I suppose Firecat Blitz is bad for the Thousand-Year Elixir version, since we can cast it from our hand to get hasted cats. It looks like all the RR flashback spells are bad. We still have the hasted deck I suppose. But, with needing extra cards for Cackling Counterpart and Firecat Blitz, it doesn't look like we get an advantage with Orvar.
Can we do a Recoup chain into another red sorcery?
Also I think the storm connection issue has been in the deck for a while, I just spotted it now going through the stage again. so we might need a better copy effect anyway. The instants I mentioned aren't safe, as they can mess up the first computation.
A lot would be easier if we could force the first computation to be too small to be broken, but that seems impossible.
Hmm, the problem with relying on Recoup as an additional mana cost, is that Recoup can give a sorcery flashback multiple times, via multiple copies of Thousand-Year Storm. (even without Bolt Bend)
Hmm, I guess the problem for the previous decks is that Retract brings Vedalken Orrery back to our hand, so we can't make use of all the copies from Thousand-Year Storm. But getting the refresh from Gravitic Punch should be okay I think? Being that low just means that the top stage doesn't work as a stage of Busy Beaver applications, but all the lower stages should work as we intended them to.
Yeah this all seems to be a dead end, unless we can force us to have another land in play, but nothing seems to work. Glacial chasm is close but also allows us to sacrifice a land. Trying a useless double red sorcery like Death by dragons is just putting us super far behind in card slots.
Too bad.
IDK if waiting until the end of the stage is soon enough? it seems like we are missing out on building up iterations of BB(x) for most of this stage.
The Spitting image version did this scaling enough as we'd get storm copies of spellweaver volute right before casting the next spitting image. so that the next 2 life lets us update from x to BB(x) more computations, instead of each 2 life only progressing us from x to 2^^^^x or so more computations.
Well, in the top stage we get one or more BB computations, plus a bunch of layers that don't update to the output of the BB. The idea was that Gravitic Punch would get us that update, so the top stage would basically just be applying BB(x) and updating the TYS count to BB(x). Then, the second stage would work properly, since during each layer we would go up to the top stage many times to apply BB(x).
The storm conversion issue in Retract/Soul Separator decks was fine because each PFE copy allows us to make a drake to increase artifact/enchantment count - allowing that count to update to the storm count with a single sorcery cast (in fact a single volute trigger).
I'm not sure that waiting until gravitic punch to get the update works; if the top layers of the stage only update the resources to some function f, then the entire stage reaches f{omega}(X) and then before we can finally update to BB(f{omega}(X)), whereas we could have instead reached BB{omega}(X).
If the top stage will only reach BB(f{omega}(X)), we basically lose a stage. But, then the f for the second stage will be faster growing than BB(X), so the second stage will be at least BB{omega}(X), and then the third stage will be at least BB{omega*2}(X), and so on. So we lose one omega, which is just fine.
Ah yes, that works fine then - same reason it's ok to use a few layers to set up one computation.
(Although I don't think the bolt bend strategy works; there's no window to cast it while meaningful PFEs are still on the stack)
That makes sense, being one iteration behind is fine. Especially if we get more total iterations.
These structures can still be counterintuitive.
So I think all of these avenues have ended up being pretty barren so I think we are back to one of the decks from a few pages ago? The best was up to 8 or 9? layers?
Unless we have any other creative ideas for optimizing the stage/hyperstage structure, it might be time for another rigorous check and then start writing it up again.
@Deedlit might be a good idea to update post #1 with some info on the whole BB/TWM engine such as this writeup and Stackfish's observation that we can have access to flash in a BB deck.
Yeah, though that wiriteup is pretty outdated, most of the core structure is still similar. Do you guys think it'd be easier to update that doc or start over (probably with a less formal tone?).
(Opponent can exit the computation with damaged boom kegs, which we can then kill by removing/stealing Darksteel forge.)
Edit:
Ok, this might not work, but considering what Tibalt's trickery is doing in other formats, we can maybe use it to get started? It does a lot of nothing for us during the rest of the combo as we won't want to mill valuable cards.
Mountain lotus Artificial Evolution, Tibalt's trickery into Omniscience and milling up to 3 cards (at least lingering souls).
With the three cards in hand we can do something like TYS, Spellweaver volute, K'rrik
flashback Lingering souls, get some more mills and free spells from trickery. (only 2 I guess...?)
The main benefit would seem to be that we might be able to start without card draw.
Edit2:
I think this can work after getting kickstarted as we can end on a sorcery (ie gravitic punches/recoup/restore) and get another batch of Tibalt's trickery and chain through the deck that way. but it needs a bit more setup than my initial line allows for.
I'm less sure about if we can withstand having about half of our cards milled though, especially if we need to hit some of the sorceries with trickery.
Tibalt's Trickery could be useful if we can afford to mill enough cards, and we can ensure it's not possible to get started without it. Would replace Show and tell + maybe some opening card draw.
Maybe a Nexus of fate could absorb the unwanted mill (though then we're unlikely to be saving cards), or maybe Recoup + Blue Sun Zenith (which already exists in some of our decks) could be used (expensive though).
Yeah it seems like we would still need card draw as I just don't think we can actually really get anywhere from land, lotus, artificial evolution, Tibalt's Trickery +3 cards as we quickly start running real low on resources.
We'd still need some card draw and just casting a big Blue Sun's Zenith would be easier and a lot cheaper.
Also wouldn't Recoup's flashback still exile the BSZ?
Either way, it seems unlikely that we can save slots with it. The only other benefit I can see would be that without Show and tell, rebuilding after Worldfire will require exactly a Dual Nature trigger, and there's no risk of getting flash at an odd time or anything like that. But the worldfire transition is already solid due to us not being able to withstand 2 worldfires and so we only get at most one show and tell after, which is not enough on its own to break things. (getting omniscience without flash is fine and getting just flash requires spending UUU on omniscience, which is too expensive.)
edit:
Looking back though the thread it looks like the best version was from Deedlit's post 3491:
(swapping Runed halo with Safe passage so we don't die when we lose all of our enchantments to start computation)
I still don't see an infinite, though I've mostly been trying to improve/combine cards, with no success.
And just to make some progress, here's the gamestate for setting up the computation:
Opponent has their 42^^^^42 or however many hacked necromancers + zombies as needed.
Opponent has the copies of our artifacts (3 Coat of Arms makes math easier) and hacked Dralnu's Crusades.
Opponent has a K'rrik with Arcbond on it, and an arcbonded zombie.
Getting started is as simple as activating our Soul Separator on Desolation giant (Thousand year elixir allows us to activate through March of the machines.)
Then resolving the token giant's trigger kills all of our artifacts and enchantments, locking us out of flash, un-animating the opponent's coat of arms, and killing our goblin boom-keg, starting an arcbond chain reaction that powers the waterfall UTM's Busy beaver program.
After the computation, we have three ways to regain flash:
1. Resolve down to an empty stack to cast Lingering souls (would require a preemptive soul separator activation for K'rrik)
2. Wait for a Spellweaver volute trigger on artificial evolution to steal a Vedalken Orrery thanks to Dack Fayden's emblem.
3. Wait for a dual nature trigger for Vedalken Orrery
Obviously 1 is a pretty terrible way, so we will mostly use the other two.
The issue in that deck is that during the hyperstage, Restore can also get Fountain of Cho, and with Amulet of Vigor that gives us white mana early.
An idea I had for an improvement is to cut the buyback instant. Since we can wait until a Gravitic Punch to update the number of permanents, we should be able to afford waiting until then to update the storm count too. Resolving one punch gets us life proportional to the number of creatures in play (which is BB(X) at that point), half of which can be converted into storm with Lingering souls (in turn updating permanents by casting instants).
Oh that's actually an easy fix now that we have basic mountain with something like Journey for the Elixir (then putting it in play with Show and tell)
I think cutting buyback runs into problems reprogramming between BBs but I have no clue how 'far' apart the desired machines are.
Edit: Basically, can we always find a O(BB^x(x)) machine within BB(x) of BB^x-1(x)?
It seems unlikely. We can get a bit more than BB(x) because we have many TYS and can hack/bounce/copy dralnu's crusades, but it still seems like not enough to find our next desired input.
Hmm, we also can do some tricky things with multiple boom kegs, killing/weakening some zombies right before proper computation starts.
We only need to do one computation per iteration of the top-level stage in order for the second level stage to act like the "real" stage without losing value in the long run.
Say the stack is something like:
- Gravitic Punch copies
- Gravitic Punch TYS triggers
- Worldfire
- Whatever triggers we need to recover after a worldire
- Stage triggers
Then resolving a Punch gets us X life, which can be used for updating storm, permanents, and setting up one computation; taking X to BB(X). It can also be used to build the top level stage but that doesn't get us to significantly more than BB(X).
After we run out of punch copies, we resolve through the hyperstage transition, use up a few triggers from the next stage, and rebuild the punch copies. So the next level stage is second-level stage is the one that repeats BB(X) a lot(overall computing BB_{omega}(X)), just like the top level stage would have been had we updated storm and permanents earlier. In the long run it's only as though the hyperstage had one fewer stage in it, which doesn't make a difference.
Edit: Since we can't have any woldfire TYS triggers, we also can't have gravitic punch TYS triggers so there will only be one of them, but I don't think that makes a difference to this logic
Actually, we can use tricks with AE to get the exact input we need, so its moot. Cutting buyback appears to cost us nothing.
Setup the output clock to add a zombie every clock cycle, so by the end we have exactly BB(x) zombies, each with 1,2,3 ... BB(x) remaining toughness. Call N the optimal input for the next computation such that N<BB(x). Then we simply resolve some of our AEs and hack a Necromancer to whenever a zombie dies, create an Assembly-worker or w/e the input is. Then hack the zombie with N+1 remaining toughness to coward. This results in a chain reaction of dying Zombies, converting exactly N of them into fresh Assembly-Workers.
Then we can simply hack the necromancer back. Or even easier, just include it in the setup.
Edit: the leftover zombies are just a bit of additional output for the next TM, as they are already in the proper sizes/damages.
Edit: this also means we don't need nearly as much colorless mana as we'd thought, making mana echoes seem like extreme overkill.
Edit3: Hmm, because of how long ago we made that buyback assumption, there might be a possibility of a big restructure of the deck. Notably, without Mana echoes I think Mimic vat is safe for computation as our opponent has no mana to activate it during computation. (still limited to 1 combat though)
The question I'm struggling with is: what can Mimic vat actually do for us? I just went to it as the most powerful card in previous decks that was excluded by mana echoes. Something like Razorfin abolisher is more back on the table.
In fact, the opponent still has access to some mana since our Show and Tell allows them to play their Wastes
Of course, there are alternatives if we need them
I haven't been able to come up with anything that seems better using Mimic vat. I wasn't ready to say that its definitely a dead end because mimic vat is such a tricky card. But if we have to also switch off of Show and Tell, it seems likely that it won't be worth it.
So I think we are just back to the above deck but with one extra card slot? Mana echoes still seems like complete overkill, but its only one card slot and there's not exactly a lot of other options.
The first machine will be a little tighter as we need to do a lot of setup, meaning lots of TYS, but just one round of drake familiar bouncing mirrormade gets us at least pretty close, and with Bloodbond march also being a free early layer I think we do get there before we even need to start activating soul separators.
Edit: after running into a few walls with Mimic Vat I think that it doesn't help, there are a few ways to get close with Razorfin abolisher, and a couple others, but I couldn't find anything that quite got there. Seems like if a mimic vat improvement is possible, it won't be easy to find.
Freeing up another slot would allow us to add Blue Sun's Zenith, getting us to 10 layers. But, I'm not sure where we would cut a card. Mana Echoes is no longer needed for its original purpose, but it seems like we still need some card for colorless mana, and Mana Echoes should do as well as any. (Unless there is a card that can provide colorless mana and fulfill some other purpose as well.) Getting rid of Panharmonicon would be nice, if we could avoid needing to cast Restore as well as Gravitic Punch and Worldfire in our hyperstage transition. But, the only alternatives I know of are green creatures. An instant would be nice, but the only instant I see that returns a land is Second Sunrise, which we couldn't make work.
As for cuts, doubling season is only needed for dack emblem+fountain of cho right? I think the earlier problems with Fuel for the cause replacing Cephalid shrine were when we were casting drake familiar between computations. Now it seems we only need to be super counterspell cautious after worldfire. Which looks like it is low enough to not cause problems. I'm not 100% that this actually works, and does make the start harder.
Bolt bend is barely needed too, but since its being used on an ability (Aegis Automaton) and not a spell, there's not a lot of other options, and none look better. If we were playing two headed giant, or three player, Alhammarret's Archive wouldn't be needed, and we could replace bolt bend with Mirror gallery.
I suppose Firecat Blitz is bad for the Thousand-Year Elixir version, since we can cast it from our hand to get hasted cats. It looks like all the RR flashback spells are bad. We still have the hasted deck I suppose. But, with needing extra cards for Cackling Counterpart and Firecat Blitz, it doesn't look like we get an advantage with Orvar.
Can we do a Recoup chain into another red sorcery?
Also I think the storm connection issue has been in the deck for a while, I just spotted it now going through the stage again. so we might need a better copy effect anyway. The instants I mentioned aren't safe, as they can mess up the first computation.
A lot would be easier if we could force the first computation to be too small to be broken, but that seems impossible.
Hmm, I guess the problem for the previous decks is that Retract brings Vedalken Orrery back to our hand, so we can't make use of all the copies from Thousand-Year Storm. But getting the refresh from Gravitic Punch should be okay I think? Being that low just means that the top stage doesn't work as a stage of Busy Beaver applications, but all the lower stages should work as we intended them to.
Too bad.
IDK if waiting until the end of the stage is soon enough? it seems like we are missing out on building up iterations of BB(x) for most of this stage.
The Spitting image version did this scaling enough as we'd get storm copies of spellweaver volute right before casting the next spitting image. so that the next 2 life lets us update from x to BB(x) more computations, instead of each 2 life only progressing us from x to 2^^^^x or so more computations.
I'm not sure that waiting until gravitic punch to get the update works; if the top layers of the stage only update the resources to some function f, then the entire stage reaches f{omega}(X) and then before we can finally update to BB(f{omega}(X)), whereas we could have instead reached BB{omega}(X).
Pull from Eternity does target, but here I believe the Bolt Bend strategy from before does actually work.
(Although I don't think the bolt bend strategy works; there's no window to cast it while meaningful PFEs are still on the stack)
These structures can still be counterintuitive.
So I think all of these avenues have ended up being pretty barren so I think we are back to one of the decks from a few pages ago? The best was up to 8 or 9? layers?
Unless we have any other creative ideas for optimizing the stage/hyperstage structure, it might be time for another rigorous check and then start writing it up again.
Also I found that Darksteel Forge allows us to go infinite, so we need another way to let us survive through the computation. I think Endure/Safe passage/Consulate Surveillance still work though?
(Opponent can exit the computation with damaged boom kegs, which we can then kill by removing/stealing Darksteel forge.)
Edit:
Ok, this might not work, but considering what Tibalt's trickery is doing in other formats, we can maybe use it to get started? It does a lot of nothing for us during the rest of the combo as we won't want to mill valuable cards.
Mountain lotus Artificial Evolution, Tibalt's trickery into Omniscience and milling up to 3 cards (at least lingering souls).
With the three cards in hand we can do something like TYS, Spellweaver volute, K'rrik
flashback Lingering souls, get some more mills and free spells from trickery. (only 2 I guess...?)
The main benefit would seem to be that we might be able to start without card draw.
Edit2:
I think this can work after getting kickstarted as we can end on a sorcery (ie gravitic punches/recoup/restore) and get another batch of Tibalt's trickery and chain through the deck that way. but it needs a bit more setup than my initial line allows for.
I'm less sure about if we can withstand having about half of our cards milled though, especially if we need to hit some of the sorceries with trickery.
Maybe a Nexus of fate could absorb the unwanted mill (though then we're unlikely to be saving cards), or maybe Recoup + Blue Sun Zenith (which already exists in some of our decks) could be used (expensive though).
We'd still need some card draw and just casting a big Blue Sun's Zenith would be easier and a lot cheaper.
Also wouldn't Recoup's flashback still exile the BSZ?
Either way, it seems unlikely that we can save slots with it. The only other benefit I can see would be that without Show and tell, rebuilding after Worldfire will require exactly a Dual Nature trigger, and there's no risk of getting flash at an odd time or anything like that. But the worldfire transition is already solid due to us not being able to withstand 2 worldfires and so we only get at most one show and tell after, which is not enough on its own to break things. (getting omniscience without flash is fine and getting just flash requires spending UUU on omniscience, which is too expensive.)
edit:
Looking back though the thread it looks like the best version was from Deedlit's post 3491:
2 Xathrid Necromancer
3 Arcbond
4 Artificial Evolution
5 Wrong Turn
6 Dralnu's Crusade
7 Opalescence
8 Desolation Giant
9 Goblin Boom Keg
10 March of the Machines
11 Vedalken Orrery
12 Dack Fayden
13 Spellweaver Volute
14 Pull from Eternity
15 Mirrormade
16 Dual Nature
17 Bloodbond March
18 Cephalid Shrine
19 Thousand-Year Storm
20 Invulnerability
21 Mana Echoes
23 Soul Separator
24 Urborg Syphon-Mage
25 Thousand-Year Elixir
26 K'rrik, Son of Yawgmoth
27 Lingering Souls
28 Drake Familiar
29 Safe Passage
30 Spellweaver Helix
31 Worldfire
32 Gravitic Punch
33 Gravitic Punch
34 Recoup
35 Mountain
36 Restore
37 Panharmonicon
38 Wormfang Behemoth
39 Stolen by the Fae
40 Chrome Mox
41 Aegis Automaton
42 Guardian of the Guildpact
43 Kurkesh, Onakke Ancient
44 Bolt Bend
45 Alhammarret's Archive
47 Show and Tell
48 Omniscience
49 Sphinx of Enlightenment
50 Marchesa's Smuggler
51 Fountain of Cho
52 Amulet of Vigor
53 Animist's Awakening
54 Blue Sun's Zenith
55 Windreader Sphinx
56 Moraug, Fury of Akoum
57 World at War
58 Doubling Season
59 Uproot
60 Drafna's Restoration
(swapping Runed halo with Safe passage so we don't die when we lose all of our enchantments to start computation)
I still don't see an infinite, though I've mostly been trying to improve/combine cards, with no success.
And just to make some progress, here's the gamestate for setting up the computation:
Opponent has their 42^^^^42 or however many hacked necromancers + zombies as needed.
Opponent has the copies of our artifacts (3 Coat of Arms makes math easier) and hacked Dralnu's Crusades.
Opponent has a K'rrik with Arcbond on it, and an arcbonded zombie.
We have resolved safe passage, and a Dack emblem.
in hand: our artifacts and enchantments, and our buyback instant, most of our creatures.
in yard: other instants and sorceries, Drake Familiar, Urborg Syphon-Mage, Desolation Giant, K'rrik.
in play: X copies of Spellweaver Volute on our instants, a token Vedalken Orrery, a Soul separator, a token Thousand-Year Elixir, a token boom keg, opalescence, March of the Machines and many leftover token copies of our other artifacts/enchantments.
Getting started is as simple as activating our Soul Separator on Desolation giant (Thousand year elixir allows us to activate through March of the machines.)
Then resolving the token giant's trigger kills all of our artifacts and enchantments, locking us out of flash, un-animating the opponent's coat of arms, and killing our goblin boom-keg, starting an arcbond chain reaction that powers the waterfall UTM's Busy beaver program.
After the computation, we have three ways to regain flash:
1. Resolve down to an empty stack to cast Lingering souls (would require a preemptive soul separator activation for K'rrik)
2. Wait for a Spellweaver volute trigger on artificial evolution to steal a Vedalken Orrery thanks to Dack Fayden's emblem.
3. Wait for a dual nature trigger for Vedalken Orrery
Obviously 1 is a pretty terrible way, so we will mostly use the other two.
An idea I had for an improvement is to cut the buyback instant. Since we can wait until a Gravitic Punch to update the number of permanents, we should be able to afford waiting until then to update the storm count too. Resolving one punch gets us life proportional to the number of creatures in play (which is BB(X) at that point), half of which can be converted into storm with Lingering souls (in turn updating permanents by casting instants).
I think cutting buyback runs into problems reprogramming between BBs but I have no clue how 'far' apart the desired machines are.
Edit: Basically, can we always find a O(BB^x(x)) machine within BB(x) of BB^x-1(x)?
It seems unlikely. We can get a bit more than BB(x) because we have many TYS and can hack/bounce/copy dralnu's crusades, but it still seems like not enough to find our next desired input.
Hmm, we also can do some tricky things with multiple boom kegs, killing/weakening some zombies right before proper computation starts.
Say the stack is something like:
- Gravitic Punch copies
- Gravitic Punch TYS triggers
- Worldfire
- Whatever triggers we need to recover after a worldire
- Stage triggers
Then resolving a Punch gets us X life, which can be used for updating storm, permanents, and setting up one computation; taking X to BB(X). It can also be used to build the top level stage but that doesn't get us to significantly more than BB(X).
After we run out of punch copies, we resolve through the hyperstage transition, use up a few triggers from the next stage, and rebuild the punch copies. So the next level stage is second-level stage is the one that repeats BB(X) a lot(overall computing BB_{omega}(X)), just like the top level stage would have been had we updated storm and permanents earlier. In the long run it's only as though the hyperstage had one fewer stage in it, which doesn't make a difference.
Edit: Since we can't have any woldfire TYS triggers, we also can't have gravitic punch TYS triggers so there will only be one of them, but I don't think that makes a difference to this logic
Setup the output clock to add a zombie every clock cycle, so by the end we have exactly BB(x) zombies, each with 1,2,3 ... BB(x) remaining toughness. Call N the optimal input for the next computation such that N<BB(x). Then we simply resolve some of our AEs and hack a Necromancer to whenever a zombie dies, create an Assembly-worker or w/e the input is. Then hack the zombie with N+1 remaining toughness to coward. This results in a chain reaction of dying Zombies, converting exactly N of them into fresh Assembly-Workers.
Then we can simply hack the necromancer back. Or even easier, just include it in the setup.
Edit: the leftover zombies are just a bit of additional output for the next TM, as they are already in the proper sizes/damages.
Edit: this also means we don't need nearly as much colorless mana as we'd thought, making mana echoes seem like extreme overkill.
Edit3: Hmm, because of how long ago we made that buyback assumption, there might be a possibility of a big restructure of the deck. Notably, without Mana echoes I think Mimic vat is safe for computation as our opponent has no mana to activate it during computation. (still limited to 1 combat though)
Edit4: Yeah using Metalworker or Everflowing Chalice for colorless seems to make Mimic vat safe.
Metalworker gets the benefit of Dual Nature/Doubling season while Everflowing chalice only gets doubling season. Though metalworker needs haste/Thousand-Year Elixir and Everflowing chalice needs 2 initial mana.
The question I'm struggling with is: what can Mimic vat actually do for us? I just went to it as the most powerful card in previous decks that was excluded by mana echoes. Something like Razorfin abolisher is more back on the table.
Of course, there are alternatives if we need them
I haven't been able to come up with anything that seems better using Mimic vat. I wasn't ready to say that its definitely a dead end because mimic vat is such a tricky card. But if we have to also switch off of Show and Tell, it seems likely that it won't be worth it.
So I think we are just back to the above deck but with one extra card slot? Mana echoes still seems like complete overkill, but its only one card slot and there's not exactly a lot of other options.
The first machine will be a little tighter as we need to do a lot of setup, meaning lots of TYS, but just one round of drake familiar bouncing mirrormade gets us at least pretty close, and with Bloodbond march also being a free early layer I think we do get there before we even need to start activating soul separators.
Edit: after running into a few walls with Mimic Vat I think that it doesn't help, there are a few ways to get close with Razorfin abolisher, and a couple others, but I couldn't find anything that quite got there. Seems like if a mimic vat improvement is possible, it won't be easy to find.
Dropping Invulnerability allows enough room to use the deck in post #3482, and add the land part, so we get:
2 Xathrid Necromancer
3 Arcbond
4 Artificial Evolution
5 Wrong Turn
6 Dralnu's Crusade
7 Opalescence
8 Desolation Giant
9 Goblin Boom Keg
10 March of the Machines
11 Vedalken Orrery
12 Dack Fayden
13 Spellweaver Volute
14 Pull from Eternity
15 Mirrormade
16 Dual Nature
17 Bloodbond March
18 Cephalid Shrine
19 Thousand-Year Storm
20 Mana Echoes
22 Soul Separator
23 Urborg, Syphon-Mage
24 Thousand-Year Elixir
25 K'rrik, Son of Yawgmoth
26 Lingering Souls
27 Drake Familiar
28 Safe Passage
29 Spellweaver Helix
30 Worldfire
31 Gravitic Punch
32 Gravitic Punch
33 Recoup
34 Restore
35 Mountain
36 Panharmonicon
37 Wormfang Behemoth
38 Distorting Wake
39 Chrome Mox
40 Aegis Automaton
41 Guardian of the Guildpact
42 Kurkesh, Onakke Ancient
43 Bolt Bend
44 Alhammarret's Archive
46 Show and Tell
47 Omniscience
48 Sphinx of Enlightenment
49 Merchesa's Smuggler
50 Fountain of Cho
51 Amulet of Vigor
52 Animist's Awakening
53 Consecrated Sphinx
54 Widespread Panic
55 Sphinx Ambassador
56 Moraug, Fury of Akoum
57 World at War
58 Drafna's Restoration
59 Uproot
60 Doubling Season
That should give us 9 layers at the end.
Freeing up another slot would allow us to add Blue Sun's Zenith, getting us to 10 layers. But, I'm not sure where we would cut a card. Mana Echoes is no longer needed for its original purpose, but it seems like we still need some card for colorless mana, and Mana Echoes should do as well as any. (Unless there is a card that can provide colorless mana and fulfill some other purpose as well.) Getting rid of Panharmonicon would be nice, if we could avoid needing to cast Restore as well as Gravitic Punch and Worldfire in our hyperstage transition. But, the only alternatives I know of are green creatures. An instant would be nice, but the only instant I see that returns a land is Second Sunrise, which we couldn't make work.
Minor change to that deck restore-> Journey for the Elixir
jfb pointed out a few posts ago, Restore can also get fountain of cho which becomes a problem with amulet of vigor
This also makes panharmonicon more required as now we need to loop in show and tell to put the mountain in play.
As for cuts, doubling season is only needed for dack emblem+fountain of cho right? I think the earlier problems with Fuel for the cause replacing Cephalid shrine were when we were casting drake familiar between computations. Now it seems we only need to be super counterspell cautious after worldfire. Which looks like it is low enough to not cause problems. I'm not 100% that this actually works, and does make the start harder.
Bolt bend is barely needed too, but since its being used on an ability (Aegis Automaton) and not a spell, there's not a lot of other options, and none look better. If we were playing two headed giant, or three player, Alhammarret's Archive wouldn't be needed, and we could replace bolt bend with Mirror gallery.
Hmm nice, maybe Fuel for the Cause does work, in which case we get:
2 Xathrid Necromancer
3 Arcbond
4 Artificial Evolution
5 Wrong Turn
6 Dralnu's Crusade
7 Opalescence
8 Desolation Giant
9 Goblin Boom Keg
10 March of the Machines
11 Vedalken Orrery
12 Dack Fayden
13 Spellweaver Volute
14 Pull from Eternity
15 Mirrormade
16 Dual Nature
17 Bloodbond March
18 Fuel for the Cause
19 Thousand-Year Storm
20 Mana Echoes
22 Soul Separator
23 Urborg, Syphon-Mage
24 Thousand-Year Elixir
25 K'rrik, Son of Yawgmoth
26 Lingering Souls
27 Drake Familiar
28 Safe Passage
29 Spellweaver Helix
30 Worldfire
31 Gravitic Punch
32 Gravitic Punch
33 Recoup
34 Restore
35 Mountain
36 Panharmonicon
37 Wormfang Behemoth
38 Distorting Wake
39 Chrome Mox
40 Aegis Automaton
41 Guardian of the Guildpact
42 Kurkesh, Onakke Ancient
43 Bolt Bend
44 Alhammarret's Archive
46 Show and Tell
47 Omniscience
48 Sphinx of Enlightenment
49 Merchesa's Smuggler
50 Crucible of the Spirit Dragon
51 Amulet of Vigor
52 Animist's Awakening
53 Blue Sun's Zenith
54 Consecrated Sphinx
55 Widespread Panic
56 Sphinx Ambassador
57 Moraug, Fury of Akoum
58 World at War
59 Drafna's Restoration
60 Uproot