PDA

View Full Version : Extra Credits: Perfect Imbalance and relevance to RPGs



Talakeal
2013-07-24, 04:38 PM
TLDR: Can anyone watch this video and tell me how it is relevant to tabletop RPGs, or cooperative team games?

http://www.penny-arcade.com/patv/episode/perfect-imbalance



Backstory:

I am currently play testing a homebrew system, and one of the players was complaining about the "nerfs" I made for the sake of balance. He then told me I needed to watch the video.

I watched it, but couldn't see the connection between this and table top RPGs, especially cooperative team games like mine is supposed to be.

When I asked the player what about it pertained to my game, he told me "It should be self-evident," and then refused to talk about it anymore (literally, as in he hangs up on me or walks out of the room and slams a door in my face anytime I bring it up).

I have watched it half a dozen times now and can't seem to make the connection, and have showed it to a couple of other players who also can't make the connection.

Scow2
2013-07-24, 04:41 PM
...Are you STILL hanging out with that lunatic?

EDIT: After watching the video... it shoots a hole in HIS arguments against the nerfs instead of your arguments for the nerfs.

I think the message he got out of it was "Don't change things just because they're OP! Players will adapt and counter it!" The problem is... we've had 10 years. We know it's broken.

Talakeal
2013-07-24, 04:45 PM
Hanging out with yes, gaming with no. Unfortunately I am sharing an apartment with him until the end of August.

Anyway, the suggestion that he show me the video wasn't his idea, it was suggested to him by someone else who is less insane, so I am thinking there might be something relevant in there that I am just overlooking.

Yukitsu
2013-07-24, 05:05 PM
To be honest, the majority of the article is about team game balance.

Beyond that, talking about nerfs in the general sense, most nerfs in my view are bad if done with the mentality of "it's overpowered" which is what most nerfs are aimed at. I'd have to know what your nerf in particular is to comment on that, but it may be possible that that is your case.

Basically, most abilities that a lot of DMs find offensive or "broken" are ones that have strict counter measures, and can easily be played against if you happen to know what to expect, and have access to those abilities. By nerfing the broken option to either be more "in line" with the rest of the abilities, or by making it bad to the point of unappealing, you're removing options from the game, which makes the gameplay more stale, as it's removing an entire level of abilities, and the layers of counter abilities beyond that.

huttj509
2013-07-24, 05:11 PM
While most evident in a competitive game, the ideas can bear fruit when you keep in mind that different characters in a cooperative game are played by different people.

I'll use as an example the board game Pandemic, since I played it recently, and it has some distinct classes.

In Pandemic, you're trying to clear 4 diseases off the world map before you're overrun with plague (fun times). There's 5 classes (before the reprint) which all have REALLY useful abilities in helping do this. Things such as moving other players, researching cures easier, transferring information easier, clearing cubes faster, etc.

The games for 2-4 players. You're not expected to have all 5 classes, because you can't. They all do great stuff, but you can't have everybody in their rote role, because you need to work with the situation on the fly.

The scope of that game is small enough that niche protection isn't hard. Everyone generally gets a chance for their specialty to feel useful in the struggle against failure (seriously, what's the success rate for that game? Still great fun). Even if the Medic uses an action and a card to place a research station, the guy who can drop them for free doesn't feel useless. There's plenty more places he's needed.


My guess is that with your changes, for a few people a class went from "well, it's not good at X, but it's great at Y" to not being as great at Y. Perhaps other classes were now better at Y, making the first class' loss of X feel like more of a sacrifice for less gain.

What makes that particularly tricky in a tabletop RPG is that the focus of the game isn't set. I've been that driver guy in a Shadowrun game where everything happened after we got to the job and before we left. Tricked out vehicles never really got used. I've seen the combat monster waiting bored because we were more focused on stealth and avoiding combat.

Classes specialized in a role need to feel awesome in that role. Their level of awesome may indeed be too high, but lowering it can serve to just highlight what they gave up to be awesome.

Talakeal
2013-07-24, 05:12 PM
To be honest, the majority of the article is about team game balance.


It is, I believe, talking about PvP team games though, rather than PvE team games. In League of Legends an OP champion can be countered by a specific champion on the other team, while in World of Warcraft there is nothing The Lich King can do about mages dealing twice the DPS of warriors.




Beyond that, talking about nerfs in the general sense, most nerfs in my view are bad if done with the mentality of "it's overpowered" which is what most nerfs are aimed at. I'd have to know what your nerf in particular is to comment on that, but it may be possible that that is your case.


The specific "nerfs" that brought it up were in regard to balancing costs (I use a point buy game). I made it so that a guy who could fly would pay the same costs regardless of how they got the ability (be it through being a wizard with the fly spell, having wings, or owning a magic carpet for example), or making various spells of the same level be "equal but different" rather than one simply being a no brainer and another being a trap option.

Yukitsu
2013-07-24, 05:18 PM
It is, I believe, talking about PvP team games though, rather than PvE team games. In League of Legends an OP champion can be countered by a specific champion on the other team, while in World of Warcraft there is nothing The Lich King can do about mages dealing twice the DPS of warriors.

Largely, tabletop RPGs are played and balanced far more along the lines of a PvP environment, since in combat, ultimately it's the DM up against the players following largely the same rules and format. WoW isn't balanced the same way as the end game AI isn't intelligent as a DM is, and follows different rules compared to the players.


The specific "nerfs" that brought it up were in regard to balancing costs (I use a point buy game). I made it so that a guy who could fly would pay the same costs regardless of how they got the ability (be it through being a wizard with the fly spell, having wings, or owning a magic carpet for example), or making various spells of the same level be "equal but different" rather than one simply being a no brainer and another being a trap option.

Depending on the general relation to all those methods of attaining flight, he may have a point. As a sort of 3.5 comparison, it's vastly easier for me to get flight as say, a warlock as compared to a rogue. This is intentional, and a major part of the identity of the warlock class. This didn't make warlock a no brainer, but it did make their flight very powerful.

Now, if your system is an M&M type point buy where every single ability is generalized into categories and you can pick and choose with no attachment or synergy, that's a different matter (and should never have been an issue in the first place), but it doesn't honestly sound like you've done that.

Talakeal
2013-07-24, 05:28 PM
Largely, tabletop RPGs are played and balanced far more along the lines of a PvP environment, since in combat, ultimately it's the DM up against the players following largely the same rules and format. WoW isn't balanced the same way as the end game AI isn't intelligent as a DM is, and follows different rules compared to the players.



Depending on the general relation to all those methods of attaining flight, he may have a point. As a sort of 3.5 comparison, it's vastly easier for me to get flight as say, a warlock as compared to a rogue. This is intentional, and a major part of the identity of the warlock class. This didn't make warlock a no brainer, but it did make their flight very powerful.

Now, if your system is an M&M type point buy where every single ability is generalized into categories and you can pick and choose with no attachment or synergy, that's a different matter (and should never have been an issue in the first place), but it doesn't honestly sound like you've done that.

The problem was it was cheaper for a rogue to simply pick up warlock casting and gain everything that comes with it than to simply pick up a magic carpet or pick a race with wings.

Yukitsu
2013-07-24, 05:36 PM
Is this a comparison to your setting, or is that for 3.5?

Because in 3.5, a rogue would never take the 6 levels just for fell flight. That would simply be obtuse.

If you're talking about your system, it may be a flaw in the way resources for wizards is handled in general.

Kornaki
2013-07-24, 05:36 PM
The principle of the video can be applied to any type of game. Your players have options, and if the game is perfectly balanced it means that all options are equally valid (in the definition of balanced in the video). This can be boring (your decisions seem less meaningful) and lowers the excitement of finding out how to invalidate your opponent's choices, because the system is designed to make that really hard.

Perfect imbalance means that all choices are pretty close to each other, but some of them can be better than others. On the flip side those things need weaknesses that can be exploited - and anyone exploiting that weakness has to have a weakness of their own that can be exploited.

Who is making these choices and how they are competing against each other doesn't really matter.

Elemental damage in dnd is an example of an attempt at such a system in an RPG. Fire spells do more damage than acid spells, but are easier to resist. So players have a strategic choice as to which spell to pick. And if players only do fire damage (in a vacuum, the better choice), the dm can counter by sending up lots of fire resistant creatures (a metagame decision by the dm) which should make them find ways of dealing with fire resistant creatures - maybe by doing acid damage, maybe by some other means.


The specific "nerfs" that brought it up were in regard to balancing costs (I use a point buy game). I made it so that a guy who could fly would pay the same costs regardless of how they got the ability (be it through being a wizard with the fly spell, having wings, or owning a magic carpet for example), or making various spells of the same level be "equal but different" rather than one simply being a no brainer and another being a trap option.

An example of perfect imbalance would look like this: the wizard's flying ability is better but only lasts for limited periods, the wings tire you out after a while and it's hard to steer/you have to keep moving (no hovering), the magic carpet maybe doesn't move as fast and you can get knocked off. Maybe one flight ability is a little bit better for its price, but it shouldn't be significantly better and the costs should be very close to being the same. Then the players have to pick which form of flight based on the metagame of how monsters will attempt to neutralize their flight, and can prepare to neutralize enemy flight ability depending on how the enemy is doing it.

Various spells of the same level being 'equal but different' is too vague to say anything about. Is it like having a fireball and an iceball and a charm person which does 5d6 psychic points of damage (which would be literally equal, but "different") or is it just an attempt to have wildly varying abilities all of which are valid options depending on the situation?

Talakeal
2013-07-24, 05:37 PM
My guess is that with your changes, for a few people a class went from "well, it's not good at X, but it's great at Y" to not being as great at Y. Perhaps other classes were now better at Y, making the first class' loss of X feel like more of a sacrifice for less gain.


Other way around actually. It was

"Class A is great at X and Y, and Class B is bad at X and great at Y".

To:

"Class A is great at X and bad at Y, and B is bad a X and great at Y".


Is this a comparison to your setting, or is that for 3.5?

Because in 3.5, a rogue would never take the 6 levels just for fell flight. That would simply be obtuse.

If you're talking about your system, it may be a flaw in the way resources for wizards is handled in general.

Generalization. It is a point buy system. The buffs you could cast with spells were so efficient that it was cheaper for a character to actually pick up spellcasting in the school that provided a certain ability than it would cost to just pick up the ability directly. The wizard would then also have the added utility of being able to cast non buff spells on top of that.

It absolutely was a flaw in way resources were handled, and I corrected the math*. Which resulted in a game that was "so bland it was absolutely no fun at all," in the words of the player.

*Basically a given character can now only have X number of buffs up at any time, and the cost to allow an additional buff is roughly equivelent to the cost to get that buff from a different source. It is the same cost either way, but if you do it through magic you can switch out the buff when needed, but on the other hand it is always available and doesn't need to be recast. It seems like an elegant solution to myself and the three players.

Yukitsu
2013-07-24, 05:45 PM
Generalization. It is a point buy system. The buffs you could cast with spells were so efficient that it was cheaper for a character to actually pick up spellcasting in the school that provided a certain ability for less than it would cost to just pick up the ability, and also have the added utility of all the other spells they could cast.

You may be well off if you severely limit the caster's resource in some measure, so while perhaps it does have tremendous versatility, which is a real draw to the caster archetype, it can't be used ad infinitum. By making it cheaper to learn the spell as compared to simply having the abilities, the class can maintain that feel, but not be outright more powerful if they can't actually use all of those abilities. That's why SP and Vancian systems are so common for actual magic, or even the sanity mechanic in CoC. Magic can very well do anything, and can be pretty easy to get, but ultimately the system limits casters (some better than others.) By making it functionally identical to taking each spell and paying for it as a generic ability, you weaken the feel of the archetype. Frankly, that's a complaint I have about an M&M PB style system in the first place.

Talakeal
2013-07-24, 05:52 PM
You may be well off if you severely limit the caster's resource in some measure, so while perhaps it does have tremendous versatility, which is a real draw to the caster archetype, it can't be used ad infinitum. By making it cheaper to learn the spell as compared to simply having the abilities, the class can maintain that feel, but not be outright more powerful if they can't actually use all of those abilities. That's why SP and Vancian systems are so common for actual magic, or even the sanity mechanic in CoC. Magic can very well do anything, and can be pretty easy to get, but ultimately the system limits casters (some better than others.) By making it functionally identical to taking each spell and paying for it as a generic ability, you weaken the feel of the archetype. Frankly, that's a complaint I have about an M&M PB style system in the first place.

Oh, no, it isn't just buying the ability and calling it a spell. That would be boring and bland. The wizard can also have a whole host of other abilities on top of that. I also do use spell points. A character could, however, pick up the skill to cast magic at a high enough level to have flight active 24 hours a day and, assuming they don't fight enemies who throw chain dispel magic at them, still have plenty of SP left over for other uses.

Basically, the whole linear fighter quadratic wizard thing was creeping into my point buy game, and it got to the point where a properly built wizard could simply replicate all the abilities of any other "class" or "race" and still have enough points left over to be an effective caster at the same time.


The principle of the video can be applied to any type of game. Your players have options, and if the game is perfectly balanced it means that all options are equally valid (in the definition of balanced in the video). This can be boring (your decisions seem less meaningful) and lowers the excitement of finding out how to invalidate your opponent's choices, because the system is designed to make that really hard.

Perfect imbalance means that all choices are pretty close to each other, but some of them can be better than others. On the flip side those things need weaknesses that can be exploited - and anyone exploiting that weakness has to have a weakness of their own that can be exploited.

Who is making these choices and how they are competing against each other doesn't really matter.

Elemental damage in dnd is an example of an attempt at such a system in an RPG. Fire spells do more damage than acid spells, but are easier to resist. So players have a strategic choice as to which spell to pick. And if players only do fire damage (in a vacuum, the better choice), the dm can counter by sending up lots of fire resistant creatures (a metagame decision by the dm) which should make them find ways of dealing with fire resistant creatures - maybe by doing acid damage, maybe by some other means.



An example of perfect imbalance would look like this: the wizard's flying ability is better but only lasts for limited periods, the wings tire you out after a while and it's hard to steer/you have to keep moving (no hovering), the magic carpet maybe doesn't move as fast and you can get knocked off. Maybe one flight ability is a little bit better for its price, but it shouldn't be significantly better and the costs should be very close to being the same. Then the players have to pick which form of flight based on the metagame of how monsters will attempt to neutralize their flight, and can prepare to neutralize enemy flight ability depending on how the enemy is doing it.

Various spells of the same level being 'equal but different' is too vague to say anything about. Is it like having a fireball and an iceball and a charm person which does 5d6 psychic points of damage (which would be literally equal, but "different") or is it just an attempt to have wildly varying abilities all of which are valid options depending on the situation?

This is exactly what I was going for actually. I am not making it "Fire ball does 5 fire damage. Cone of Cold does 5 cold damage. Magic missile does 5 magic damage. Hitting him with your staff does five blunt damage" Do whatever you like because it all works out the same in the end.

Rather it is: "Fireball does 10 fire damage to all enemies in a burst and adds a burning effect. Cone of Cold does 10 damage to all enemies in a line and adds a slow effect. Magic Missile does 15 points of damage with no secondary effects. Quarter staff does 10 points of blunt damage to an enemy and doesn't require spell points," and then I am nerfing things like "Charm person does 20 points of damage and adds a dominate effect and costs no more spell points than the first three spells."

Yukitsu
2013-07-24, 05:59 PM
Rather than limiting what magic can do, I'm far more inclined towards fixing the balance of spell points in that particular case. It sounds like a lot of problems stem from an ability to use spells constantly without any fear of running out rather than the ability to gain fly as a spell in particular. It simply sounds as though your caster SP calculations were well off their mark.

Talakeal
2013-07-24, 06:03 PM
Rather than limiting what magic can do, I'm far more inclined towards fixing the balance of spell points in that particular case. It sounds like a lot of problems stem from an ability to use spells constantly without any fear of running out rather than the ability to gain fly as a spell in particular. It simply sounds as though your caster SP calculations were well off their mark.

At a base level I would agree. The problem was when you add metamagics into the mix you get a wizard giving the entire party 24 hour flight for a single spell slot.

Rather than nerfing a mages ability to perform utility spells or combat actions each round I chose to instead simply limit buffing ability.

Ravens_cry
2013-07-24, 06:28 PM
At a base level I would agree. The problem was when you add metamagics into the mix you get a wizard giving the entire party 24 hour flight for a single spell slot.

Rather than nerfing a mages ability to perform utility spells or combat actions each round I chose to instead simply limit buffing ability.
Or make more party buffs so that you are helping the whole party, not just you.

Talakeal
2013-07-24, 06:34 PM
Or make more party buffs so that you are helping the whole party, not just you.

It was actually party buffs that were the big problem. Wizards became mandatory in every party and encounters that weren't balanced with a fully buffed party in mind would get taken down in seconds.

If no one in the party wanted to play a wizard, or the wizard missed a night, the party would get stomped by enemies that were balanced for a buffed party.

And of course, if someone wanted to power game, they could make an all wizard party and have everyone be a flying, fire breathing, energy immune, incorporeal, amporphous, multi headed, master of all skills with +10 bonuses to every attribute and a halo of animate weapons and armor.


To use the fly example, the player who paid the ECL for a winged race or spent the gold for a flying carpet sure would feel like a fool when the party mage snapped his fingers and made the entire rest of the party fly for free.

erikun
2013-07-24, 06:40 PM
TLDR: Can anyone watch this video and tell me how it is relevant to tabletop RPGs, or cooperative team games?

http://www.penny-arcade.com/patv/episode/perfect-imbalance
I've seen this before, but don't seem to recall any relevance it had to non-video game RPGs. In fact, I had the impression that it distinctly had no relevance to tabletop RPGs. Give me several minutes to re-water the episode to refresh my memory.

...

Okay, first - and this is a big first - the video is talking about balance in the context of player vs player interactions in a large-scale metagame. This doesn't have relevance in your (likely) cooperative campaign game, and even if it was a player vs player game, would still not have relevance. The video discusses how dynamic metagames play out, with several players specifically targetting the weaknesses with the vastly popular options, those choices becoming the popular options, and then several different players specifically targetting the weaknesses with the now-different popular options.

(I would also argue that the depth of such does not seem as good as Extra Credits has indicated. It seems you'd only need three heroes: Hero A being slightly better, Hero B the counter, and Hero C the counter's counter. From there, you'd just go back to using the superior Hero A. The concept is known as Yomi Layers (http://www.sirlin.net/articles/yomi-layer-3-knowing-the-mind-of-the-opponent.html), for your reading pleasure.)

Anyways, the majority of the video has nothing to do with your game. The game isn't PvP, and even if it was, the players do not swap out characters nor is there a large enough player base (I hope) to make the discussion really relevant. I would not say this applies in a "PvDM" sense because, once again, the players aren't likely to be swapping characters frequently enough to select metagame counters to the DM's choices. Well, not unless you're creating a very different game than most RPGs I've seem.


Generalization. It is a point buy system. The buffs you could cast with spells were so efficient that it was cheaper for a character to actually pick up spellcasting in the school that provided a certain ability than it would cost to just pick up the ability directly. The wizard would then also have the added utility of being able to cast non buff spells on top of that.

It absolutely was a flaw in way resources were handled, and I corrected the math*. Which resulted in a game that was "so bland it was absolutely no fun at all," in the words of the player.

*Basically a given character can now only have X number of buffs up at any time, and the cost to allow an additional buff is roughly equivelent to the cost to get that buff from a different source. It is the same cost either way, but if you do it through magic you can switch out the buff when needed, but on the other hand it is always available and doesn't need to be recast. It seems like an elegant solution to myself and the three players.
It seems to me that your player/friend/playtester is trying to say "perfect balance is boring" in the most obtuse passive-agressive way possible.

Anyways, most point-buy games tend to make the flight spell itself rather cheaper than the granted-flight ability, but make the spellcasting ability quite expensive to make up for it. As such, it's cheaper to just pick up flight (if that's all you want) or just pick up flight and darkvision, but if you want a dozen different abilities, you'd be better off picking up spellcasting.

Reading through the other responses, I have to ask how a spellcaster can provide flight to an entire group of 4-6 people for 24 hours with a single spell slot? Chained persistent spells are intended to be quite expensive in D&D, and if spellcasters in your system can do so very easily, then I think the biggest problem is (1) The low cost of metamagic, or (2) The availability to reduce said cost to insigificant levels.

Talakeal
2013-07-24, 07:03 PM
I've seen this before, but don't seem to recall any relevance it had to non-video game RPGs. In fact, I had the impression that it distinctly had no relevance to tabletop RPGs. Give me several minutes to re-water the episode to refresh my memory.

...

Okay, first - and this is a big first - the video is talking about balance in the context of player vs player interactions in a large-scale metagame. This doesn't have relevance in your (likely) cooperative campaign game, and even if it was a player vs player game, would still not have relevance. The video discusses how dynamic metagames play out, with several players specifically targetting the weaknesses with the vastly popular options, those choices becoming the popular options, and then several different players specifically targetting the weaknesses with the now-different popular options.

(I would also argue that the depth of such does not seem as good as Extra Credits has indicated. It seems you'd only need three heroes: Hero A being slightly better, Hero B the counter, and Hero C the counter's counter. From there, you'd just go back to using the superior Hero A. The concept is known as Yomi Layers (http://www.sirlin.net/articles/yomi-layer-3-knowing-the-mind-of-the-opponent.html), for your reading pleasure.)

Anyways, the majority of the video has nothing to do with your game. The game isn't PvP, and even if it was, the players do not swap out characters nor is there a large enough player base (I hope) to make the discussion really relevant. I would not say this applies in a "PvDM" sense because, once again, the players aren't likely to be swapping characters frequently enough to select metagame counters to the DM's choices. Well, not unless you're creating a very different game than most RPGs I've seem.


It seems to me that your player/friend/playtester is trying to say "perfect balance is boring" in the most obtuse passive-agressive way possible.

Anyways, most point-buy games tend to make the flight spell itself rather cheaper than the granted-flight ability, but make the spellcasting ability quite expensive to make up for it. As such, it's cheaper to just pick up flight (if that's all you want) or just pick up flight and darkvision, but if you want a dozen different abilities, you'd be better off picking up spellcasting.

Reading through the other responses, I have to ask how a spellcaster can provide flight to an entire group of 4-6 people for 24 hours with a single spell slot? Chained persistent spells are intended to be quite expensive in D&D, and if spellcasters in your system can do so very easily, then I think the biggest problem is (1) The low cost of metamagic, or (2) The availability to reduce said cost to insigificant levels.

That was my analysis of the video as well. I was just hoping there was something useful there I was missing.


As for my spell system (bear with me here):

A wizard can get an extra spell point for one character point.
All spells cost one spell point to cast, succeed or fail.

If a character rolls a critical success (a natural 20 followed by a confirmation roll) the spell point is refunded.

Spells require a skill test to cast.

Skills value correspond as follows:

-5: Untrained character with poor stars.
-: Average untrained character
+5 Trained character with poor stats or untrained character with exceptional stats.
+10 Average full trained character. (players usually start around here).
+15 Veteran
+20 Master
+25 Epic hero.
+30 God

All spells have "seeds" which are the base DC to cast them. The variables such as range, number of targets, duration, casting time, etc. modify this DC.

Fly has a base seed DC of 15. +10 for affecting an unlimited number of targets. +10 for a duration of one month. -1 for range of touch. -2 for a casting time of 1 hour. Final Result is 32.

The mage had a modifier in the high teens when the spell was cast, and had the party bard singing for a +3 to the test. So he could cast the spell roughly a third of the time.

So, barring enemy counter spells, the party could have perpetual flight for 3 spell points. The wizard could have self buffed flight in the same manner with a virtually 100% chance of failure, and a less skilled wizard could pull of the spell self only with about the same odds.

If the party had merely taken "wings" as a trait it would have cost EACH of them between 6-18 character points depending on how manueverable they wanted to be.

Yukitsu
2013-07-24, 07:09 PM
Seed systems are invariably a problem. I would seriously reconsider using any variant of a seed system. The reason being is that it's impossible to estimate how much any one player will or will not have in terms of their ability to hit certain DCs, meaning that you're balancing to the average, and when you balance to the average, you're stuck with some people far too weak and incompetent, and some players who are doing things like easy 24 hour buffs.

Ravens_cry
2013-07-24, 07:20 PM
It was actually party buffs that were the big problem. Wizards became mandatory in every party and encounters that weren't balanced with a fully buffed party in mind would get taken down in seconds.

If no one in the party wanted to play a wizard, or the wizard missed a night, the party would get stomped by enemies that were balanced for a buffed party.

And of course, if someone wanted to power game, they could make an all wizard party and have everyone be a flying, fire breathing, energy immune, incorporeal, amporphous, multi headed, master of all skills with +10 bonuses to every attribute and a halo of animate weapons and armor.


To use the fly example, the player who paid the ECL for a winged race or spent the gold for a flying carpet sure would feel like a fool when the party mage snapped his fingers and made the entire rest of the party fly for free.
A winged race has the advantage of not being dispelable or AMF'd, the item bearer has the option of not spending spell slot or similar resources. I am not saying all buffs must be party buffs, but I would like to see more rather than less.
Furthermore, so what if everyone chooses to play one class and layer on the buffs? One of the advantages of a live DM is you can adjust things if needed.

erikun
2013-07-24, 07:31 PM
All spells have "seeds" which are the base DC to cast them. The variables such as range, number of targets, duration, casting time, etc. modify this DC.

Fly has a base seed DC of 15. +10 for affecting an unlimited number of targets. +10 for a duration of one month. -1 for range of touch. -2 for a casting time of 1 hour. Final Result is 32.

The mage had a modifier in the high teens when the spell was cast, and had the party bard singing for a +3 to the test. So he could cast the spell roughly a third of the time.

So, barring enemy counter spells, the party could have perpetual flight for 3 spell points. The wizard could have self buffed flight in the same manner with a virtually 100% chance of failure, and a less skilled wizard could pull of the spell self only with about the same odds.
Or the mage could just drop the "+10 for affecting unlimited number of targets" and have a DC 22, resulting in a pretty much guaranteed result for 3 spell points per person. For flight for one month. While I assume he could just restore spell points within 24 hours.

I think there are some major problems with both your spellcasting system and your point-buy values for different skills if granting flight for an unlimited number of targets for one month, for as little as 3 character points. Compared to the same flight costing 18 character points without spellcasting.

kyoryu
2013-07-24, 07:39 PM
(I would also argue that the depth of such does not seem as good as Extra Credits has indicated. It seems you'd only need three heroes: Hero A being slightly better, Hero B the counter, and Hero C the counter's counter. From there, you'd just go back to using the superior Hero A. The concept is known as Yomi Layers (http://www.sirlin.net/articles/yomi-layer-3-knowing-the-mind-of-the-opponent.html), for your reading pleasure.)


It's also known as Rock-Paper-Scissors, as well as Not Having Dominated Strategies.

The important thing in an RPS cycle is that the cycle exists - not that the payoff for each choice is exactly even. Even if Rock gets 2 'points' on a win, while paper gets only one point, it's still a workable game, just with a different optimal strategy.

The game only breaks down (and even then it's a partial breakdown) when one of the options is *always* inferior to another, regardless of what the opponent does.

The game really breaks down if one option *always* is superior, and has no counter.

There's weaker versions of these types of dysfunctional games, generally where the advantage of a strategy that's not actually dominated is lower than the error cost of making the decisions appropriately.

To put it more simply, it doesn't matter if the wizard is slightly more effective than the fighter. It's really irrelevant, especially within the context of a cooperative game. What matters is that the Fighter is better at some things than the wizard. When that ceases to be true, then you've eliminated the Fighter as a valid choice.

When you get to the point that the wizard can replicate all the functions of *every* other class, then you've got a real issue.

Talakeal
2013-07-24, 07:49 PM
Or the mage could just drop the "+10 for affecting unlimited number of targets" and have a DC 22, resulting in a pretty much guaranteed result for 3 spell points per person. For flight for one month. While I assume he could just restore spell points within 24 hours.

Spell points recover each month, not each day, which is why the month long duration is optimal.

It is a six person party, so dropping the unlimited targets would mean 90~ success rate, or 6.6 spell points per mission.



I think there are some major problems with both your spellcasting system and your point-buy values for different skills if granting flight for an unlimited number of targets for one month, for as little as 3 character points. Compared to the same flight costing 18 character points without spellcasting.

The system overall works really well. Significantly better than any published RPG system I have played if I do say so myself. There were just a few little snarls and loopholes which I had to iron out, which is what brought about the whole "perfect balance is bad" argument.

Basically I just had to put a few limits on "buffs" and spells which create permanent resources (such as turning lead into gold, a contingent spell, or a spell that allows another mage to recover spell points) to smooth it out.


A winged race has the advantage of not being dispelable or AMF'd, the item bearer has the option of not spending spell slot or similar resources. I am not saying all buffs must be party buffs, but I would like to see more rather than less.


All spells in the game are able to be cast on others. I don't believe I have any "self only" spells, except for maybe things like divinations which don't have a physical effect.

The guy with wings can operate in AMFs fine. But will tire quickly and can be grounded by having their wings bound or injured.

The guy with the carpet can fly all the time, but has to keep track of the item which can be stolen or taken from him.

The guy with the spell has to keep track of spell points and worry about counter spells, but they can switch out flight for some other ability when it isn't an optimal strategy.


All three of these seem roughly balanced to me. Why should one cost significantly more or less than any of the others? If someone wants to do one for flavor reasons, why punish them by having it be arbitrarily more costly?



Furthermore, so what if everyone chooses to play one class and layer on the buffs? One of the advantages of a live DM is you can adjust things if needed.

In theory nothing. In practice, it forces players into one roll (the guy who doesn't WANT to be a mage is going to be forced to either be at a huge disadvantage or play a role he doesn't want to play), and it ruins the mood of the setting by instead of having magic being rare and mysterious it is ultra prevelent, with the majority of PCs (and worthwhile opponents) being forced into wizardry.

Ravens_cry
2013-07-24, 08:03 PM
In theory nothing. In practice, it forces players into one roll (the guy who doesn't WANT to be a mage is going to be forced to either be at a huge disadvantage or play a role he doesn't want to play), and it ruins the mood of the setting by instead of having magic being rare and mysterious it is ultra prevelent, with the majority of PCs (and worthwhile opponents) being forced into wizardry.
Only if the players choose to go that route. Also, wizardry being rare is itself variable depending on setting, and settings can be altered. The rareity and an all wizard party isn't even mutually exclusive. In Ars Magica or Mage: The Awakening, everyone is a mage of some sort, even though magic is rare in those settings.

NichG
2013-07-24, 08:29 PM
The point of controlled intentional imbalance is more about psychology than game theory I think. Simply put, it can be fun to feel like you've 'gotten away' with something. So from the point of view of the game designer, it can be attractive to put in gimmicks that appear 'broken' to the players but aren't actually a big problem to the overall stability of the game. This is of course a challenging thing to do - its far too easy to make abilities that can be countered, but only by one or two things (thus resulting in things like 'all enemies have to be able to fly above a certain CR' or 'all bosses have to have immunity to ability damage, drain, etc' to avoid shivering ray and the like).

In my mind, what you want in a tabletop RPG is to create a balance between this feeling of 'getting away with something' or more generally being supremely competent at something, pre-cooked tactics, and scenario-specific tactics. This is more important than some kind of rock-paper-scissors balance between PC classes.

What I mean here by pre-cooked tactics is something like a strategy that is basically done by the time you leave character creation - some inherently powerful trick that is going to basically be good in all situations. In my mind these are nearly strictly bad, though it can be good to have some decent ones of these for players who aren't very tactically minded. In D&D analogy, these tactics are the equivalent of the spellcaster preparing Gate and gating in a Wish-granting creature - there are very few situations that cannot be solved by this pattern, so the need to analyze the scenario is reduced and most of the 'game' is actually found in the character generation phase rather than in actual play.

Scenario-specific tactics on the other hand are when the PC abilities interact very differently with different situations. This is where you want a lot of imbalance, back and forth between different abilities, classes, etc. You want something where in one situation the best answer may be 'the wizard casts Fly' but in another situation the guy with built-in flight is doing better, and in a third situation the guy who can jump really far is doing better. Ideally these differences in immediate ability are not between characters, but between things that each character can choose to do (e.g. so you don't have the issue of the rogue having to sit out the battle because everything is immune to sneak attack, because you have given the rogue some other thing to do that does work well against, say, undead but not against living targets).

Emmerask
2013-07-24, 08:45 PM
TLDR: Can anyone watch this video and tell me how it is relevant to tabletop RPGs, or cooperative team games?

http://www.penny-arcade.com/patv/episode/perfect-imbalance



Backstory:

I am currently play testing a homebrew system, and one of the players was complaining about the "nerfs" I made for the sake of balance. He then told me I needed to watch the video.

I watched it, but couldn't see the connection between this and table top RPGs, especially cooperative team games like mine is supposed to be.

When I asked the player what about it pertained to my game, he told me "It should be self-evident," and then refused to talk about it anymore (literally, as in he hangs up on me or walks out of the room and slams a door in my face anytime I bring it up).

I have watched it half a dozen times now and can't seem to make the connection, and have showed it to a couple of other players who also can't make the connection.

For games like dota and lol cyclic imbalance somewhat works because you play hero xyz for ~40min then you can move on. And even then its not the whole truth pa talks about really, because the meta does not really evolve further at certain points. If you watch some pro games you will notice that ~2months after that last patch/hero introduction 99% of the games will feature roughly the same lineups.

The thing with rpgs is that in most cases you will want to play your character for a lot longer then 40minutes so you cant adapt by just changing to char xyz. So no its not really applicable to rpgs...

valadil
2013-07-24, 08:47 PM
I was aware of the concept, but never heard the term perfect imbalance tied to it. Thanks for sharing.

TheThan
2013-07-24, 11:13 PM
OH boy game theory!

I look at it in the terms of player power and fairness. After all nobody likes being out done by the party wizard because he’s more powerful and the rest of the party.


Anyway as they said creating intentional imbalance is not easy, and it’s fairly subtle. That’s the thing, it is subtle. If you have one class or set of abilities that are widely out of scope with each other, then your game is not in balance and the class or set of abilities that are imbalanced in a negative light are not going to see play when people notice the balance difference.

Now when its imbalance that’s only marginally out of scope with each other, each imbalance element will have advantages and disadvantages, which in a screwy way keeps them in balance because now you have to weigh those advantages and disadvantages. This is why the traditional “four (or five) man band” party is so popular; because each player character has certain advantages over the others. No-one can pick locks and disarm traps except the rogue. No-body can magically heal the party except the druid or cleric. Nobody can throw out massive aoe effect spells like the wizard or sorcerer, and nobody and take hits and deal crazy amounts of damage like the fighter and barbarian. Each character supports each other and works as part of a team, filling a specific role. The sum of their parts is more powerful and effective than each individual.

Let me give an example of perfect imbalance:
The fantastic four is made up of four characters: Mr. Fantastic, The invisible Woman, The Human Torch and The Thing. Each one has their own unique set of advantages and disadvantages. They are more powerful as a team than they are by themselves. This is perfect imbalance, The Human Torch can’t lift as much as The Thing, nor can he take as much abuse. But The Thing can’t fly and shoot fire, like the Human Torch can, it’s a trade off.

Now to create an example of perfect balance:
Imagine if the justice league was made up of only Superman, Supergirl, and PowerGirl. Now all three of these characters share the exact same power set: super strength, flight, heat vision, invulnerability the works. This is perfect balance; they have the same advantages and the same disadvantages. You can replace each one with another one and have no change in general capabilities. They’re pretty much the same.

Now to create an example of poor balance:
Imagine Superman teaming up with wildcat. Now wildcat is a good fighter, but superman trumps him in literally every measurable way, he’s stronger, faster, more durable, just as experienced and skilled in fighting. Also he’s got powers that wildcat does not have: flight, heat vision, super breath, etc. why would anyone looking to “win” pick wildcat when they could pick superman? Superman is obviously better. Writers of such a story are forced to go out of their way to make wildcat relevant at all since superman completely overshadows him.

Or to use a humors example:
BMX bandit and angel summoner (http://www.youtube.com/watch?v=sbzUfV3_JIA)

huttj509
2013-07-25, 01:25 AM
The point of controlled intentional imbalance is more about psychology than game theory I think. Simply put, it can be fun to feel like you've 'gotten away' with something. So from the point of view of the game designer, it can be attractive to put in gimmicks that appear 'broken' to the players but aren't actually a big problem to the overall stability of the game. This is of course a challenging thing to do - its far too easy to make abilities that can be countered, but only by one or two things (thus resulting in things like 'all enemies have to be able to fly above a certain CR' or 'all bosses have to have immunity to ability damage, drain, etc' to avoid shivering ray and the like).

[snip]


I think this idea is key to what happened for the OP's situation. While the player involved felt like he was 'getting away with something,' it was actually too much for the system. As a result, the system was adjusted (as it should be, point of playtesting and adjusting). However, the player then felt 'punished' for finding a break point.

Ideally, playtesters can separate that sort of psychology from the testing, but it can be really difficult when the playtesting is wrapped into the normal gaming time/mindset.

Heck, I ran into that with a bit of a card game playtest where in some feedback I was consciously trying to separate "these heroes/villians have some balance issues" from "I'm annoyed we lost so handily." Both were true, I just didn't want to overstate the former due to the latter.

kyoryu
2013-07-29, 04:37 PM
It's kind of neat, because as usual, Sirlin demonstrates an instinctive understanding of game theory (in the mathematical sense) while missing a few key issues.

Specifically: Chess is not highly memorization-based/solved/whatever because of its "perfect balance". It has those attributes because it is a type of game known as a sequential perfect knowledge game. Basically, it's a game where the players take their turns sequentially, and there is no hidden knowledge (randomization, hidden cards, etc.).

Any sequential perfect knowledge game is solvable - that is, there is a definable optimal sequence of moves to make. The most obvious example of this is tic-tac-toe, as everyone (including WOPR) knows, you can't win it against a player that knows the right moves to make. Chess is non-trivial, of course, because its solution space is so much bigger.

(side note: the sequential perfect knowledge game that is probably the least solved is Go, and that's just because its solution space dwarfs that of Chess. And even parts of it are "solved")

Sirlin's classic example of Yomi, wake-up games, are not sequential perfect knowledge games. They're simultaneous games, in that, for all practical purposes, each side makes their choice at the same time. Rock/paper/scissors is another simultaneous turn game, and the two are actually closely related. Sirlin's main point with some of this stuff is that unequal payouts in simultaneous move games can create more interesting mental games on the first move, which is utterly valid. The Nash Equilibrium (essentially, but not precisely, the optimal strategy for the game) for RPS is trivial - play each 1/3 of the time. If you use unequal payouts, then it becomes much harder to solve for the Nash Equilibrium but is still possible. It ends up being something like play the "better" choice 40% of the time, its counter 40% of the time, and the remaining choice 20% of the time. I'm probably way off with that, but you get the idea.

Where the idea of "perfect balance" in games is truly harmful is to imagine a game of rock paper scissors where, no matter what you and your opponent choose, you each get a point.

Now, as far as metagames, what they often do is put a RPS-type cycle at the top level of the game - which can be fine, but taken to too much of an extreme can mean that the "metagame" is the only game that actually matters, as the match is effectively determined once your heroes/decks/whatever are chosen.

There's a lot more in-depth stuff that's applicable to games in terms of game design in mathematical game theory. Understanding dominated strategies, and all sorts of other good stuff.