General [Design Discussion] Perfect Imbalance

[FlowerSunRain]

A good point, and cyclical is indeed simplistic at best (and yes frustrating). But a cube is far too multi-variable to analyze as-is. So cyclical ABCA is a good model for understanding. XvY is another. And any new meta will almost guaranteed have to be multidirectional to a degree. Before people even know what's what about X v. Y, they have to play around with the options out there, and this can be kinda "circular". Maybe more experienced players will finish this almost immediately and move to XvY, or ForceThisThingAlways, or whatever. Almost no model is a complete farce, just more accurate in some situations and at some times.

Every game will have a "feeling things out" period where strategies arise to beat other strategies which are in turn defeated by other strategies, but none of that is predicated on "perfect imbalance" as a design goal. A lack of system knowledge handles that just fine. As knowledge mastery is approached, the game presents other problems (execution) and this is perfectly acceptable. Trying to keep game in a stage of constant "discovery" via "a fluxuating metagame" doesn't actually happen. Metagames always settle or at least move linearly in response to increased mastery. In order to keep the metagame changing, game designers have to actually change the game. This is certainly a way you can achieve that result, but it can never result in a finished game. When your product doesn't have to stand up to repeated play in a singular form, you can take a lot more liberties with your design.

 

[sigh]

Every game will have a "feeling things out" period where strategies arise to beat other strategies which are in turn defeated by other strategies, but none of that is predicated on "perfect imbalance" as a design goal. A lack of system knowledge handles that just fine. As knowledge mastery is approached, the game presents other problems (execution) and this is perfectly acceptable. Trying to keep game in a stage of constant "discovery" via "a fluxuating metagame" doesn't actually happen. Metagames always settle or at least move linearly in response to increased mastery. In order to keep the metagame changing, game designers have to actually change the game. This is certainly a way you can achieve that result, but it can never result in a finished game. When your product doesn't have to stand up to repeated play in a singular form, you can take a lot more liberties with your design.

Actually exactly what we are talking about, but maybe from different perspectives. The takeaway from both our arguments is "the way to keep the meta changing is to continuously change the cube". My variables don't include a constant offset. Just resources and knowledge. So yes, exactly as you say, as mastery is developed, a resource change must occur to keep change happening.

The only other way is to somehow stop every single player from learning..... Or, changing your players often (another way to change the knowledge side)

 

[FlowerSunRain]

1) Is a constantly changing meta desirable? I play games for their specific quality. I usually don't want them to be constantly changing. My constant tinkering with my cube design is not to "keep the meta fresh", but rather to improve the gameplay of the cube. My theoretical end goal is to finish it and never change it again.
2) The idea of "imperfect balance" (as presented in the youtube video) is that the meta-corrects WITHOUT outside influence: the game is designed to do so. If you need to keep changing the game to achieve that effect, then a game isn't designed in a "perfectly imbalance" fashion, it is just constantly being redesigned with different imbalances.

 

[Rasmus Källqvist]

Every game will have a "feeling things out" period where strategies arise to beat other strategies which are in turn defeated by other strategies, but none of that is predicated on "perfect imbalance" as a design goal.

Isn't this essentially that we want to avoid the roshambo rock-paper-scissors? The kind of meta game that boils down to "Am I playing rock? Great, most of the field is scissors today!".

 

[FlowerSunRain]

Isn't this essentially that we want to avoid the roshambo rock-paper-scissors? The kind of meta game that boils down to "Am I playing rock? Great, most of the field is scissors today!".

Things are always weird in the beginning of learning of complex game and doesn't necessarily reflect how great the design is. If your game is rock-paper-scissors at an early stage of learning the game, its not desirable, but it might be fine because players will get better and get past it. If your game is rock-paper-scissors approaching the end level of mastery, you've got a problem.

 

[Rasmus Källqvist]

Things are always weird in the beginning of learning of complex game and doesn't necessarily reflect how great the design is. If your game is rock-paper-scissors at an early stage of learning the game, its not desirable, but it might be fine because players will get better and get passed it. If your game is rock-paper-scissors approaching the end level of mastery, you've got a problem.

Yes, I agree.

1) Is a constantly changing meta desirable? I play games for their specific quality. I usually don't want them to be constantly changing. My constant tinkering with my cube design is not to "keep the meta fresh", but rather to improve the gameplay of the cube. My theoretical end goal is to finish it and never change it again.

I could reason that it is desirable because a solved meta is undesirable and that changing the meta avoids that problem. But I think I'm making some assumptions? I usually try to fall back on the thumb rule that we want to maximize the number of meaningful decisions we can make, and a solved meta means that there is a right and wrong way to draft, build, strategize and play.

I guess a goal for the draft part is to let people just express creativity and ingenuity without having to worry about the master recipe and just respond to what is going on at the table during that play session.

 

[sigh]

1) Is a constantly changing meta desirable? I play games for their specific quality. I usually don't want them to be constantly changing. My constant tinkering with my cube design is not to "keep the meta fresh", but rather to improve the gameplay of the cube. My theoretical end goal is to finish it and never change it again.
2) The idea of "imperfect balance" (as presented in the youtube video) is that the meta-corrects WITHOUT outside influence: the game is designed to do so. If you need to keep changing the game to achieve that effect, then a game isn't designed in a "perfectly imbalance" fashion, it is just constantly being redesigned with different imbalances.

1) maybe not. I enjoy having a new challenge. So do most type 2 players, etc etc. I think the issue with coming full stop on cube design will be a very "samey" feeling environment. If this is the sort of environment that suites you and yours, then that sounds perfect.

2) My earlier posts haven't actually covered the initial question in depth, just major ways the meta will change independent of the source material. Perfect imbalance is referring to something along the lines of internal excitation. This could come from players being human, moods etc etc, and the shuffling of the cube. The order cards fall in will make every draft at least a little different. I DON'T actually think imbalance will achieve the effect desired in the OP UNLESS the imbalance does change. So yes, and yes, I agree with you.

In effect, I think my takeaway from the discussion so far is: the idea in the video of meta flux from imbalance isn't super duper. You'll get a skewed meta with people vainly fighting against it, imo. The idea of a meta that can constantly change is, however, A-OK, just not via that method, but via cube change at a constant rate.

EDIT: Rasmus beat me to it a little . I think we three are off the imbalance ship, and onto the "does shifting the meta make your cube more fun?" ship

 

[FlowerSunRain]

I could reason that it is desirable because a solved meta is undesirable and that changing the meta avoids that problem. But I think I'm making some assumptions? I usually try to fall back on the thumb rule that we want to maximize the number of meaningful decisions we can make, and a solved meta means that there is a right and wrong way to draft, build, strategize and play.

I guess a goal for the draft part is to let people just express creativity and ingenuity without having to worry about the master recipe and just respond to what is going on at the table during that play session.

1) I think a well designed cube draft has enough moving parts that having total knowledge of "correct play" requires more mastery then one will acquire in their lifetime, so it is irrelevant.
2) I don't think avoiding solved metas is difficult. If you want to avoid a solved meta you have 100s of other people's cubes and 1,000s of other games you can play. It is a non-existent problem.

I just enjoy games for what they are. The constant tinkering is to keep people invested in a single game. But I have different moods and I like playing different games for those different moods and would prefer a single definitively great version of the game for when that mood comes.

In other words, I don't want to have to relearn Caylus everytime I want to play Caylus.

 

[vennythekid]

Isn't this essentially that we want to avoid the roshambo rock-paper-scissors? The kind of meta game that boils down to "Am I playing rock? Great, most of the field is scissors today!".

Unfortunately, in Magic, the overarching strategies (aggro, control, midrange) form a roshambo relationship with each other, unless we can develop a set of fundemental strategies different from these that don't form that relationship.

Tangent: are these three strategies inherent to the game of Magic, or is it just because every Magic player who has read strategy has been taught in this way (literally every strategy article ever references this roshambo) and therefore thinks about the game in these terms?

From what you guys have said (and I might be interpreting this wrong), it sounds like we're searching for a "meta" where every strategy has a 50-50 matchup versus every other strategy, assuming you have the skill to draft the strategy and the also the skill to play the matchup.

To respond to the solved meta question: Even if this meta is "solved" isn't the dynamic of playing it out interesting and exciting?

 

[Grillo_Parlante]

1) Is a constantly changing meta desirable? I play games for their specific quality. I usually don't want them to be constantly changing. My constant tinkering with my cube design is not to "keep the meta fresh", but rather to improve the gameplay of the cube. My theoretical end goal is to finish it and never change it again.
2) The idea of "imperfect balance" (as presented in the youtube video) is that the meta-corrects WITHOUT outside influence: the game is designed to do so. If you need to keep changing the game to achieve that effect, then a game isn't designed in a "perfectly imbalance" fashion, it is just constantly being redesigned with different imbalances.

The difference between "keeping the meta fresh" via constant tinkering and "improving the gameplay of the cube" seem a bit semantical. You can never really separate these two concepts completely. One of the bi products of constant tinkering is that the meta never has to settle and function on its own. If your theoretical goal is to get to a point where its "finished" than it has to have some sort of self perpetuating system in place once the designer's hands have been removed from the product.

As you stated:

When your product doesn't have to stand up to repeated play in a singular form, you can take a lot more liberties with your design.

I would also point out that the idea of "constant" change isn't native to a self-regulating system. People react at their own pacing, and in small playgroups especially, it can take months for a meta change to occur.

I'm also not sure why the discussion is starting to hinge on whether a format is completely hands off or not. I consider my own format finished, but if I went in a couple times a year to make one or two cuts or adds I wouldn't consider it a failure, I would consider it a resounding success. If the system is self-perpetuating itself so well that it needs such minimal maintenance, thats fairly extraordinary for a cube format.

 

[FlowerSunRain]

Tangent: are these three strategies inherent to the game of Magic, or is it just because every Magic player who has read strategy has been taught in this way (literally every strategy article ever references this roshambo) and therefore thinks about the game in these terms?

No. They are based on a sound principal inherent to gameflow, but because Magic uses specific cards that interact in specific ways with each other these relationships mutate wildly.

From what you guys have said (and I might be interpreting this wrong), it sounds like we're searching for a "meta" where every strategy has a 50-50 matchup versus every other strategy, assuming you have the skill to draft the strategy and the also the skill to play the matchup.

I think moving towards this is ideal. I also thinks its impossible. Further, I think that moving in this direction will sometimes require compromises in other areas of design that are undesirable. Strategic balance is an important consideration, but its not everything. I'd rather two decks have a 60/40 matchup that is fun then a 50/50 matchup that is a coin flip.

To respond to the solved meta question: Even if this meta is "solved" isn't the dynamic of playing it out interesting and exciting?

I think it is.

 

[FlowerSunRain]

The difference between "keeping the meta fresh" via constant tinkering and "improving the gameplay of the cube" seem a bit semantical. You can never really separate these two concepts completely. One of the bi products of constant tinkering is that the meta never has to settle and function on its own.

I see that as a failure. I want my product to stand up to repeated plays. I don't want to change it. Every time I change a card it is because I, as the designer, fucked it up. If I could plug my design goals into a machine that would remove human error and give me my ideal cube that I would never have to change, I would totally do that.

 

[Rasmus Källqvist]

1) I think a well designed cube draft has enough moving parts that having total knowledge of "correct play" requires more mastery then one will acquire in their lifetime, so it is irrelevant.
2) I don't avoiding solved metas is difficult. If you want to avoid a solved meta you have 100s of other people's cubes and 1,000s of other games you can play. It is a non-existent problem.

Both of those points feel like assumptions and/or just based on taste, would you care to elaborate on why it is a "non-existent problem"?

 

[FlowerSunRain]

Again, there are thousands of games and game variants out there. If you solved the meta of ONE variant of ONE game you can just play one of the many, many other options. Constantly changing a specific game to keep people playing it is important to game producers who want to keep an income stream going, but of no particular benefit to game players who will likely be just as happy playing something else. It is solving a non-existent problem: there is no shortage of other games to play.

 

[Rasmus Källqvist]

But not you've made the assumption that you're experience is the definitive experience and that we shouldn't have this discussion. I'm not playing other peoples cubes, so I'm not primarily interested in those meta games. It's possibly arbitrary, but as I make the choice of not wanting a solved meta I'm interested in whats going on in this thread. For me, figuring out the meta is part of the sub game. If you don't find this interesting I don't see why you'd engage.

 

[Grillo_Parlante]

Again, there are thousands of games and game variants out there. If you solved the meta of ONE variant of ONE game you can just play one of the many, many other options. Constantly changing a specific game to keep people playing it is important to game producers who want to keep an income stream going, but of no particular benefit to game players who will likely be just as happy playing something else. It is solving a non-existent problem: there is no shortage of other games to play.

We're not talking about constantly changing one specific game though: we are talking about designing the game so it can self perpetuate itself as long as its able, hopefully indefinitely.

The fact that you could just play a different game seems more like a self-imposed intellectual dead end, and a bit unrealistic given how expensive most cube formats are to build.

 

[ahadabans]

Tangent: are these three strategies inherent to the game of Magic, or is it just because every Magic player who has read strategy has been taught in this way (literally every strategy article ever references this roshambo) and therefore thinks about the game in these terms?

At the risk of just repeating my feelings on this, I simply can't resist replying. SQUIRREL!

Anyway, I think the answer is both. Roshambo is inherent in the the game to some extent. And that is fairly simply to prove out. If you have a low mana curve, you will be able to play more cards quicker. That is a very clear early advantage where mana is constrained. Likewise, later in the game when mana is not constrained, the higher curve deck is going to be playing better threats card for card. Also a very clear advantage. This gets muddied a bit when you factor in CA engines, "power" cards way above the curve or cards that generate ongoing value (walkers, etc), and what not, but the idea is still applicable. Slow decks do better later. Fast decks do better early. Midrange is generally better against fast decks because it's harder to go under midrange.

With that said, the traditional aggro > control > midrange > aggro model I fully believe is greatly overstated in most cubes simply because deck consistency is usually not there. Even if you push aggro to the moon, are guys really drafting decks with nine 2 power one drops and an average CMC of 2? Or control decks with 1 finisher, 8 wraths and a ton of conterspells and draw? You just can't build constructed decks in cube. At least not ones that will have the same consistency (IMO anyway). So I think that breaks the roshambo. Moreover, even in the most polarized scenario, it's not like aggro is 100% over control. 60/40? Maybe better depending. Take that edge off for cube though due to reduced consistency and what are we looking at 55/45? Maybe not even that much. To me, we might as well be saying aggro = control = midrange = aggro in cube.

To me, most of the focus of discussing aggro/midrange/control is in how the decks play versus how predictable these match-ups play out statistically speaking. And when people say "control needs love", it's usually not because control isn't winning it's because no one is making control decks so it feels underrepresented. My 2 cents anyway.

 

[FlowerSunRain]

But not you've made the assumption that you're experience is the definitive experience and that we shouldn't have this discussion. I'm not playing other peoples cubes, so I'm not primarily interested in those meta games. It's possibly arbitrary, but as I make the choice of not wanting a solved meta I'm interested in whats going on in this thread. For me, figuring out the meta is part of the sub game. If you don't find this interesting I don't see why you'd engage.

Because I hate that video.

 

[Grillo_Parlante]

Because I hate that video.

Yeah well....like I said, wish they had a transcript. Than I could just cut out and repackage everything to save the pain.

 

[Rasmus Källqvist]

Because I hate that video.

Yarrrrr! (That's fine!)

 

[MCMcEmcee]

I like having generically powerful, obvious first pick cards in the pool. Sometimes a new player doesn't know all of the themes/synergies, sometimes someone just ends up with a train wreck, and it's an overall better experience if they at least know they are live to draw their first pick and maybe win some games/matches.
That said, the generically powerful cards can ALSO be archetype anchors/payoffs/enablers, eg. Kitchen Finks, Archangel of Thune, Treasure Cruise, etc. Or the generically powerful cards can have a lot of play to them to get even more mileage out of them, eg. Vendilion Clique, Cryptic Command, Tasigur, etc.

Also, if your cube has gotten enough reps that it has actually been solved, you are very lucky to have such a dedicated play group!

 

[Grillo_Parlante]

I like having generically powerful, obvious first pick cards in the pool...Kitchen Finks, Archangel of Thune, Treasure Cruise....Vendilion Clique, Cryptic Command....

....if your cube has gotten enough reps that it has actually been solved

I think you may have just solved your own format

 

[safra]

I like having generically powerful, obvious first pick cards in the pool. Sometimes a new player doesn't know all of the themes/synergies, sometimes someone just ends up with a train wreck, and it's an overall better experience if they at least know they are live to draw their first pick and maybe win some games/matches.
That said, the generically powerful cards can ALSO be archetype anchors/payoffs/enablers, eg. Kitchen Finks, Archangel of Thune, Treasure Cruise, etc. Or the generically powerful cards can have a lot of play to them to get even more mileage out of them, eg. Vendilion Clique, Cryptic Command, Tasigur, etc.

Also, if your cube has gotten enough reps that it has actually been solved, you are very lucky to have such a dedicated play group!

This is where I'm at right now. I'm trying to avoid quote-unquote 'dull' powerful cards like Brimaz and instead push my power level with strong on-theme cards. The Collected Company theme is a good example; I'm running 3-drop monsters with additional casting costs, or flash, or just efficient stats. But the idea is that as powerful as these cards are in goodstuff decks, they're more powerful, or more flexible, in synergy decks. Eternal Witness is an incredibly goodstuff card that unlocks even more potential when you start to get tricky about things like flickering or reanimating it. Kitchen Finks is as goodstuff as most of us here get, but it enables half a dozen archetypes at the same time. It doesn't matter how on-theme bad cards are, I don't want to play them - so the ones I do want to play should be as on-theme as possible imo.

The easiest way to push that is to make your environment's strongest goodstuff cards archetypal ones; that way your best goodstuff decks (and don't tell me resto+finks+thragtusk isn't a goodstuff deck) are tapping into that gestalt synergy too.

 

[MCMcEmcee]

I think you may have just solved your own format

Yeah I just draft white aggro a lot

 

[sigh]

I think a good way to have a "different" feel each session is to simply have a slightly too-big cube. 405 instead of 360, for instance. If you really value the consistency etc etc, then this isn't the idea for you.

For a group whose less competitive, newer and smaller, oversizing can create variance that leads to "freshness" no matter what. Especially fun if there are dichotomy's like the Troll and the Edict. If one fails to show up, the imbalance is made for that session. In this way you get your "constantly changing slight imbalance" every session.