Epic Experiment has been a favorite of mine since its release. I tried to make it work so many times pre-Theros but it never quite came together. The math of the deck rarely worked out; consider the following deck construction criteria:

• Your deck must be mostly instants and sorceries and...
• They must actually do something on an empty board because..
• You probably can't play many, if any, creatures because they'll dilute your Experiment.
• You need a lot of mana to operate yet you are penalized for having too many land.
• You can't play Counterspells unless you want to screw around with Quicken, so...
• You probably just pack a ton of burn and hope for the best.

These do not inspire a whole lot of confidence.

These are not exactly a set of rules I want to build around.

But I'm going to try, at the very least so I can set this concept aside forever and put my troubled mind to rest.

Let's get one thing out of the way quickly: Goblin Electromancer and Guttersnipe don't belong in this deck. Relying on 2-toughness creatures in any sort of combo deck that can't run countermagic is a losing proposition. I love those two guys, but we must set them aside for now. That's a different deck.

Let's focus on burn.

The list of burn cards in Standard that are worth playing is pretty unappealing. I stopped my initial search for burn spells at CMC 4 because I honestly don't see this deck surviving long enough to hit more than 6 land drops. The list is as follows:

We can consider Rakdos Charm and Toil // Trouble if we want to dig deep and are OK with conditional damage spells, but let's put Rakdos Charm in the 'board for now. That's 32 cards, plus 4 Epic Experiment. Add 24 lands and you have what might, loosely, be considered a deck of Magic cards.

Anyone can assemble a pile of burn and call it a deck, but we need to justify building around Epic Experiment as opposed to just playing burn and creatures. I'm going to make a few assumptions to do some math.

• We will presume that over the course of the first 6 turns we draw 6 lands and 1 Epic Experiment.
• We will presume that we have not drawn any extra cards in any way.
• We will assume a perfectly random distribution of cards in our Epic Experiment and no knowledge of what may or may not be on top of our deck.
• We'll be playing first so we have enough mana to kill threats as they hit the board.

Over the course of the first few turns of the game, we'll use our burn to stay alive by killing their guys and aim to fire off an Epic Experiment for a minimum of X=4. Anything less is just a waste and anything more is assuming too much longevity. Now we have to do some math to figure out if this deck is worth playing at all.

On turn 6 ( on the play ) we'll have drawn a total of 12 cards, 6 of which will be lands and one of which will be an Epic Experiment. The remaining 5 will be some random burn spells. Thus, our remaining deck contains 48 cards of which 27 are eligible for the Experiment. Let's go with the best case scenario when X=4 and you hit 4 burn spells. If the average burn spell in this deck does 2.875 damage, that's only 11.5 damage.

Hardly a great pay-off for a best-case scenario.

If we swap Shock for Lava Axe and increase X to 5, we get up to 16.25 damage in this insane Magical Christmas Land where our experiment hits 100% of the time. That's a respectable total considering we'll probably have enough burn in-hand to do the last 3.75 damage.

Sadly, this is not Magical Christmas Land. The probability of hitting one of 27 burn spells on our first reveal for Experiment is around 55%. I'm doing some serious napkin math here, but I'd be glad to base my math around hitting 2.75 live burn spells in the average case. With an average burn spell value of 3.25 once we swap out Shock for Lava Axe, that puts us just shy of 9 damage per Epic Experiment. Hardly impressive for a 7-mana sorcery.  At that CMC, there are probably better things to do.

Now, there will be cases when the Experiment reveals some devastating mix of Warleader's Helix and Lava Axe, but I am honestly not equipped to do the math to estimate the probability of such an outcome. What I know is that the average scenario results in about 9 damage for 7 mana and I'm not wiling to build a deck around that concept, no matter how cool it would be to watch. I will applaud anyone who can build this deck and win matches with it, but based on the napkin math I'm not willing to call this a successful brew.

Experiment Over.  Time to grab a beer and drink this idea out of my head forever.  Unless...hmm...Modern....