Brainstorm
Force of Will
Lion's Eye Diamond
Counterbalance
Sensei's Divining Top
Tarmogoyf
Phyrexian Dreadnaught
Goblin Lackey
Standstill
Natural Order
Brainstorm Realist
I close my eyes and sink within myself, relive the gift of precious memories, in need of a fix called innocence. - Chuck Shuldiner
The entire problem is the very nature of thresholds. Again, the algorithm itself could care less if the format is "good" or "fun" or well, anything. Pillerfield Ox is no more or less important than Underground Sea. Is that also the case for players? You claim the algorithm is objective, but this is a fallacy for a number of reasons. One, no "objective standard" was responsible for it's creation or design, only subjective criteria executed by an "object." There is no objective answer to what is "strict" or "lenient." So, even in that case, you still just have someone, or some group's, subjective valuation.
However, let's pretend that we are somehow in a format the is in the following state: 33.3% of top 8's are deck A, 33.3% are deck B, and 33.3% are deck C sustained over, lets say a year. Nothing else ever puts up results. Step one, what are we to make of this "format?" Should something be banned? Let us pretend, for the sake of this exercise, that A, B and C share no cards in common. So, what should have action taken on it? A, B, or C. How should we know, based off the data?
Lets take another case. Lets say our ban threshold is 80% sustained top 8 representation (whatever that means). What happens when something shows up 79.9% and sustains it? Is this a "good" format? Who knows, the algorithm doesn't care, it only takes input and renders output. If we round it, what happens if it is still just outside the "rounding" threshold? How close does "close counts" factor in? Why or why not?
Furthermore, to even construct this algorithm, but again we need to consider what the thresholds not only are, quantitatively, but what they serve, as a qualitative output. Because playing the game is not a quantitative exercise, it's a qualitative experience. What good is a format that has been it's card appearance distribution statistically normalized that is no fun for anyone to play?
Is a format "fun" because statistically speaking, if no card shows up more than 24.67% of the time? What about 37.45% of the time?
So, yeah, maybe I'll buy your algorithmic approach if you could somehow answer how or why these aren't issues, or how they can be clearly addressed via an algorithm. I'd surmise that an algorithmic approach would inevitably make absolutely 0% of anyone happy, unless they actually don't care about playing the game and only care out the statistical output.
Sorry, but Dataism doesn't save us, just like creative narratives don't. In the end, it's still just a subjective process of valuation. Trying to fashion an algorithm on top needlessly (and pointlessly) formalizes it, serving literally no one's interest as far as I can tell. Unless, again, formal statistical information is all that matters. In which case, why play Magic at all?
"The Ancients teach us that if we can but last, we shall prevail."
—Kaysa, Elder Druid of the Juniper Order
Have any of you guys played 3-card blind. Imagine a format where every year you have a final tournament, and then after the Final ALL the non-basic land cards from the winning deck get banned for X years. You'll have an ever changing format, the best cards will be banned in a revolving door style fashion and non of the cards could ever be that expensive since demand will be cyclical and the good ones are predisposed to be banned every other year. Fun.
Fun fact, I actually think Death Metal is completely enjoyable and I don't like it at all.
Yes, I'd much rather listen to "pure noise," as some liken Wolf Eyes, than DM, it's actually vastly more enjoyable to me.
That's just the thing though, there is no accounting for taste. This thread is a massive testament to that. The thing is, I really do try to not frame my subjective biases as facts of the matter. Some people though have a great deal of trouble separating the two, figuring they must be one and the same. I mean, I'm sure I fail from time to time as well.
The point of my likening there is that somehow, some people imagine that Legacy should be some harmonious, diverse place where anything and everything can "go." Where there is some optimal equilibrium where nothing has any "upper hand" and everything is unique, viable, there is both equality of opportunity and equality of outcome.
That's a nice idea, but it's not how anything works. You could say, "but that's how it should be!" Except it never will be. Pillerfield Ox won't make a Legacy top 8. Oh, but "No, no, that's not the point, we just want more things "competitive."" Ok, go on then, explain what that means, exactly. What's the threshold? What should be viable and what shouldn't and why?
There is no answer to this. The banned list is just Wizard's collective subjective valuation of what Legacy should be. It is an amalgamated, inscrutable, mess of contradictory, contingent, biased choices, some made empirically, some made spuriously teleologically, some just on some economic or historical basis.
That's what Legacy is. That is what Legacy will always be, because Legacy can't exist, nor can any Magic format that has any subjective play value, as a rigid formal system. It will, always, grow in potentially playable cards and constrict in actually playable cards in "natural" cycles. That is it's sort-of dialectical process, within itself, where it the metagame gradually moves toward a sort of "equilibrium." Sometime that state is undesired, for, again, (largely) inscrutable reasons on Wizard's part and the equilibrium is forced to shift.
What state is "optimal" equilibrium? Who knows. But Wizards is partial, seemingly, to the Brainstorm-Fetch paradigm being the "normative" state. Maybe a fetchless state is "more fun" but no one actually presented evidence that this would be the case. Because there is no way to qualify "fun." The notion, on the other hand, is presented as if it will lead to something like fabled diversity, which is somehow de facto surmised to be "better" and to be achievable.
Well, maybe I am just biased, but I am certainly unconvinced. Difference does not equal better. Poor mana is not, de facto, more interesting, than better mana, because, guess what, I like casting my spells. Go figure. And, most importantly, Wizards seems to agree.
If you want to make the case that bad mana Eternal is "better," well, go on and run events, get people to play and put out some data. Maybe then I'd take the notion seriously. But I don't by that this should be the clear and obvious teleological aim of Legacy. So, no, while I love Deathrite and Dig Through Time, they are not "more interesting" than Fetchlands as a matter of facts, only as a matter of some people's subjective valuation.
You know? I don't know, some shit like that. Maybe I am just old and tired? Or maybe not maybe?
"The Ancients teach us that if we can but last, we shall prevail."
—Kaysa, Elder Druid of the Juniper Order
No, there are all sorts of fields of human endeavor where, in complex decision making, algorithms do better than human intuition.
LMental: You read my mind when you said the algorithm could allow a slow trickle of unbanning to verify whether previously broken cards remained broken. The tricky thing is that Black Lotus should never get unbanned, but how does the algorithm know the difference between Black Lotus and Earthcraft when nobody plays either because both are banned?
I tend to agree with this, I think fetchlands enable more strategies than they disable. The arguments for banning fetches are good and it was an interesting perspective the first time it was brought up, just from my view. But, the problems in the format can be handled in more elegant ways than banning fetches, I think.
Yes, and I don't think regulating the Legacy banlist is one of those, especially when you consider the resources you need to spend in relation to what quality of outcome you expect. However, I'm merely an engineer and physicist and not an expert on these matters, only decently educated on some related basics (like most engineers and physicists would be). Edit: well actually I spent almost ten years designing IT-solutions so I guess I do have some amount of experience in designing algorithms but by no means am I an expert in algorithmic decision making, that was my point..
It would be nice if they broke it down by ban era on mtgtop8 but looking at things by years:
2016: Hymn/Blue Hymn (CB)/Pseudo-Hymn (Probe-Therapy): 30%
2017: 24%
2018: 24%
2019: 12%
Last 4 months: 7%
The only decks we‘re counting are the problem ones here Grixis Delver [pre 2019], Shardless/Czech/Grixis, and Counterbalance combo.
You have to play around that trash, because you can’t afford not to. It gets old quick b/c there‘s never a good enough proactive card to break the cycle. If there is it‘s either monopolized by CB (see SDT) or banned. While it may have sucked to endure some feelsbad losses to Wrenn/Oko, you didn‘t have to ruin your deck to play against Hymn bs. Your hand actually mattered with Wrenn for the first time since Dig b/c you could actually get a card from your hand onto the stack when you chose a spot, rather that having to play like crap to avoid the worst excesses of discard. Also your spells actually cost real resources for opponents to interact with rather than getting derped out by CB‘s infinite value ceiling.
You know more than I do, then, about algorithmic methods, and I agree they’re a bridge too far for Legacy. However, the topic came up because some fraction of Legacy players appear dissatisfied with certain homogeneities in the format, and therefore as a way of exploring the Eternal card pool without those homogeneities and as a way of putting certain theories (e.g., banning fetches) to the test, I suggested an alternative Eternal format with no fetches. If such a format ever became popular, it would make sense to exclude cards on the Reserved List, since needing staples from that list is antithetical to format popularity, so I suggested that too. Finally, to prevent certain format expectations — such as “the best deck will always be blue and have Brainstorm” — from being baked into the format as central tenets, which has alienated a fraction of Legacy players, I suggested that an algorithmic approach to banning would be suitable. It could be a breath of fresh air if it were part of the premise of a new format even while being unwelcome in a format with as much tradition as Legacy.
Even with my limited knowledge of algorithmic methods I could propose a specific algorithm that would not take much in terms of resources, either computational or in terms of data collection, but that would ban the most egregiously ubiquitous cards and then leave an equilibrium. I gave a rough draft above. I provided a criterion for cards being highly ubiquitous. The other piece would be a criterion for a format being in balance. By weighing the two criteria, the algorithm could determine whether the most ubiquitous card ought to be banned.
I do not assert that the fraction of Legacy players mentioned above, unhappy with Brainstorm and/or with fetches, is a majority. But it clearly exists, and the format hypothesized could be more suited to their subjective preferences while also being more accessible to newcomers (because no Reserved List cards) and able to support more of a player base than Legacy in the long term.
@BirdsOfParadise: I see, well that does make sense, and I would like to see, out of curiosity, what a format without Brainstorm would look like (also not minding it as much any more, for some reason).
About the algorithm I'm basically just waiting for someone to tell me how wrong I was. Yes, designing an algorithm is easy, but having one that considers the business aspects of WotC and the player base's relationship to changes, just seems very difficult. A simple model based on just play data, like card representation on MTGO, would be very much achieveable.
Edit: I read the discussion again and I think we were discussing different things. I replied to the statement that an objective banning algorithm would be prefereble to subjective, man-made decisions. Which I disagreed with. Then I commented on how an algorithm cannot easily replace the subjective decision making. I wasn't commenting on how a simpler algorithm may be implemented, although this would seem natural and it was easy to interpret that way.
Last edited by pettdan; 11-23-2019 at 06:15 AM.
Maybe certain format expectations are precisely the reason why people play Legacy instead of something else like Modern and Pioneer.
In fact you yourself have to admit:
So with you faboulous algorithm you could reach the wonderful result of alienating the player base (who is the other majority who plays it because, guess what, they like it as it is) from the format.
You know the player base is not the same as those who write here, don't you?
You don't seem to understand that every algorithm would always be as subjective as the human minds who decided what its purpose would be.
It's been fun to come back here, but I think the best moment recently was when someone finally shifted from "ban all fetchlands" to "ban all BLUE fetchlands". At least it was more honest.
It is sometimes presented here that "most players" don't agree with some argument and prefer the status quo. Usually an argument attempting to point out a problem with the format and solve it. I have noticed in a couple of surveys over the past six months that this was not the case, a majority of people voting didn't support what was referred to as the opinion of "most players".
So, my point is, be very careful suggesting or implying what most people think. Don't belittle arguments by claiming you know what most players think.
The quote above wasn't that dramatic, but it builds upon that understanding that is questionable.
I can only speak for myself and 7 other legacy players I talked to last night.
Told them, that the old discussion of banning fetches and brainstorm came up again after the W6 ban.
All I got, was raised eyebrows, laughter and questions, if people don't understand the format, or why they even play it, if they don't like it.
I believe, that less then 1% of the legacy community wants them gone.
The second, fetches and brainstorm leave the format, there will be at least 8 legacy players less. This I know for sure.
If someone does not like all the cantripping and shuffling, he's free to move to less interactive formats.
I hope from the bottom of my heart, that wizards never ever bans fetches or BS in Legacy.
@Grizzlenasty: well on this specific point I'm mostly on your side, but still, trying to use objective reasoning is better than shooting down arguments with "everyone thinks you're wrong"-type of arguments (that I noticed were often wrong, that was my point). I don't think anyone disagrees with that. I think for fetches [edit: or the discussion of banning them rather] , we had some decent criticism of it on this page. My own point was that I think they enable more than they disable in terms of strategies..
Last edited by pettdan; 11-23-2019 at 03:31 PM.
Well the main takeaway is that legacy players can basically be separated into people that like the format because of the large card pool, who would enjoy blue not having a monopoly on card selection (and stack interaction to a lesser extent) and opening up other colors, and those that like legacy because it offers a level of card selection not available in any other format, who therefore only care about playing blue because it offers it. There is basically no common ground between these two positions hence there is no real solution. Honestly it is way too late to nerf blue out of the format because the format has been warped so long that r&ds solution was to give many non blue strategies powerful dumb stuff to counter not having relative card selection.
Is it just I who feel like this has changed? Warning, this post is based on feelings.. I don't have numbers to back it yet, but it used to feel like blue was the be all end all. Back when Miracles and later Grixis Delver, at some point Grixis Control, were the best decks, it felt like you had to play a blue deck to stand a chance. I felt like it cleared up a bit last spring.. Several people were commenting on, in different forums, that Legacy was the best it had ever been.. Then we had some powerful releases, quick meta game changes then W6 slowly took over and stabilized as the unthreatened king of the format. With it gone, I imagine a wider set of strategies are again viable. Hogaak powered up graveyard shenanigans. Depths was powerful for a long time, elvish reclaimer helped a lot there [in keeping it a competitive non-blue strategy without ruining the format, I think]. Moon strategies are going to come back. Karn enables several types of sol land based strategies (Painter, Bomberman). Death and Taxes looks good, my guessing. Maverick seems good, I just watched this week's Legacy Premier League (amazing quality, watch it) and it looked great vs Grixis Delver. I hear Nic Fit (with Oko) is doing great, don't expect it to be tier 1 but it's nice to hear.. There is Urza and Emry allowing artifact strategies too, independently of Karn (and probably Brainstorm). . Oko is a bit worrisome but my guess is it'll be fine, we'll have to wait and see. It didn't crush the format yet though, did it? Reanimator and Storm are probably decent to great options, Sneak and Show is back.. Echo of Eons + Narset is basically a new archetype.. And there are Mystic Forge artifact storm-like decks..
It's too early to say but it seems like all these strategies are viable. It didn't feel like brainstorm was necessary this last spring and now both Veil of Summer and Once upon a Time have been printed and they help green decks interact with both counterspells, discard and removal (all of grixis control basically) and adds consistency.. Oh yeah, there's also a new mulligan rule that helps decks like DnT and Maverick get off to a good start at least, closing the distance in consistency to the Brainstorm decks..
Ah yeah, we also got Ouphe to keep artifact strategies in check, force of vigor to let graveyard decks interact with hate, force of negation to increase interaction with fast combo and graveyard shenanigans, I'm probably missing some but all of these cards seem to lead to more interaction without punishing any decks out of existence.. Narset does interact with csntrips and csn be played in Tezz, Karn decks as a blue Chains of Mephistopheles substitute..
There are currently 438 users browsing this thread. (0 members and 438 guests)