BarbieTheRPG
2017-03-28, 08:50 PM
I have been banging my head against the wall now for a long time trying to think about what I want the combat system to look like in a system I have been writing. Then I came to the realization that I've been having troubles not because I can't find mechanics to make the results I want, but because I don't really know what I want.

So, assuming that combat is a major deal in the game in question, what does a good combat system need to have, in your eyes? These can be as abstract and theoretical or as concrete as you want.
Speed. In combat, the fast can strike first and most often, but strength/stamina needs to play a role in some attacks, while it has a diminished effect versus slashing/piercing attacks. Magical attacks should fit one of those two, unless it's mental.

For me, I think the following are non-negotiable:


1. There must be meaningfully different options that support different play preferences.
Martials strike first. They have a "combat score" that enables them to not only strike first, but strike more than once within an action round. Multiple strikes do multiple damage. Mages need to concentrate their power so it's slow but can be as effective as multiple attacks.


2. There must be time or turns enough that a player can see something wrong with his choices and then rethink his tactics in order to win.
Well, this is just giving players the chance to rethink their action based on what has already occurred.


3. You must be able to do things that are cool and special and make the player feel powerful.
Allowing multiple attacks and having a "called shot" effect that allows characters to disable/maim foes works here.


4. It must not be clear who will win a combat before the first die is cast.
Reputation as a stat: so & so fighter trained at ___ and was the very best and defeated ____ so everyone knows about him. Take his/her Reputation and roll versus the opponent's Rep to see who grabs situational advantage (bonuses). If you have a fighter who's killed many opponents, and this is known to his/her opponent, it creates a psychological advantage that can be mimicked by a graded mechanic.

Vitruviansquid
2017-03-28, 09:55 PM
Lots of great responses to read through.

I think some people are getting a bit hung up on my non-negotiable #4. The fact is, I simply worded it in a sloppy manner.

Of course if a dragon fights a farmer, the fight should be decided before the first die is cast. What I meant was that you should not already know that a wizard will defeat a fighter by virtue of the wizard being a better class or a counter to the fighter or having a trick that the fighter cannot do anything against.

On the topic of mordhau and coup de grace, I like the term "coup de grace" because it describes a class of action, and not a specific action. Describing a class of action is good because it allows the player or GM to fill in what exactly happened in a way that is appropriate to the context. When you describe specific actions, you run into situations where they may not make sense, like how in DnD 4e a fighter has an attack that trips opponents, and can somehow inexplicably use it on a serpent opponent. Of course don't name anything outrageously poorly so that you mix up the meaning, but I thought that extra consideration for naming was important to mention.

On the topic of damage types and resistances, I think they definitely have the potential to make a mechanically interesting system. Unfortunately for most systems, their use of resistances and damage types are limited to making verisimulitude. I like the idea of damage types as a mechanic to do things like dissuade focused builds (as mentioned earlier), to make players feel good about putting the right tool in the right place, and give players some meaningful choices between different weapons (for example, swords might work against virtually anything while hammers are great against certain enemies but terrible against others, and such).

NichG
2017-03-28, 10:12 PM
In terms of #4, I tend to like a version of it that goes something like 'choices and events which can happen during combat should be important enough to overcome most power gaps that are expected to occur in play'.

Personally I hate when systems use randomness to achieve that. I want it to come from cleverness and skill. The idea that 'I shouldn't be able to win against this foe, but I figured out how to do it anyways' is very appealing to me.

A system where I could essentially give the GM a script with some precomputed default set of actions and still do as well as if I actually paid attention doesn't really make me feel that it's worth playing out or paying attention to the fight itself. In such a system, the meaningful gameplay seems to me to be before the fight actually begins - making sure to only pick fights where your character build guarantees a win.

It doesn't have to be 'I could always lose' - even something where some choices end up being situationally more resource efficient or things like that it's enough. But there has to be an actual thing where me engaging mentally could improve the outcome in some form.

Frozen_Feet
2017-03-29, 01:49 AM
@Vitruviansquid: again, remember the difference of system versus scenario design.

Think of Chess again. A Queen has play power of 9 and a Pawn has play power of 1. If those were the only two pieces on board, the Queen would win. But during course of normal play, where they aren't the only pieces, a Pawn can capture a Queen, and it is hard to say if the Queen will capture any specific Pawn.

A roleplaying game is not set before the scenario designer (usually the GM) has picked which pieces to use, how many, and their initial positions. In any sufficiently non-random game, this will naturally lead to transparently unwinnable scenarios (f.ex. "the Fighter can't win because Wizard can do X"). Think carefully about whether this is something you need or want to fix on the system's side.

RazorChain
2017-03-29, 04:21 AM
I want a system based on attack roll vs a defense roll and armor is subtracted from damage

I want a system with hit locations and injury mechanism where you can stun your opponent with heavy blows, cripple/maim by aiming for bodyparts. Knock somebody out by aiming for his head and where injury gives you penalties to your fighting abilities.

I want a combat system where movement and positioning matters. Striking somebody from his flank or rear gives him penalty to defend and using a weapon with better reach matters.

I want a system where fatigue and endurance play part, either as a resource to push yourself in battle or as a penatly when you get tired. The best is to use both.

I would like to see maneuvers that you can use in combat...defensive strikes, deceptive attacks, charge etc. and where you can chose to go on the offensive and sacrifice some defense or go defensive and sacrifice some offence. These are not feats that have to be learned but everyone can use but you could get better at it as your character gets better.

Environment should matter, like fighting on ice, knee deep in water or your opponent has higher ground. These can either give you bonus or penalty.

A system where close combat hand to hand, melee combat and ranged combat go hand in hand. If I want to use a beat to bludgeon my opponents weapon aside and follow up with a kick in the knee then it should be allowed in the system. I want to be able to get past my opponent defense and go for a armed grapple like in Talhoffer's Fechtbuch (a 15th century fighting manual) and your shield gives you advantages against ranged attackers as you can block incoming missiles.

jok
2017-03-29, 10:55 AM
The problem with this approach is that inputs end up being more important than outputs. i.e. The designer spends so much time trying get the right number for each variable that the output (hit/miss the target; damage/not damage the target) that the overall outcome is wrong.

This type of simulation only really works with long range gunnery duels between targets with limited ability to hide -WW2 naval combat is the only example I can think of. Although space ship combat is also possibly another one.

If you want to simulate outcomes the better models use a target number based on the situation, and then apply a small number of modifiers, no more than 5. Usually 3 modifiers are enough.
For example in a d20 environment. You might say:
Target number is X.
Easy situation - no modifier
Moderate difficulty +2 to the target number
Difficult: +4 to the target number
Extreme difficulty: +8 to the target number.

Why is that a problem for the designer? This would be a selling point.
A games like Shadowrun has potentially alot more then 5 modifiers. Between the environment, skills, magic, cyberware
and wireless cyberspace for a given situation. And every potential modifier is a part of the setting and can have narrative importance. In my opinion Shadowrun is on the edge of too rule complex. Things like detailed hit locations and interesting criticals or critical malfunctions of gear or detailed wounds would bloat too much.
The hypothetical app supported ruleset would allow for all of this and more, while keeping the required fiddling pretty low.

A typical attack would need the following tags:
State of attacker relative to target; targets; firemode/strike type; distance of targets; state of targets relative to attacker

Cluedrew
2017-03-29, 01:22 PM
I would not want to use the system you are describing with a phone. I would like a full sized keyboard and mouse for all that input. Even with clever use of tracking state, auto-completion and combined commands to update the system that is a lot of inputs. I mean you could do it but it would probably require a level of master to use quickly. And even then I'm not sure it would go as fast as one (more in your head) done mostly in your head with similar mastery.

To Psikerlord: I forgot to say thanks for the link. Thanks, its an interesting idea, I'm already floating related ideas to see if any fit in my work.

Quertus
2017-03-29, 02:35 PM
Speed. In combat, the fast can strike first and most often, but strength/stamina needs to play a role in some attacks, while it has a diminished effect versus slashing/piercing attacks. Magical attacks should fit one of those two, unless it's mental.

For me, I think the following are non-negotiable:


Martials strike first. They have a "combat score" that enables them to not only strike first, but strike more than once within an action round. Multiple strikes do multiple damage. Mages need to concentrate their power so it's slow but can be as effective as multiple attacks.

If you base initiative on the time it takes to perform an action, Telepaths* strike first, moving at the speed of thought. Then come the slow flesh bags, limited by the meat. Sure, then the casters, further limited by gathering external resources, can have a turn. Unless, of course, that's not how their casing mechanics work.

If, on the other hand, you base turn order on combat experience, skilled fighters and combat mages* can better read the scene and act first, while academia mages like Quertus, or civilians like his player, go last.

Actually, one of the combat mechanics I most enjoyed was from Shadowrun, where fast characters often got to act multiple times before regulars. Fast characters felt fast. So I certainly appreciate speed being able to make a character feel cool.

But, honestly, when I read the first word of your post, I thought you meant resolution speed.

* and chronomancers.


Allowing multiple attacks and having a "called shot" effect that allows characters to disable/maim foes works here.

Does it? I guess it depends on what people feel makes a character feel "cool". I've tried a few times to start a thread on the subject. Here are a few things I've considered cool in combat:

For me, Shadowrun iterative actions making fast characters feel fast felt cool. The pumped Street Samurai goes on 36, then on 26, then on 16, ... then some normals get to go, then (s)he goes again on 6. So much cooler than WoD Celerity, and other such belated iterative actions.

It felt really cool having my character use the ropes he was bound with as a makeshift garote to kill his captors. It really felt cool to have a monster that was too powerful to defeat in a head-on confrontation, and drop a boulder on it (this would obviously work in CoC, too).

I enjoy it when other people add flavorful descriptions to their characters' actions, involving back flipping off chandeliers into tea cups before putting daggers through people's eyes or something. My best experience with this was not from Exalted, or 7 Seas, or any other system designed with this in mind, but from D&D, which gives no benefit to such characterization. People did it because they wanted to, rather than it being forced out of them by the mechanics. That was good times.

Of course, my worst experience with it was also D&D, when the players were penalized for trying to make the game more fun with creative descriptions of their actions. I may have been the DM.

At the start of combat, Armus moves to protect someone tougher than himself (in D&D, that would be someone with better AC and more HP). If you can figure out why, you've seen through one of the likely 6 or so deceptions Armus will use to win a fight before it's even begun.

I want combat to have a metagame. I want tactics and deceptions to matter.

I love having cool toys. Lots of them. And I love the tactical minigame of choosing the beat tool for the job. But. If I'm in the mood to use a hammer, I want art Kat 90% of the encounters to be kind enough to look like a nail.

That is, if I'm in the mood to watch things burn with my trusty flamethrower, I want to watch things burn! This means several things. For one, very few things in the setting should have immunity to fire (or immunity to ranged, it immunity to technology, or immunity to AoE, or evasion, or protection from red). Second - and this is general best practice for encounter and adventure design - I don't want everything I face that session to all happen to have such immunities. Lastly, it means I want to have 3+ encounters per session, so that even if one encounter won't burn like a good nail, I can still hammer most of the encounters to ash. Sure, it may have been more efficient to utilize a different, optimized tactic of picking the perfect tool, but it was still possible.

But what does this mean for the system? Little, beyond perhaps that the ability to completely or even functionally eliminate a particular attack form shouldn't be common & trivial.

ExLibrisMortis
2017-03-29, 02:50 PM
Although I thoroughly enjoyed learning about "mordhau" and "coup de grace", I'm gonna touch on the other half of this post.

As I write this, someone is in the other room is playing a video game which has this rock scissors paper style of interaction, where pikemen deal +50% damage to cavalry, but only take 50% damage from cavalry, or something like that. And it feels very... artificial. The units aren't designed to be better by virtue of how they function (pikes have longer reach or whatever), but because of seemingly arbitrarily inflated numbers. Sure, if it's historic units, and you understand the underlying reasons, it can be ok, I guess, but I have no basis to understand why, say, a Hydralisk might deal extra damage vs, say, an AT-AT. I want their relative effectiveness to be encoded in the way they operate, not tacked on +-X% damage modifiers.
I agree for TTRPGs, but the level of abstraction that only considers one variable "strength" is valid in some games (campaign-level rather than battle-level games). For a TTRPG, you probably need to figure out why pike beats horse, and implement that. In D&D, reach is a part of it (pikes have 3 squares reach), as is bracing (double damage when readied against a charge). It results in a good damage bonus, which, at an abstract level, is all you care about.

jok
2017-03-29, 03:13 PM
I would not want to use the system you are describing with a phone. I would like a full sized keyboard and mouse for all that input. Even with clever use of tracking state, auto-completion and combined commands to update the system that is a lot of inputs. I mean you could do it but it would probably require a level of master to use quickly. And even then I'm not sure it would go as fast as one (more in your head) done mostly in your head with similar mastery.

To Psikerlord: I forgot to say thanks for the link. Thanks, its an interesting idea, I'm already floating related ideas to see if any fit in my work.

I agree that a phone would probably be too small. Tablets or PC then. A decent WI-FI tablet for 100$ should do it.

The virtual tabletop Fantasygrounds has alot of automatization for D&D 5e already build in. A game from the ground up build with app support in mind would open up alot of possibilities.

One would have to be smart about the inputs and the UI and menus ofcause.

CharonsHelper
2017-03-29, 03:38 PM
Then there's another thing which might be just me. A lot of systems implement rock scissors paper by having concepts like, "immunity to fire". My personal preference, in video games, was a video game that had the monsters be kind enough to take a single point of damage when they were "immune".

This means that themed characters, like a Fire Mage, suddenly don't get to play, when they encounter something that is immune to their shtick.

Personally, I hate everyone else sitting out while the net runner / face / heavy / whatever is doing their thing. No matter which of the three roles I'm in - GM, featured player, or inactive player - it just isn't fun for me when a group game... isn't.

But I like "immunity to fire", because it discourages focused builds. I like the 2e D&D toolkit fighter, with his vast array of weapons, picking the right one for the job, so much more than the 3e min-maxed single combat style fighter. I like the "all the spells" wizard so much more than the focused fire mage / enchanter / whatever. I like having rock, paper, and scissors, and being able to choose the tool that's best for the job at hand.

So - more like Pokémon where attacking something with an element that they're strong against deals 1/2 damage, and x2 damage with something they're weak against? (Though even Pokémon has a few immunities - ghost vs normal/fighting & dark vs psychic.)

Against a fire creature it'd be worth pulling out that water attack which you're sub-par with than continuing to use fire moves, but the fire moves wouldn't be totally useless either.

Plus - in the case of Pokémon most of the elements have an obvious reason of being more/less effective rather than being totally for out of game balance reasons.

(However - I'm not sure if I'd like a system which used the above system without it being a main pillar. It works in Pokémon because it's the primary pillar of the combat system, with a few lesser ones.)

erikun
2017-03-29, 04:20 PM
Capable of attacking a specific target. If a sniper is attacking, there needs to be some method of attacking the sniper directly in order to stop them.
If no one has a weapon capable of effectively engaging at the sniper's range, why should the system automatically allow the targets to return fire?
They cannot.

The list was for situations that the system must be able to resolve, if they come up and are relevant. If the party encounters a sniper on a ledge with no path up to it, and they don't have ranged weapons, then they have no way of attacking the sniper. (I made this specific example in my post.) Similarly, if the players are up on a ledge and the NPCs have no ranged weapons, then the NPCs would have no method of attacking any of the PCs.

On the other hand, if the party is in an open room and there are a bunch of enemies there, then they should be free to run up and attack one of them. Doing so means that choices of targets and targeting priority is meanful for the players. There should not be some sort of "total enemy HP level", or some sort of "victory rolls" that the party needs to roll. If that sort of a situation happens - if attacking one enemy is roughly the same as attacking another, if there is no way to eliminate an enemy past "winning the fight" - then it stops being a sensible combat system for a RPG. It starts turning into some sort of board game or video game, where the goal isn't "what would my character do?" but "how do I score the most points against the enemy?"

Quertus
2017-03-29, 04:58 PM
So - more like Pokémon where attacking something with an element that they're strong against deals 1/2 damage, and x2 damage with something they're weak against? (Though even Pokémon has a few immunities - ghost vs normal/fighting & dark vs psychic.)

Against a fire creature it'd be worth pulling out that water attack which you're sub-par with than continuing to use fire moves, but the fire moves wouldn't be totally useless either.

Plus - in the case of Pokémon most of the elements have an obvious reason of being more/less effective rather than being totally for out of game balance reasons.

(However - I'm not sure if I'd like a system which used the above system without it being a main pillar. It works in Pokémon because it's the primary pillar of the combat system, with a few lesser ones.)

Pokémon, I'm told, encourages you to play exactly two elements, by its combination of resistances and resource management. Playing only one is stupid, because you will almost certainly lose if you encounter something with resistance to your element. Playing more than 2 is stupid, because then you have a hard time getting the energy to power your pets, and you defeat yourself.

I want... Hmmm... The opposite? I want the most possible valid play styles. But, of those, I personally prefer having the full toolkit. Which is hard to do in Pokémon, or MtG, but easy to do in (2e) D&D.

CharonsHelper
2017-03-29, 05:27 PM
Pokémon, I'm told, encourages you to play exactly two elements, by its combination of resistances and resource management. Playing only one is stupid, because you will almost certainly lose if you encounter something with resistance to your element. Playing more than 2 is stupid, because then you have a hard time getting the energy to power your pets, and you defeat yourself.

Maybe the card game (never played it) but the video game encourages you to diversify.

Pauly
2017-03-29, 07:54 PM
Why is that a problem for the designer? This would be a selling point.
A games like Shadowrun has potentially alot more then 5 modifiers. Between the environment, skills, magic, cyberware
and wireless cyberspace for a given situation. And every potential modifier is a part of the setting and can have narrative importance. In my opinion Shadowrun is on the edge of too rule complex. Things like detailed hit locations and interesting criticals or critical malfunctions of gear or detailed wounds would bloat too much.
The hypothetical app supported ruleset would allow for all of this and more, while keeping the required fiddling pretty low.

A typical attack would need the following tags:
State of attacker relative to target; targets; firemode/strike type; distance of targets; state of targets relative to attacker

The problem is that complex models are really bad at predicting outcomes, with the exception being physical models such as aerodynamic testing. And even then physical models only work well in stable controllable environments.

The starting point for a model of combat is the skill of the attacker and defender, so a complex model just doesn't work. All your model will tell you is the model designers assumptions, which is what we see in economic and financial modelling, government spending models.
Another problem for complex models is the exact interaction between specific variables. Lets say you are shooting and you have a +2 modifier for cover and a +2 modifier for rain. What Is the modifier for shooting at a target in cover when it is raining? Game designers will tell you +4, but it may be +2 (there is no extra benefit), +3 (there is a small additional benefiy), +4 (there is a linear relationship), or +5 (there is a synergic benefit).
Unless you have all the possible combinations modelled perfectly then your system is wrong, and chaos theory takes your model into the weeds.

The problem with the app isn't that the computer or the interface. The problem is the model by its very nature is incable of delivering realistic (even by fantasy RPG standerds) outcomes.
Outcome based systems using a small number of modifiers get to the end result faster and more accurately than complex systems.

jok
2017-03-30, 12:25 AM
The problem is that complex models are really bad at predicting outcomes, with the exception being physical models such as aerodynamic testing. And even then physical models only work well in stable controllable environments.

The starting point for a model of combat is the skill of the attacker and defender, so a complex model just doesn't work. All your model will tell you is the model designers assumptions, which is what we see in economic and financial modelling, government spending models.
Another problem for complex models is the exact interaction between specific variables. Lets say you are shooting and you have a +2 modifier for cover and a +2 modifier for rain. What Is the modifier for shooting at a target in cover when it is raining? Game designers will tell you +4, but it may be +2 (there is no extra benefit), +3 (there is a small additional benefiy), +4 (there is a linear relationship), or +5 (there is a synergic benefit).
Unless you have all the possible combinations modelled perfectly then your system is wrong, and chaos theory takes your model into the weeds.

The problem with the app isn't that the computer or the interface. The problem is the model by its very nature is incable of delivering realistic (even by fantasy RPG standerds) outcomes.
Outcome based systems using a small number of modifiers get to the end result faster and more accurately than complex systems.

This was never intendet to simulate reality.
Besides models for economics or weather are still used and are useful, even if they might be biased...

Pauly
2017-03-30, 01:02 AM
This was never intendet to simulate reality.
Besides models for economics or weather are still used and are useful, even if they might be biased...

Complex models almost never beat 'the long term trend will continue', and even then if one gets it right many are wrong you are left with it being a blind squirrel finding a nut. And there are many cases of complex models being completely wrong.

So unless a complex model can reliably provide a superior outcome to a simple model, then it is best use the simple model.

That's before considering ease and speed of use. Many players don't want to spend their brain processing capacity on being aware of and and choosing between myriads of possible modifiers even if the heavy calculation is being done by the computer. The players still have to know the modifiers to make an informed decision.
Compare that to if you do X it will be an easy shot for you, but you will be an easy target and if you do Y you will be a very difficult target, but your shooting will be moderately difficult.

CharonsHelper
2017-03-30, 07:37 AM
This was never intendet to simulate reality.
Besides models for economics or weather are still used and are useful, even if they might be biased...

The problem is that they ignore tons of variables - and those variables are often the most important things.

Often someone will look back and see that one model worked - and ignore the three dozen which failed horribly. (Reminds me of basketball brackets.)

Max_Killjoy
2017-03-30, 09:28 AM
The problem is that they ignore tons of variables - and those variables are often the most important things.

Often someone will look back and see that one model worked - and ignore the three dozen which failed horribly. (Reminds me of basketball brackets.)

Most humans in general seem to have a very high "positive outcome" bias. It's not the 10 times that a tactic or prediction or method failed that they remember, it's the 1 time that succeeded that sticks in their head.

CharonsHelper
2017-03-30, 09:49 AM
Most humans in general seem to have a very high "positive outcome" bias. It's not the 10 times that a tactic or prediction or method failed that they remember, it's the 1 time that succeeded that sticks in their head.

Or the reverse when something especially bad happens. Hence the myth of the full moon causing more crime which some police officers swear by - but the actual numbers prove to be untrue.

Because they remember that time it was crazy & the full moon, but they forget the dozen times the same year that crime was normal/low on the full moon.

Max_Killjoy
2017-03-30, 10:46 AM
Or the reverse when something especially bad happens. Hence the myth of the full moon causing more crime which some police officers swear by - but the actual numbers prove to be untrue.

Because they remember that time it was crazy & the full moon, but they forget the dozen times the same year that crime was normal/low on the full moon.

Correlation bias in general. But yes, I was referring to the subset that amounts to irrational optimism.

Knaight
2017-03-30, 10:59 AM
The problem is that they ignore tons of variables - and those variables are often the most important things.

Often someone will look back and see that one model worked - and ignore the three dozen which failed horribly. (Reminds me of basketball brackets.)

Weather models are accurate more times than not*, and the claims of inaccuracy are usually along the lines of "they said there was an 80% chance this would happen, and it didn't", which is where the difference between 80% and 100% gets relevant.

*Modern ones anyways, 1950's computational fluid mechanics was less than impressive.

CharonsHelper
2017-03-30, 11:14 AM
Weather models are accurate more times than not*, and the claims of inaccuracy are usually along the lines of "they said there was an 80% chance this would happen, and it didn't", which is where the difference between 80% and 100% gets relevant.

*Modern ones anyways, 1950's computational fluid mechanics was less than impressive.

I was thinking more for the economics ones.

Knaight
2017-03-30, 11:19 AM
I was thinking more for the economics ones.

Those I'm not particularly interested in defending.

NichG
2017-03-30, 12:49 PM
The reason complex models are more frequently wrong has more to do with overfitting than that they're missing variables and aren't yet complex enough.

If you have some set of historical data and you propose a 1-parameter model, fit that parameter, and find that the result matches the historical data well then there's a good chance your model captured something fundamental. If you propose a 100-parameter model, the space of possible 100 parameter models is so huge that the chance that you found something that worked by coincidence rather than because it captured something real increases rapidly with model complexity.

There are methods which let you test this, which generally involve holding back some of the data and seeing if the model still works on that held-out data. But the thing is, lets say you do this test and your model fails - are you going to stop proposing models? If you then propose a new model, that held-out data isn't independent anymore. That's one problem things can run into. Another problem has to do with the independence assumptions underlying the selection of the hold-out set. If you're making a forecasting model, maybe you would take one interval of 10 years to fit the model and another interval of 10 years to test it. Or maybe you'd take alternating months. It turns out that these things are not equivalent - if you take alternating months, you're much more likely to miss overfitting because nearby months aren't statistically independent of eachother. However, for very long-term processes (compared to human timescales), you can't ever really get something statistically independent to test against, so there's always some danger that even if your model works on the hold-out set it's actually just overfit.

The hope in scientific approaches is that by deriving things from first principles you decrease the free parameter count and give yourself less rope to hang yourself. The hope in machine learning approaches is that you can use regularization and cross-validation to systematically make approaches which work even as you asymptotically approach an infinite number of free parameters (basically, using ways of mathematizing Occam's Razor and letting the model figure out how many parameters it can actually support).

Anyhow, point being, if nothing else added complexity in models drastically increases the chance of making a mistake in methodology. So it's no surprise that you see a lot of cases where adding complexity makes things worse. It's hard (but not impossible) to actually do it in a way that makes things better.

For a game designer though, this is all more or less irrelevant because there's no ground truth outcome to compare with. The important thing when adding complexity in a game is to make that complexity stay meaningful to the player. If there's a complex simulation under the hood, but it's all so compartmentalized that the player need not try to understand it (nor benefit much from doing so), that complexity is superfluous. In order for a game to support a high degree of complexity, it also needs to do a good job in clarity and explanation.

BayardSPSR
2017-03-30, 03:00 PM
So unless a complex model can reliably provide a superior outcome to a simple model, then it is best use the simple model.

Yes!

At the same time, though, games aren't just models. You could have an RPG combat system that boiled down to a single d% roll, or even one that was completely deterministic, that was a fine model of reality (or "reality"). I'd be happy with something like that, since I usually like my RPGs focused on the RP. But I suspect quite a few people would be dissatisfied with the lack of game there, and would prefer a "model" that didn't emphasize verisimilitude or simplicity.

Max_Killjoy
2017-03-30, 03:07 PM
Yes!

At the same time, though, games aren't just models. You could have an RPG combat system that boiled down to a single d% roll, or even one that was completely deterministic, that was a fine model of reality (or "reality"). I'd be happy with something like that, since I usually like my RPGs focused on the RP. But I suspect quite a few people would be dissatisfied with the lack of game there, and would prefer a "model" that didn't emphasize verisimilitude or simplicity.

IMO, it's always a balancing act between verisimilitude (a swordfight should feel at least kinda like a swordfight, and not like craps or doing taxes) and practicality (a perfect simulation using dice could take hours to work out a single exchange of blows) and avoiding utter predetermination AND utter arbitrariness.

Vitruviansquid
2017-03-30, 11:09 PM
Since I am not designing a game to go with a phone or computer, a large number of modifiers is bad.

I am also not a fan of many modifiers because most dice systems (including d20) have a habit of breaking at either extreme ends.

I am also not a fan of many modifiers because modifiers tend to allow synergies that make results unpredictable. While synergies can be fun, I am on the side that prefers balance and ease of use for the GM.

Frankly, I don't really see a good use of designing a system in order to have a large number of modifiers. To use many modifiers just because you are able to process them seems like a kind of folly to me.

CharonsHelper
2017-03-31, 08:42 AM
I am also not a fan of many modifiers because modifiers tend to allow synergies that make results unpredictable. While synergies can be fun, I am on the side that prefers balance and ease of use for the GM.

It's basically a variation of the design conundrum of customization vs balance. Customization is awesome - but it always leads to far more balance difficulties for the designer as the capabilities (or lack thereof) of the PCs become more and more unpredictable.

A well designed customization system can limit the power level differences, but they'll still be there to some degree.

Max_Killjoy
2017-03-31, 09:00 AM
It's basically a variation of the design conundrum of customization vs balance. Customization is awesome - but it always leads to far more balance difficulties for the designer as the capabilities (or lack thereof) of the PCs become more and more unpredictable.

A well designed customization system can limit the power level differences, but they'll still be there to some degree.

IMO, that requires enough granularity that the customization doesn't swamp the basic mechanics.

Systems that use a single d6 have a lot less room for details than systems that use, oh, 2d10 or 3d8 or whatever. +1 is huge on a 1-6 linear distribution, not so much on a 3-24 curve.

Then you get these systems using a die pool instead of a fixed curve, which can have two axes of adjustment (if each die in the pool is pass/fail, both the threshold for "pass" and the number of "passing" dice needed can fluctuate).

Zombimode
2017-03-31, 09:16 AM
I want a system based on attack roll vs a defense roll and armor is subtracted from damage

I want a system with hit locations and injury mechanism where you can stun your opponent with heavy blows, cripple/maim by aiming for bodyparts. Knock somebody out by aiming for his head and where injury gives you penalties to your fighting abilities.

I want a combat system where movement and positioning matters. Striking somebody from his flank or rear gives him penalty to defend and using a weapon with better reach matters.

I want a system where fatigue and endurance play part, either as a resource to push yourself in battle or as a penatly when you get tired. The best is to use both.

I would like to see maneuvers that you can use in combat...defensive strikes, deceptive attacks, charge etc. and where you can chose to go on the offensive and sacrifice some defense or go defensive and sacrifice some offence. These are not feats that have to be learned but everyone can use but you could get better at it as your character gets better.

Environment should matter, like fighting on ice, knee deep in water or your opponent has higher ground. These can either give you bonus or penalty.

A system where close combat hand to hand, melee combat and ranged combat go hand in hand. If I want to use a beat to bludgeon my opponents weapon aside and follow up with a kick in the knee then it should be allowed in the system. I want to be able to get past my opponent defense and go for a armed grapple like in Talhoffer's Fechtbuch (a 15th century fighting manual) and your shield gives you advantages against ranged attackers as you can block incoming missiles.

There are 5 letters that deliver the combat experience that you describe: GURPS

Vitruviansquid
2017-03-31, 11:10 AM
IMO, that requires enough granularity that the customization doesn't swamp the basic mechanics.

Systems that use a single d6 have a lot less room for details than systems that use, oh, 2d10 or 3d8 or whatever. +1 is huge on a 1-6 linear distribution, not so much on a 3-24 curve.

Then you get these systems using a die pool instead of a fixed curve, which can have two axes of adjustment (if each die in the pool is pass/fail, both the threshold for "pass" and the number of "passing" dice needed can fluctuate).

Granularity is easy to come by. Increase die size, increase numbers.

Going for more granularity in order to increase the numbers of modifiers, however, then runs you into the problem where modifiers stop being individually impactful, which is bad in its own way.

Max_Killjoy
2017-03-31, 11:16 AM
Granularity is easy to come by. Increase die size, increase numbers.

Going for more granularity in order to increase the numbers of modifiers, however, then runs you into the problem where modifiers stop being individually impactful, which is bad in its own way.

You're absolutely right -- a d100 system with a bunch of +1 and -1 modifiers can be just as tedious, and the modifiers meaningless instead of overwhelming.

KorvinStarmast
2017-03-31, 12:07 PM
There are degrees here.
1) Mechanical core
2) Simplicity
3) Verisimilitude
4) Depth
Not complexity. Depth. The two are different.
5) Accounting for the enviroment
6) Modularity
I tip my cap to your answer, in detail. *golf clap*