How to fix riddle of steel

[OgreBattle]

Well, I like the idea of 1on1 duels where your actions are a big part of getting the win, but it's been said lots of times that RoS has many problems, such as alpha-strike rapier thrusts to the nuts being the winning move.


So, has anyone gone about making a better balanced RoS or their own 1on1 dueling game?

[fectin]

Part of the problem is that real duels /are/ that fast.

The best solution might be a fencing-style priority system: if you both alpha strike, you both lose (no-one "goes first"). Immediately after a successful parry though, you have priority, and do go first. That leads to a lot of back and forth.

[OgreBattle]

currently the way to beat an alpha strike in RoS is to buy initiative and hit that person first, causing shock damage that reduces their attack power or perhaps just kills them.

But the method for generating that is kind of clunky and I dont remember exactly how it goes.


I figure the attack/defense strength/weakness relationships should go like...

Attack vs Attack scenarios:
Speedy strike>Alpha strike, outspeeds and hit causes shock to make foe miss/die
Defensive strike>Speed strike, absorbs speedy attack and strikes back with good accuracy
Alpha Strike>defensive strike, crushes defense leading to a double hit

Attack vs Defend scenarios:
Alpha strike>counterattack, the alpha strike hits too hard
alphastrike=defend, unstoppable vs immovable can go either way a bit
defend>defensive strike, a defensive strike wont penetrate a full defense
defend>speedy strike, the light attack is easily blocked
counterattack>speedy strike, the light attack is easily turned around
counterattack=defensive strike, the light attack is easily turned around

so it's basically rock paper scissors spock lizard, you decide your action then reveal. (I've been playing a lot of Pokemon battles lately and found it satisfying)

what I really like about RoS is that your action pool only refreshes after two exchanges, so if your alpha strike fails and your opponent has pool left his next attack is pretty much guaranteed to hit. If you successfully defend, you take the initiative. Counterattacks are a defense which give you a bonus on your next action.... but people using alpha strikes just means you need to dedicate your entire action pool against it, making this "two exchanges" thing really just one. All of those "set up" maneuvers like "beat" (knock opponent's weapon away) or feinting just don't seem to be useful compared to an alpha strike, so how can I fix that?

[Korgan0]

Firstly, the fact that there exists different mechanics for attribute and skill checks is pure bullshit.

Secondly, people using ranged weapons need to get on the same initiative count as melee weapon users, given that it'll often take you 3 or so rounds to fire a single arrow, when the dudes with swords have acted six times. I'm not sure how you could do this without losing the granularity of melee combat that's RoS's main selling point, but there you have it.

Thirdly, the entire system of terrain rolls needs to be overhauled to make multiple-on-multiple combats (aka anything involving a party) something other than the clusterfuck it already is.

Fourthly, magic needs to die in a fire. It has no place in RoS.

[angelfromanotherpin]

Riddle of Steel actually got what was basically a second edition recently, with a free pdf, if you don't mind signing up for their site. I'd start by looking at that.

[Archmage]

If you aren't trying to expand the game's system to handling multiple participants and are in fact focused on one-on-one duels, there are worse places to look for inspiration than David Sirlin's Yomi.

The thing about RPS, even with five options, is that single-iteration RPS boils down to total luck. If your option is chosen at random and the other guy's is chosen at random your rate of W:L:D is 1:1:1. If you have to play more than one round, the best "strategy" is still to choose truly at random unless you have really good reason to think your opponent is going to play rock every time or something--which, since you can't really know that for certain, "play randomly" is probably still the way to go.

What you have to do to make RPS interesting is create unequal payoffs. If winning with rock is worth 2 points, winning with scissors is worth 1 point, and winning with paper is worth 0.5 points, you at least have an interesting mind-games scenario going on where you can develop a playstyle of being risky and going for 2-point rock or something. Even that's not tactically deep enough to stay interesting except maybe as a game theory exercise, so enter Yomi.

Yomi handles the unequal payoffs by making the various options situational; some characters have strong incentives to do certain kinds of moves because of special abilities or move properties and people who have small hands often want to block because blocking lets you draw cards. However, the tactical landscape is still only interesting because you have known information about your opponent's character's strengths and weaknesses (which lets you decide what move is optimal given the matchup and what your opponent's "best move" is likely to be) and then you can factor in observations about your opponent's playstyle (especially in a best two-out-of-three tournament game).

If you don't know anything about your opponent's capabilities you're still basically guessing randomly what to do. And if you are a player and the DM naturally runs all the opponents then all of the mind-games and fake-outs are very directly "player vs. DM," which can feel kind of adversarial.

[fectin]

Also, that's only true if each round is a complete reset. If each round affects the next, you do have some strategy.

[OgreBattle]

Also, that's only true if each round is a complete reset. If each round affects the next, you do have some strategy.

thats what bugs me. In RoS the only way to respond to an alpha strike is to also throw all of your pool against it, so that whole "two exchanges in a round" thing is thrown out the window.

How can you get that "two exchanges per round" to function properly?


Oh yeah I've checked out blade of throne of iron a bit, but it seems to have the same "the first turn is the only turn" prblem.

Will check out Yomi... and play more Pokemonshowdown

[Josh_Kablack]

If you aren't trying to expand the game's system to handling multiple participants and are in fact focused on one-on-one duels, there are worse places to look for inspiration than David Sirlin's Yomi.

Might want to check out Yomi's online version first.

[Grek]

What you have to do to make RPS interesting is create unequal payoffs. If winning with rock is worth 2 points, winning with scissors is worth 1 point, and winning with paper is worth 0.5 points, you at least have an interesting mind-games scenario going on where you can develop a playstyle of being risky and going for 2-point rock or something. Even that's not tactically deep enough to stay interesting except maybe as a game theory exercise, so enter Yomi.

This doesn't actually work. The correct solution there is to play with weighted randomness - play a choice X out of n many options with P(X)=U(X)/sum of U(Xn) from 1 to n, where U(X) is how many points you get for winning with X and P(X) is the probability you chose X.

[Archmage]

What you have to do to make RPS interesting is create unequal payoffs. If winning with rock is worth 2 points, winning with scissors is worth 1 point, and winning with paper is worth 0.5 points, you at least have an interesting mind-games scenario going on where you can develop a playstyle of being risky and going for 2-point rock or something. Even that's not tactically deep enough to stay interesting except maybe as a game theory exercise, so enter Yomi.

This doesn't actually work. The correct solution there is to play with weighted randomness - play a choice X out of n many options with P(X)=U(X)/sum of U(Xn) from 1 to n, where U(X) is how many points you get for winning with X and P(X) is the probability you chose X.

I accept your game theory skill and likely general math ability as greater than my own. Even if that solution is correct, I think the general point still stands, which is that in order for RPS to be interesting there needs to be a tactical evaluation as to whether to choose rock, paper, or scissors over another option out of the three.

The Yomi solution is to approximate the fighting game trio of attack, throw, and block in an RPS triangle (where attacks beat throws, throws beat blocks, and blocks/dodges beat attacks). However, the rewards are not all equivalent. It's a card game, so you have a hand of options available and play one based on the current tactical landscape:

Successfully blocking an attack negates damage. You get to keep your block card for future use, and you draw another card as a bonus (generating card advantage, because attacks and throws are generally discarded for good and can't be retrieved).
Successfully dodging an attack lets you counter with an attack of your own. Great when punishing your opponent with damage is more important to you than the potential for a card draw (to balance those options, blocks return to your hand, but dodges are discarded; so blocks are better early game when you're trying to build up a hand to create a combo and wreck someone and dodges are better later when you already have the hand and want to deliver the finishing blows).
Successfully throwing someone who tries to block or dodge results in their block/dodge card getting discarded (so you stay even on cards) and also deals damage (so you get ahead on life). This is the only way to get permanently remove a block card from someone's hand, and it's actually possible to get "blockscrewed" and wind up having trouble building your hand if your decision to block is too predictable.
Successfully attacking someone who tries to throw results in both cards getting discarded (even on cards) and deals damage (ahead on life).

Then the game complicates things with special character abilities and attack speeds (if you both throw or both attack whoever's faster wins the "tie").

Individual matches probably take too long to play out for it to be good for an RPG, it can't handle multiple combatants at all, and I'm not really sure if you want a card-based engine, but as Josh said, it's worth a look if you want duel mechanics. The online version is free to play (albeit with some limitations).

[Josh_Kablack]

it can't handle multiple combatants at all,

Well, the forthcoming 2nd edition of Yomi will have a 2 vs 2 mode and a Free For All mode...but those are very different from an RPG engine which can handle multiple combatants.

[OgreBattle]

what would be a way to make "two exchanges before your action pool refreshes" work?

[tussock]

Work in what way? How about, a limit on maximum success in the first exchange, but not in the second. That means all-out strait away tends to be wasted and only deliver a small penalty, but sucking someone in and then going all-out when they're "off balance" works fine.

You could also lower the price of defence, 2:1 cost for attack. That means it's cheaper to gamble on defence-only in the first exchange, and you can defend both phases against someone who puts everything into attack on the "full payout" 2nd phase if you notice them waiting out the 1st gambling on your pure defence there. Might push people toward saving most stuff for the 2nd phase, but then a 1st phase attack can cheaply add penalties to them if it's predictable.


Meh, no, none of that. In the end all you can generate at best is a rock-paper-scissors game of ways to split your pool, why not just RPS for it? Less victories needed for the actor with a bigger pool. "Rock is offense, Paper is esquive, Scissors is trickery".

[Josh_Kablack]

However if I were to write a game where one on one dueling had to be an interesting strategic minigame, here's a rough outline of what I'd want:

You have three general strategies in a RPS relationship

1. Offense (aka Rushdown)

This is dealing damage/ wounds and winning the duel fast. It has a lot of high risk and fatiguing or self-damaging options.

In an RPG context this is an all-out, Berserker style. Hit hard, hit fast and hope someone can heal you after you win

2. Defense

This is reducing and negating the damage you take and thereby prolonging the duel.

In an RPG context, this is a cautious White Mage style. Forget dealing damage, you'll just prevent it while the enemy tires themselves out.

3. Positioning(aka Powerup aka Super Combo Meter)

This is setting up for a big finishing move.

In an RPG context, this is a dedicated Buffer. If you just have enough time to prepare, your numbers get bigger than anyone else's at everything and you win.

At the macro level, strategies interact as R/P/S:
Defense beats Offense, since it holds off the onslaught and Offenses fatigue and self-damage eventually cede the edge.
Positioning beats Defense, since Defensive moves and damage negation do not prevent setup for a super move / finisher or casting a billion buffs. And in the long run, those situational buffs add up to the best values in the game
Offense beats Positioning since doing a bunch of damage nownownow means that the positioning guy doesn't have time to set up his finisher or powerup his combo meter.

Still at the macro level, you must allow characters to shift between these during fights. (possibly with a slight delay to make thinking ahead important) So you may open a fight in Defensive turtling but then shift to all-out offense when you see your opponent going for pure Positioning / Powerup. Yet I want a system where, maybe he would open with a powerup because he's best at Defense and wanted to bait you into doing exactly that. (and yes, that sort of thing means multi-round fights, which tilts things towards padded sumo)

However to make it interesting and not just yet another R/P/S clone you need to add two more strategic levels below the macro RPS triangle:

At the medium level you want to have two or three sub-strategies for each of Offense / Defense and Positioning. And you want to make sure that each of the sub-strategies within the macro strategy have more synergy with each other than they do with themselves. So if you break offense up into say "hitting hard", "hitting many times" and "going first", you need to make sure that it's generally better to split between "hitting hard" and "hitting many times" than it to just dump everything into "hitting many times". Any cases where any particular mono-substrategy outshines the sub-strategy need to be edge cases that occur only rarely in a typical campaign...and you need aggressive playtesting of the dueling subsystem to find them.

And then at the micro level, you need individual moves.
And while each of your moves fits into a particular tactic for both a Macro and Medium strategy, they should also have direct use or strong synergy with objectives of a different Macro level strategy. So you get stuff like "Crippling Strike", where you hit someone in a way that deals damage and also reduces their ability to use Positioning - an offensive, 'hit hard" ability. But since "Crippling Strike" also slows them down, making your Defensive ability of "Run Away" much more useful.

However, to make that sort of thing work in an RPG context you need a couple hundred distinct moves, and you need to destructively playtest their interactions with people competing to optimize for dueling. Neither of which is easy.


Then to make it interesting and plausible in an RPG, you probably also want to add some rules for establishing initial terrain and which particular moves that terrain allows / disallows / enhances/ penalizes. This is especially true if you go with the easier WSoD name "Positioning" name instead of the more videogamey "Combo Meter".

[kzt]

The problem you run into is that in virtually all martial arts there is no "defense" once you get out of the basics. For example, all the moves that look like blocks in Karate are actually used to damage the opponent as they attack while also preventing them from doing damage to you.

Miyamoto Musashi
"The primary thing when you take a sword in your hands is your intention to cut the enemy, whatever the means. Whenever you parry, hit, spring, strike or touch the enemy's cutting sword, you must cut the enemy in the same movement. It is essential to attain this."

[Josh_Kablack]

That's a bullshit strawman in a discussion of hypothetical game mechanics for a hypothetical fighting game.

It's also flagrantly untrue for the violence specialists I encounter in everyday life

[kzt]

That's a bullshit strawman in a discussion of hypothetical game mechanics for a hypothetical fighting game.

It's also flagrantly untrue for the violence specialists I encounter in everyday life

They are actually a multi-purpose tool.

Edit: Also, if you are using a riot shield you are not planning to be in a really serious fight or have an intent to use deadly force. For example, your local violence specialists didn't seem to be using many riot shields on June 4, 1989.

[OgreBattle]

The problem you run into is that in virtually all martial arts there is no "defense" once you get out of the basics. For example, all the moves that look like blocks in Karate are actually used to damage the opponent as they attack while also preventing them from doing damage to you.

Well yes the goal is to be in a position to begin attacking, If you just turtle you're only delaying your own demise, but you still get to block and duck n' weave. It's about striking at the moment your foe drops his guard.

Check out the UFC Champion Anderson Silva

*if image doesn't show up, direct link: http://ctasmith.hostzi.com/silvavid.gif

He is known for his crazy evasive abilities. He gained his dodging skills from training with Lyoto Machida, who has a karate background that's given him an excellent judge of distance for dodging and counterattacking.

In RoS's pacing Anderson Silva took the defensive and used a counterattack maneuver on his 1st exchange (his foe misses) and on his second the bonus he got for evading the attack applies to his punch to the gob.

[Grek]

What you have to do to make RPS interesting is create unequal payoffs. If winning with rock is worth 2 points, winning with scissors is worth 1 point, and winning with paper is worth 0.5 points, you at least have an interesting mind-games scenario going on where you can develop a playstyle of being risky and going for 2-point rock or something. Even that's not tactically deep enough to stay interesting except maybe as a game theory exercise, so enter Yomi.

This doesn't actually work. The correct solution there is to play with weighted randomness - play a choice X out of n many options with P(X)=U(X)/sum of U(Xn) from 1 to n, where U(X) is how many points you get for winning with X and P(X) is the probability you chose X.

I accept your game theory skill and likely general math ability as greater than my own. Even if that solution is correct, I think the general point still stands, which is that in order for RPS to be interesting there needs to be a tactical evaluation as to whether to choose rock, paper, or scissors over another option out of the three.

Definitely agreed there. I'm just saying, that specific solution (make rock do twice as much damage as scissors which does twice as much as paper) isn't enough to be a proper tactical evaluation. You need some setup that isn't a solved game. And in this situation, I think a combination of the Prisoner's Dilemma and Chicken is you're looking for. First we choose to either Attack, Block or Move, then we get the choice to Continue or to divert your Attack into a Block, your Block into a Move or your Move into an Attack (but not in the reverse order). The final outcome table looks like:

Final Outcome	Attack	Block	Move
Attack	Both take damage	Attack fails, blocker gains small advantage	Movement fails, mover damaged
Block	Attack fails, blocker gains small advantage	Roll damage, defender takes difference	Movement successful
Move	Movement fails, mover damaged	Movement successful	Both movements successful

But that actually gives us much more in depth tactics than are apparent at first glance: Consider the case of Warrior Row attacking Commander Col's block:

Attack v. Block	Keep Blocking!	Switch to Move!
Keep Attacking!	Row Fails; Col gains advantage	Row hits Col big time
Switch to Block!	Minor damage to both Row & Col	Col moves successfully

[OgreBattle]

Your action modifying your following action could also help


ex: "Assault" increases the speed of your next attack, encouraging you to continue being aggressive

"Block open" increases the attack power of your next attack.