Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games
  • Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.

Phase-based combat in 1v1 fighting

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
Imagine a standard TB combat system, where the player chooses actions for their character and the character attempts to execute those actions. The outcome is determined by the character's stats, the enemy's stats, and RNG.

With traditional TB the characters alternate turns. I attack, you defend. You attack, I defend. And so on.

However one twist you can put on the system is to resolve actions simultaneously using a phase-based approach. Each phase has two parts -- in part one we both give our characters orders, and in part two they compete to execute those orders while also defending against each other. This is a less common approach, but has been done before.

So imagine a simple boxing match between two fighters, call them Ali and Frazier. The fighters have several stats, e.g. punching power, speed, stamina, and ability to resist damage. They also have a few types of actions they can take, e.g. punches like jab, cross, hook, and uppercut.

A round in a boxing match lasts 3 minutes, so if you break each round into say 30 second increments, that gives you 6 phases of combat per round. Now let's say the player can assign 3 actions per phase, and also set some overarching general strategies, such as how much distance the fighter should try to keep from the opponent.

Therefore an example phase of combat might be like this:

Ali, being the longer and faster fighter, does best when using straight punches and keeping to the outside. His player instructs him to Jab, Jab, Cross and sets his distance to "outside".

Frazier, on the other hand, is more suited to fighting up close and throwing for power. His player instructs him to Jab, Hook, Uppercut and sets his distance to "inside".

Make sense so far? Now my questions are, given all this:

- How does the combat system resolve which fighter is able to control the distance?

- How does the combat system resolve which punches hit?

- How do we handle major changes to the game state that occur partway through a phase?

For instance let's say Frazier's second punch (Hook) stuns Ali badly and nearly knocks him down -- a major event that would prompt a change in strategy for both fighters. Ali now wants to focus on defending himself and avoiding further damage, while Frazier wants to throw caution to the wind and unleash as much offense as he can. So in a case like this does it make sense to interrupt the phase and allow for new inputs from the players?

I can visualize how all this would work in traditional TB but I'm just not clear on how different it would / should be with a phase-based implementation.
 
Last edited:

LESS T_T

Arcane
Joined
Oct 5, 2012
Messages
13,582
Codex 2014
I think you should look at Toribash:







I've only played a few matches so don't have a lot of insights about this, but matches some of your description pretty well.
 

DraQ

Arcane
Joined
Oct 24, 2007
Messages
32,828
Location
Chrząszczyżewoszyce, powiat Łękołody
If you're controlling one character only, any sort of TB is effectively pointless at best and actively counterproductive at worst.
Remember that TB, whether alternating or simultaneous execution, is not free of cost - the cost is first and foremost lowered responsiveness (reactions are delayed up to at most turn's length in PB, 2x turn's length in TB) which is prohibitive in things like mechanically deep 1on1 melee.

Maybe if you could extend the control to the degree where player would control even minute aspects of things like footwork it could become attractive despite this added cost but it would also become awfully fiddly, require hiring and unconditionally obeying martial arts experts consultants just to not fuck it up royally and the built in delays would hurt even more - the AI would be hard as fuck to write as well.
 
Last edited:

Lord Rocket

Erudite
Joined
Feb 6, 2008
Messages
1,089
- How does the combat system resolve which fighter is able to control the distance?

- How does the combat system resolve which punches hit?

- How do we handle major changes to the game state that occur partway through a phase?

- RNG (skill/combat rolls)

- RNG (skill/combat rolls) <-- why should these be different from the way normal TB does them? Are you making a turn-based physics simulation game or something?

- The simplest solution would just be to queue up one action per phase and thus neatly sidestep the whole thing, but in your example what (I think) you're describing is Frazier getting a critical hit, so in that case you could add some kind of 'interrupt for new orders on critical result' type mechanic
 

Zednick

Educated
Joined
Apr 19, 2017
Messages
40
I think this will be of interest to you:



It's card-based game where each player selects a card, and the outcome is decided by the interplay of cards. It has one action per turn, which is probably what you want for simplicity sake. Ali doesn't think, 'In a minute and a half I might throw an uppercut.'

The best turn-based combat board game I know of is Title Bout Boxing, which also has an aging computer version. The mechanics in the digital version aren't displayed for the user though, IIRC.
 

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
If you're controlling one character only, any sort of TB is effectively pointless at best and actively counterproductive at worst.
Remember that TB, whether alternating or simultaneous execution, is not free of cost - the cost is first and foremost lowered responsiveness (reactions are delayed up to at most turn's length in PB, 2x turn's length in TB) which is prohibitive in things like mechanically deep 1on1 melee.

Maybe if you could extend the control to the degree where player would control even minute aspects of things like footwork it could become attractive despite this added cost but it would also become awfully fiddly, require hiring and unconditionally obeying martial arts experts consultants just to not fuck it up royally and the built in delays would hurt even more - the AI would be hard as fuck to write as well.

I don't really understand the point you are trying to make. Do you mean to say that TB doesn't work without a party? If so then I don't agree with that assertion at all.

The concept I am working on is actually for mixed martial arts (MMA) -- which is quite a bit more complex than boxing -- but I don't think the complexity has to extend to the nth degree to be meaningful. In most cases a fighter doesn't have more than five or six reasonable options to choose from, but the sport is still very tactically rich.
 

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
I think this will be of interest to you:



It's card-based game where each player selects a card, and the outcome is decided by the interplay of cards. It has one action per turn, which is probably what you want for simplicity sake. Ali doesn't think, 'In a minute and a half I might throw an uppercut.'

The best turn-based combat board game I know of is Title Bout Boxing, which also has an aging computer version. The mechanics in the digital version aren't displayed for the user though, IIRC.


I will check this out.

I think really the heart of my question is about the interplay aspect of the mechanics. Meaning how do Ali's actions affect Frazier's and vice versa. It might not be a big issue in the end, idk.
 

Ninjerk

Arcane
Joined
Jul 10, 2013
Messages
14,323
It would have to be constructed differently for your example but the following game has something like what you are describing.


Your little soldiers "lock on" to enemies if they are set to engage and they spot an enemy. Different soldiers have different "reaction times"--shotgun dudes shoot the fastest but obviously have limited range while on the other end snipers can engage from long distances but take quite a good deal of time to engage.
 

Jokzore

Arbiter
Joined
Mar 18, 2017
Messages
623
Aside from the previously mentioned toribash, which should answer all those questions, isn't this how more traditional roguelikes work? Dungeons of Dredmor for example.

EDIT:
Make sense so far? Now my questions are, given all this:

- How does the combat system resolve which fighter is able to control the distance?

- How does the combat system resolve which punches hit?

- How do we handle major changes to the game state that occur partway through a phase?

- Both control the distance because both move simultaneously. In your example they'd be at the same distance they started because one wants to move further out and the other further in. The biggest problem here would be what happens if your system allows one to move and the other to hit in the same phase/round. In that case one would have to be given the advantage.

-I think that's resolved independently. If you hit you hit. Maybe it's both, maybe it's neither, maybe its just one.

-There's two (and a half) solutions i see here:

1) Both players have to lock in all 3 of their moves. They have to rely on their ability to predict what their opponent will do.

2) Each new move is chosen after the previous one is resolved. This allows you to react to what's happening.

2.5) You can have a hybrid system where you allow a player to have a limited amount of times they can interrupt a sequence and change their moves. This would mean they still have to predict their opponents moves but also gives them the opportunity to react in case something disastrous happens.
 
Last edited:

PrettyDeadman

Guest
I think you should look at Toribash:







I've only played a few matches so don't have a lot of insights about this, but matches some of your description pretty well.


It's a good game. I played it from 1.7 I believe, evne was like top 3 at some point. But I feel like it lost its steam way before it was released on steam.
Still, if anyone wants to play it pm me.
 

DraQ

Arcane
Joined
Oct 24, 2007
Messages
32,828
Location
Chrząszczyżewoszyce, powiat Łękołody
I don't really understand the point you are trying to make. Do you mean to say that TB doesn't work without a party?
I'm saying that in this case TB has definitive cost without adding anything.

TB adds latency to player's tactical decisions (you don't get to react to anything before the next turn - that's effectively input lag corresponding to turn's nominal duration) on game's side and incurs time overhead on player's side (sweet time player uses to analyse situation, and time spent on advancing to the next turn). If the game uses any sort of ordering of actions instead of true simultaneous resolution, there is also overhead resulting from serializing actions that would normally be parallel (scaling proportionally to the number of participants) and logical/mechanical distortions resulting from the fact that actions that would be simultaneous become sequential (that may be exploitable and require additonal patching/workarounds like overwatch system.

Those are very real costs of TB. Whether or not TB is worth it depends on whether or not it can offer anything to make up for those costs or make them irrelevant - for example:
  • In party based game TB cancels out inaccuracies and timing differences resulting from trying to control multiple characters at once. It also eliminates overhead of having to spam pause constantly.
  • In a game where turn durations are much longer in game time than time spent in each turn in user's time (days, months, even minutes), TB will not incur overhead and actually speed up the game considerably without forcing the game to work at slapstick pace (that's one of the reasons some people really like tile-based crawlers - you can traverse the levels as fast as you can click).
  • In a game where required and sufficient reaction times are long enough to be reasonable turn durations the lag doesn't matter
Individual melee combat is fast paced, with desirable reaction times on the order of tens of ms. Without party to allow positioning and cooperative tactics take the edge off split second decisions and allow player higher level on which they can affect the situation all TB does is reducing depth, making combat clunky and limiting scale of events game can handle.
What's the upside?
 

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
DraQ are you advocating for RTwP in this case? I am personally not a fan of RTwP due to clusterfuck it generally devolves into, but a 1v1 fight might be manageable enough for it to work.
 

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
What I want to avoid at all costs is making a "fighting game". For one because I don't want to, and also because nothing I could possibly do in that space would be competitive with EA.

I like the idea of approaching MMA as a tactics game with RPG mechanics. However traditional TB feels too clunky in this context because with 1v1 you can't really model initiative, AoOs, or any of the other abstractions that make TB fun and somewhat compatible with real life. It would be boring, akin to two guys just standing in front of each other swinging hammers.

Phase-based however gets us closer to the truth. You choose a course of action that you think will put you closer to winning, then match that against the opponent to see who can impose their will. Visualize a fight as a sequence of phases -- 30 seconds or 10 seconds or 1 second for that matter. It's a good way to model the thing we are trying to model.

So I think there potential there but I'm not sure how to properly abstract it in a way that is fun and meaningful.
 

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
-There's two (and a half) solutions i see here:

1) Both players have to lock in all 3 of their moves. They have to rely on their ability to predict what their opponent will do.

2) Each new move is chosen after the previous one is resolved. This allows you to react to what's happening.

2.5) You can have a hybrid system where you allow a player to have a limited amount of times they can interrupt a sequence and change their moves. This would mean they still have to predict their opponents moves but also gives them the opportunity to react in case something disastrous happens.

I like these ideas and I will admit that the 30 second phase / 3 actions per phase approach was just a blind guess on my part. Certainly not something I am married to.

To back it up a step, the situation is that a standard MMA fight by US rules lasts for 3 x 5 minute rounds, i.e. 15 minutes total of fighting. So the design question is to what extent do you want to break that up and give the player opportunities to re-evaluate and issue new commands?

Give the player too many inputs and the game becomes an auto-pausing slog. Give them too few and they lose agency, being relegated to watching the game play itself.

Granting multiple actions per phase felt like a way to increase the player's inputs without increasing pauses, and possibly open up more complex strategies. But it also could make things too abstract to really be useful.

Keep in mind that you could -- and I'm guessing almost certainly should -- increase the speed of the clock so that a 5 minute round doesn't actually take 5 minutes. Just another thing to consider in terms of pacing.
 

adrix89

Arbiter
Joined
Dec 27, 2014
Messages
700
Location
Why are there so many of my country here?
Shigatari has 1vs1 duel combat based on actual timing, go check it out. It's a bit like the combat in Dwarf Fortress Adventure Mode.
I think this system could work well if you added a spatial component and could even work with multiple opponents.

 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,346
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
You could also try Kongai on Kongregate. I think it was designed by Sirlin too (the guy who did Puzzle Strike and yomi).
 

Stormcrowfleet

Aeon & Star Interactive
Developer
Joined
Sep 23, 2009
Messages
1,020
It's farm from being an RPG, but as far as modern setting goes, Breach and Clear has a phase thing where all action are simultaneous.
 

Grumpy Grognard

Inn Between Worlds
Developer
Joined
Dec 5, 2013
Messages
103
Location
Grizzled Gnoll's Gorge
Imagine a standard TB combat system, where the player chooses actions for their character and the character attempts to execute those actions. The outcome is determined by the character's stats, the enemy's stats, and RNG.

With traditional TB the characters alternate turns. I attack, you defend. You attack, I defend. And so on.

However one twist you can put on the system is to resolve actions simultaneously using a phase-based approach. Each phase has two parts -- in part one we both give our characters orders, and in part two they compete to execute those orders while also defending against each other. This is a less common approach, but has been done before.

So imagine a simple boxing match between two fighters, call them Ali and Frazier. The fighters have several stats, e.g. punching power, speed, stamina, and ability to resist damage. They also have a few types of actions they can take, e.g. punches like jab, cross, hook, and uppercut.

A round in a boxing match lasts 3 minutes, so if you break each round into say 30 second increments, that gives you 6 phases of combat per round. Now let's say the player can assign 3 actions per phase, and also set some overarching general strategies, such as how much distance the fighter should try to keep from the opponent.

Nice description. Odd Gods has a phase-based combat system, pretty similar to what you described. I described it ain a bit of detail over in this thread:

http://www.rpgcodex.net/forums/inde...pg-about-the-1990s.118969/page-3#post-5363817

Briefly though: our phases are 3 'ticks' long (a tick is equivalent to a fast, light attack or other ability, or moving one step/grid square), and we have unarmed combat, so there's some common ground to start with.

Therefore an example phase of combat might be like this:

Ali, being the longer and faster fighter, does best when using straight punches and keeping to the outside. His player instructs him to Jab, Jab, Cross and sets his distance to "outside".

Frazier, on the other hand, is more suited to fighting up close and throwing for power. His player instructs him to Jab, Hook, Uppercut and sets his distance to "inside".

Make sense so far? Now my questions are, given all this:

- How does the combat system resolve which fighter is able to control the distance?

- How does the combat system resolve which punches hit?

When you say 'control the distance', I assume you mean - who gets a hit in if say, two fighters attack in exactly the same way in the same phase?

There's obviously a few ways to solve it - but for OG, we decided a stalemate like this is a feature, not a bug. Effectively, those jabs cancel each other out. But they're not without consequences.

A stalemate doesn't deal damage, but has a cost or a tactical out come. It may result in a grapple (and subsequent detachment mechanics in the next combat phase), stamina loss to both combatants, a mild knockback to separate the opponents, and so forth.

The other factor here is movement. Plain old positioning tactics, flanking and so on come into play - you didn't mention movement explicitly, but in our system you could move two squares forward, and jab. Or jab 3 times if you've got the stamina, so you could still have a straightforward battle of attrition.

You could have the resolution purely stat-based, of course (frazier and ali jab, but ali's jabs are faster and thus always out-jab Frazier, but maybe Frazier could perform an ali-speed jab if he invested more stamina, which could be a cool risky proposition). FWIW, I'd argue that phase-based combat's primary strength is deterministic tactics. In a purely deterministic system, Frazier could never out-jab Ali. He'd have to play to his own strengths - clever defending against the fast attacks, feinting, trying to get ali to over commit and leave an opening, etc.

- How do we handle major changes to the game state that occur partway through a phase?

It can be tricky... In our system, we just let them happen and deal with it systemically. e.g. if two people try to move into the same spot, they bump into each other. A heavier/stronger character will 'win' a bump in that scenario, and potentially trigger a status effect, like a knockback, or an off-balanced state etc.

For instance let's say Frazier's second punch (Hook) stuns Ali badly and nearly knocks him down -- a major event that would prompt a change in strategy for both fighters. Ali now wants to focus on defending himself and avoiding further damage, while Frazier wants to throw caution to the wind and unleash as much offense as he can. So in a case like this does it make sense to interrupt the phase and allow for new inputs from the players?

I can visualize how all this would work in traditional TB but I'm just not clear on how different it would / should be with a phase-based implementation.

I'm not against phase-interrupts in principle, and you could allow for it in that case - my concern that it would function as a 'get out of jail free' card when shit goes bad.

In OG, we just let that scenario play out, basically. If you get knocked down, you're pretty screwed - but only if you stay down. Being knocked down means you can't do a lot of stuff - but anyone can roll left and right, get up, attempt to attack from the ground, trip your opponent, wrestle etc.
 

Egosphere

Arcane
Joined
Jan 25, 2018
Messages
1,909
Location
Hibernia
I thought of a similar phase based system for something like Jagged Alliance 2/Fallout, and I think it could work.
Instead of taking turns to direct your team, open fire, go prone etc. you simply state all of your commands and then watch the phase unravel as your team and the opposition tries to execute their plans simultaneously. It could add a lot of depth to the gameplay. For example, if your team member ( call him A ) targets an opponent ( call him B ), then what your opponent does during that phase will affect his chances. If B dashes across the map, hitting him will be harder ( unless A has high Perception, Dexterity, great weapon handling skills etc. ). But if B stands tall in order to throw a grenade at you, then A has a clear shot and the chances of getting a critical hit go through the roof. You could employ many tactics in this scenario: suppressie fire, throwing grenades in order to get a breather and rearrange positioning etc.
 

J1M

Arcane
Joined
May 14, 2008
Messages
14,616
DraQ are you advocating for RTwP in this case? I am personally not a fan of RTwP due to clusterfuck it generally devolves into, but a 1v1 fight might be manageable enough for it to work.
Draq holds the fringe opinion that all games would be improved by increasing simulation elements. It's like a weird religion.
 

Zanzoken

Arcane
Joined
Dec 16, 2014
Messages
3,558
Fair enough. I think simulation is good but it's only one aspect of design. Even if you like crunchy systems with lots of complexity (which most of us here probably do) you still need abstraction for pacing, balance, difficulty, etc.

The more I think about it though, the more I think RTwP might be a better approach for this game. 1-v-1 is much less tactically demanding than party-based or war games, which means 1) it's easier to manage and 2) since there is less to do and think about you need a faster pace so the combat doesn't get boring. I also like the idea of the Speed stat affecting the rate at which the fighters can perform actions.

And to be perfectly honest I am more interested in the management sim aspects of the game than the fighting itself. That's not to say the combat shouldn't be good -- it absolutely should. But the core vision is not to role play a fighter -- it's to be in the position of owning your own MMA gym and managing / coaching a stable of fighters that you are trying to build into champions. RTwP would likely result in less time and emphasis spent on fighting -- in fact, I would almost certainly want to include an option to sim them entirely.
 

Norfleet

Moderator
Joined
Jun 3, 2005
Messages
12,250
Individual melee combat is fast paced, with desirable reaction times on the order of tens of ms. Without party to allow positioning and cooperative tactics take the edge off split second decisions and allow player higher level on which they can affect the situation all TB does is reducing depth, making combat clunky and limiting scale of events game can handle. What's the upside?
Yeah, turn-based combat when there is only one character you control tends to just be unnecessarily clunky at best and exploit-ridden at worst. Everyone loves Fallout, but Fallout is pretty much a textbook case of it being both clunky and exploit-ridden. The only way I can see TB being particularly great in a single-character situation is if the game revolves around duels and the actions and turns are sliced really fine as a result.

I think you should look at Toribash:
Yeah, that pretty much fits what I described: 1v1 where the turns are sliced really fine. At the level of abstraction of a typical game, TB adds nothing to single-character play.
 
Joined
Sep 18, 2013
Messages
1,258
I don't really understand the point you are trying to make. Do you mean to say that TB doesn't work without a party?
I'm saying that in this case TB has definitive cost without adding anything.

TB adds latency to player's tactical decisions (you don't get to react to anything before the next turn - that's effectively input lag corresponding to turn's nominal duration) on game's side and incurs time overhead on player's side (sweet time player uses to analyse situation, and time spent on advancing to the next turn). If the game uses any sort of ordering of actions instead of true simultaneous resolution, there is also overhead resulting from serializing actions that would normally be parallel (scaling proportionally to the number of participants) and logical/mechanical distortions resulting from the fact that actions that would be simultaneous become sequential (that may be exploitable and require additonal patching/workarounds like overwatch system.

Those are very real costs of TB. Whether or not TB is worth it depends on whether or not it can offer anything to make up for those costs or make them irrelevant - for example:
  • In party based game TB cancels out inaccuracies and timing differences resulting from trying to control multiple characters at once. It also eliminates overhead of having to spam pause constantly.
  • In a game where turn durations are much longer in game time than time spent in each turn in user's time (days, months, even minutes), TB will not incur overhead and actually speed up the game considerably without forcing the game to work at slapstick pace (that's one of the reasons some people really like tile-based crawlers - you can traverse the levels as fast as you can click).
  • In a game where required and sufficient reaction times are long enough to be reasonable turn durations the lag doesn't matter
Individual melee combat is fast paced, with desirable reaction times on the order of tens of ms. Without party to allow positioning and cooperative tactics take the edge off split second decisions and allow player higher level on which they can affect the situation all TB does is reducing depth, making combat clunky and limiting scale of events game can handle.
What's the upside?

This is bordering on the insane. Fast paced split second decisions feed from a well of skill, knowledge and experience. TB and PB are the only systems that can simulate/abstract the complex interaction of those dynamics and present the player with combat options based on those factors. Whether with single character or a party, there are always several factors that play into any one decision and all of those are filtered by character's stats. TB (and PB) alone provide a sane framework to translate all those variables into an interface that the player can navigate. This is the exact opposite of reducing depth and limiting the scale of events. A real-time game can never simulate the level of stat-driven detail and dynamism without requiring an insane amount of feedback -in turn necessitating an impossible level of real-time perception- and becoming a mess of input.

As for making combat clunky, that is an argument that I find very odd but is possibly ultimately very personal, very subjective. Clunkiness, in my experience, results from a poor combination of systems and interface, regardless of the combat system used.

"Fast-paced" is a non-argument. Whether any form of combat, in terms of "realism" relative to the characters, would be fast or slow is simply not even a part of any discussion about combat systems. But to degree that the depth of any combat interaction can be simulated is the entire basis of it. So whether it fits with anyone's personal gaming preferences or not, TB/PB can simply not be surpassed without a level of tech that is in Total Recall or The Matrix.
 
Joined
Sep 18, 2013
Messages
1,258
Yeah, turn-based combat when there is only one character you control tends to just be unnecessarily clunky at best[/u] and exploit-ridden at worst. Everyone loves Fallout, but Fallout is pretty much a textbook case of it being both clunky and exploit-ridden. The only way I can see TB being particularly great in a single-character situation is if the game revolves around duels and the actions and turns are sliced really fine as a result.

Could you explain what this "clunkiness" exactly means, how Fallout's combat was "clunky" and just how something you describe as inherently clunky in a single character game somehow becomes not-clunky when you throw more characters into the mix? Examples would help a lot.

Fallout's combat was certainly exploity but unless the entire town turned on you so you had to wait on 10+ NPCs, I don't know how anything about the game sets it apart from any other game to be distinctly clunky in comparison. I suspect "clunky" has just become a catch-all word.
 

As an Amazon Associate, rpgcodex.net earns from qualifying purchases.
Back
Top Bottom