PDA

View Full Version : Game Theory / Have you seen Rock-Scissors-Paper done right?



Quertus
2021-10-08, 08:58 PM
OK, so it's a little "Game Theory", because I have to explain what I consider Rock-Scissors-Paper done right vs done wrong looks like. And this may invoke future debate / additions.

Short answer: emergent property vs forced.

So, an example that does it wrong is "Battle for Middle Earth". Your units have Damage amounts… and then those numbers get arbitrarily changed based on what you're attacking.

An example of it done right is Magic: the Gathering. A Timber Wolf (1/1 first strike) beats a Ball Lightning (6/1 Trample Haste) beats a Wall of Wood (0/3 wall defender) beats a Timber Wolf. There's no additional metadata of "this is how these two cards interact" necessary to make one superior over another.

So, in Middle Earth, the units trump one another because they trump one another; in MtG, cards *usually* trump one another as emergent properties of rules interactions.

The reason I ask is, D&D historically had several really bad "Rock-Scissors-Paper" minigames, from psionic combat to BECMI immortal combat to BECMI elemental "a beats b beats c beats d". (As you could guess, I'll not limiting my question to just "minigames with 3 items").

I'm wondering how many people ever made something actually good. But more examples of stupid designs are always fun to laugh at, too.

So, Playground, what examples of Rock-Scissors-Paper minigames have you seen, and how many of them were actually engaging and done well?

HidesHisEyes
2021-10-10, 06:41 AM
It’s not something I’ve ever given thought to. Can I ask why it’s bad to have meta rules creating the rock-paper-scissors dynamic instead of it emerging from existing rules? I wouldn’t necessarily disagree, just interested in your thinking on why one is superior to the other.

JNAProductions
2021-10-10, 12:32 PM
It’s not something I’ve ever given thought to. Can I ask why it’s bad to have meta rules creating the rock-paper-scissors dynamic instead of it emerging from existing rules? I wouldn’t necessarily disagree, just interested in your thinking on why one is superior to the other.

I would say that if you need to specifically call out interactions, it probably results from a lacking core system.

It's better (in my opinion, and I'd imagine most others) to have a robust core system with depth to it, rather than to have a shallow core system that is propped up by a bunch of specific rules.

The examples Quertus gave aren't causes, they're symptoms.

MoiMagnus
2021-10-10, 02:55 PM
While I'm not an expert on Pokemon (the videogame), I think that's a good example of both emergent and forced Rock-Scissor-Paper.

As for "forced", the creature type is obviously one of those: water beats fire, bugs beats psychic, etc.
As for "emergent", the creature various statistics are one (speed, attack, defence, special attack, special defence, etc).

And I think both show the main advantages and problems of both approaches:
+ The statistics have a lot more depth, especially with how they are entangled with every other part of the game. While I'm not a competitive player myself, I'd guess they're part of what make the game interesting to those players.
+ On the other hand, for more casual players, those same statistics can be quite unreadable, and more like a "black box" that you don't look into. (Except to have a nice instant gratification when you see those numbers go up when you level up, but that's not the subject here)
+ While much simpler, the creature types are much "flashier". But having literally a tag on them saying "I'm weak against some stuff and strong against other stuff", it allows for simple strategising that don't require much though, meaning you can focus on the core of the game: putting your favourite Pokemon together and have them fight for you. Need to fight the plant Gym leader? Just take the cutest Fire Pokemon you found, level him up a little bit, and he should be able to go through most of the opposition (whatever what the actual enemy Pokemon are).

I would not place emergent R/S/P as strictly superior to forced R/S/P. If think forced R/S/P have their place too.

But I will fully agree that if your goal is to add depth to a strategy game, emergent R/S/P is the way to go. And even in RPGs like D&D, I tend to dislike quite a lot forced R/S/P (e.g. I've never really liked the mini-game of damage type and associated resistance & immunities). Though to be honest, I'm not sure I appreciate emergent P/S/C all that much either (e.g the concept of glass canon character), so my problem might be more about R/S/P and myself than anything else.

NichG
2021-10-10, 03:50 PM
There's something like this conceptually in the interplay between things like patrols, stealth, passive access controls, and siege weapons. Stealth means that even if you arm your guards better, it doesn't matter if they don't spot the infiltrator. Passive access controls mean that even if you evade the guards you might not be able to get in. Siege weapons can deconstruct anything, but require time and leave their crews vulnerable, so either infiltrators or active defenders can mess with them. You could turn that into forced RPS, but emergent RPS comes from the fact that each thing has a prerequisite condition in order to exert influence, and something can disrupt the prerequisite without having to directly face the thing's power.

- It doesn't matter how good your weapons are if you don't have a target
- It doesn't matter how sneaky you are if you can't physically access the place you want to get to
- It doesn't matter how good your lock or how hard it is to scale your wall if those things can be destroyed
- It doesn't matter how big your attack is if it takes so long to go off that you're killed before you can use it

That sort of design (all power has interactive prerequisites it depends on) should frequently give rise to emergent RPS.

Pex
2021-10-10, 11:31 PM
There's also this.

https://www.youtube.com/watch?v=x5Q6-wMx-K8

Metastachydium
2021-10-11, 06:29 AM
Another example from Total War: Rome 2: pikemen, archers and melee cavalry. Pikemen are nigh impossible to even approach by regular melee units, so their armour doesn't have to be that good which makes them extremely vulnerable to archers which, in turn, are weak in melee so a unit type that can close the distance to them quickly can decimate them with relative ease.
Of course, in practice, cavalry can beat pikes if they manage to outflank them, pikes will beat archers if they somehow manage to close with them and cavalry is useless against archers placed on a wall or otherwise hard-to-approach position, but the pattern is there.

ahyangyi
2021-10-11, 07:08 AM
Examples are abundant. Mostly, as long as there are two ways to do one thing and the two ways are not symmetrical, one can construct a Rock-Scissor-Paper out of it.

For example, consider:
* a Strength Fighter with medium attack, AC, and high damage with access to Power Attack;
* a Dexterity Fighter with high attack, AC but low damage, and
* a Damage Reduction Fighter with low attack, AC, damage but a decent damage reduction.

It's easy to make up numbers so that the Dexterity Fighter wins over the Strength Fighter, who in turn wins over the Damage Reduction Fighter, who in turn wins over the Dexterity Fighter.

This example is completely emergent. The Power Attack thingy is optional, since the example also works without Power Attack (it's just so iconic that I feel it's appropriate to include it here).

The real reason that we have a Rock-Scissors-Paper situation here is that there are two ways to avoid damage (dodging and damage reduction in this case), and they work differently. This, alone, is sufficient for emergent Rock-Scissor-Paper situations to happen. And this is why I say they are abundant.

---

The fundamental problem with Rock-Scissor-Paper situations, in my opinion, is that they tend to favor versatility over specialization. But sometimes that runs into conflict with character concepts. What if the character concept is really about high damage reductions? Then if your game plays too heavily with the Rock-Scissor-Paper, the player who's stuck as a Rock will be able to do nothing against a Paper encounter.

( And again, spellcasters already tend to be more versatile than martials... )

There is a reason that while older editions of D&D have lots of conditional damage reductions (DR 10/slash etc), but newer versions are moving away from that. A system forcing the players to carry a few more side-arms and doing nothing else is not in high demand.

And back to Pokemon, I think there is a reason that a player commands a team of six pokemons. Were it a game where a player roleplays one pokemon (which would be more similar to regular RPG), the elemental system would have been much more annoying.

Vahnavoi
2021-10-11, 07:57 AM
By these weird definitions, RPS itself is "forced" and thus "RPS done wrong". :smalltongue:

@Ahyangyi:

Pokemon has six controllabe creatures because like many other computer roleplaying games, it's actually emulating a tabletop party where multiple players play different characters. Pokemon "done right" on the tabletop would have one player per pokemon and possibly seventh player as the trainer.


--;

But, Pokemon metagame is developed enough that you can take some terminology from there and use it to talk about the topic: namely, checks, counter and hard counters.

A "check" is a strategy which prevents the opponent from implementing their strategy. For example, if Pokemon A always beats Pokemon B from equal position, that's a check.

A "counter" is a strategy which defeats the opponent's strategy. For example, if Pokemon A always beats Pokemon B from disadvantaged position (such as after taking a hit), that's a counter.

A "hard counter" is a strategy which is strictly superior to the opponent's strategy. For example, if Pokemon A always beats every version of Pokemon B, that's a hard counter.

Finding these in a complex game can be very difficult. In Pokemon, every new version and every alteration to the ruleset creates ripples throughout the metagame and it's only because of very large number of people playing that some of these interactions can be found.

Similarly, there are hundreds of such interactions in d20 D&D variants. It just isn't worth the effort to find most them outside those which are directly stated in the rules (and thus are "doing it wrong"), firstly because only a a tiny sliver of possible characters ever gets played for any given playgroup, secondly because a game master mixing and matching the opposition can just create new such interactions out of wholecloth.

ahyangyi
2021-10-11, 08:51 AM
Good point, though I guess the more sequential nature (1-3 pokemons active, others stay in the Poke balls) felt different than the adventuring party.

But after reading your response I can draw the parallels. We are fine that some fighters just have no way to deal with certain traps before a door but certain rogues and wizards can trivialize them, we are also fine that certain Pokemons just can't win against a well-typed wall plus some Stealth Rock...

---

Though it's also interesting that none of the current fan-made tabletop Pokemon systems I know (5e pokemon, PTA, PTU) use that "one player one pokemon" rule.

When I played one of them (PTU) I really wanted to just focus on one pokemon and avoid juggling poke balls, and that's possible but apparently not what these systems are designed for.

Composer99
2021-10-11, 09:35 AM
Eh, I'm not really seeing how emergent RPS-like mechanics are superior of necessity to forced RPS-like mechanics in an RPG (or any game). It seems to me that it depends on what you're trying to accomplish, and how easily and elegantly you can do so with either sort of mechanics.

For instance, in D&D 5e, a fire elemental possessing immunity to fire damage as an explicit property, and certain attacks, spells, etc., being explicitly identified as dealing fire damage, would be a forced RPS-like mechanic, if I'm not misunderstanding something. But that mechanic easily and elegantly models what you would expect from the fiction of a fire elemental, and I at least am hard-pressed to think of a mechanic that could do a better job of it.

kyoryu
2021-10-11, 04:00 PM
And back to Pokemon, I think there is a reason that a player commands a team of six pokemons. Were it a game where a player roleplays one pokemon (which would be more similar to regular ), the elemental system would have been much more annoying.

Yeah, this is the big thing. Build-time RPS works best when you have multiple units. When you're locked into one, it gets kind of dull as the in-game stuff stops mattering very much.

If you're going to do RPS in an RPG, I think it works best if it's a decision that can be made on a frequent basis inside of combat.

That doesn't mean that everything has to be equal - like, if you do "weighted RPS" (where victories give you different amounts of points depending on what you chose) it can still work and just changes strategies.

oxybe
2021-10-11, 06:06 PM
Pokemon has a lot with it's PSR gameplay that even though a lot of players can do fine with some simple memorization (Fire good v Grass) the competitive metagame is where the complexity lies.

1) Each Pokemon can have up to 2 types and it's strengths and weaknesses combine. A flying/water is 4x as weak v lightning then a mono flying type. there are 18 types to consider.

2) Each Pokemon may know up to 4 moves, of which they don't necessarily need to be of it's own typing. using a move of the same type as your mon gives it a damage bonus, but Charizard, a flying/fire type can learn ground type moves as means to potentially keep lightning types in check.

3) Stats matter quite a bit and not all pokemon are competitively viable due to stats and movepool for the most part. if you see a weird pick, like the world championship pachirisu, it was added for very specific reasons and in most other competitive setups it's just a bad pikachu clone who is now fighting a landorus.

4) held items can have a huge effect on gameplay, either bolstering an already strong pokemon offensive or defensive capabilities, covering a weakness, or giving it a trick to use in a scrap.

4) competitive/professional play is 2v2 and not 1v1 like in the game's storymode. You also don't use the full 6 but rather pick 4 from your 6-man list to bring into a match beforehand, with both sides aware of the other's 6 mons. it's a very different dynamic. you also can have duplicates of mons or items on your team.

5) depending on the competitive season, not all mons available to catch may be eligible for tournament play.

these are all explicit rules and a lot of it is laid at the feet of the trainer, so it's easy to get lost in how deep it can get.

Jay R
2021-10-11, 06:28 PM
In any game in which non-bludgeoning weapons do half damage against skeletons, sword beats mace, mace beats skeleton, skeleton beats sword.

[Yes, there are lots of complications. The basic point stands.]

Vahnavoi
2021-10-12, 02:44 AM
A good example of RPS done with dice are the various dice games done with intransitive dice. (https://en.m.wikipedia.org/wiki/Intransitive_dice)

To give one example: there's a set of dice in which dice A rolls higher than B 2/3 of the time, B rolls higher than C 2/3 of the time, C rolls higher than D 2/3 of the time and D rolls higher than A 2/3 of the time.

The article also gives two more useful terms for this discussion, in addition to "check", "counter" and "hard counter" above:

A situation is "fair" for the person who has initiative if they can pick a strategy that's succesful half the time.

A situation is "advantageous" for the person who has initiative if they can pick a strategy that's succesful more than half the time.

A situation is "disadvantageous" for the person who has initiative if they can only pick strategies which succeed less than half of the time.

In Chess, there's a term, "Zugzwang" (https://en.m.wikipedia.org/wiki/Zugzwang), for being in the last position. In English, you could call it "compelled move" or, in the sense combinatorial game theory uses it, "compelled loss".

Totally Guy
2021-10-12, 07:11 AM
I have just released a game which uses rock paper scissors as approaches to a task.

The system is non-random and free of point spending but still unpredictable and exciting. Check my signature.

Easy e
2021-10-12, 11:47 AM
Pretty much every Napoleonic, Ancient and Naval wargame ever made have had to feature Rock-Paper-Scissors of some sort in their design as it is core to the warfighting of those periods.

Quertus
2021-10-12, 08:37 PM
It’s not something I’ve ever given thought to. Can I ask why it’s bad to have meta rules creating the rock-paper-scissors dynamic instead of it emerging from existing rules? I wouldn’t necessarily disagree, just interested in your thinking on why one is superior to the other.

The very first post in this thread stymied me - I found I couldn't answer why I believed this. So I became a lurker in my own thread while I went home and rethought my life.

And, well, this is where I gained traction regarding my thought process:



Another example from Total War: Rome 2: pikemen, archers and melee cavalry. Pikemen are nigh impossible to even approach by regular melee units, so their armour doesn't have to be that good which makes them extremely vulnerable to archers which, in turn, are weak in melee so a unit type that can close the distance to them quickly can decimate them with relative ease.

OK, but why is it better to implement that emergent properties, than as "archers deal +20 damage to Pike"?

Well…



Of course, in practice, cavalry can beat pikes if they manage to outflank them, pikes will beat archers if they somehow manage to close with them and cavalry is useless against archers placed on a wall or otherwise hard-to-approach position, but the pattern is there.

Because it lacks the emergent gameplay of putting your archers on/behind walls.

OK, but… why does that matter?

Well, because our intuition tells us so. And because castles exist IRL.

OK, but… why not just implement that as "castles give +20 defense to archers" or something?

Well… that's complicated. There's a matter of elegance, of how one needs to add extra rules of "musketeers get +X bonus for walls" to add musketeers under such a system, whereas it's automatically handled by physics under the other. And a matter of comprehensibility and debugging, of "what is this '+3 bonus for trees', who does it apply to, and why?".

Then there's the ease of translating back and forth between the fiction and the mechanics layer. That's easier to do when your mechanics are "ranged, light armor, move 3" than when they're "+20 damage to pikemen, +10 defense vs cavalry per wall level".

But what if it's something that doesn't have a real-world analog, like… Pokemon. We don't have historic experience that walls give grass types extra defense against fire, we don't have intuition to work from, right? So it doesn't matter if we can translate the mechanics back to a fiction, does it?

Well, imagine a world that had never had archers, cavalry, or pikemen. In one implementation, the players could discover that they have complex interactions with things like swaps, walls, smoke, etc. In the other, it's just a simple game of rock scissors paper.

OK, but… is one better than the other, for this world that has no concept of archers physics outside this game?

And… I guess not? Or, more precisely, yes, but which is better will depend on the game.

I happen to believe that, for an RPG, the game is better when your stat line isn't just "Rock: crushes scissors, brains Spock".

Which, as it turns out, pretty well matches the second post of the thread:



I would say that if you need to specifically call out interactions, it probably results from a lacking core system.

It's better (in my opinion, and I'd imagine most others) to have a robust core system with depth to it, rather than to have a shallow core system that is propped up by a bunch of specific rules.

The examples Quertus gave aren't causes, they're symptoms.

So… maybe that's all I'm really looking for, is robust systems.

Thing is, though, if that's true, then I'm uncertain if… line… "advanced rock paper scissors theory-crafting" will actually be as helpful a tool as I had hoped to apply to issues like "martial / caster disparity", or ensuring everyone gets their time in the spotlight.

So… I'm hoping that there's more to it than that.

KorvinStarmast
2021-10-14, 02:05 PM
If you want to see Rock, Paper, Scissors done well, Starcraft was the best example of that I've seen. (Computer RTS)
I can't speak for Starcraft II being as good as the original, or the original + Broodwar, since I have only dabbled in it. (My son is pretty good in both, and his take is "the original did RPS better" so there's that).

The original I spent too many hours playing.
You wanna piece or me, boy? (Starcraft Marine quip)
In the pipe, five by five

Stonehead
2021-10-14, 03:03 PM
Two immediately come to mind, and weirdly both have "average" as a corner.

Flat DR vs multiple attacks. Flat damage reduction blocks a flat amount of damage per attack, so the more attacks you have, the more damage gets blocked. So damage reduction beats multiple attacks, and assuming things as a whole are balanced, then multiple attacks will beat "average" fighters, who will then beat damage reduction.

Also, accuracy vs evasion. Because accuracy is (usually) capped at 100%, excess accuracy is wasted. So again, assuming things as a whole are balanced, then accurate beats evasive which beats "average" which beats accurate.


It’s not something I’ve ever given thought to. Can I ask why it’s bad to have meta rules creating the rock-paper-scissors dynamic instead of it emerging from existing rules? I wouldn’t necessarily disagree, just interested in your thinking on why one is superior to the other.

My understanding of game design is that the problem is complexity vs depth. I'm not going to do these ideas justice because I'm not a professional or anything, but broadly speaking, depth is good, and complexity is bad. This is a very specific meaning of complexity, which is a little different from normal usage, so even for people who like "complex" games, this form of complexity is usually bad. It's something like the amount of discrete rules there are, or how hard it is to learn a game. Depth is something like how many different meaningful choices you have, or how much strategy goes into your decisions. So, "emergent" rps is good because it adds depth without adding any complexity. If your game has accuracy and flat dr, then the above triangles will naturally add depth without adding any complexity to the game. "Forced" rps isn't necessarily bad, because it does add depth, it just also adds complexity. So meta-rules creating rps elements isn't necessarily bad, it's just that naturally emerging triangles is better (and more fun to talk about).

kyoryu
2021-10-14, 03:59 PM
My understanding of game design is that the problem is complexity vs depth.

I don't know if this is a formal definition, but I generally think of them this way:

Complexity: The number of possible options available, or the amount of work to figure out the return on an option
Depth: The number of options available that are not dominated (IOW that are not clearly inferior to another option)

Man_Over_Game
2021-10-14, 07:24 PM
So, Playground, what examples of Rock-Scissors-Paper minigames have you seen, and how many of them were actually engaging and done well?

Fire Emblem does a pretty great job. Swords -> Axes -> Spears -> Swords. Pretty simple.

Thing is, each weapon type also has different base stats. It's harder to hit a swordsman, but they are easy to kill. Axemen have more HP, deal the most damage, but also miss the most.

So you might, for instance, decide to attack an Axeman with a Spearman if you expect a counter attack against your Swordsman (who might die in one hit).

Really, the best examples of RPS done right is where using it in a bad matchup might be situationally beneficial. For that reason, versatile mechanics (like MtG) generally have better RPS design than the rigid design of, let's say, Final Fantasy.

If you have a good reason to use a Fire spell against enemies that are resistant to fire, you're probably using a pretty good RPS system. Otherwise, all the player has to think about are absolutes, a Yes/No, as opposed to comparing the infinite choices between the two sides.


By these weird definitions, RPS itself is "forced" and thus "RPS done wrong".

Obviously, a strict Win/Lose RPS system has its place, but it's important to ask why it's beneficial first. It works in classic RPS, but that's because it's supposed to be quick and settled in an instant. For games where you're playing for more than 10 minutes, players want something more to think about and grab onto.

One thing I picked up on with game design: Any bad mechanic is a good mechanic if it's in the right place with the right reasons.


Dead Cells and its chaotic, randomized level design supports the fast and action-paced combat that keeps you on the move.
Hades, on the other hand, has very boring, square arenas that are littered with traps. The boring design gives the player lots of room to leverage the dodge and knockback mechanics while enemies cover the screen with telegraphed AoEs, and the traps add a level of punishment that keeps the player constantly aware of the environment.
CRAWL intentionally hides information from the players when it's not relevant (like what gear you have is hidden except when you're in a shop) so that it's difficult to tell who's in first place and so that players are constantly staying engaged in the fight instead of checking the inventories of your competition.
In Rock, Paper, Scissors, the game's simplicity with its Yes/No counter system means you can constantly replay it without feeling bored, which feels fun each time since it's a multiplayer game.


So....yeah. Classic RPS is a bad mechanic. It can be a good mechanic, too. It just depends on what it's doing for you. If the answer is "I dunno"...probably not a great fit. Final Fantasy is an excellent example of what not​ to do.

Vahnavoi
2021-10-15, 07:23 AM
For complexity versus depth, a version of Occam's razor may help: "between two rulesets which can implement the same space of game situations (=create same amount of depth), the one with fewer rules (=less complexity) is better".

kyoryu
2021-10-15, 10:33 AM
For complexity versus depth, a version of Occam's razor may help: "between two rulesets which can implement the same space of game situations (=create same amount of depth), the one with fewer rules (=less complexity) is better".

Some people want that complexity, though.

That principle works for me, though.

Stonehead
2021-10-15, 12:57 PM
Some people want that complexity, though.

I think it's more that they (we) want a level of depth that's impossible with a low complexity. Like, how deep of a game do you think you could make that only has 3 rules? I'm sure you could come up with some pretty fun, interesting game theory problems, but nothing that would even come close to approaching the depth of your average board game. It's not necessarily that people like complicated rules, it's that they like the depth that can only arise from complex rules.

It's kinda like expensive cars. It's not that people want to pay more money for their car, they want a level of quality that costs a lot of money. Although, I'm pretty sure there are people who buy expensive cars because the price point is a status symbol, maybe the same is true for complex rules systems as well.

Man_Over_Game
2021-10-15, 01:07 PM
I think it's more that they (we) want a level of depth that's impossible with a low complexity. Like, how deep of a game do you think you could make that only has 3 rules? I'm sure you could come up with some pretty fun, interesting game theory problems, but nothing that would even come close to approaching the depth of your average board game. It's not necessarily that people like complicated rules, it's that they like the depth that can only arise from complex rules.

It's kinda like expensive cars. It's not that people want to pay more money for their car, they want a level of quality that costs a lot of money. Although, I'm pretty sure there are people who buy expensive cars because the price point is a status symbol, maybe the same is true for complex rules systems as well.

You can have complexity and simplicity in the same game, generally through removing absolutes. Rock-Paper-Scissors is a game of absolutes, there isn't really a reason to use Rock when Paper is better. A game like Tic-Tac-Toe has a lot more depth, even without that many more rules, since a bad move can still work out in your favor (assuming players aren't playing perfect games, anyway).

By making the Yes/No results into a spectrum, where even an inefficient move might be a valid play, you can create depth and complexity by just leveraging simple mechanics.

D&D has this same thing through multiclassing. You could multiclass a Rogue with a Fighter, instead of leveling just into Rogue or Fighter. Now that character has to choose between how many levels of each class they should invest into for what features, which is a lot more complicated than just dumping all of your levels into the 1 class you have.

By allowing 2 simple mechanics to mix, you can create infinite possibilities, and from those possibilities you create depth.

It's like mixing colors. Red's a simple color, and so is Yellow. Yet, by choosing how much of each you pull from, you can create an infinite number of Oranges.

That way, folks who only want to use Red or Yellow still can, and those that want more can pursue more complex combinations and patterns of Oranges until the system matches their skill level. So your game is both brain-dead easy and incredibly complex.

At least, that's the theory anyway. You don't need crunch and rules for complexity, you just need to remove absolutes and allow players to make a mixture of several options at once.

kyoryu
2021-10-15, 01:10 PM
I think it's more that they (we) want a level of depth that's impossible with a low complexity. Like, how deep of a game do you think you could make that only has 3 rules? I'm sure you could come up with some pretty fun, interesting game theory problems, but nothing that would even come close to approaching the depth of your average board game. It's not necessarily that people like complicated rules, it's that they like the depth that can only arise from complex rules.

It's kinda like expensive cars. It's not that people want to pay more money for their car, they want a level of quality that costs a lot of money. Although, I'm pretty sure there are people who buy expensive cars because the price point is a status symbol, maybe the same is true for complex rules systems as well.

No, I'm saying that there are some people that, given two games with equal depth, would prefer the one with more complexity.

And yeah, I think displaying their system mastery is a status symbol in some cases. In other cases, they enjoy crunching through the complexity.

Stonehead
2021-10-15, 01:38 PM
You can have complexity and simplicity in the same game, generally through removing absolutes. Rock-Paper-Scissors is a game of absolutes, there isn't really a reason to use Rock when Paper is better. A game like Tic-Tac-Toe has a lot more complexity, since a bad move can still work out in your favor (assuming players aren't playing perfect games, anyway).

By making the Yes/No results into a spectrum, where even an inefficient move might be a valid play, you can create depth and complexity by just leveraging simple mechanics.

D&D has this same thing through multiclassing. You could multiclass a Rogue with a Fighter, instead of leveling just into Rogue or Fighter. Now that character has to choose between how many levels of each class they should invest into for what features, which is a lot more complicated than just dumping all of your levels into the 1 class you have.

By allowing 2 simple mechanics to mix, you can create infinite possibilities, and from those possibilities you create depth.

It's like mixing colors. Red's a simple color, and so is Yellow. Yet, by choosing how much of each you pull from, you can create an infinite number of Oranges.

That way, folks who only want to use Red or Yellow still can, and those that want more can pursue more complex combinations and patterns of Oranges until the system matches their skill level. So your game is both brain-dead easy and incredibly complex.

At least, that's the theory anyway.

I think this is an issue where the technical game-design term "complexity" doesn't perfectly match the colloquial usage, because in game design terminology, you described a rule set with very low complexity, and very high depth. But, in your average every-day terminology, it's reasonable to say it's a complex system. In strictly formal terminology, 2 simple mechanics creating infinite possibilities is like the perfect example of low-complexity/high-depth.

I'm not saying we should all use the formal definitions of words, I don't really care. So if anyone wants to describe your example as both simple and complex, that's fine. They don't have to use formal terms. I'm just trying to make my point understood.


No, I'm saying that there are some people that, given two games with equal depth, would prefer the one with more complexity.

And yeah, I think displaying their system mastery is a status symbol in some cases. In other cases, they enjoy crunching through the complexity.

Maybe. You're probably right, there are lots of people in the world after all. The general wisdom though, is that complexity is a necessary sacrifice. I think at least some of it is people using "complexity" in different ways though. And again, I'm not trying to force people to use one specific definition of a word, I myself say I like complex games in casual conversations.

I'm sure there are people who start licking their lips the more sub-systems you add to a game, but complexity is basically never a design goal itself. If it was, you'd see rules who serve no purpose other than to add more complexity to the system. Usually, additional rules add gameplay depth, or realism, or help to model a specific play style or something.

I think subsystems are actually a pretty good example of this. In probably 90%* of rpgs, there are 2 main systems. One for combat and one for everything else. There's two fairly distinct modes of play, in combat you have a map (usually), you have to wait your turn, and your options of what to do (at least, mechanically) are relatively limited; whereas outside of combat, gameplay takes the form of any player describing what they'd like to do, sometimes rolling a check, and the DM describing the outcome. Having 2 different systems like this does add a lot of complexity to the game, but the designers decided that it was worth it, because they wanted combat to play out in a more structured, tactical way, that they didn't want to impact the rest of the game. Some older games have a lot of different subsystems. There are whole rulesets, just as strictly defined as combat, for bartering, or long distance travel, or city construction. That adds way more complexity to the game. To fans of those older games, it's worth it in exchange for the extra depth it adds to the game, although general design trends have moved away from that model. Some newer games don't even draw a distinction between combat and non-combat. These games value simplicity over the strategic depth of tactical combat, and some of them don't even care about strategy at all.

If complexity was a goal itself, not a result of following other goals, then we'd see games full of different subsystems whose purpose is only to add more subsystems, not for realism, or strategy or anything else. Climbing a cliff would use a different resolution mechanic than jumping over a hole. Sailing a ship would look as different to flying a plane as stabbing a troll looks to talking to it in current DnD.

The only reason I brought this all up, is because it's the advantage emergent rps mechanics have over forced rps mechanics. One adds complexity and depth, while the other adds only depth.

TLDR: You're right that I shouldn't use absolute terms when describing these things. There are countless different people under the sun. However, complexity (in the formal game-design meaning) is almost never a design goal in and of itself.

Man_Over_Game
2021-10-15, 01:44 PM
I think this is an issue where the technical game-design term "complexity" doesn't perfectly match the colloquial usage, because in game design terminology, you described a rule set with very low complexity, and very high depth. But, in your average every-day terminology, it's reasonable to say it's a complex system. In strictly formal terminology, 2 simple mechanics creating infinite possibilities is like the perfect example of low-complexity/high-depth.

I'm not saying we should all use the formal definitions of words, I don't really care. So if anyone wants to describe your example as both simple and complex, that's fine. They don't have to use formal terms. I'm just trying to make my point understood.

I absolutely agree with you.


TLDR: You're right that I shouldn't use absolute terms when describing these things. There are countless different people under the sun. However, complexity (in the formal game-design meaning) is almost never a design goal in and of itself.

The only exceptions I could see are when complexity is a selling part of the game (like learning 3.5e or lawyer-talk). Which I guess does line up with what Kyoru was saying.

Even games like Keep Talking and Nobody Explodes rely on complexity as a key element to the fun, so even a bad mechanic like overcomplexity can still be a great tool in the right place.

Telok
2021-10-16, 12:10 AM
The only exceptions I could see are when complexity is a selling part of the game (like learning 3.5e or lawyer-talk). Which I guess does line up with what Kyoru was saying.

Having had some lawyer exposure by writing computer programs for them I can assure you the vast majority of them do not like complexity in communications. What you're probably thinking of is documents that sound complex because litigation requires absolute specificity and covering all potential corner cases. Really all most of those documents are doing is saying about six things each in ten different ways.

Likewise, complexity wasn't a selling point of d&d 3.x. Quite the opposite, things like base attack bonus and the three saves were sold as simplifications of AD&D look-up tables. Likewise d&d 5e is sold as a simplification from 3e & 4e, but added weirdness like the ability for a melee attack with a weapon, an attack with a melee weapon, and a melee weapon attack to all be different exclusive things that different class abilities may or may not work with seemingly at random sometimes.

I'd actually argue that it may often matter more if complexity is obvious or hidden than the actual amount of it (barring extreme outliers of course)...

Sorry, interrupted by small child, train of thought major derailment, all ideas become casualties, half post end of line

Man_Over_Game
2021-10-16, 11:24 AM
Having had some lawyer exposure by writing computer programs for them I can assure you the vast majority of them do not like complexity in communications. What you're probably thinking of is documents that sound complex because litigation requires absolute specificity and covering all potential corner cases. Really all most of those documents are doing is saying about six things each in ten different ways.


Someone was telling me that a big cause for the overcomplexity of lawyer-talk is so that it's inaccessible for the public. When one person can afford a lawyer that can navigate the laybrinth of legalities, and another can't, it ensures that one of the two stays on top. When the system reliably catches 95% of the problems, folks will think that's good enough and stop wasting resources on the remaining 5%, essentually securing a window for those that can access it. It's kinda the same way how the IRS is having problems taxing the very wealthy in the states. Wealth disparity, class systems, and all that jazz.

In essence, a positive feedback loop that promotes those who are already on top, similar to how the victor in a fighting game is the one who's better able to manipulate the dozens of mini-systems involved. Same thing could be said about Keep Talking and Nobody Explodes.

I'm not smart enough or educated enough to know any better about the topic, I was just meaning to explain the thought process, although I agree with you. I haven't done that kind of programming before, but I have done both programming and generating a bunch of by-laws, and I can definitely see why real folks would prefer something simpler.

HidesHisEyes
2021-10-16, 11:41 AM
So… maybe that's all I'm really looking for, is robust systems.

Thing is, though, if that's true, then I'm uncertain if… line… "advanced rock paper scissors theory-crafting" will actually be as helpful a tool as I had hoped to apply to issues like "martial / caster disparity", or ensuring everyone gets their time in the spotlight.

So… I'm hoping that there's more to it than that.

I think you’ve pretty much cracked it. I think there’s something we value which is better served by emergent interactions between mechanics that are already there than new mechanics with the specific intention of creating a RPS dynamic.

I suppose the value is… something to do with discovery? Creativity? Having a set of elements to play with and being able to understand them and navigate them in a logical way to get the results you’re after. As opposed to the game just telling you “this is how you do this thing, this is how you do that thing”.

This reminds me of Chris McDowall’s slogan “put the core to work”, meaning have a robust, flexible mechanical core that can be used to resolve all sorts of situations, instead of having subsystems for everything. He usually says it in support of rules-lite minimalism, but I think in this context it applies to more complex systems too.

Pauly
2021-10-16, 03:50 PM
Someone was telling me that a big cause for the overcomplexity of lawyer-talk is so that it's inaccessible for the public. When one person can afford a lawyer that can navigate the laybrinth of legalities, and another can't, it ensures that one of the two stays on top. When the system reliably catches 95% of the problems, folks will think that's good enough and stop wasting resources on the remaining 5%, essentually securing a window for those that can access it. It's kinda the same way how the IRS is having problems taxing the very wealthy in the states. Wealth disparity, class systems, and all that jazz.

In essence, a positive feedback loop that promotes those who are already on top, similar to how the victor in a fighting game is the one who's better able to manipulate the dozens of mini-systems involved. Same thing could be said about Keep Talking and Nobody Explodes.

I'm not smart enough or educated enough to know any better about the topic, I was just meaning to explain the thought process, although I agree with you. I haven't done that kind of programming before, but I have done both programming and generating a bunch of by-laws, and I can definitely see why real folks would prefer something simpler.

Having been a lawyer in a former career it isn’t that.
Some reasons include:
- The law has a jargon, based on the wording of legislation and precedents. It’s kind of like casting a magic spell, for the spell to work properly you have to use the correct incantation. Paraphrasing or shortening the magic phrases doesn’t work because it shows that you don’t fully understand the relevant law. Some of these magic phrases are from 100 years ago or more, so sound stilted and complex to modern ears, but to lawyers it’s just jargon. For example in my jurisdiction lawyers never said “it was his own fault”, instead they said “he was the author of his own misfortune” because the latter was the magic incantation and the former had no magic power even though they mean the same thing.
- specificity as outlined by Telok.
- Complexity. A simple action may involve several precedents and several different pieces of legislation. To be thorough, and to prevent themselves being sued for incompetence, lawyers will walk you through each individual step in the process every time a new issue arises.

You do get lawyers who send overly complex hard to decipher letters. That is a clear sign that they have a weak case. Lawyers who have a good case will hit you over the head with it in the simplest and bluntest language.

Man_Over_Game
2021-10-16, 07:11 PM
Having been a lawyer in a former career it isn’t that.
Some reasons include:
- The law has a jargon, based on the wording of legislation and precedents. It’s kind of like casting a magic spell, for the spell to work properly you have to use the correct incantation. Paraphrasing or shortening the magic phrases doesn’t work because it shows that you don’t fully understand the relevant law. Some of these magic phrases are from 100 years ago or more, so sound stilted and complex to modern ears, but to lawyers it’s just jargon. For example in my jurisdiction lawyers never said “it was his own fault”, instead they said “he was the author of his own misfortune” because the latter was the magic incantation and the former had no magic power even though they mean the same thing.
- specificity as outlined by Telok.
- Complexity. A simple action may involve several precedents and several different pieces of legislation. To be thorough, and to prevent themselves being sued for incompetence, lawyers will walk you through each individual step in the process every time a new issue arises.

You do get lawyers who send overly complex hard to decipher letters. That is a clear sign that they have a weak case. Lawyers who have a good case will hit you over the head with it in the simplest and bluntest language.

I'm glad I was wrong, thanks for correcting me. It was way too cynical of a perspective to want to keep around.

BloodSquirrel
2021-10-18, 11:53 AM
My understanding of game design is that the problem is complexity vs depth. I'm not going to do these ideas justice because I'm not a professional or anything, but broadly speaking, depth is good, and complexity is bad. This is a very specific meaning of complexity, which is a little different from normal usage, so even for people who like "complex" games, this form of complexity is usually bad. It's something like the amount of discrete rules there are, or how hard it is to learn a game. Depth is something like how many different meaningful choices you have, or how much strategy goes into your decisions. So, "emergent" rps is good because it adds depth without adding any complexity. If your game has accuracy and flat dr, then the above triangles will naturally add depth without adding any complexity to the game. "Forced" rps isn't necessarily bad, because it does add depth, it just also adds complexity. So meta-rules creating rps elements isn't necessarily bad, it's just that naturally emerging triangles is better (and more fun to talk about).

I think "depth" is best described as the size of your game's valid possibility space. Complexity is the cost to create depth, but good or bad design decisions can lead to an enormous gap in how much depth your complexity buys you.

Case in point: If you have three damage types in your FPS, and you give your player three guns (one with each damage type), and then put one group of monsters in each room that is immune to two damage types, then you have spent complexity without creating depth. The valid possibility space is the same as if you only had one gun- in each room there, is only one valid gun to use. In face, you may have reduced your depth by adding complexity. If you had three guns with different firing rates/optimal ranges/etc then, without the damage types, you would have had three valid guns to use in each room. Now you only have one. More complexity with even less depth.

A game with lots of depth is one where there can a lot of different meaningfully distinct game states that don't immediately resolve to "you lose" or "you win". A game that runs entirely on strict RPS has almost no such possibility space- If your opponent chooses rock, then paper is an insta-win and scissors is an insta-lose.

But that's where I think the OP's premise is flawed. The comparison is being made between a system where multiple mechanics are interacting and a system that consists of a single mechanic because the result is similar. I haven't played Middle Earth, but it's easy to speculate that a game could have a forced RPS mechanic that interacts with other mechanics to create more emergent strategies. Easy example: If you have a rock monster in your hand in MtG, and your opponent puts out a scissor monster, do you put the rock monster out this turn? Do you wait, in case he has multiple scissor monsters, hoping that he'll spend his mana putting another one out instead of doing something more dangerous to you with it? Should you put the scissor monster you have in your hand out instead to see if he has any kind of "kill target creature" card in his hand? Do you wait because you know that he need blue mana to summon scissor monsters, and you want to let him put out more blue lands instead of switching to white lands to summon a paper monster?

A game that relies entirely on a single forced RPS system is going to be pretty shallow, but would be just as true for MtG if it relied solely on creature cards with attack/defense stats with no abilities, artifacts, spells, enchantments, etc.

kyoryu
2021-10-18, 12:18 PM
The example I use for complexity vs. depth is Rock Paper Scissors Dynamite.

Everyone gets RPS - each play beats one other, and loses to one other, and ties against itself. There's a Nash Equilibrium at "play a symbol randomly". All choices are valid, and useful.

Dynamite adds dynamite to the game. Dynamite gets its wick cut by scissors, cancels itself, and blows up rock and paper. So... what does this do to the game?

Well, if you look at it, dynamite and paper lose to scissors... paper loses to dynamite, but they both beat rock. Any time that you might play paper, you may as well play dynamite. You'll always get as good of or a better result. This is called a "dominated strategy" *

So, you have more possible choices. But, after crunching the numbers, it turns out that really, you have three viable choices - Rock, Dynamite, Scissors. So the depth of RPSD is exactly the same as RPS, it just has higher complexity.

* technically a choice needs to be better than another one in all cases to dominate it. I call this "weakly dominated" in that each scenario is as good or better - while it's still not actually "dominated", there's no reason to ever choose paper, so the effect is the same.

Man_Over_Game
2021-10-18, 12:30 PM
The example I use for complexity vs. depth is Rock Paper Scissors Dynamite...
So, you have more possible choices. But, after crunching the numbers, it turns out that really, you have three viable choices - Rock, Dynamite, Scissors. So the depth of RPSD is exactly the same as RPS, it just has higher complexity.


I think "depth" is best described as the size of your game's valid possibility space. Complexity is the cost to create depth, but good or bad design decisions can lead to an enormous gap in how much depth your complexity buys you....

A game that relies entirely on a single forced RPS system is going to be pretty shallow, but would be just as true for MtG if it relied solely on creature cards with attack/defense stats with no abilities, artifacts, spells, enchantments, etc.

So...Why many word when few word do trick?

Stonehead
2021-10-19, 10:40 PM
I think "depth" is best described as the size of your game's valid possibility space. Complexity is the cost to create depth, but good or bad design decisions can lead to an enormous gap in how much depth your complexity buys you.

Case in point: If you have three damage types in your FPS, and you give your player three guns (one with each damage type), and then put one group of monsters in each room that is immune to two damage types, then you have spent complexity without creating depth. The valid possibility space is the same as if you only had one gun- in each room there, is only one valid gun to use. In face, you may have reduced your depth by adding complexity. If you had three guns with different firing rates/optimal ranges/etc then, without the damage types, you would have had three valid guns to use in each room. Now you only have one. More complexity with even less depth.

A game with lots of depth is one where there can a lot of different meaningfully distinct game states that don't immediately resolve to "you lose" or "you win". A game that runs entirely on strict RPS has almost no such possibility space- If your opponent chooses rock, then paper is an insta-win and scissors is an insta-lose.

But that's where I think the OP's premise is flawed. The comparison is being made between a system where multiple mechanics are interacting and a system that consists of a single mechanic because the result is similar. I haven't played Middle Earth, but it's easy to speculate that a game could have a forced RPS mechanic that interacts with other mechanics to create more emergent strategies. Easy example: If you have a rock monster in your hand in MtG, and your opponent puts out a scissor monster, do you put the rock monster out this turn? Do you wait, in case he has multiple scissor monsters, hoping that he'll spend his mana putting another one out instead of doing something more dangerous to you with it? Should you put the scissor monster you have in your hand out instead to see if he has any kind of "kill target creature" card in his hand? Do you wait because you know that he need blue mana to summon scissor monsters, and you want to let him put out more blue lands instead of switching to white lands to summon a paper monster?

A game that relies entirely on a single forced RPS system is going to be pretty shallow, but would be just as true for MtG if it relied solely on creature cards with attack/defense stats with no abilities, artifacts, spells, enchantments, etc.

https://www.celjaded.com/wp-content/uploads/2014/07/CelJaded-Magic-Unglued.jpg Hmmm.

Actually, I do remember MaRo saying that one of the reasons Magic was so successful, is because "vanilla" magic is still pretty fun. Even with no special abilities, the decisions of when to cast who, and when to attack or block is still pretty complex. Definitely wouldn't be the massive hit it is without special effects, but interesting enough to keep you entertained for a few hours.

Tangents aside, I think it's still worth pointing out that forced rps mechanics can still add fun depth. Like, in your fps example, if instead of being immune to 2/3 damage types, the enemies took half damage or something, then assuming there's some overhead to switching weapons, you've added some depth to your gameplay. Is it worth the extra time to switch weapons in order to deal more dps? The main advantage of "emergent" rps mechanics isn't that it's the only one that adds depth, but that it's the only one that adds basically no complexity. Hard coding type resistances (usually) adds both.

Vahnavoi
2021-10-20, 04:55 AM
It makes the issue easier to understand once you get that emergent interactions don't add depth - they are depth.

Zuras
2021-10-27, 12:24 PM
Are we talking about RPGs or games in general here? Because the issue in RPG design is less Rock-Paper-Scissors and more analogous to Magic’s “Color Pie”, where the goal is to provide every class a useful role.

KorvinStarmast
2021-10-27, 12:26 PM
This reminds me of Chris McDowall’s slogan “put the core to work”, meaning have a robust, flexible mechanical core that can be used to resolve all sorts of situations, instead of having subsystems for everything. He usually says it in support of rules-lite minimalism, but I think in this context it applies to more complex systems too. I love his thought on that, but I will also say that I think it's 'easy to say and hard to do' in practice. It sure is a good target to aim at.
(Have you got a few games that you think take this advice and apply it well?)

Yakk
2021-10-30, 09:57 AM
I think it's more that they (we) want a level of depth that's impossible with a low complexity. Like, how deep of a game do you think you could make that only has 3 rules?
1. Each player takes turns placing their stones on a 19x19 grid; if you can't play, you lose.
2. Connected groups of allied stones (horizontally/vertically) need a connected empty space to live. Enemy dead stones leave the board after you play; moves that kill your own stones are illegal.
3. No repeating board positions.

I think that replicates go, even down to victory conditions. Someone can probably make a better version of it.

But that is sort of cheating, because go is poetry in game design.

(The "who can stop the enemy from playing" rule is very close to every single go scoring mechanism; the scoring rules basically avoid having to actually play it out, as I understand it.)

MoiMagnus
2021-10-30, 11:13 AM
1. Each player takes turns placing their stones on a 19x19 grid; if you can't play, you lose.
2. Connected groups of allied stones (horizontally/vertically) need a connected empty space to live. Enemy dead stones leave the board after you play; moves that kill your own stones are illegal.
3. No repeating board positions.

I think that replicates go, even down to victory conditions. Someone can probably make a better version of it.

But that is sort of cheating, because go is poetry in game design.

(The "who can stop the enemy from playing" rule is very close to every single go scoring mechanism; the scoring rules basically avoid having to actually play it out, as I understand it.)

Yeah, even with very few rules, one can make a game with a lot of depth.
(Chess has more rules than Go, but one could easily remove the few weird rules and obtain a high depth game with a handful of rules other than the list of pieces and their movement)

And while that's not really a game, the Game of Life is also a good example of absurdly high depth and variety from just a couple of rules.

However, all those examples have one thing in common: they are abstract games.
An by that I don't just mean that they lack a theme, you could quite well add some worldbuilding behind chess that justify all the movement rules.
I mean that the rules are first and foremost designed to make an interesting "mathematical problem", rather than trying to simulate how a given world behave in a way that happens to also lead to interesting strategies.

[Additionally, those games are way above the threshold of "how much system mastery influence your strength at the game" of what is commonly considered as acceptable for a TTRPG.]

Stonehead
2021-10-31, 10:25 PM
1. Each player takes turns placing their stones on a 19x19 grid; if you can't play, you lose.
2. Connected groups of allied stones (horizontally/vertically) need a connected empty space to live. Enemy dead stones leave the board after you play; moves that kill your own stones are illegal.
3. No repeating board positions.

I think that replicates go, even down to victory conditions. Someone can probably make a better version of it.

But that is sort of cheating, because go is poetry in game design.

(The "who can stop the enemy from playing" rule is very close to every single go scoring mechanism; the scoring rules basically avoid having to actually play it out, as I understand it.)

Haha, you win this round.

Aedilred
2021-11-01, 07:47 AM
RPS is the guiding principle of a number of Napoleonic tabletop wargames, from what I understand (it's not my period, so I can't cite specific examples with confidence, but have heard Napoleonic gamers talk about it quite a bit), with the idea that you have the three combat arms (infantry, cavalry, artillery) each of which has its specific strengths and weaknesses and have to be used cooperatively, with the right tool for the right job at the right time. You also have, in some systems, the ability for infantry to choose between line and square formation, which provide protection against artillery and cavalry respectively.

They're not RPGs, obviously, but if you're looking for whole game systems built around the idea of RPS they're probably not a bad place to look.

The first Total War Game, Shogun, was one of the best examples of this type of thing I can recall in video-game format - no artillery, obviously, but archers largely filled the same role. Everyone had access to more or less exactly the same troop types so battles were a question of, firstly, bringing the right troops, and secondly, using them effectively. Notably this was before the Mongol Invasion expansion, which ruined Total War forever introduced ninjas and kensai which completely threw out the balance of the game.