I always wondered how a 4 horsemen deck could survive post-sideboard. Turns out it doesn’t and the answer is turn the deck into show and tell. Brilliant.
The Judges will lock you up if you are caught selling Four Horsemen, but they can do nothing if you keep the playsets separately and tell the players to assemble the deck themselves.
You say that, but I do occasionally encounter people suspiciously lurking outside my LGS wanting to buy/sell cards. One of them did ask "You the guy with the thoughtsiezes?" Clearly a criminal looking to assemble four horsemen on the down-low.
I haven't watched it yet, but I was about to say "again?" And Brian mentioned that he decided to play it again. Then Brian mentioned that it was 4 years ago. Wow, time flies.
I remember the first 4 horseman video like it was yesterday, I was a freshman in college who couldn’t sleep and would watch one of your videos before going to bed
56:10 I think keeping a ponder with two blanks and a brainstorm actually sees fewer cards than shuffling when you already have a brainstorm in hand. Keeping the ponder: Brainstorm sees 1 new card, 2nd brainstorm sees 1 new card. 2 total new cards seen. Shuffling the ponder: Sees 1 new card, then brainstorm sees 3 new cards. 4 total new cards seen.
This is the very first deck I ever watched on your channel. Never watched legacy before this and I've been hooked since. Thanks for the great daily content Bosh!
My favourite part about this 4 horseman deck is that it now actually plays Syr Konrad, who is a man on a horse - a horseman! I believe the original 4 horseman deck used Sharuum the Hedgemon + blasting station
Love the change moving the ads to the end of the round instead of the start of the round, as someone who watches almost all of your videos it makes it easier to skip ads I've seen tons of times ty
At least you can respond to The emrakul trigger by milling more to get all your narcomebas in play before you get the reshuffle. So then you're just looking for konrad and dread return.
I ran 4 Horsemen in an SCG event back when the wincon was Sharuum, Blasting Station. Was able to get away with it with Leyline of Anticipation and Quicken, but maybe Borne upon a Wind could be your way around the slow play games? I also have a vivid memory of hard casting an emrakul against oops all spells in that event to make prize.... What a weird weird event that was....
Something about non deterministic really peaks my interest. It’s probably best for rules simplicity and running tournaments that is is basically not allow but part of me this it be cool if we found some way to make it not shadow ban.
It might be the Narcomoebas? It was some stupid Internet forum joke 10+ years ago. Syr Konrad wasn’t printed yet, just a happy coincidence that they printed an actual horse for the deck long after you couldn’t play it.
It is referring to the original wincon of the deck, which involved Dread Returning four times for four attacking creatures ftw (don't know what those creatures were, you'd have to ask Finn).
My magic history is hazy, but my understanding was that the original 'four horseman' was a very different deck that is actually unplayable with current rules - the four horseman were the original three eldrazi titans plus blightsteel colossus - massive, game-ending creatures that all have the 'if this is put into your graveyard, shuffle the graveyard into library' text. The deck created a loop where you would mill all but four cards of your library, then put creatures from the top of your deck into play, and this could be repeated an unlimited number of times. Thus, in theory, the players argued they would always get their four reshuffle effects on the bottom eventually, with the end result being all those creatures in play. You could play the deck until the heat death of the universe without it happening, though, unlike this version that is somewhat practical to actually execute the combo. The current deck inherited the name because it's much better but the same rules that made the previous deck unplayable also prevent shortcutting the combo in the dread return + emrakul deck.
@BoshNRoll The original Four Horsemen combo used a single copy each of Emrakul, Dread Return, Sharuum the Hegemon, and Blasting Station. It required reshuffling until you milled these four cards in an even more specific manner. Search up: "Finn" and "[deck]The Four Horsemen" or "[Deck] The Four Horsemen - Orb/Monolith" for the original mtgsalvation/mtgthesource threads from 2011. Finn is the creator of the deck (to my knowledge). Before that specific arrangement was settled on, there were other 'horsemen' cards considered as well - early on, 'horsemen' cards were often considered to be Emrakul + things to reanimate (with multiple Dread Returns being played, going through packs of reshuffled Narcomoebas). That being said, having 4 Narcomoebas was a bit more relevant in the original as they also served as the fuel for Blasting Station, making the eventual tortuously-slow kill slightly faster.
This comment brought back some old memories so i had to respond with a comment of my own. I remember way back when Wake Thrasher came out I was playing a merfolk tribal deck one friday night at a game store and some cool older dude playing next to me goes "Did you just say that merfolk was a 9/9? 😂
This deck traditionally uses the basalt/orb mill engine; but would it also work to use the cephalid breakfast (nomads/illlusionist) engine with the four horseman finish (konrad/emrakul)?
It's rather a shame that you can't play this in paper without getting a billion slow play violations. Gitrog Monster in cEDH has similar problems and I've heard players have tried contacting high level judges about answers regarding what the "game state" means because of how that combo plays out but they supposedly never got a straight answer. Both combos are super cool :(
Love the videos, but what was thought process in the last match with the ponder and “Brainstom does literally see another card so I won’t shuffle”? You had the opportunity to see 3 new cards.
Nope and that's the reason why this deck is soft banned at tournaments. Mathematically speaking it is impossible to guarantee that the playline of get the narcomoebas into play ->get dread return and konrad/another wincon creature into graveyard ->cast dread return because there is always the chance to hit your graveyard shuffler at any point before the last step which requires you to restart the hunt for dread return and wincon. Since you can't shortcut you have to manually play it out, which then becomes a fight against the round clock.
This deck can't shortcut since it is random. Any none random combination that advances the game is okay. This combo may or may not on each iteration. Personally I think it should be okay since I'm really sure that if you simulated this combo it would "always" hit. Though I guess if you simulated it 100000000000000000000000000000 times it would fail in one simulation. Disclaimer I did no math but am familiar with randomness and hugely mega unlikely outcomes.
With any effect that lets you cast sorceries at instant speed (quicken, t3feri) yes you may shortcut. I have played the deck in paper many times and my rule for not getting slow play calls was that I can go through the loop and shuffle once without changing the board state and then I'd have to pass the turn.
Yeah I make every combo player play it fully out as a way to punish them for being dicks and equally make them lose to time. I find most combo players quit when everyone makes them click through it every time. They don't like the combo themselves either, they just like to act clever and get people to concede via "good manners"
I understand the appeal of not folding to Endurance or Surgical, but why is this deck better for that than Breakfast? Can't you do the exact same thing with Nomad and Illusionist?
Really cool video as always thanks Bosh. Sweet deck, right into my alley for combo decks. The paper ruling issue questions me quite a bit though? I play a bunch of MTG but never got that invested into competitive settings but I've played a fair amount of competitive YGO. In Yu-Gi-Oh creating an infinite loop is forbidden as long as it doesn't impact the gamestate (quite an oversimplification but whatever). This however is a combo that results in a change in gamestate after each action and ends in one player's victory. It's not just milling your library forever with no payoff. So I wonder what, paper tournament rules wise, makes it qualify as a loop. Can't the player just choose to perform every action putting them on the stack, resolving, etc. Without defaulting to shortcuts? In YGO as long as the player would not take unreasonable time to think between each action while performing the loop this would not warrant slow play warnings... Also am I mistaken for thinking this kind of combo would fly in cEDH? I thought this kind of combos had reliable representation in that format, but admittedly I don't know much about cEDH.
A loop is created when you can say "If I do this X times, the game will look like this." This is not a loop because you don't know how many times it will take to get where you want, and "eventually I'll hit Dread Return and Konrad before Emrakul" doesn't solve for what other cards are in your deck and graveyard at that point. There's no shortcut to get there. In the rules you must advance the game. Milling over your whole deck and shuffling it back in does not advance the game, much like simply tapping and untapping Basalt Monolith over and over doesn't advance the game. It lands in a spot where this is clearly doable, and doesn't even really take that long, but it gets squeezed between "you must advance the game with your actions" and "this isn't a loop."
@@BoshNRollThanks a bunch for the answer that makes it clearer! The definition of slow play in MTG seems also vastly different from what would be considered slowplay in YGO so I guess this also has to do with it It's quite unfortunate because both this and Gitrog cEDH (that was the deck I was thinking of in cEDH) are so cool "on paper" but get drowned in slow play warnings over that technicality. But I also kinda understand that the rules are like this for a good reason and if you allow this kind of non-deterministic loop to be short-cutted it opens the door for a whole lot of degenerate unfun stuff.
If you have a good judge yes because it's not infinite, you will eventually run out of green mana or green cards to exile, i.e. at that point it's no slower than echos storm (also not infinite as echo exiles)
Can anyone explain why slow play warnings are essentially guaranteed in paper, if you play it out (as you would have to)? It seems like your loop would be "I tap and untap Basalt Monolith, and mill one each time", and you just pick up your deck and start milling. That seems like you could go pretty quickly, only pausing when you see a creature. Do you get slow play warnings if you stop playing spells, and the only thing changing is your deck and your graveyard? Or if you have to shuffle too many times, is that automatically slow play? It honestly seems like this would play faster than, say, a typical storm deck.
The problem sets in not when you mill through your library, but when you mill an Emrakul, shuffle your graveyard into your library, and then have to do it all again; the Emrakul exists so that you can hide your graveyard from graveyard removal on command, but also means that you can have Emrakul hit your graveyard before all the right pieces for your combo land in your graveyard. Once you have to shuffle Emrakul and your graveyard into your library and start the process over again, you are repeating a loop that doesn't necessarily change the game state, and that isn't guaranteed to hit a specified endstate (there's always a mathematical chance that you NEVER hit Konrad/Narcomoebas/Dread Return before Emrakul). That's where the slow play warnings start piling up.
@@Vorpalsword138 I've got all that. But you are taking actions and changing the game state (though not the board state except for the Monolith), and not necessarily slowly. My understanding of "slow play" is thinking a lot between actions, not quickly taking lots of simple (and maybe boring) but relevant actions. With an exception for declaring non-deterministic loops, apparently. It seems like it might be something where even if you're not claiming it's a loop you can repeat an arbitrary number of times, and you're just playing it out, the ruling is that you could think of it like a loop, maybe because it's infinite. Is that it? If so, would it still be slow play if it's not actually infinite? Say if you had to exile a card from your graveyard each time Emrakul triggered. Or is it just that you get slow play warnings only when it does go badly? I could see that. So if you played 4 Konrad, 4 Dread Return, and 1 Emrakul (a 71% success rate, I think), with a little luck you might be able to go a whole tournament without a warning?
@@ArsenicDrone Ok, so, I'll try to explain this, but its kinda stupid. The magic rules for slow play say that you can't stall, and states that you must play at a rate that will allow the game to conclude within the timeframe. Stalling is define as taking actions that do not move the gamestate forward. The first time you mill your deck, it is deterministic. You can prove that after you fully mill you will have your whole deck in your graveyard, 2 emrakul triggers on the stack, and 4 narcomoebas in play. This is a **loop**, is it a deterministic set of actions with an exact predictable outcome. After this, however, repeatedly milling your own deck only for it to get shuffled back in does not advance the gamestate. You cannot loop this action because it is non-deterministic, you can't know how many times you need to do it. The fact you can't know how many times you'll need to do it, and that you aren't advancing the gamestate until you do, means that it runs afoul of the stalling rule - you are not playing in a manner that WILL result in the game ending in the allotted time. The fact that you will inevitably - even reasonably quickly - hit the exact combination of cards in time doesn't actually matter... the way the rules are written, you will get a warning for taking actions that do not advance the gamestate if you try to do this after demonstrating the reshuffle. All the opponent has to do is call a judge until you get a match loss, and eventually a DQ. And since this deck is kinda famous, this is pretty well known.
@@robmccarty755 Thanks for the explanation. So it's sort of like the second paragraph in my second post (everyone knows it can be looped infinitely, even if you're not declaring that), along with a definition where if you get back to an identical (or at least indistinguishable) game state after you shuffle, that's "not advancing the game state"? And you'll get a warning upon your first (or maybe second) shuffle back to having the same library contents? I wonder what the ruling would be in the hypothetical case where it was not infinite, if Emrakul's trigger exiled a card from your graveyard (say of your choice, although random would be an interesting case too). Or really if anything changed outside of your library (the exile zone, opponent's library, the life totals, etc) each loop.
@@ArsenicDrone Basically, yeah. If Emrakul was exiling itself, and the contents of your library were changing, that would be a result that changes the gamestate, or if your opponent lost a life each shuffle or something that would do it, etc. Sadly, not the case. I really do think this deck is a very cool deck. Its original version was unbelievably awful and way too rng to ever be viable, and really did require the idea of going infinite on reshuffles to stack your deck. This version actually combo kills pretty quickly once it gets together, it really doesn't take many reshuffles to get the cards you need in the graveyard. I wish the rules would be adjusted in a way that permit this deck to exist, because I like the creativity behind its construction.
Yes, you just remove the emrakuels, add Oracle, and now you have a bad thoughtlash/breakfast deck. Self milling isn't illegal, infinite non-deterministic state looping is. If you don't have a way to reshuffle it becomes deterministic, i.e. all your cards are miller after X (52) activations. The PROBLEM is it's BS because this deck does have a deterministic win, it's impossible to not win once you are looping, the tourney paper shortcut rules though don't allow that. This is just the most common version but there are other niche decks that are shadow banned the same way. And it's sad how fast the judges will issue issue a slow play if you bring this deck while your opponent over there storming for 20 minutes straight or running Nadu isn't. Likewise if your opponent taking 5 minutes a phase because he's "thinking" or shuffling. Slow play really just means "judge doesn't like you" in reality
I can't help but feel with them constantly adding cards to the pool that there isn't a way to avoid slow play violations with this deck... because I understand you cannot shortcut it because it is not deterministic but there should be a way to still add to the board every now and again to be able to over rule slow play violations... because it is a very cool and interesting take in deck building and wish I could see it back in coverage especially with how stale and over dominated legacy and modern have become recently... what a shame
Why has the deck never been run with a creature that only shuffles itself? You'd lose out on the graveyard hate resilience, but as far as I'm aware surgical extraction is the only hate card it actually stops, and it only does so when you have 3 excess mana since they can wait for the untap trigger to go on the stack.
Does Teferi Time Raveler make the combo deterministic? In response to Emrakul's shuffle trigger, you mill the rest of your deck and flashback dread return at instant speed.
Because Nadu advances the game with cards like Springheart Nantuko / land drops and eventually hits a point where it is deterministically infinite. This deck does neither of those things.
You should probably be yielding to your orb triggers and monolith activations, especially with a deck known for slow play issues. I know, at least for me, it saves a very real amount of time, especially with the MTGO lag. EDIT - I see you yielded to orb later so ignore that
It doesn't seem that bad even in paper. The mill loop is defined since it's just tap, untap, mill 1, so you can mill cards very quickly. From there you just have to shuffle your deck a few times until you get coin flip. Once you get past that point, the kill is deterministic, isn't it?
Yes, but that isn't how the slow play rules are written. Taking actions that do not advance the game state are against the rules. Repeatedly milling your own deck only for it to get shuffled back in does not advance the gamestate. The fact that you will inevitably - even reasonably quickly - doesn't actually matter... the way the rules are written, you will get a warning for taking actions that do not advance the gamestate if you try to do this after demonstrating the reshuffle. All the opponent has to do is call a judge until you get a match loss, and eventually a DQ. And since this deck is kinda famous, this is pretty well known.
@@robmccarty755 well if you do it 60 times in a row holding priority, you will end up with 4 narcamoebas in play. That advances the game state, doesn't it?
@@dave-kt7sj Yes but that part doesn't win you the game. Here's how this goes in the rules. You tap and untap your monolith, mill one. You tap and untap your monolith, mill one. You tap and untap your monolith, mill one. You hit an Emrakul. Your opponent calls a judge. He says to the judge "The actions my opponent is taking will result in the deck being shuffled an infinite number of times. This does not advance the gamestate." The judge says that he's right, and that makes it stalling. The judge asks you if you can tell him how many times you want to perform this action, and what the board looks like after you have finished. You tell him that you will do this (for the sake of argument, lets say 50) times, and that will result in your entire deck in the graveyard, 2 emrakul triggers on the stack, and four narcomoebas in play. The judge tells your opponent that that sounds like a deterministic result, which makes it a loop - effectively, an exception to the slow play rules. The action is legal. All four narcomoebas are still now in play. You tap and untap your monolith, mill one. You tap and untap your monolith, mill one. You hit Emrakul. Your opponent says "JUDGE!" He makes the same argument... this does not advance the game state. Once again, the judge asks how a number of times you want to do this, and what the resulting game state will be... and you have no idea. You don't know how many times you need to do this before you reach a state with your dread return and horsy in the graveyard. The judge explains that since you cannot demonstrate an exact number, this is not a deterministic result, and therefor not a loop, and therefore not an exception to the stalling rules. You get a warning for slow play. You tap and untap your monolith, mill one. JUDGE! You were previous warned for slow play - you just got a game loss. Next game, You tap and untap your monolith, mill one. JUDGE! Match loss. Do it again next match? DQ.
If somebody wants to know why this isnt possible in paper: As soon as you trigger the emrakul and then mill a card you get a warning for slow play. (Because you spend all this time milling yourself to come back to the original board state with no changes to the game state) Mill a 2nd card? Game loss. Do this in another game? Match loss. Do it in a third game? DQ.
Interesting, I was wondering why this combo was illegal while other non-deterministic combos are allowed (all the others I'm aware of even have some chance of "fizzling"). If I understand correctly, as long as something has happened to advance the game state even if you still have to tediously execute every iteration of the combo manually (and possibly even have some chance of "fizzling") it's alright (normal time rules for the event applying, of course)? So, hypothetically, if you had a combo that created a 1/1 token that dies at the end of the turn every iteration (guaranteed) but giving them haste to actually win the game was non-deterministic, that would be fine?
Still dind't get where is slow play here. "you spend all this time milling yourself to come back to the original board state with no changes to the game state" Could you please explain why the deck pilot need to mill himself in you example (i.e. the goal of those mills) and why those mills have "no changes to the game state"? I honestly don't see this action of the deck's pilot which the player NEED to do for the win AND which makes "no changes to the game state" from rules perspective at the same time.
@@AlexRawen So, lets start from the very beginning. What is progressing the gamestate? Thats material changes within a current game. Like: Drawing cards, losing or gaining life, permanents leaving orentering the game, ending a Phase (for example from Mainphase to combat) or changeing the amount of cards in a library or a graveyard. What happens when you mill yourself with the Mesmeric Orb? You change the game state - cards go from the lib into the graveyard. Now you hit Emrakul. Emrakul shuffles everything into the lib. We now reached a gamestate with 0 graveyard cards, X handcards, Y life and Z cards in library. You keep milling yourself. Hit Emrakul again. Shuffle your yard into your lib. What is the game state? Again its "0 graveyard cards, X handcards, Y life and Z cards in library." the very same as before you milled cards. The whole action did not change the gamestate but took several minutes. Thats, by MTG tournament rules, slow play. You are not progressing the game "in atimely fashion". Infact you dont progress anything. Thats why you cant play 4 horsemen. The Emrakul shuffle is not a deterministic loop. (In theory you could be unlucky and flipp Emrakul for 50min, never hitting the Dread Return) Therefore you can not shortcut it. You have to execute the process again and again - without advancing the gamestate. Conclusion: You will get disqualified for Magic Tournament Rule 5.5 Slow Play
Tbh horsemen should really be considered a shortcutable state by the game as it never ends up having to actually stop unlike eggs/kci, I don't see why it cannot be shortcut aside info gained by the opponent based on mills as aside that it is predetermined to create it's outcome.
My understanding from other comments is that would still be considered illegal. As soon as you shuffle the deck with Emrakul and mill a card you'd get a warning for slow play since repeatedly shuffling your deck does not advance the game state and that is all you've accomplished after taking all this time to perform those manual action. Basically, it falls in an awkward spot where doing it manually is slow play since each iteration accomplishes nothing, but it's non-deterministic so you can't just say "I do the thing until it works".
@@adamgalloy9371 Your understanding is correct. Under tournament magic rules, choosing to mill another card after you have demonstrated emrakul will inevitably shuffle it back is by itself against the slow play rules as they are written.
At the end of a Nadu loop you can say "After N loops I have X mana floating and the game looks like this." If there's any confusion about whether Nadu is deterministic, it's a user error. MTGGoldfish recently made a video where they explained the Boseiju/Otawara endgame incorrectly that introduced variance that doesn't have to be there.
You're mistaken, you're allowed to play Four Horseman, but you're not allowed to "shortcut" your non-deterministic combo. This simply means that you need to have enough time left in the tournament to resolve all the actions, because you're not allowed to hand wave that you'll eventually get your combo. You shouldn't get a slow play warning if your turn is only four minutes, you're under the wrong impression because you're mistaken on what combo got the slow play warning. I'll explain:
Not quite. They ruled a while back that shuffling your library over and over "doesn't advance the board state." An X card shuffled library is considered the same as an X card shuffled library, even if its theoretically in a different configuration. That's where the slow play infractions come in, because you are just looping actions without advancing the board state. You can still play the deck, but its tricky. You have to make sure that in between each Emrakul shuffle, you are somehow changing the board state. Bringing narcomebas into play, casting extra cards from your hand, etc. Anything to change the card count in your library or the cards in play, so its not the same X-card library after each iteration.
@@eSporks you didn't even let me finish typing. That is the exact definition of gatekeeping. Please try to provide a better argument than Wizard's official statement, because I doubt you've even spent one minute thinking about your response.
@@victorianchan8638 Brother, you posted your comment. We're allowed to respond to it. Furthermore, in the deck tech, Bosh openly explains how "You will rapidly accumulate slow play warnings and be removed from the event."
@@midnalight6419 that's why I corrected his mistake! You really should learn to read, or at least watch the next few minutes of the video, where BoshNRoll says "it's actually very fast Combo". Thing is though, you're refuting someone who's posted Wizards' official stance on the Four Horseman, and BoshNRoll has gotten the name of the Combo wrong, it's a Grinding Station Combo. But you knew that; right? Have a nice day.
@@eSporks That isn't how that ruling works, repeated looping actions have to not be able to change the game state for them to be stalling, if the rules worked the way you described then repeated use of effects that can do nothing based on dice or coin odds would be considered stalling if you got the same odds of doing nothing multiple times in a row. Simply by a judge checking the list if the deck it would be determined that it isn't stalling since it isn't arbitrarily repeating an action that will create the same gamestate as before regardless of repetitions. Your solution also doesn't function as a blanket statement of having repeated actions not be stalling and seems to be based on how rules related to sensei's divining top are done since activating top to look at the top 3 cards of a library is considered fine as long as game state changes because it's possibly done for a purpose unlike looking at the top 3 only to do so again and again on repeat since it's still going to be the same three cards and ultimately none of it matters aside the last configuration. That simply doesn't apply to all repeatable actions and especially not ones that don't do anything as no judge is going to let you slow down the game by tapping and untapping a basalt monolith inbetween each action when doing so serves no purpose. Hope this helps you understand what stalling is and is not.
