Choking Tethers (reprint) Elvish Fury (reprint - buyback is back!) Goblin War Party (2R, Sorcery, Choose -Create three 1/1 goblins OR -Creatures get +1/+1 and haste, Entwine 2R) Headless Specter (1BB, 2/2 flyer, Hypnotic Specter once Hellbent) Imposter of the Sixth Pride (1W 3/1 changeling) Lava Dart (sweet reprint - flashback is back!) Martyr's Soul (2W, 3/2 spirit, convoke, when enters if you control no tapped lands put two +1/+1 counters) PROHIBIT (REPRINT THANK GOD) Savage Swipe (BEAR TRIBAL - G, sorcery, target creature gets +2/+2 if power 2. Creature fights target creature) Stream of Thought (U, sorcery, target player mills 4, you shuffle up to 4 cards, replicate 2UU) Venemous Changeling (2B 1/3 deathtouch changeling) Wall of a Thousand Cuts (3WW 3/5 defender, flying, W: attack as if no defender)
Lava Dart and Prohibit are definitely Modern playable. I do think that Prohibit likely precludes Counterspell, unless there are some Limited considerations to putting Counterspell at another rarity. That's probably a bit disappointing but I'm pumped for the cards. Excited for more cards today!
Choking Tethers (reprint) Elvish Fury (reprint - buyback is back!) Goblin War Party (2R, Sorcery, Choose -Create three 1/1 goblins OR -Creatures get +1/+1 and haste, Entwine 2R) Headless Specter (1BB, 2/2 flyer, Hypnotic Specter once Hellbent) Imposter of the Sixth Pride (1W 3/1 changeling) Lava Dart (sweet reprint - flashback is back!) Martyr's Soul (2W, 3/2 spirit, convoke, when enters if you control no tapped lands put two +1/+1 counters) PROHIBIT (REPRINT THANK GOD) Savage Swipe (BEAR TRIBAL - G, sorcery, target creature gets +2/+2 if power 2. Creature fights target creature) Stream of Thought (U, sorcery, target player mills 4, you shuffle up to 4 cards, replicate 2UU) Venemous Changeling (2B 1/3 deathtouch changeling) Wall of a Thousand Cuts (3WW 3/5 defender, flying, W: attack as if no defender)
Lava Dart and Prohibit are definitely Modern playable. I do think that Prohibit likely precludes Counterspell, unless there are some Limited considerations to putting Counterspell at another rarity. That's probably a bit disappointing but I'm pumped for the cards. Excited for more cards today!
Flusterstorm is the new BAB promo (with new art), and a new/fixed version of Astral Slide:
Astral Drift 2W Enchantment Whenever you cycle Astral Drift or cycle another card while Astral Drift is on the battlefield, you may exile target creature. Return that creature to the battlefield under its owner's control at the beginning of the next end step. Cycling 2W
Not impressed honestly. I mean storm not really an issue. We need broad counters not specific counters.
Flusterstorm, as the BAB status suggests, is probably not just targeted at Modern players. It's a chase card that will have mass eternal format appeal and is likely included to drive down the price and incentivize other format specialists to buy the product. It's also the best possible reprint outlet. Good Modern safety valve, but probably not super Modern relevant right now. I'm still optimistic for other countermagic options.
Drift is sweet. We'll need more cycling support to make Drift decks viable, because Drift is dead in a lot of high-profile matches, but it's a great inclusion with strong brewing potential. Net positive for Modern overall.
I hope that almost nobody will use that combo...it is just atrocious. Question also is, if enough people could get their hands on lattice before it exploded in price.
Not going to directly weigh in on the Karn question because there simply isn't enough data out there, but I will say that Karn highlights the advantage of digital cards and errata. Design and balance mistakes happen all the time in all games, especially Magic, and especially once you move out of Limited/Standard to older formats. In a world where you are stuck with paper cards and the text printed on them, you can't retroactively fix these cards. This applies to both broken cards, e.g. everything that has ever been banned in any format, and very bad cards, e.g. cards that totally miss the mark and don't see play.
Digital Magic and errata/patches give Wizards an alternate solution to this problem. It allows them to make small/large adjustments to cards in order to make them better or worse, depending on the context. Take Karn, for example. Any of the following adjustments would avoid the Lattice combo, if one's goal is to remove that synergy:
1. Change to "Activated abilities of nonland artifacts your opponents control can’t be activated."
2. Change to "-2: You may choose an artifact card you own with converted mana cost 4 or less from outside the game or in exile, reveal that card, and put it into your hand."
3. Change to "-X: You may choose an artifact card you own with converted mana cost X or less from outside the game or in exile, reveal that card, and put it into your hand."
4. Change Lattice to "All nonland permanents are artifacts in addition to their other types. All nonland cards that aren’t in play, spells, and nonland permanents are colorless. Players may spend mana as though it were mana of any color."
This errata/patching approach would improve balance for so many banned cards, underplayed decks, and broken metagames. It's also the model used by all major online games. I believe Wizards will eventually go in this direction, and/or are actively trying to figure out how to do this without creating a huge gap between online and paper Magic. Banning Nexus in BO1 and nowhere else is just the start of this journey, as was the functional Pridemate change.
Of course, this approach is not without risk. Wizards can become too balance-happy and issue errata for things that simply don't need it based on mass panics or outcry. Imagine if Wizards nerfed Teferi to only untap one land, or Wilderness Reclamation to untap only a few lands, because of MTG Arena complaints. Or increased CoCo's mana cost by 1 at the height of its power. Tweaks like this, and countless others we could speculate on, might assuage a short-term hysteria but would limit long-term feasibility of those cards/decks, especially if Wizards is balancing across formats. After all, if Wizards messed up the design to begin with, they can certainly mess it up again down the road. Additionally, part of the beauty of non-rotating formats is in finding design mistakes and exploiting them for powerful decks, and too much errata would make that challenging. Patching/errata is not a perfect solution by any means, but I do think if done correctly it would really open up Magic play and design space in ways that would benefit players and formats like Modern.
I mean that makes sense with Graduation and Finals on the Horizon. Is it different from the previous year or something?
Yes. 2018 was not nearly as bad/slow as it is right now. Its the format. I mean even looking around at most of the forum, traffic certainly appears way down.
I would attribute the slowdown to a wide combination of factors including the rise of Discord, increased Reddit traffic, heightened Standard/Arena popularity, increased Arena streaming, entrance into a slower Modern season after the GP stretch, uncertainty about what to play due to MH, the decline of traditional forums like ours (both generally and specifically), and a host of other factors. We have seen posts every few months for literal years that Modern is losing players and/or popularity in an often isolated or small N context. I remember it in basically every metagame from Eldrazi Winter to present. The format is now, and continues to be, largely fine.
I disagree with them on Counterspell, but I like them exploring cards that are actually worth discussing (compared to that CFB article a while ago discussing why Sneak Attack would be bad for Modern - no *****). FoW is definitely the most controversial inclusion in the Horizons discussion and although I don't see it happen I also don't think FoW would be so bad for Modern.
Sigh... For real? I just have no faith in their analysis if they're concerned about getting Lava Spike countered for UU compared to allowing control to tap out and still be "safe".
On a totally separate note.
What is your take on FoW?
Personally, I think it's sort of silly to need a T0 counter in a T4 format. I also think it's sort of silly to need a free counter before we even attempt a 2cmc one. I can get the difference between using FoW vs fair and unfair decks; I've borrowed probably just under enough legacy decks to really grasp the concept fully, but I think I can grapple with it a bit.
I have no issue with FoW-like cards, but would prefer alternate FoW designs to preserve some Modern and Legacy differences. Mix and match examples of FoW alternatives include:
-FoW for non-creature spells only. This gives Modern's least broken permanent type a leg up over the truly broken stuff.
-FoW where the alternate cost can only be paid in the first 1-3 turns, like a reverse Serra Avenger. This lets us respond to the broken early plays while not creating truly dominant late game states for Ux strategies.
-FoW Mana Leak instead of FoW hard counter. In many ways, this acts like the Avenger design with different functionality in the late game.
FoW itself is probably fine, but adjustments like these might make it more Modern appropriate. I know Ux control has a stigma as being pretty bad or overhyped, but if these existing, viable decks had consistent access to a powerful FoW tempo play, they could become really dominant. Adjustments like I proposed would assist Ux strategies and police broken stuff, while not also pushing too many other decks out. I would also hope other fair strategies get comparable tools so blue doesn't become a virtually automatic best choice for those preferring interactive decks.
So far, the ban mania about Neoform has not panned out from a results perspective. We are two Challenges in the books since WAR and only a single T32 placement to show for it, which came two weeks ago at around 30th place. We also have a 2nd place IQ finish at a Newington SCG event (http://www.starcitygames.com/decks/129156), but no other T8s at the 5 other SCG events that weekend including the larger Classic. I also found 5 other non-SCG paper events reported last weekend, also with no Neobrand showings.
Yes, it's early in the WAR release and prospective pilots might need to obtain the cards needed for Neobrand. Yes, the deck isn't fully refined and needs some tuning before reaching an optimal version. Yes, the metagame is still evolving and we won't have a good picture of it until after the Mythic results, WAR, and MH have settled. Even acknowledging those factors, however, the deck has seen a very mediocre and unimpressive competitive debut. I am much more excited by the UWx staples and their consistent performance.
I've been running some goldfishes with a recent 5-0 League list (https://www.mtggoldfish.com/deck/1871192#paper) to try and establish a goldfish baseline for T1, T2, or T3 wins. I'm using the London Mulligan rule. A Reddit poster also did their own N=100 analysis with the Vancouver Mulligan rule and a different list (they didn't specify the list) to calculate T1 and T2 wins. It's not perfect data, but both goldfish sets help us estimate a baseline. The Reddit user found a T1 win in 11% of games and a T2 win in 48% of games. I'm currently at N=50 goldfishes, all on the play, and have found a T1 win of 8% and a T2 win of 16%. I've added T3 wins at 36% for a total pre-T4 win rate of 60% in the current N=50 sample. Note this only represents goldfish rates, so the true pre-T4 win rate will almost necessarily be lower once interaction is added. But it is helpful to see a baseline.
I'll add in N=50 games on the draw to get the sample to N=100 and then do some confidence interval calculations, increasing N if needed depending on how wide the intervals look.
We'll see our first weekend of paper War results in a few days. This will include a variety of regional events and an SCG Classic. We'll also get another Modern Challenge on MTGO. These are the sort of results that matter more for assessing a deck's strength than the data we currently have. We won't have a Modern GP until late June and won't have a Modern SCG Open until late May. On the one hand, this means we won't see the most important, large-scale event results for a while. On the other hand, it will give the metagame time to stabilize and the lists time to refine themselves. By the time we do have results, it will be a lot clearer if the deck is just another glass cannon strategy or if it's a new T4 rule violator. Remember that a deck must be both top-tier AND win too consistently before T4 to violate the rule.
So far, our only relevant result is a 30th place finish at an MTGO Challenge (https://magic.wizards.com/en/articles/archive/mtgo-standings/modern-challenge-2019-04-30), which isn't the most convincing datapoint. I encourage everyone to be very skeptical of Twitch streamers, article authors, pros, and Twitter/Reddit personalities who speak publicly on this issue. These people aren't necessarily wrong, dishonest, or misleading. But their position gives them extreme incentive to oversell, misrepresent, or outright lie about big issues if it results in clicks, views, popularity, revenue, etc. If such figures are not citing their sources or making supported arguments, that should be a bright red flag.
what precedents do we have to work with? the only that come to mind are infect and amulet bloom where turn 4 or something akin to it were cited as reasons for bans.
even then you have to consider context. such as infect with probe being able to check if its okay to go for it, and bloom having overwhelming mid to lategame backup plans. if im reading things correctly what the neoform combo has going for it is pact of negation for protection and it doesnt use the GY to go off. tbh without more information ive no clue how those compare (in relation to improving the decks overall efficacy or whatever).
Back in 2015, I estimated that Amulet Bloom won approximately 23% of its games (estimated from a coverage and MTGO sample) before T4. This included concessions by an opponent. The confidence interval of that range, calculated in two independent samples, was 15%-30% in the smaller N sample, and 17%-29% in the larger sample. Seething Song Storm averaged about a 25% pre-T4 kill using similar methods. Based on this, I'm comfortable saying if a deck exceeds 20% for pre-T4 wins and is also top-tier, it's probably in trouble.
When I played Cheeri0s a lot on MTGO in 2017, I had about a 75% MWP overall with about 200 games total in the sample, winning on T2 or T3 in an outrageous 40% of games. Obviously, nothing from Cheeri0s ate a ban, so the "true" win turn in a larger sample was undoubtedly lower. The deck was also not strong enough to catch on at a top-tier level. This shows that single-player samples don't necessarily reflect the MTGO-wide sample. We need more data to draw meaningful conclusions about potential bans and T4 rule violators.
I did a coverage-based analysis for both Counters Company and Affinity in a sample of approximately 60 Affinity games and 40 Company games. In those analyses, I found that Affinity averaged a pre-T4 win rate of approximately 5.3% with a 95% CI somewhere between 0% and 14.2%. For Company, it was 3.5% with the CI at 0%-10.7%. Neither of these decks had bans, so I'm comfortable saying the deck needs to be winning at least 10%+ of its games before T4 in order to be on Wizards' radar.
Finally, I remember an analysis I did of over 100 Caleb Scherer MTGO Storm games back when Baral first came out and Storm looked a little scary. During that time, we calculated an approximate 12%-14% (don't remember the exact #) pre-T4 win-rate. This iteration of Storm also did not eat a ban, so I'm comfortable with 12%-14% being a safe range.
Overall, based on all this, I'd estimate that a deck is in danger if it is BOTH top-tier AND wins approximately 20%+ of games before T4.
Reporting that people are pre-turn-3 killing opponents 50% of the time on MTGO is not ban mania. I counter your "knee-jerk ban-mania" claim with "anti-ban knee-jerk" claim! And double down for using the tired glass canon meme. Yes, combo-decks that are disrupted mid-combo fail to win, we know that.
The key points with this new deck are that:
It draws the entire library.
It kills on the turn it goes off.
Due 1 and 2, it can run pact of negation for protection.
Due to 1, it can ensure it has pact in hand.
There is a reason Enter the Infinite is 12 mana. It basically is ad-nauseam, but with a faster clock, an earlier fundamental turn. Sure, people can start packing dispel or some 1-mana creature bounce as a reaction, but that would not be healthy. (And still fails if they happen to have pact of negation in hand.)
None of this is relevant to the bannability of the deck. All four of those qualities apply to Ad Nauseam, an eminently acceptable combo deck. The only question is whether or not the deck can consistently win before T4. Based on the previous Bloom ban example, I would estimate a deck would need to win before T4 in 20%+ of games in a large N, large T context in order to be bannable. Cherrypicked streamer results don't get there. My Cheeri0s win rate was about 75% with more than 25% of my wins happening before T4. Obviously, nothing from Cheeri0s got banned. My personal examples don't necessarily reflect the wider metagame realities. Let's increase the number of auditable events and the time span before we entertain ban discussion like Cohen's.
@ktk, I am sorry for this and I said that you are right, but I already watched multiple games today with turn 0,turn 1,turn 2 kills.
The first two pictures are from a game where the opponent turn 0 killed kanister(I mean turn 1 before the opp plays a land), then turn 1 killed kanister.
There is no way this deck does not have consistency in pre-turn 4 kills.
Meanwhile, lot's of cards look silly on the banlist.
Again, it doesn't matter if the deck is ultimately broken. That shouldn't change our method of ban analysis. We need auditable results that make up a large dataset over multiple larger events/fields. Just because a knee-jerk ban mania response is right every 20 or so cards doesn't make it a good response or approach to banlist analysis. Streamers and social media perosnalities have significant incentive to overyhype these kinds of claims. It drives views, clicks, ad revenue, future partnerships, speculation targets, etc. We cannot rely on this as an indicator of anything. Wait for neutral, large N results
Well, we do not have to speak about bans, that's what you did not get. Most of us(me included) prefer no bans, unbans instead! But, I did have the pictures, and this is a thread about the state of the meta. I am telling you that the state of the meta in MODO right now, is closing in on lot's of pre-turn 4 kills. You don't have to turn every discussion about the meta into possible ban mania comments(unless you are referring to Justin Cohen and other notorious twitter posters).
This transition from an issue discussion to a hyperbolic, personal one ("You don't have to turn every discussion...") is unfortunate and inaccurate. Even a cursory review of my post history shows that I have not mentioned ban mania as an issue in this thread in months. Talking about the metagame is fine. Talking about bans or unbans is also fine. Speculating on bans with extremely limited data over an extremely limited time period is, however, ban mania. There is no good reason for posters in this thread or Magic personalities elsewhere to buy into this uncritical method of ban analysis/prediction. It doesn't matter if this bad method is right once every 20+ cards. It's still a bad method. Once we have data from events like Challenges and MOCS's or other similarly sized venues, we can revisit with a more informed perspective.
@ktk, I am sorry for this and I said that you are right, but I already watched multiple games today with turn 0,turn 1,turn 2 kills.
The first two pictures are from a game where the opponent turn 0 killed kanister(I mean turn 1 before the opp plays a land), then turn 1 killed kanister.
There is no way this deck does not have consistency in pre-turn 4 kills.
Meanwhile, lot's of cards look silly on the banlist.
Again, it doesn't matter if the deck is ultimately broken. That shouldn't change our method of ban analysis. We need auditable results that make up a large dataset over multiple larger events/fields. Just because a knee-jerk ban mania response is right every 20 or so cards doesn't make it a good response or approach to banlist analysis. Streamers and social media perosnalities have significant incentive to overyhype these kinds of claims. It drives views, clicks, ad revenue, future partnerships, speculation targets, etc. We cannot rely on this as an indicator of anything. Wait for neutral, large N results
I am unsurprised that the current hot glass cannon deck leads to ban mania and speculation. There have been literally dozens of ban suggestions in the last 18 months, and at this point more ban hysteria is par for the course. Whether or not the deck ends up being banned is irrelevant. Ban discussion should only be based off auditable results in large fields and datasets. We can't change our ban analysis method just because a deck sparks panic. We also can't say a deck is bannable based off a few streamer games. Let's at least get some MOCS and Challenge events in the books before assessing the bannability of a deck.
Again, the ban mania reaction is inappropriate independently of whether or not a deck is ultimately banned. Bad analysis methods are bad whether or not they are right one in twenty (or more) times. If people speculate about bans on every top deck in Modern, which they did throughout all of 2017 and 2018, they are eventually going to be right. And they were this year: KCI did get banned. But that doesn't mean it's a good method when ever other ban candidate (Stirrings, Company, Temple, GDS, Vial, Cavern, Tron lands, Bridge, Moon, SSG, Opal, etc.) was safe. We need to condition ourselves to avoid this kneejerk reaction and wait for results before making statements like Justin Cohen's. Unfortunately, there is significant incentive to make those statements today, whether financial (people are speccing on Riders or some other piece,) social (people want the upvotes/shares), or just personal (people enjoy the praise and accolades of "Getting it right" or "Saying it first").
There are some oddities in either Karsten's data or his confidence interval calculation, and I'm not sure how he's finding the Ad Naus/Hardened Scales lower CI to be >50% on a 95% interval. But overall, it's great data. Humans definitely appears a strong contender again, and Tron appears very middling.
Here's the Modern-only data we wanted. Wizards noted that no player had a perfect 10-0 Modern record, so no one earned 30 points. The list below represents all decks with an 8-2 or better record. Wizards did not separate out 9-1s and 8-2s, and I don't want to do that parsing myself when the 8-2+ bracket is already a valuable grouping. Here's the Wizards' list: https://magic.wizards.com/en/events/coverage/2019MC2/24-27-modern-match-point-decklists-2019-04-28
And here's my breakdown of conversion/representation rates of top decks. It's ordered from highest Day 2-->24+ Point conversion and includes both the Day 2% share and the 24+point% share. You'd read it like so: Ad Nauseam made up 1.9% (6 players) of the Day 2 metagame and 7.69% (2 players) of the 24+ point metagame, representing a 33.33% conversion rate of Day 2 --> 24+ players and an over-representation of +5.79% percentage points.
Humans is the clear winner here, with the second highest 24+ point share (15.4%), the fourth highest conversion rate (12.9%) and the highest of decks with >10 Day 2 pilots, and the second highest over-representation (+5.7% points). Izzet Phoenix has comparable numbers: highest 24+ point share (19.2%), comparable Day 2 N (62 vs. 53), equal conversion rate (12.8%), and the highest over-representation of Day 2 to 24+ points (+7%).
All decks above Amulet Titan in the list had expected or better conversions into the 24+ points. Hardened Scales, Tron, GDS, BG Rock, Esper Control, Bogles, Infect, and the "Other" decks all fared much worse. Notably, 49 players played "Other" decks that weren't listed here representing 9.5% of Day 1 and 10% of Day 2, but ZERO of them (0%) made 24+ points. I'm comfortable viewing all these listed decks as the most viable 19 strategies in Modern, with everything else in a lower bracket of viability.
Also, Ad Nauseam looked really good in this event. Karsten identified it as having the highest MWP in the event and these metrics show it having the best conversion rate (33.3%) and one of the best representations (+5.8% relative to Day 2). Definitely a sleeper pick I would look into.
Choking Tethers (reprint)
Elvish Fury (reprint - buyback is back!)
Goblin War Party (2R, Sorcery, Choose -Create three 1/1 goblins OR -Creatures get +1/+1 and haste, Entwine 2R)
Headless Specter (1BB, 2/2 flyer, Hypnotic Specter once Hellbent)
Imposter of the Sixth Pride (1W 3/1 changeling)
Lava Dart (sweet reprint - flashback is back!)
Martyr's Soul (2W, 3/2 spirit, convoke, when enters if you control no tapped lands put two +1/+1 counters)
PROHIBIT (REPRINT THANK GOD)
Savage Swipe (BEAR TRIBAL - G, sorcery, target creature gets +2/+2 if power 2. Creature fights target creature)
Stream of Thought (U, sorcery, target player mills 4, you shuffle up to 4 cards, replicate 2UU)
Venemous Changeling (2B 1/3 deathtouch changeling)
Wall of a Thousand Cuts (3WW 3/5 defender, flying, W: attack as if no defender)
Lava Dart and Prohibit are definitely Modern playable. I do think that Prohibit likely precludes Counterspell, unless there are some Limited considerations to putting Counterspell at another rarity. That's probably a bit disappointing but I'm pumped for the cards. Excited for more cards today!
EDIT: Images posted by another Reddit user - https://imgur.com/a/7QWHclB
**SPOILERS**
Choking Tethers (reprint)
Elvish Fury (reprint - buyback is back!)
Goblin War Party (2R, Sorcery, Choose -Create three 1/1 goblins OR -Creatures get +1/+1 and haste, Entwine 2R)
Headless Specter (1BB, 2/2 flyer, Hypnotic Specter once Hellbent)
Imposter of the Sixth Pride (1W 3/1 changeling)
Lava Dart (sweet reprint - flashback is back!)
Martyr's Soul (2W, 3/2 spirit, convoke, when enters if you control no tapped lands put two +1/+1 counters)
PROHIBIT (REPRINT THANK GOD)
Savage Swipe (BEAR TRIBAL - G, sorcery, target creature gets +2/+2 if power 2. Creature fights target creature)
Stream of Thought (U, sorcery, target player mills 4, you shuffle up to 4 cards, replicate 2UU)
Venemous Changeling (2B 1/3 deathtouch changeling)
Wall of a Thousand Cuts (3WW 3/5 defender, flying, W: attack as if no defender)
Lava Dart and Prohibit are definitely Modern playable. I do think that Prohibit likely precludes Counterspell, unless there are some Limited considerations to putting Counterspell at another rarity. That's probably a bit disappointing but I'm pumped for the cards. Excited for more cards today!
EDIT: Images posted by another Reddit user - https://imgur.com/a/7QWHclB
Flusterstorm, as the BAB status suggests, is probably not just targeted at Modern players. It's a chase card that will have mass eternal format appeal and is likely included to drive down the price and incentivize other format specialists to buy the product. It's also the best possible reprint outlet. Good Modern safety valve, but probably not super Modern relevant right now. I'm still optimistic for other countermagic options.
Drift is sweet. We'll need more cycling support to make Drift decks viable, because Drift is dead in a lot of high-profile matches, but it's a great inclusion with strong brewing potential. Net positive for Modern overall.
Not going to directly weigh in on the Karn question because there simply isn't enough data out there, but I will say that Karn highlights the advantage of digital cards and errata. Design and balance mistakes happen all the time in all games, especially Magic, and especially once you move out of Limited/Standard to older formats. In a world where you are stuck with paper cards and the text printed on them, you can't retroactively fix these cards. This applies to both broken cards, e.g. everything that has ever been banned in any format, and very bad cards, e.g. cards that totally miss the mark and don't see play.
Digital Magic and errata/patches give Wizards an alternate solution to this problem. It allows them to make small/large adjustments to cards in order to make them better or worse, depending on the context. Take Karn, for example. Any of the following adjustments would avoid the Lattice combo, if one's goal is to remove that synergy:
1. Change to "Activated abilities of nonland artifacts your opponents control can’t be activated."
2. Change to "-2: You may choose an artifact card you own with converted mana cost 4 or less from outside the game or in exile, reveal that card, and put it into your hand."
3. Change to "-X: You may choose an artifact card you own with converted mana cost X or less from outside the game or in exile, reveal that card, and put it into your hand."
4. Change Lattice to "All nonland permanents are artifacts in addition to their other types. All nonland cards that aren’t in play, spells, and nonland permanents are colorless. Players may spend mana as though it were mana of any color."
This errata/patching approach would improve balance for so many banned cards, underplayed decks, and broken metagames. It's also the model used by all major online games. I believe Wizards will eventually go in this direction, and/or are actively trying to figure out how to do this without creating a huge gap between online and paper Magic. Banning Nexus in BO1 and nowhere else is just the start of this journey, as was the functional Pridemate change.
Of course, this approach is not without risk. Wizards can become too balance-happy and issue errata for things that simply don't need it based on mass panics or outcry. Imagine if Wizards nerfed Teferi to only untap one land, or Wilderness Reclamation to untap only a few lands, because of MTG Arena complaints. Or increased CoCo's mana cost by 1 at the height of its power. Tweaks like this, and countless others we could speculate on, might assuage a short-term hysteria but would limit long-term feasibility of those cards/decks, especially if Wizards is balancing across formats. After all, if Wizards messed up the design to begin with, they can certainly mess it up again down the road. Additionally, part of the beauty of non-rotating formats is in finding design mistakes and exploiting them for powerful decks, and too much errata would make that challenging. Patching/errata is not a perfect solution by any means, but I do think if done correctly it would really open up Magic play and design space in ways that would benefit players and formats like Modern.
I would attribute the slowdown to a wide combination of factors including the rise of Discord, increased Reddit traffic, heightened Standard/Arena popularity, increased Arena streaming, entrance into a slower Modern season after the GP stretch, uncertainty about what to play due to MH, the decline of traditional forums like ours (both generally and specifically), and a host of other factors. We have seen posts every few months for literal years that Modern is losing players and/or popularity in an often isolated or small N context. I remember it in basically every metagame from Eldrazi Winter to present. The format is now, and continues to be, largely fine.
I have no issue with FoW-like cards, but would prefer alternate FoW designs to preserve some Modern and Legacy differences. Mix and match examples of FoW alternatives include:
-FoW for non-creature spells only. This gives Modern's least broken permanent type a leg up over the truly broken stuff.
-FoW where the alternate cost can only be paid in the first 1-3 turns, like a reverse Serra Avenger. This lets us respond to the broken early plays while not creating truly dominant late game states for Ux strategies.
-FoW Mana Leak instead of FoW hard counter. In many ways, this acts like the Avenger design with different functionality in the late game.
FoW itself is probably fine, but adjustments like these might make it more Modern appropriate. I know Ux control has a stigma as being pretty bad or overhyped, but if these existing, viable decks had consistent access to a powerful FoW tempo play, they could become really dominant. Adjustments like I proposed would assist Ux strategies and police broken stuff, while not also pushing too many other decks out. I would also hope other fair strategies get comparable tools so blue doesn't become a virtually automatic best choice for those preferring interactive decks.
Yes, it's early in the WAR release and prospective pilots might need to obtain the cards needed for Neobrand. Yes, the deck isn't fully refined and needs some tuning before reaching an optimal version. Yes, the metagame is still evolving and we won't have a good picture of it until after the Mythic results, WAR, and MH have settled. Even acknowledging those factors, however, the deck has seen a very mediocre and unimpressive competitive debut. I am much more excited by the UWx staples and their consistent performance.
I'll add in N=50 games on the draw to get the sample to N=100 and then do some confidence interval calculations, increasing N if needed depending on how wide the intervals look.
So far, our only relevant result is a 30th place finish at an MTGO Challenge (https://magic.wizards.com/en/articles/archive/mtgo-standings/modern-challenge-2019-04-30), which isn't the most convincing datapoint. I encourage everyone to be very skeptical of Twitch streamers, article authors, pros, and Twitter/Reddit personalities who speak publicly on this issue. These people aren't necessarily wrong, dishonest, or misleading. But their position gives them extreme incentive to oversell, misrepresent, or outright lie about big issues if it results in clicks, views, popularity, revenue, etc. If such figures are not citing their sources or making supported arguments, that should be a bright red flag.
Back in 2015, I estimated that Amulet Bloom won approximately 23% of its games (estimated from a coverage and MTGO sample) before T4. This included concessions by an opponent. The confidence interval of that range, calculated in two independent samples, was 15%-30% in the smaller N sample, and 17%-29% in the larger sample. Seething Song Storm averaged about a 25% pre-T4 kill using similar methods. Based on this, I'm comfortable saying if a deck exceeds 20% for pre-T4 wins and is also top-tier, it's probably in trouble.
When I played Cheeri0s a lot on MTGO in 2017, I had about a 75% MWP overall with about 200 games total in the sample, winning on T2 or T3 in an outrageous 40% of games. Obviously, nothing from Cheeri0s ate a ban, so the "true" win turn in a larger sample was undoubtedly lower. The deck was also not strong enough to catch on at a top-tier level. This shows that single-player samples don't necessarily reflect the MTGO-wide sample. We need more data to draw meaningful conclusions about potential bans and T4 rule violators.
I did a coverage-based analysis for both Counters Company and Affinity in a sample of approximately 60 Affinity games and 40 Company games. In those analyses, I found that Affinity averaged a pre-T4 win rate of approximately 5.3% with a 95% CI somewhere between 0% and 14.2%. For Company, it was 3.5% with the CI at 0%-10.7%. Neither of these decks had bans, so I'm comfortable saying the deck needs to be winning at least 10%+ of its games before T4 in order to be on Wizards' radar.
Finally, I remember an analysis I did of over 100 Caleb Scherer MTGO Storm games back when Baral first came out and Storm looked a little scary. During that time, we calculated an approximate 12%-14% (don't remember the exact #) pre-T4 win-rate. This iteration of Storm also did not eat a ban, so I'm comfortable with 12%-14% being a safe range.
Overall, based on all this, I'd estimate that a deck is in danger if it is BOTH top-tier AND wins approximately 20%+ of games before T4.
None of this is relevant to the bannability of the deck. All four of those qualities apply to Ad Nauseam, an eminently acceptable combo deck. The only question is whether or not the deck can consistently win before T4. Based on the previous Bloom ban example, I would estimate a deck would need to win before T4 in 20%+ of games in a large N, large T context in order to be bannable. Cherrypicked streamer results don't get there. My Cheeri0s win rate was about 75% with more than 25% of my wins happening before T4. Obviously, nothing from Cheeri0s got banned. My personal examples don't necessarily reflect the wider metagame realities. Let's increase the number of auditable events and the time span before we entertain ban discussion like Cohen's.
This transition from an issue discussion to a hyperbolic, personal one ("You don't have to turn every discussion...") is unfortunate and inaccurate. Even a cursory review of my post history shows that I have not mentioned ban mania as an issue in this thread in months. Talking about the metagame is fine. Talking about bans or unbans is also fine. Speculating on bans with extremely limited data over an extremely limited time period is, however, ban mania. There is no good reason for posters in this thread or Magic personalities elsewhere to buy into this uncritical method of ban analysis/prediction. It doesn't matter if this bad method is right once every 20+ cards. It's still a bad method. Once we have data from events like Challenges and MOCS's or other similarly sized venues, we can revisit with a more informed perspective.
Again, it doesn't matter if the deck is ultimately broken. That shouldn't change our method of ban analysis. We need auditable results that make up a large dataset over multiple larger events/fields. Just because a knee-jerk ban mania response is right every 20 or so cards doesn't make it a good response or approach to banlist analysis. Streamers and social media perosnalities have significant incentive to overyhype these kinds of claims. It drives views, clicks, ad revenue, future partnerships, speculation targets, etc. We cannot rely on this as an indicator of anything. Wait for neutral, large N results
Again, the ban mania reaction is inappropriate independently of whether or not a deck is ultimately banned. Bad analysis methods are bad whether or not they are right one in twenty (or more) times. If people speculate about bans on every top deck in Modern, which they did throughout all of 2017 and 2018, they are eventually going to be right. And they were this year: KCI did get banned. But that doesn't mean it's a good method when ever other ban candidate (Stirrings, Company, Temple, GDS, Vial, Cavern, Tron lands, Bridge, Moon, SSG, Opal, etc.) was safe. We need to condition ourselves to avoid this kneejerk reaction and wait for results before making statements like Justin Cohen's. Unfortunately, there is significant incentive to make those statements today, whether financial (people are speccing on Riders or some other piece,) social (people want the upvotes/shares), or just personal (people enjoy the praise and accolades of "Getting it right" or "Saying it first").
https://magic.wizards.com/en/events/coverage/2019MC2/modern-win-rates-and-interesting-decklists-2019-04-28
There are some oddities in either Karsten's data or his confidence interval calculation, and I'm not sure how he's finding the Ad Naus/Hardened Scales lower CI to be >50% on a 95% interval. But overall, it's great data. Humans definitely appears a strong contender again, and Tron appears very middling.
Here's the Modern-only data we wanted. Wizards noted that no player had a perfect 10-0 Modern record, so no one earned 30 points. The list below represents all decks with an 8-2 or better record. Wizards did not separate out 9-1s and 8-2s, and I don't want to do that parsing myself when the 8-2+ bracket is already a valuable grouping. Here's the Wizards' list:
https://magic.wizards.com/en/events/coverage/2019MC2/24-27-modern-match-point-decklists-2019-04-28
And here's my breakdown of conversion/representation rates of top decks. It's ordered from highest Day 2-->24+ Point conversion and includes both the Day 2% share and the 24+point% share. You'd read it like so: Ad Nauseam made up 1.9% (6 players) of the Day 2 metagame and 7.69% (2 players) of the 24+ point metagame, representing a 33.33% conversion rate of Day 2 --> 24+ players and an over-representation of +5.79% percentage points.
1. Ad Nauseam: 1.9% (n=6) --> 7.69% (n=2) (33.33% conversion / +5.79% representation)
2. Jund: 1.3% (n=4) --> 3.85% (n=1) (25% conversion / +2.55% representation)
3. Affinity: 1.6% (n=5) --> 3.85% (n=1) (20% conversion / +2.25% representation)
4. Titanshift: 2.2% (n=7) --> 3.85% (n=1) (14.29% conversion / +1.65% representation)
5. Humans: 9.7% (n=31) --> 15.38% (n=4) (12.9% conversion / +5.68% representation)
6. Izzet Phoenix: 12.2% (n=39) --> 19.23% (n=5) (12.82% conversion / +7.03% representation)
7. Eldrazi variants: 5.3% (n=17) --> 7.69% (n=2) (11.76% conversion / +2.39% representation)
8. Burn: 2.8% (n=9) --> 3.85% (n=1) (11.11% conversion / +1.05% representation)
9. Dredge: 6.6% (n=21) --> 7.69% (n=2) (9.52% conversion / +1.09% representation)
10. Whir Prison: 3.4% (n=11) --> 3.85% (n=1) (9.09% conversion / +0.45% representation)
11. W/U Control: 7.5% (n=24) --> 7.69% (n=2) (8.33% conversion / +0.19% representation)
12. Amulet Titan: 3.8% (n=12) --> 3.85% (n=1) (8.33% conversion / +0.05% representation)
13. Hardened Scales: 5.9% (n=19) --> 3.85% (n=1) (5.26% conversion / -2.05% representation)
14. Tron: 12.8% (n=41) --> 7.69% (n=2) (4.88% conversion / -5.11% representation)
15. Grixis Shadow: 5.3% (n=17) --> 0% (n=0) (0% conversion / -5.3% representation)
16. The Rock: 3.8% (n=12) --> 0% (n=0) (0% conversion / -3.8% representation)
17. Esper Control: 1.6% (n=5) --> 0% (n=0) (0% conversion / -1.6% representation)
18. Infect: 1.6% (n=5) --> 0% (n=0) (0% conversion / -1.6% representation)
19. Bogles: 0.9% (n=3) --> 0% (n=0) (0% conversion / -0.9% representation)
20. Other: 10% (n=32) --> 0% (n=0) (0% conversion / -10% representation)
Humans is the clear winner here, with the second highest 24+ point share (15.4%), the fourth highest conversion rate (12.9%) and the highest of decks with >10 Day 2 pilots, and the second highest over-representation (+5.7% points). Izzet Phoenix has comparable numbers: highest 24+ point share (19.2%), comparable Day 2 N (62 vs. 53), equal conversion rate (12.8%), and the highest over-representation of Day 2 to 24+ points (+7%).
All decks above Amulet Titan in the list had expected or better conversions into the 24+ points. Hardened Scales, Tron, GDS, BG Rock, Esper Control, Bogles, Infect, and the "Other" decks all fared much worse. Notably, 49 players played "Other" decks that weren't listed here representing 9.5% of Day 1 and 10% of Day 2, but ZERO of them (0%) made 24+ points. I'm comfortable viewing all these listed decks as the most viable 19 strategies in Modern, with everything else in a lower bracket of viability.
Also, Ad Nauseam looked really good in this event. Karsten identified it as having the highest MWP in the event and these metrics show it having the best conversion rate (33.3%) and one of the best representations (+5.8% relative to Day 2). Definitely a sleeper pick I would look into.