Also, guys, doing it my way up above, you have a max of 12 iterations of the combo before you get called on slow play, and more likely 8-10. So, going by even numbers for the 2 card combo with Sidisi and the 3 card combo with Sharuum+Blasting Station, your chances of fizzling on each:
Sidisi chance of failure per iteration (2/3)
Sharuum chance of failure per iteration (3/4)

Sharuum first:

6: 17.8%
8: 10.0%
10: 5.6%
12: 3.2%

It gets acceptable once you're below about 10% if you ask me, so this is *barely* acceptable.

Sidisi:

6: 8.8%
8: 3.9%
10: 1.7%
12: 0.8%

WOW that's a lot better! Amazing what 8% chance increase per iteration has!

So, how much worse is Sidisi than Blasting Station?

Well, Sidisi has several vulnerabilities: Dies to topdecked sweepers, if your opponent has JtMS out, they get 2 cards which can be enough to combo off and kill you, etc. It may not be the greatest chance for your opponent to win (unless they have lethal on board) but I don't know how that compares to the chances of slow play with Blasting Station. So, my personal suggestion is that if you're playing at a Regular REL event, use Station as they are supposed to relax on competitive judging rules, and if at something competitive, make your best judgement call.

Nother thought: Splash red for Lava Dart? We're only 2 colors atm so we can sorta afford it. It gives you an extra iteration, maybe 2 depending on how many mountains you have. It's an instant even.

Edit: If you hit your Narc, it's 2 more iterations.

Hey, so I got the itch to look into Four Horsemen again after seeing some new cards out of the latest few sets, but nothing new seemed to improve things much.

I did revisit the math on adding Viscera Seer to the deck, though, and I think it looks promising.

If Emmy is first: 120 fails, 0 successes. - The remaining five cards {BDdSV} have 5!=120 possible ways to be ordered, but none of it matters since Emmy was first.

If Emmy is second: 120 fails, 0 successes. - The other five cards can be split a bunch of ways {B}{DdSV} gives 1!4!=24 ways, {D}{BdSV} gives 1!4! too, {d}{BDSV}, {S}{BDdV}, and {V}{BDdS}. Together that's 5(1!4!)=120 failures, since there's no way to get both Dread Return and a target in front of Emrakul if Emmy is the second card flipped.

If Emmy is third: 96 fails, 24 successes. - Again, there are 120 ways to order the non-Emmy cards, and again most of those orders fail. However, Viscera Seer + Dread Return will succeed and can do so with only two cards; it can occur either as {VD}{BdS} or {Vd}{BDS} which can each occur 2!3!=12 ways.

If Emmy is fourth: 36 fails, 84 successes. - Again, there are 120 ways to order the non-Emmy cards, but this time most of those orders can succeed and in fact it'll be easier to count the failures than the successes. The only way that we can get 3/5 of our non-Emmy cards and not get either Viscera+Dread or Sharuum+Dread+Blasting is by not getting Dread Return or by getting two Dread Return without Viscera Seer. That's {BSV}{Dd} 3!2!=12 failures, {BDd}{SV} 3!2!=12 fails, and {DdS}{BV} 3!2!=12 failures.

If Emmy is fifth: 0 fails, 120 successes. - If you have a Viscera Seer in your top four, then you're guaranteed at least one of the two copies of Dread Return is above Emmy; if Viscera Seer is the only under-Emmy card, then you're guaranteed Sharumm with a Blasting Station and two Dread Returns above Emmy.

If Emmy is fifth: 0 fails, 120 successes. - If Emmy is the bottom card you can do whatever you want at sorcery speed.

Grand total that's 348 successful stacks out of the total 720 possible stacks, for 48.3% odds of hitting a good stack from a randomized deck. It's not amazing, but it's definitely an improvement over the traditional 25% odds of stacking {BDS}E from {BDES}.

Also important for the discussion of increasing dead cards:

BDES: 25% odds. - Playing the old combo without any additional pieces has the fewest dead cards but the lowest non-zero success odds.

BbDES: 30% odds. - Doubling up on any single combo piece adds a potential dead card to the deck but boosts the success rate by 5%.

BbDdES: 36.6% odds. - Doubling up on any two different combo pieces raises odds by 6.6% with two dead cards.

BbDdESs: 45.7% odds. - Doubling up on all three combo pieces raises the odds by a further ~9% but with a total three dead cards.

BDdESV: 48.3% odds. - Adding Viscera Seer and a second Dread Return does more with two dead cards than doubling up every combo piece does with three.

tl;dr - Adding a second Dread Return to the deck along with a single Viscera Seer raises the odds of a randomized deck giving a successful stack of milled combo pieces from 25% to 48.3% for the cost of two "dead" cards. This is even better than the ~46% odds that running 2xSharuum+2xDread+2xBlasting would give.

The low manacost on Viscera Seer also makes it fairly easy to hardcast, so it's not the worst spell to have in hand either.

Ant the combo will now be deterministic. Mill everything, get 4 narcos, dread return Sharuum, get Desecrated Tomb, mill Dread or Guile 20 times, get 20 Bats with Haste thanks to Anger, attack for the win. Unburial Rites is there as a protection measure for the Dread Return, same with Sun Titan

If you add 4x Ancient Tomb and 1x of Mountain and 1x Plain (dual most likely) the rest of the deck is completely free to explore. You could go basically all out on search and counterspells with:

Force of Will
Spell Pierce

Brainstorm
Careful Study (To remove graveyard things from hand)
Ponder

And even add 4x Fragmentize in the SB which basically deals with most permanent forms of graveyard hate. You can also easily splash anything that is Blue, Red, White and Black, which will also allow you to cast Unburial Rites from hand.

Maybe that will be a new deck and not Four Horseman, but it is a cool new idea anyway. I call it Four Batman!

Sidisi chance of failure per iteration (2/3)

Sharuum chance of failure per iteration (3/4)

Sharuum first:

6: 17.8%

8: 10.0%

10: 5.6%

12: 3.2%

It gets acceptable once you're below about 10% if you ask me, so this is *barely* acceptable.

Sidisi:

6: 8.8%

8: 3.9%

10: 1.7%

12: 0.8%

WOW that's a lot better! Amazing what 8% chance increase per iteration has!

So, how much worse is Sidisi than Blasting Station?

Well, Sidisi has several vulnerabilities: Dies to topdecked sweepers, if your opponent has JtMS out, they get 2 cards which can be enough to combo off and kill you, etc. It may not be the greatest chance for your opponent to win (unless they have lethal on board) but I don't know how that compares to the chances of slow play with Blasting Station. So, my personal suggestion is that if you're playing at a Regular REL event, use Station as they are supposed to relax on competitive judging rules, and if at something competitive, make your best judgement call.

Edit: If you hit your Narc, it's 2 more iterations.

I did revisit the math on adding Viscera Seer to the deck, though, and I think it looks promising.

If Emmy is first: 120 fails, 0 successes.- The remaining five cards {BDdSV} have 5!=120 possible ways to be ordered, but none of it matters since Emmy was first.If Emmy is second: 120 fails, 0 successes.- The other five cards can be split a bunch of ways {B}{DdSV} gives 1!4!=24 ways, {D}{BdSV} gives 1!4! too, {d}{BDSV}, {S}{BDdV}, and {V}{BDdS}. Together that's 5(1!4!)=120 failures, since there's no way to get both Dread Return and a target in front of Emrakul if Emmy is the second card flipped.If Emmy is third: 96 fails, 24 successes.- Again, there are 120 ways to order the non-Emmy cards, and again most of those orders fail. However, Viscera Seer + Dread Return will succeed and can do so with only two cards; it can occur either as {VD}{BdS} or {Vd}{BDS} which can each occur 2!3!=12 ways.If Emmy is fourth: 36 fails, 84 successes.- Again, there are 120 ways to order the non-Emmy cards, but this time most of those orders can succeed and in fact it'll be easier to count the failures than the successes. The only way that we can get 3/5 of our non-Emmy cards and not get either Viscera+Dread or Sharuum+Dread+Blasting is by not getting Dread Return or by getting two Dread Return without Viscera Seer. That's {BSV}{Dd} 3!2!=12 failures, {BDd}{SV} 3!2!=12 fails, and {DdS}{BV} 3!2!=12 failures.If Emmy is fifth: 0 fails, 120 successes.- If you have a Viscera Seer in your top four, then you're guaranteed at least one of the two copies of Dread Return is above Emmy; if Viscera Seer is the only under-Emmy card, then you're guaranteed Sharumm with a Blasting Station and two Dread Returns above Emmy.If Emmy is fifth: 0 fails, 120 successes.- If Emmy is the bottom card you can do whatever you want at sorcery speed.Also important for the discussion of increasing dead cards:

BDES: 25% odds.- Playing the old combo without any additional pieces has the fewest dead cards but the lowest non-zero success odds.BbDES: 30% odds.- Doubling up on any single combo piece adds a potential dead card to the deck but boosts the success rate by 5%.BbDdES: 36.6% odds.- Doubling up on any two different combo pieces raises odds by 6.6% with two dead cards.BbDdESs: 45.7% odds.- Doubling up on all three combo pieces raises the odds by a further ~9% but with a total three dead cards.BDdESV: 48.3% odds.- Adding Viscera Seer and a second Dread Return does more with two dead cards than doubling up every combo piece does with three.tl;dr- Adding a second Dread Return to the deck along with a single Viscera Seer raises the odds of a randomized deck giving a successful stack of milled combo pieces from 25% to 48.3% for the cost of two "dead" cards. This is even better than the ~46% odds that running 2xSharuum+2xDread+2xBlasting would give.The low manacost on Viscera Seer also makes it fairly easy to hardcast, so it's not the worst spell to have in hand either.

We can have a package of something like:

Combo (20 cards)

Battlefield:

4x Basalt Monolith

4x Mesmeric Orb

Graveyard:

4x Narcomoeba

1x Anger

1x Desecrated Tomb

1x Dread

1x Dread Return

1x Guile

1x Sharuum the Hegemon

1x Unburial Rites

1x Sun Titan

Ant the combo will now be deterministic. Mill everything, get 4 narcos, dread return Sharuum, get Desecrated Tomb, mill Dread or Guile 20 times, get 20 Bats with Haste thanks to Anger, attack for the win. Unburial Rites is there as a protection measure for the Dread Return, same with Sun Titan

If you add 4x Ancient Tomb and 1x of Mountain and 1x Plain (dual most likely) the rest of the deck is completely free to explore. You could go basically all out on search and counterspells with:

Force of Will

Spell Pierce

Brainstorm

Careful Study (To remove graveyard things from hand)

Ponder

And even add 4x Fragmentize in the SB which basically deals with most permanent forms of graveyard hate. You can also easily splash anything that is Blue, Red, White and Black, which will also allow you to cast Unburial Rites from hand.

Maybe that will be a new deck and not Four Horseman, but it is a cool new idea anyway. I call it Four Batman!