I think The Celestus would work to turn artifact targeting instants / sorceries into lifegain. Tapping one of those will even trigger lifegain from all of them, but that is not an extra layer because updating their number also gives us new untapped copies. It is also sorcery speed only. That seems fine with just Wizard's Spellbook below it to create copies via targeting. We just let all copy triggers resolve and later tap them one at a time. But we wouldn't be able to incorporate it into longer sequences.
Kardur, Doomscourge would give us an extra lifegain layer out of each combat: We attack with X creatures and can destroy those one at a time to get Y lifegain triggers each. Though we'd need a way to keep the opponent alive.
Okay, so no Inscription of Abundance. But, if we give the opponent's creatures lifelink, we can give the opponent life through fights. Prizefight seems like a good card to use in the opening, since it gives us a Treasure token. We can also use The Tarrasque, to hook into Moraug and get triple nested 9 layers again.
Edit: So comparing the various options, it looks like:
If we just look at what we get from Moraug, Kardur, Doomscourge requires one less card, so it would be the better option by itself. But, if we can make Prizefight work, then The Tarrasque will require one less additional card for Moraug.
Assuming a 5 exile start again, we can exile 12 instants/sorceries afterwards. We can also give copies of The Celestus over to the opponent, allowing for 53 triggers of giving the opponent life. That should get us to 10 -> 10 -> 10 -> 10 -> 10 -> 66 -> 8. Then Kardur plus Moraug should get us to
Forsaken Monument goes infinite since we can counter any colorless spell we cast to get it into the graveyard and reuse it from there. No opportunity to have destruction of an artifact as a bottleneck.
Edit: I might as well mention another possibility: We can put life gain at the beginning of the layer sequence, and hook into red mana later. My first idea was to go Tibalt's Trickery into Sacred Fire to gain life, and then use Dragonspark Reactor to turn life into damage dealt to an opponent's creature. That looks like it only gets 7 layers though. The weakness appears to be Dragonspark Reactor, which makes us use all our life for one iteration of the Toralf stage. If we could pay a fixed amount of life per iteration, that would get us to 8 layers again.
- Tap Spellbook to exile Invoke the Winds. Resolve the copy to untap Spellbook.
- Tap Spellbook to exile Turntimber Symbiosis. Resolve Turntimber Symbiosis two put Jin-Gitaxias, Progress Tyrant into play.
- Tap Spellbook to exile Unexpected Windfall. Resolve the copy of Turntimber Symbiosis to put Orvar, the All-Form into play. Resolve the copy of Invoke the Winds to untap Spellbook. Resolve two copies of Unexpected Windfall to discard two cards, draw four cards and create four Treasure tokens. (4 Treasure tokens, 4 cards)
- Use the floating mana and 2 Treasure tokens to cast Mirror Box. (2 Treasure tokens, 3 cards). Cast For the Family to create a copy of Orvar, the All-Form. Cast For the Family again to create two copies of Jin-Gitaxias, Progress Tyrant (0 Treasure tokens, 1 card)
- Tap Spellbook to exile Snow Day. Cast Unexpected Windfall first to get 2 copies. Snow Day targets Orvar to create two more copies. Resolve Turntimber Symbiosis to put Ashaya, Soul of the Wild onto the battiefield. Resolve Invoke the Winds to untap Spellbook. Resolve Snow Day to discard a card and draw two cards. (2 cards) Resolve 3 copies of Unexpected Windfall to discard 3 cards, draw 6 cards, and create 6 Treasure tokens. (6 Treasure tokens, 5 cards)
- Tap our 3 nontoken creatures for 3 green mana, and cast 3 copies of Massive Might. Two copies target Orvar, to double the number of Orvars twice to 16. The next copy targets Jin-Gitaxias to create 16 copies. We next cast Ancestral Anger, getting 16 copies, to draw 17 cards. (5 Treasure tokens, 18 cards)
5 Treasure tokens is not very many, but if we can create 10 -> 10 -> 10 -> 10 -> 10 Xorns before we need to tap Wizard's Spellbook again, all the remaining exiles will add to the Conway chain expression.
Hmm, I was thinking that having two instant/sorceries that created Treasure tokens would double up on our number, but maybe it doesn't; it looks like the relevant numbers are the number of Orvars on the battlefield when Invoke the Winds is cast, and the number if Jin-Gitaxiases when Unexpected Windfall is cast. Since we can't use one to increase the other, it's still just an increment of one per Wizard's Spellbook activation. Still, we can bump up to 19 if we can get started with 4 exiles.
If we ever manage to get started without exiling Unexpected Windfall to Wizard's Spellbook we can easily go infinite. And even if all the other exilable instants/sorceries are literal blanks we can probably get going: We could exile at least a targeting instant from the G stage and still have an infinite with the extra treasures later. With a few repetitions of that, Invoke the Winds and Tibalt's Trickery there shouldn't be anything stopping us.
Invoke the Winds is pretty much the only card that I feel comfortable having in the deck despite it going infinite if we don't exile it at the start.
Ok now that everything is out, and maybe as a bit of a warmup for the vintage writeup, does Capenna change much? It would be nice for there to have been even one mv 0 spell, then we could go off turn two.
I still have to go through all the Capenna cards. Glittering Stockpile is interesting. But, the extra layer it gets doesn't stack with Xorn though. I was thinking it might allow us to double up with Wizard's Spellbook, but again the token creation depends on how many Orvars we have when we cast Invoke the Winds, and the untapping depends on how many Jin-Gitaxiases we have when we cast it. So we can't use one to increase the other, sadly.
Edit: Dragonspark Reactor is an alternative to Markov Retribution, but it looks like it gets the same number of layers. It could get one more layer if we had something that increased the number of its tokens that are created, but the only thing I found was Kaya, Geist Hunter, and sorcery speed messes that up.
Depopulate is a sweeper that also draws a card, It would be great if we could make a higher order stage instead of 'only' being able to chain them together. But I don't really see a way, it doesn't seem like enough.
Oh, I somehow missed Goldhound when I was looking for 1 mana acceleration. I was hoping it wouldn't come to this ^^'
Playing that turn 1 would allow us to get started on turn 2. But it also very much goes infinite with our current stages. In particular the stage with Orvar, the All-Form + creature targeting instants would produce mana of any color, which makes it go infinite and incompatible with the other stages. Currently we use Orvar and creature etb effects to make progress and reuse instants. Changing that seems extremely hard.
If we get to a choice between starting on turn 2 with say 2 stages and turn 3 with 4 stages I would go with turn 3.
There is a possibility to use Goldhound for the start and still get the current combo. We'd need to sacrifice Goldhound for the start and then make sure it gets stuck in the graveyard. Currently it should only get out with Devious Cover-Up, which we need to move Spara's Adjudicators to the library in the U stage. An alternative for that would be Bury in Books, which would leave creatures in the graveyard stuck.
The problem with Bury in Books is that it can't get Plains into the library, so I don't know how the U stage could produce and accumulate W to build a big W stage.
We could try moving another creature to produce white. But adding Rakish Revelers lets us produce red and start a bit of the Toralf, God of Fury stage with Toralf's Hammer. Shattered Seraph lets us produce an additional blue and go infinite that way, plus it gains life. I don't know any other creatures we need in hand to produce white mana.
Trying to animate the plains to get it into the library with Bury in Books would also allow us to target it in the G stage and get Orvar copies.
But wait. That problem is easier than I initially thought and was basically already solved with Secluded Courtyard. We just need a nonbasic, untapped, white producing land that doesn't go infinite. Brightclimb Pathway should work. (It doesn't go infinite in the W stage because Path to the World Tree can't fetch it.)
Depopulate is a nice creature sweeper, to make it work we need to have creature stage. I don't know how to make one outside of Mimic Vat and Soul Foundry. (which doesn't mean there isn't a way, of course!)
Errant, Street Artist seems somewhat cheap, as creature copying is cheap and blue mana is cheaper than our red spells at the end.
Riveteers Ascendancy is interesting, although again creature copying is cheap, so it doesn't seem like we can use it late - plus it could get Goldhound.
Without randomness we can use 1 more card and play Biblioplex Assistant. But if that is something we want to do we should also play 2 extra copies of Tibalt's trickery so no matter how much that spell mills we cast the same card. Mill randomness at the start seems unavoidable, but I find that more acceptable than randomness throughout the entire combo.
Starting on turn 2 we can't use Scaled Herbalist to get a 4th mana, so there is no extra red from the land drop. That means we can't go straight into Culmination of Studies after 5 exiles. I think it is likely that 5 exiles with Vault Robber can still get us started. That would be one less exilable spell, but we should work out the start again to make sure.
Okay, so we drop to 17 if we are willing to accept large amounts of randomness, and 16 if we aren't, assuming that we can get started with 5 exlies and Vault Robber.
I took a look at Aegar, the Freezing Flame and Flumph again to see if they could be used to remove life gain from the Toralf stage, but I can't quite figure it out. It seems like we want some way to separate the drawing of the card from the resolution of Toralf's trigger - with Aegar, if we could finish it an opponent's creature with excess damage, from something that isn't a Giant, Wizard, or spell, then Aegar could work? But Toralf isn't a Giant or Wizard, so we would need something like Arachnoform, and the damage would have to come from some ability from a noncreature. I didn't find anything like that.
Edit: Well I guess there are some sagas that could do that, but we need a way to draw a card without triggering Toralf, which makes Aegar hard to make work.
Cast For the Family targeting Goldhound. Leave it on the stack.
Tap the remaining land and sacrifice the Goldhound to cast Tibalt's Trickery, targeting the For the Family. Counter it, mill 3 cards, then exile and cast Wizard's Spellbook. (3 cards in hand)
4th Exile: Lorehold Command. Don't cast Tibalt's Trickery. Command targets Orvar to create a token. Get Ashaya, Soul of the Wild. Untap the Spellbook. Command gives all creatures indestructible and haste.
Tap Orvar, Jin and Ashaya for G. (3 G, 3 cards in hand)
Cast For the Family 3 times: Orvar, Orvar, Jin. (8 Orvars, 8 fresh Jins)
Get the Orvar tokens, put Xorn, Vault Robber, Aberrant Mind Sorcerer and 5 other creatures into play. Get For the Family back from GY. Give those creatures haste and tap them for G. Untap Orvar and tap it for G. Untap the Spellbook.
Cast For the Family targeting Aberrant Mind Sorcerer. 24 tokens are enough to bring back and cast For the Family 8 times (7x Orvar, then Jin) and get at least one copy back to hand in the end. (24*2^7=3*2^10=3072 Orvars)
Resolve Tibalt's Trickery and cast Bloody Betrayal targeting Orvar (6144 Orvars). Get 3072 Jin copies of Bloody Betrayal. Use one of the untaps on the hasty Vault Robber. Use the rest to generate 3072 G.
Cast For the Family targeting Aberrant Mind Sorcerer. 6144 tokens are enough to bring back and cast For the Family another 3069 times targeting Orvar. We end up with 3*2^11*2^3069=3*2^3080 Orvars. (number has 928 digits)
Cast For the Family targeting Xorn, get token copies. Then exile the first creature (Goldhound) with Vault Robber to create 3*2^3080+2 treasures.
We still have 3074 Blood tokens and a few instants/sorceries in hand, so using some of that mana to get access to the entire deck and go through the full combo will not be a problem.
Randomness really isn't a concern, we've had to face worse odds in previous decks. (shuffling cards in with Devious cover-up and having them on top in the exact right order.) Even though its happening throughout the combo, it still doesn't affect our average damage as the best case grows even faster. Minimizing randomness is good but not it if reduces our upper bound.
Very nice start! I like how we can initially present the opening hand as a bad RG stompy list with 4x For the family. (minor correction: the trickery from the 3rd exile mills Lorehold Command and Bloody Betrayal)
I had an improvement for more orvars/jins but Wizard's spellbook only works at sorcery speed so it doesn't work.
1 layer more than our current activations of the Spellbook. But I don't know if we can hook it into combat somehow.
Since we can only activate Spellbook as a sorcery we need a new Spectral Steel ability each time we want to use it this way. So max 4 times. Then 4 more casts of Lithoform Engine (at instant speed via Tibalt's Trickery).
We will need some card to create that first Orvar token. It is quite annoying that Croaking Counterpart can't target Orvar. >_<
Options like Necroduality seem to be worse for the start than just running a second copy of Orvar. And with that I can't quite manage to get started with 5 exiles, but 6 exiles shouldn't be a problem. So we lose out on 1 exilable spell from running an additional creature and 1 from the slower start.
The other possibility would be running Double Major, which could take care of the initial token creation and replace Mirror Box, while being exilable. To tie the 6 exiles option above we could take up to 8 exiles for the start. But 7 exiles might be doable.