When I say I have an 18 land deck, I mean I have 18 effective mana sources. Most grixis decks only want 4 to 6 lands in the long game and have 12-14 actual lands. We want double that. We don't need twice as many lands, but we need the effective lands drawn per game to double. At the same time, we can't really cut the number of spells drawn or we'll get run over. The solution is to draw more cards.
Once you're on that game plan, you need to keep drawing cards to avoid petering out as the game draws onward.
It's also worth mentioning that "deck thinning" from fetches is more relevant than the article someone linked earlier would indicate at face value--those percentages are based purely on an assumed 1 card/turn draw rate. The reality is that, by those numbers, most of the time our deck beats that percentage by about a factor of 2.5. I won't go too deep into the math but the thinning effect is "amortized" over fewer effective turns because we draw more cards per turn. It is actually enough to be relevant to a noticeable extent in long games, particularly when "draw x" spells come into play.
Cast out is solid anytime you specifically need an out to planeswalkers, noncreature artifacts and enchantments, or more exile effects (to beat dredge and the like).
While I appreciate all of the attempts to bring Math towards the mana base of this deck, I think most of you miss the point entirely to begin with:
It's not about curving out. I certainly don't play 26 lands (really 18 and 8 fetches) because I want to curve out. Rather, think of it like this: I'm playing 18 lands in my deck, because in most games I see about 3/4 of my deck and I want to see between 12 and 14 lands in play in most of my games. I then add fetch lands until the color fixing works out.
The core engine of the deck, in its original design, was to actually hit land drops over a protracted game, with the end goal of leveraging an advantage in available mana to do things like activate a colonnade with double cryptic backup, or to support backbreaking sphinx's revelations/white sun's zeniths. The goal is not necessarily to actually hit 4 lands on 4 or 6 lands on 6. The goal is really more like "we want to have 8 lands in play on turn 12, nearly 100% of the time".
The problem with adding serum visions and cutting lands is not really one of "hitting land drops"--the reality is, we're playing 25 or 26 lands for the most part, or maybe 24 with a full set of serum visions. That's generally about +3 or +4 lands compared to the average modern deck with the same cantrip suite. The problem with serum visions is tipping over the critical mass of "do I have enough actual THINGS to do in a long game?". In my personal opinion and experience, when you hit the level of playing 14 absolute do-nothing draw spells (4 charm, 4 serum visions, 4 think twice, 2 rev), you're diluted to the point of only having 20 to 22 actual business spells in your deck, as compared to the 24-26 in a build without serum visions at all. The difference between 20 and 26 spells is huge in the context of a long game. In the current metagame where opponents are often times loading up on just 2 or 3 distinct threats and lots of ways to recur them, it's absolutely defensible. When you start talking about collected company decks that effectively play 40 individual cards that you have to answer, some of which recur each other? That can get harder. That's part of why bant eldrazi and eldrazi tron are harder matchups than traditional midrange decks--they play a higher density of threats that actually have to be answered.
Don't hear me saying that cantrip builds are wrong--I think they're likely the correct choice in the current metagame. What I'm trying to point out here is that people on the cantrip-heavy builds (and ancestral vision or mystical teachings plans) are fundamentally changing the objective of the deck from "survive, leverage x-spells and mana sinks at some future date" to "shed the deadweight and move down to match my opponents tempo in the early game, and eventually overpower them with efficient draw spells". The further you move away from 26 ish lands and towards a lower land count and more serum visions, the worse white sun's zenith and sphinx's revelation become as engine tools (although revelation still retains some relevance as being effectively a win condition in certain matchups). It's not a discrete "this is where these cards are bad", but more of a "this is a smooth sliding scale that varies greatly across matchups and metagames, but in general the trend exists". This is why we eventually see wafo tapa on no zenith, no secures, and efficient cantrips with a dependence on snapcaster and creature lands to close out games.
Final point... Obligatory "I told you so" on cast out. Everyone who's played with it so far, feel free to sound off with your experiences, which mostly appear to be neutral or positive.
the kobe list is a midrange deck. They fixed the problem with the mentor-based midrange decks by removing mentor, and instead playing as many 4/4 lifelink creatures as they could. Seems like a reasonable approach, but it'll never actually beat a grindy control deck. If you look at his sideboard, he very carefully set up to beat the expected field. Good matchup against death's shadow, good matchup vs the perennial tier 1's and sideboard cards for the expected flavor of the months.
can I get a current working (UG) list? I started putting this deck together a long while ago, and stopped when early harvest wasn't immediately available, but I have some free time in my life right now so I wanted to branch back down the obscure combo builds again for a bit.
you don't want to pitch extra copies of as foretold--multiples in play is where the deck starts to go off.
too many CIPT lands is going to kill the deck--if you're playing a lot of cheap spells and looking to win by snowballing a leveraged mana advantage from the enchantments by casting more spells, not more expensive spells, cipt lands means you just don't do anything in the early game and don't have the expensive spells to ever make up the tempo.
dodging color requirements is a big deal--note that we're splashing esper charm off of two sources and likely playing more in the sideboard (like paths, timely reinforcements, etc). ancestral, serum visions, snapcaster all play nice mainphase, while mana leak, remand, and cryptic play nice on opposing turns.
Remember, the goal isn't to be the most busted as foretold deck--it's to be the most busted as foretold deck that can actually still win games when it doesn't find the key card.
So that's kind of the point here--seeing how far we can push the envelope. If we have a spell set that says it wants 25 lands, does playing 22 with some as foretolds do as well at actually casting things?
restore balance is a clear trap. If we were to play a sweeper, it should be supreme verdict or anger of the gods because those are actually castable. Remember, the goal is to have a deck that can actually *play* magic if it doesn't draw as foretold, not an aether vial deck that can't do anything powerful without the aether vial.
Goal: abuse as foretold not as a combo enabler, but as a backbone mana engine in an otherwise land-light control deck, taking advantage of low land count and high cantrip density to efficiently find answers.
Sample deck: not tested, not intended to be tuned, but rather as a proof-of-concept:
Assumptions: bolt/snap/electrolyze/serum visions/ancestral vision is a core component. there is a minimum amount of velocity you need to maintain in order to justify playing a 3 mana "do nothing" like as foretold. esper charm aids in either maintaining velocity or effectively executing an attrition plan to enable creature-land beatdown to close out the game. white remains critical for access to sideboard cards, particularly path to exile for larger creatures in the format.
I think the problem with surgical vs no surgical against company has more to do with how you approach the matchup than it does with how good the extraction effect itself is.
For some setups, surgical could be a good part of the plan. For other plans, it might as well be a dead card.
I'm not a fan of sideboarded leylines. I think there are higher impact sb slots available and that leyline is only good as a mainboard card because it is such a broadly applicable tool.
The other side is that you don't really know how much player skill changes the outcome. Donny is pretty good with this deck. I would say that he has at a minimum one game every stream where he makes a mistake that costs him an otherwise eminently winnable game, usually a mistake which I or the chat calls him on, and which he recognizes immediately after the fact. If you add those random game wins up over time, he's probably giving up a match every other league based on "unforced errors".
This doesn't count "judgement call" errors or disagreements in sideboarding technique--this is things like literally missing a clean window to gain life against burn postboard, or sequencing lands badly.
It makes me wonder, if the knowledge well goes deeper, how many of the things I would consider "judgement calls" are just plain wrong? Wafo doesn't try to beat burn or dredge at all.
My point is, the powerlevel is there for the deck. Fatal push was huge. We're only a small format shift or another good modal spell away from tier 1.
@codyx i actually stopped doing regular tournament reports because I've missed a bunch of events. I have a ton of notes from GP richmond because I played over 80 matches from thursday to monday morning, but they aren't in order. My stats sheet from the GP (for modern... we won't talk about my sealed or legacy trainwrecks here) was 40-11 against unique opponents, 42-15 in total (so i went like 4-20 something in other formats). I didn't feel a writeup was merited because almost two thirds of my matches were death's shadow decks and my configuration was supercharged to beat them, and I otherwise didn't really play against any bad matchups. Basically merfolk, eldrazi, burn, death's shadow. Five or more matches against each of those. Four tron players, three of whom demolished me.
I find that overall, tournament reports are about as useful as a "this in, this out" sideboarding guide: it depends entirely on how well you understand what is going on. That's why when I give my reports, I always explain WHY i sideboarded the way I did, and even that could be useless: the context of what's in my deck (mb leylines, zenith, snares) can often radically change what is or is not a reasonable sb plan. Like I said a page or two back about valakut decks: two mainboard cards can be the difference between an almost unwinnable matchup and one you have to try hard to lose.
I don't ever surgical the dredgers. When I'm on extraction effects I most often win by decking the dredge player. Take their ghasts and amalgams, and all of a sudden they're a bad draft pile that probably mills itself for about 3 a turn.
The collected company matchups are always awful to play for this deck because CoCo is so flexible--it's on the CoCo pilot to decide how to play the match, and we have to adjust accordingly. If they choose to nickel and dime us for card advantage (only commit one card at a time, noting that a lot of their cards are 2 for 1's; never aggressively presenting a potential combo opportunity) it can be hard to get anywhere with the wraths + haymakers plans. On the other hand, if they aggressively play for the combo, spot removal is hard to make use of because they'll just rebuild from nowhere again. It's a matchup that highlights the classic reason why threats are better than answers: there are no wrong threats, only wrong answers. In a game of chicken here, we lose more than half the time because we're the ones who have to both "guess right" AND "draw right". You can't commit to both spot removal and sweepers, or they'll just bury you in card advantage with renegade rallier, eternal witness, saffi, kitchen finks, chord, and collected company. This is why I've always tried to fight these matchups by attacking their source of card advantage (almost always graveyard-dependent, so cage/rip/surgical), and by presenting on-board threats that they can't actually trump (baneslayer angel, elspeth).
@kodieyost While I agree with you in principle that rest in peace is more brutal against dredge in general, surgical extraction and extirpate are far stronger for the esper draw-go builds against dredge simply because we don't really pack a way to "turn the corner" and punish dredge for overextending into the graveyard to begin with. Once you surgical the amalgams, you're left with trying to beat bloodghasts and hard-cast dredge threats. Beating hard-cast dredge threats is fairly easy now that they all die to wraths (no more regenerating trolls!), and the first couple of ghasts can be path'd or just traded for--after a certain point, dredge also runs out of fetchable lands, which makes wrath + surgical a way to clear them once and for all as well.
Imo, valakut is easy to beat. It just takes two mainboard slots or 3-4 sb slots.
2 leyline or halo main is enough to beat valakut because the rest of their strategy can be answered without our taking significant damage.
I prefer to fight coco with RIP + wraths, but these new infinite mana builds mean you probably want to actually extract the combo pieces or even just the coco + chord.
I have never really liked EE in this deck--i wouldn't play them unless blood moon becomes much more common. It's a flexible removal spell that can double as a sweeper against tokens or aggro. However, I tend to think that we have more powerful options than EE for the same targets.
You'll note that I have a full set of surgical + baneslayers and purges in list A and a pair of RIP + the same in list B. I'm not throwing the match away, but in the open I think surgical vs rip improves the expected matchup against the field more than a more flexible board like what I would play in the GP.
@xalthy I'm in the middle of a security clearance investigation so most of my life is on hold for a few weeks, and then I'm teaching full time until the end of the school year, so probably not until summer starts.
Once you're on that game plan, you need to keep drawing cards to avoid petering out as the game draws onward.
It's also worth mentioning that "deck thinning" from fetches is more relevant than the article someone linked earlier would indicate at face value--those percentages are based purely on an assumed 1 card/turn draw rate. The reality is that, by those numbers, most of the time our deck beats that percentage by about a factor of 2.5. I won't go too deep into the math but the thinning effect is "amortized" over fewer effective turns because we draw more cards per turn. It is actually enough to be relevant to a noticeable extent in long games, particularly when "draw x" spells come into play.
It's not about curving out. I certainly don't play 26 lands (really 18 and 8 fetches) because I want to curve out. Rather, think of it like this: I'm playing 18 lands in my deck, because in most games I see about 3/4 of my deck and I want to see between 12 and 14 lands in play in most of my games. I then add fetch lands until the color fixing works out.
The core engine of the deck, in its original design, was to actually hit land drops over a protracted game, with the end goal of leveraging an advantage in available mana to do things like activate a colonnade with double cryptic backup, or to support backbreaking sphinx's revelations/white sun's zeniths. The goal is not necessarily to actually hit 4 lands on 4 or 6 lands on 6. The goal is really more like "we want to have 8 lands in play on turn 12, nearly 100% of the time".
The problem with adding serum visions and cutting lands is not really one of "hitting land drops"--the reality is, we're playing 25 or 26 lands for the most part, or maybe 24 with a full set of serum visions. That's generally about +3 or +4 lands compared to the average modern deck with the same cantrip suite. The problem with serum visions is tipping over the critical mass of "do I have enough actual THINGS to do in a long game?". In my personal opinion and experience, when you hit the level of playing 14 absolute do-nothing draw spells (4 charm, 4 serum visions, 4 think twice, 2 rev), you're diluted to the point of only having 20 to 22 actual business spells in your deck, as compared to the 24-26 in a build without serum visions at all. The difference between 20 and 26 spells is huge in the context of a long game. In the current metagame where opponents are often times loading up on just 2 or 3 distinct threats and lots of ways to recur them, it's absolutely defensible. When you start talking about collected company decks that effectively play 40 individual cards that you have to answer, some of which recur each other? That can get harder. That's part of why bant eldrazi and eldrazi tron are harder matchups than traditional midrange decks--they play a higher density of threats that actually have to be answered.
Don't hear me saying that cantrip builds are wrong--I think they're likely the correct choice in the current metagame. What I'm trying to point out here is that people on the cantrip-heavy builds (and ancestral vision or mystical teachings plans) are fundamentally changing the objective of the deck from "survive, leverage x-spells and mana sinks at some future date" to "shed the deadweight and move down to match my opponents tempo in the early game, and eventually overpower them with efficient draw spells". The further you move away from 26 ish lands and towards a lower land count and more serum visions, the worse white sun's zenith and sphinx's revelation become as engine tools (although revelation still retains some relevance as being effectively a win condition in certain matchups). It's not a discrete "this is where these cards are bad", but more of a "this is a smooth sliding scale that varies greatly across matchups and metagames, but in general the trend exists". This is why we eventually see wafo tapa on no zenith, no secures, and efficient cantrips with a dependence on snapcaster and creature lands to close out games.
Final point... Obligatory "I told you so" on cast out. Everyone who's played with it so far, feel free to sound off with your experiences, which mostly appear to be neutral or positive.
too many CIPT lands is going to kill the deck--if you're playing a lot of cheap spells and looking to win by snowballing a leveraged mana advantage from the enchantments by casting more spells, not more expensive spells, cipt lands means you just don't do anything in the early game and don't have the expensive spells to ever make up the tempo.
dodging color requirements is a big deal--note that we're splashing esper charm off of two sources and likely playing more in the sideboard (like paths, timely reinforcements, etc). ancestral, serum visions, snapcaster all play nice mainphase, while mana leak, remand, and cryptic play nice on opposing turns.
Remember, the goal isn't to be the most busted as foretold deck--it's to be the most busted as foretold deck that can actually still win games when it doesn't find the key card.
restore balance is a clear trap. If we were to play a sweeper, it should be supreme verdict or anger of the gods because those are actually castable. Remember, the goal is to have a deck that can actually *play* magic if it doesn't draw as foretold, not an aether vial deck that can't do anything powerful without the aether vial.
Sample deck: not tested, not intended to be tuned, but rather as a proof-of-concept:
4 flooded strand
1 bloodstained mire
2 watery grave
2 steam vents
1 hallowed fountain
1 sacred foundry
2 island
1 swamp
2 creeping tar pit
2 wandering fumarole
4 ancestral vision
4 lightning bolt
4 fatal push
4 snapcaster mage
3 mana leak
3 remand
4 as foretold
2 electrolyze
2 cryptic command
Assumptions: bolt/snap/electrolyze/serum visions/ancestral vision is a core component. there is a minimum amount of velocity you need to maintain in order to justify playing a 3 mana "do nothing" like as foretold. esper charm aids in either maintaining velocity or effectively executing an attrition plan to enable creature-land beatdown to close out the game. white remains critical for access to sideboard cards, particularly path to exile for larger creatures in the format.
For some setups, surgical could be a good part of the plan. For other plans, it might as well be a dead card.
I'm not a fan of sideboarded leylines. I think there are higher impact sb slots available and that leyline is only good as a mainboard card because it is such a broadly applicable tool.
This doesn't count "judgement call" errors or disagreements in sideboarding technique--this is things like literally missing a clean window to gain life against burn postboard, or sequencing lands badly.
It makes me wonder, if the knowledge well goes deeper, how many of the things I would consider "judgement calls" are just plain wrong? Wafo doesn't try to beat burn or dredge at all.
My point is, the powerlevel is there for the deck. Fatal push was huge. We're only a small format shift or another good modal spell away from tier 1.
I find that overall, tournament reports are about as useful as a "this in, this out" sideboarding guide: it depends entirely on how well you understand what is going on. That's why when I give my reports, I always explain WHY i sideboarded the way I did, and even that could be useless: the context of what's in my deck (mb leylines, zenith, snares) can often radically change what is or is not a reasonable sb plan. Like I said a page or two back about valakut decks: two mainboard cards can be the difference between an almost unwinnable matchup and one you have to try hard to lose.
@kodieyost While I agree with you in principle that rest in peace is more brutal against dredge in general, surgical extraction and extirpate are far stronger for the esper draw-go builds against dredge simply because we don't really pack a way to "turn the corner" and punish dredge for overextending into the graveyard to begin with. Once you surgical the amalgams, you're left with trying to beat bloodghasts and hard-cast dredge threats. Beating hard-cast dredge threats is fairly easy now that they all die to wraths (no more regenerating trolls!), and the first couple of ghasts can be path'd or just traded for--after a certain point, dredge also runs out of fetchable lands, which makes wrath + surgical a way to clear them once and for all as well.
2 leyline or halo main is enough to beat valakut because the rest of their strategy can be answered without our taking significant damage.
I prefer to fight coco with RIP + wraths, but these new infinite mana builds mean you probably want to actually extract the combo pieces or even just the coco + chord.
I have never really liked EE in this deck--i wouldn't play them unless blood moon becomes much more common. It's a flexible removal spell that can double as a sweeper against tokens or aggro. However, I tend to think that we have more powerful options than EE for the same targets.
@xalthy I'm in the middle of a security clearance investigation so most of my life is on hold for a few weeks, and then I'm teaching full time until the end of the school year, so probably not until summer starts.