B&R Announcement - May 18, 2020

@chubbyrain1 said in B&R Announcement - May 18, 2020:

How much is an extra card worth in MTG?

The concept is messy because Wizards knows the cost is high and that's why they keep trying to mitigate the effect with the mulligan rules, but going off the data collected from 17lands.com Resources on MTGA, it's about 12% after the London mulligan. It was previously 16% with the Vancouver mulligan. For context, the best players tend to be in the 60% for win rates, the best decks tend to be in the mid 50% range. So without mitigating the card disadvantage, you are basically turning an average player on an average deck into one of the best players on one of the best decks.

Now please don't waste people's time pointing out irrelevant details like the difference between limited and Vintage. That's not the point. The point is an extra card has a huge effect on a deck's win percentage and the Companions were trying to mitigate that advantage solely by their deck building restrictions. Or they could have made the cards week, like a basically vanilla 5/5 or 4/5 (and those still see competitive play). Lurrus's deckbulding restriction is not very significant in the context of Vintage and the card is pushed relative to other three drops if you've ever tried to play it maindeck in any format. Lurrus compounds the advantages of the extra card by being a strong card in it's own right. The 11% advantage that we calculated from our data set is likely an accurate representation of the cards power despite the fact that the format is completely warping around itself with up to 75% of decks running it.

I don't know how to spell it out any clearer.

This analysis is incomplete and I think somewhat incompatible. It measures win rates against an opponent with a 7 card hand size, but does not account for the diminishing returns as you go down the scale. For instance the 12% you refer to is in reference to a 6 card hand vs a 7 card hand, but that same statistic would likely not hold up when measuring a 4 card hand vs a 5 card hand. the ratios are undoubtable a curve that suffers from some sort of diminishing returns as you go down the scale and your hand starts to become less important than what you top deck, and I imagine as players start getting into the very low numbers the difference starts to flatten.

Likewise, we have no precedent for an 8 card hand size until just now, sorta. Companion has its own wrinkles that do not make it really an 8 card hand size, but for the sake of simplicity let's roll with it. I would imagine that the win ratios here are theoretically a bell curve of sorts, where the higher up the chain you go the more if eventually flattens. Take it to a more extreme level like what does a 19 card hand look like vs an 20 card hand, and I imagine it is not nearly as sharp a decline as a 6 vs 7.

@vaughnbros said in B&R Announcement - May 18, 2020:

Potentially throw in some combo cards, like Necropotence and Flash, and we may be able to get back to a 3-deck format.

The format is already a multideck format, Lurrus is not a deck type any more than decks running force of will are a deck type.

@protoaddict said in B&R Announcement - May 18, 2020:

The format is already a multideck format, Lurrus is not a deck type any more than decks running force of will are a deck type.

With that, I agree.

......but Force of Will is -1 card and Lurrus (not specifically) is +1 card.........

@protoaddict

So your argument is that we should compare 7-card vs 8-card hands to 6-card vs 7-card hands and not to 4-card vs 5-card hands or 19-card vs 20-card hands? Or put more succinctly: the difference in win rates between between sets of opening hands of +1 card will be more similar the closer in hand size those sets are?

Thank you for validating my approach, even though you stated you were doing the opposite.

@chubbyrain1 said in B&R Announcement - May 18, 2020:

@protoaddict
So your argument is that we should compare 7-card vs 8-card hands to 6-card vs 7-card hands and not to 4-card vs 5-card hands or 19-card vs 20-card hands? Or put more succinctly: the difference in win rates between between sets of opening hands of +1 card will be more similar the closer in hand size those sets are?
Thank you for validating my approach, even though you stated you were doing the opposite.

You are either not understanding what I said, or willingly trying to turn it into something it is not.

The mulligan analysis at the time worked to analyze the 2 sets of mulligans, and thus started at the same starting point every time of 7 cards. It did not even go as deep as to say what happens when one player mulligans once versus one mulligans once and one twice. If you really wanted to determine the value of what an extra card is worth you would have charted out all the permutations of +1 hand and +2 hand size size ( 1-2, 1-3, 2-3, 2-4, etc) and constructed a weighted average both on the play and on the draw.

The new paradigm of having a companion means that this analysis falls short because there is no standard for an 8 card hand to measure by, and each permutation has a different W-L ratio. This is just at a very high level because in this example we also have a fixed card, which is something else that this analysis does not account for because it never happened before.

It's all about diminishing returns and utility. Ancestral recall is typically better when you have 1 card in hand than when you have 7, because you can only apply so many cards in a typical turn in most games of magic. The value Lurrus gives you is almost undoubtedly greater than 0 in matches against non Lurrus decks, but we don't have any real frame to know if its 12%, 5%, 1%, or 20% without a more comprehensive analysis.

There is also a bit of a gamblers fallacy at play here. Just because you won with Lurrus in your deck does not mean you won because of Lurrus, and in games where you lost but did not cast Lurrus it does not mean he didn't cost you the game because of the changes you had to make to your list. If you want to attach that metric to this analysis you need deeper and more specific data points.

@vaughnbros said in B&R Announcement - May 18, 2020:

@chubbyrain1

I don't think you have to make it clearer because I don't think anyone is disagreeing that Lurrus is absolutely broken right now, but my argument is that is because he has so little competition in the companion slot.

And yet, @Protoaddict exists...

What you seem to be suggesting is that Wizards had to be aware of the implications of the mechanic in terms of improved win %'s solely on getting a free card. Therefore, they had to be aware that its existence would eventually lead to every deck needing to play a companion.

They weren't. They thought the deckbuilding restrictions were enough to balance the companions. They screwed up. The plan was not to make Companions an integral part of Magic but kind of fringe strategies similar, yet obviously better than Battle of Wits, that could create different ways of playing the game, not completely redefine competitive Magic over one set.

@protoaddict You also missed the part where I said this is "messy", don't "waste people's time pointing out irrelevant details", and the point is that the effect is considerable (not that the numbers are equivalent) and Wizards did not counteract the benefits of the extra card by having an appropriately detrimental deckbuilding restriction or a significantly weak companion.

But to talk about why you are specifically wrong here, you are literally arguing that the relationship is a bell-shaped curve and that a point right next to another point in the curve is incomparable. You are also arguing that a "fixed card" is going to be worse than a random card (and the mulligan analyses will underestimate the effect of +1 card because the missing card is not random due to how mulligans now work). There was no argument that people won "because of the extra card". No one is trying to make that argument because it's a dumb argument to make. Did Alex Bertoncini win because he drew extra cards in many of his games? He assuredly won some of them. Does smoking cause lung cancer or did a patient's smoking cause their lung cancer? Causation doesn't really work the way you are representing. And then you argued that because something is new, we can't form valid comparison, which is simply untrue in every discipline. This is mind-numbing.

None of what you mentioned undercuts the core concept of the comparison I made, much if fallacy, much is irrelevant and some even supportive.

@chubbyrain1 said in B&R Announcement - May 18, 2020:

They weren't. They thought the deckbuilding restrictions were enough to balance the companions. They screwed up. The plan was not to make Companions an integral part of Magic but kind of fringe strategies similar, yet obviously better than Battle of Wits, that could create different ways of playing the game, not completely redefine competitive Magic over one set.

I have a very hard time believing that they didn't know that a set of 10 legendary creatures that they compared to "Commanders" wasn't going to have a large impact on the game.

@chubbyrain1 There's one thing missing from this analysis, which I think is very important. The "hand size" calculations consider 7 random cards vs 8 random. This is not accurate. It's 7 random cards vs 7 random cards + 1 specific card. That's very different. Consider pre-companion calculations of 6 random cards vs 7 random cards. That gives you a certain WR%. Now consider 6 random cards vs 6 random cards + 1 specific card where the one specific card is always ancestral recall (or something at least as strong as hyper-value-engine Lurrus). I bet the +1 deck has a considerably higher win % than just random 7 vs random 6.

@thewhitedragon69

I mean, I really just tried to make the argument that:

  1. The benefit of having an extra card in hand is considerable.
  2. The drawback of having a Companion such as Lurrus as an extra card (in the case of Lurrus) is often not very significant.
  3. They screwed up, similar to however many other cards where they underestimated the benefits or underestimated the costs. (Most famously with Skullclamps +1/-1 and most recently with Oko's +1).

You can make the point that a non-random card is often going to be better than a random card - certainly a good card you can build around like Lurrus or Yorion (the bane of other formats), but even mostly vanilla creatures like Jegantha, Kaheera, and Umori are seeing play - however, it doesn't add much to the argument I'm making and I really didn't want to go down the rabbit hole that is analyzing every difference in a minor comparison that is secondary to a larger picture. It's like every spoiler season where someone makes an inane comparison to card X and then the entire thread is about card X. Like most of Shennanigans thread was people comparing the card to Ancient Grudge and being comically wrong about Force of Vigor.

As far as actually using these cards to create diversity, if they were less powerful or more restrictive, they could have done so. Otherwise, you are arguing that the Artifact Lands created diversity because they were Artifacts or Skullclamp created diversity because you have to play 1/1 creatures to maximize the 1/1 toughness. The more powerful a strategy is to other strategies, the less diverse and the companions literally constrain deckbuilding as part of their play. "But Lurrus is not a deck" Neither was Skullclamp... It was played in Elf in Nail, Genesis Chamber Affinity and others. Printing more companions is not viable because it's going to be impossible to balance the deckbuilding restrictions across formats. You'd have to employ the tweaks and nerfs that other games use. Good luck getting that to happen...

last edited by chubbyrain1

@thewhitedragon69 said in B&R Announcement - May 18, 2020:

The "hand size" calculations consider 7 random cards vs 8 random. This is not accurate. It's 7 random cards vs 7 random cards + 1 specific card. That's very different.

The analysis he posted does not compare 8 vs 7. The data compared 7 vs 6 and 7 vs 5, on the play and on the draw, before and after the mulligan change. To my point, regardless of if the 8th card is a fixed element or a random one, the comparison is a rough estimation at best as every step in the mulligan permutation chart has different win/loss ratios. It is not a small difference.

@chubbyrain1 said in B&R Announcement - May 18, 2020:

As far as actually using these cards to create diversity, if they were less powerful or more restrictive, they could have done so. Otherwise, you are arguing that the Artifact Lands created diversity because they were Artifacts or Skullclamp created diversity because you have to play 1/1 creatures to maximize the 1/1 toughness. The more powerful a strategy is to other strategies, the less diverse and the companions literally constrain deckbuilding as part of their play. "But Lurrus is not a deck" Neither was Skullclamp... It was played in Elf in Nail, Genesis Chamber Affinity and others. Printing more companions is not viable because it's going to be impossible to balance the deckbuilding restrictions across formats. You'd have to employ the tweaks and nerfs that other games use. Good luck getting that to happen...

How are you defining diversity? Is just the presence of Lurrus enough for you to say that this PO deck and this Breach deck are too similar? The downturn in decks that cannot use Lurrus such as Workshops and Dredge had started before this printing, and frankly can be accredited to other things. This may be exacerbating those issues but may not be the thing you need to remove to fix them.

Even in the events where decks containing Lurrus went 8 for 8, you still had a meta that had PO, 4C, Breach, Doomsday, "Lavina", Oath, Xerox and more. Everyone of those decks also had a section of cards that each and everyone of them played like lotus, ancestral, brainstorm, etc, but no one says those decks are the same because they have uniform power cards, so why are we not applying that standard to Lurrus?

75% of decks playing blue heavy decks to accommodate Force of Will is not diversity, lol. The only thing less diverse than a typical Vintage is a Magic tournament in terms of gender.

That said, the game outside of Vintage/Legacy is incredibly diverse. Standard and Modern force diversity through two different concepts, rotation in standard's case and heavy policing of the most powerful strategies in Modern. Vintage and Legacy due to the "grandfather" clause that apparently we are abiding by willingly now, basically all but guarantees that diversity will never happen. There will always be a best Force of Will deck, a best Bazaar deck, and a best Workshop deck. There used to be a best Ritual deck too, but apparently they didn't make the cut point for the "grandfather" clause and died of a heart attack in their mid 40s before their kids could procreate.

@Protoaddict have you played any games with Lurrus?
Lotus, recall, brainstorm ,etc are all one-ofs randomized in a 60 card deck. Lurrus is available to you every game, whenever you want it, in a place no one can touch it. Companions are fundamentally different than all other cards. That's why people are looking at them differently.

@illig719 said in B&R Announcement - May 18, 2020:

@Protoaddict have you played any games with Lurrus?
Lotus, recall, brainstorm ,etc are all one-ofs randomized in a 60 card deck. Lurrus is available to you every game, whenever you want it, in a place no one can touch it. Companions are fundamentally different than all other cards. That's why people are looking at them differently.

And cards in your sideboard are fundamentally different too. And card that tutor for cards in your sideboard are different too. The game has a lot of fundamentally different cards that you can look at with different lenses, but the commonality is you have to choose to include it in your deck and that it has interactions with other cards.

I mean, the moxes are a one of each right, but are they really? Depending on the composition of your list it's in a lot of ways a 5 of. What about fetchs and dual lands? Shouldn't there be a penalty to playing a 4 color deck with like 15 lands? The one life you pay for a fetch that also provides protection against strip mine effects and thins your list and provides a shuffle for brainstorm does not seem like much of a penalty to me, and they are largely restriction proof as well since there are so many ways to mix and match your lands in 15 cards that many decks don't even run the full compliment. Same issue as the moxen, if you restrict every fetch land you can still run 4 that fetch island. If you restricted every fetch and every dual you MIGHT see some decks play a less greedy mana base which would subsequently help reduce how much lurrus you see to start with and make for more substantial descisions in play and deck building.

I don't see that just having access to a card is inherently a problem, the issue is when the card is too good. That being said one of the restrictions on Lurrus is that you can only have 1 companion, and right now there are no other valuable companions. Forcing players to make that choice would go a long way towards us not seeing him be so prominent in the top 8s.

The only fundamentally different thing about companions are that they are restriction-proof. The format generally manages to reach an equilibrium because the really broken new cards that they print can be restricted if they are warp the format. If restricting lurrus meant you only got to play him 1 out of 4 games then that would be a uncontroversial action at this point. Any card which had eliminated dredge and shops would be getting restricted in short order.

The reason any of this is controversial is that only largely unprecedented action works against a card that is warping the format as a 1-of.

last edited by mana_duane

For the past 27 years magic has been a game played by drawing 7 random cards. Gaining access to an 8th known card that cannot be interacted with is fundamentally different.

My view on all this is very practical :

  • Vintage has always been my favorite format and i try to play at least one (usually very small and when i can bigger) paper tournament every month.
  • For the last two years power creep in the format has been quite obvious and even more recently. IMHO we are now at a point where a game is mostly decided by our opening hand so the only really important décision is keep or mulligan. i quited several competitive tournaments just because playing was just boring because of that. Not every games are like that but much too often. This was bound to happen sooner or later and that means really strong décisions (whatever) should be taken if Wizard wants a future for Vintage. I put it short so it might sound quite dark view but IMHO if we are not there we are very close to it : power level is much too high now.
  • I don't mind big changes in Vintage if it will allow it to stay a format that can be played.
  • As for compagnion, i don"t mind if in a few years there is one or two decks that uses such strategy (in the same way there is a dredge deck in the format) but if the goal is that every one is playing one compagnion then i will just stop playing Vintage and will find another format. If i wanted to play with a commander i would play Commander format, plain and simple.
  • If every one agree that Lurrus is broken then i see no reason to wait. Ban it as compagnion and do not let the few vintage players left in the world struggle with an biased meta.
    This is just my feeling about all this and still hope it will go for the best.
last edited by albarkhane

@mana_duane said in B&R Announcement - May 18, 2020:

The only fundamentally different thing about companions are that they are restriction-proof.

No, SOME of them are restriction proof. Lurrus and Lutri are the 2 companions that have the distinction of restricting how many other copies you can play in your main deck, and just so happen to be the 2 that are most obviously vintage viable. All the others can be played in multiple and in the case of Gyruda it is actually pretty fundamental to the cards function. There is nothing about the companion rule that makes them immune to restriction.

@illig719 said in B&R Announcement - May 18, 2020:

For the past 27 years magic has been a game played by drawing 7 random cards. Gaining access to an 8th known card that cannot be interacted with is fundamentally different.

A) Drawing 7 random cards has almost never been the way the game is played. There have been I think 4 different mulligan rules since the dawn of the game and the printing of serum powder, plus the introduction of the play or draw rule. Best argument you can make is that a 7 card hand has been the starting point in magic, but it certainly has changed multiple times.

B) All the companions can be interacted with. They can be countered. They can be named off meddling mage. You can play a card like Drannith Magistrate. Void winnower affects a few of them. These may not be the interactions you want, but they are interactions all the same.

@albarkhane said in B&R Announcement - May 18, 2020:

If every one agree that Lurrus is broken then i see no reason to wait. Ban it as compagnion and do not let the few vintage players left in the world struggle with an biased meta.

Not everyone agrees. I have some confidence it will be banned regardless of what everyone or anyone thinks because it is the easiest way to deal with the perceived problem some people have, but just looking at this thread you can see some dissenting voices other than my own.

All I know is that if Lurrus gets the ban hammer, it will create a precedent with (possibly) many cards to follow. I personally care not either way, I'll still play when it fancies me. I've had many cards I've loved restricted and many I've hated get un-restricted. Just part of the game. If it's gotten to the point we need to ban, it just becomes part of the game.

last edited by Serracollector

@serracollector I agree with this. I don't think it'll result in many, if any future bannings. It will be precedent in a way, but precedent is not needed for WotC to fix any future gaffs. It seems to me that since we've gone 25 years with bannings only to themes, vanguards, dexterity cards, ante cards, and lonely sharazad, that WotC is trying to avoid banning whenever possible. If they do end up with a ban here, I think we'd likely go many more years before we see something akin to it down the road. It's safe to say the testing they do is to try to avoid bannings, and if it's something that needs a ban in Vintage, it's likely reeeeally busted. In this case, the mechanic itself was disfunctional by being +1 card at almost no cost. I doubt that kind of mistake would happen in the near future.

I find that precedent is an odd form of measurement, and in all metrics. The way that you feel today is not how you will feel tomorrow. If that is the case, I would challenge to you (in a non-threatening way) that if you feel the same tomorrow it is likely that you haven't learned enough today.

  • 206
    Posts
  • 12872
    Views