Serum Visions vs. Sleight of Hand: A Comparative Analysis

[Matt]

Now that the card Serum Visions is confirmed via scan (http://www.magicdeckvortex.com/CARDS/serum_visions.jpg), some discussion about its place in GAT builds that run Sleight of Hand has arisen. The following is an analysis of both cards' relative merits by an exhaustion of possibilities.

For the uninformed:

Serum Visions

Sorcery
Draw a card.
Scry 2 (Look at the top two cards of your library. Put any number of them on the bottom of your library and the rest on top in any order.)

Sleight of Hand

Sorcery
Look at the top two cards of your library. Put one of them into your hand and the other on the bottom of your library.



CONCERNS OF NOTE:

--First we will consider the cards' merits in the absence of other information about the status of the top cards of your deck.

--The shuffling ability of tutors and fetchlands is not relevant here, because for Sleight you don't know what the third card is going to be, and for Visions you DO know but will either leave the card on top or put it on bottom anyway.

--All sub-cases are eqaully likely with respect to each other.

--Winner of each sub-case in parentheses.



Case 1: one of top 3 cards is the one you want to keep - Winner: Tie

Sub-case A: Top card is the one you want, next two are no good. (VISIONS)
Visions: gets the good card now and a random card next turn
Sleight: gets the good card now and a bad one next turn.

Sub-case B: Second card card is the one you want, top and third are no good. (SLEIGHT)
Visions: gets a bad card now and the good card next turn.
Sleight: gets the good card now and a bad card next turn.

Sub-case C: Third card is the one you want, top two are no good. (TIE)
Visions: gets a bad card this turn, good card next turn.
Sleight: gets a bad card this turn, good card next turn.



Case 2: two of top 3 cards are ones you want to keep - Winner: Tie?*

Sub-case A: Top two cards are the ones you want, third card is no good. (VISIONS)
Visions: gets a good card now, good card next turn.
Sleight: gets a good card now, bad card next turn.

Sub-case B: Top and third card are the ones you want, second card is no good. (TIE)
Visions: gets a good card now, good card next turn.
Sleight: gets a good card now, good card next turn.

Sub-case C: Second and third cards are the ones you want, top card is no good. (SLEIGHT)
Visions: gets a bad card now, good card next turn, good card on third turn.
Sleight: gets a good card now, good card next turn, random card on third turn.



Case 3: three of top 3 cards are cards you want to keep - Winner: Visions
Visions: gets good card now, get second good card next turn, and a good card after that.
Sleight: gets good card now, get second good card next turn, and a random card after that.



Case 4: none of top 3 cards are cards you want to keep - Winner: Visions
Visions: gets a bad card now and a random card next turn.
Sleight: gets a bad card now and a bad card next turn.




*Considering cases 2A and 2C, Visions gives you four good cards and one bad card, Sleight gives you three good cards, one bad card, and one random card. This is partially offset by the fact that Sleight gives you its good cards faster than does Visions, and so Visions' overall superior card quality may prove ineffective at offsetting Sleight's speed if you don't get to draw that third card (by either natural draw or Ancestral Recall or whatever). This is however a very marginal advantage and I judge it to be less important than Serum Visions' overall superior card quality.



Thus, Serum Visions is superior to Sleight of Hand on its own merits (i.e., cast without any other knowledge about the status of your library). But, what about in combination with Brainstorm?



Case 1: Brainstorm, putting two bad cards back. (VISIONS)
Visions: One bad card in hand, one on bottom of library, next turns' draw is known.
Sleight: One bad card in hand, one on bottom of library, next turns' draw is random.



Case 2: Brainstorm, putting one good card and one bad card back. (VISIONS)
Visions: One good card in hand, one bad card on bottom of library, next turns' draw is known.
Sleight: One good card in hand, one bad card on bottom of library, next turns' draw is random.



Case 3: Brainstorm, putting two good cards back. (VISIONS)
Visions: One good card in hand, next turns' draw is good, third turns' draw is good or random.
Sleight: One good card in hand, next turns' draw is random, third turns' draw is random.

So Serum Visions also has a superior interaction with Brainstorm.



CONCLUSION: Any deck that would run Sleight of Hand should be running Serum Visions first. It is assumed that all such decks will be running four Brainstorm as a matter of course.

[Smmenen]

Could you explain the timing of "Scrye"?  Does it happen immediately after the card text, or concurrently?  

Steve

[Matt]

You do everything on a card in the order it's written (usually doesn't matter, but the first time I played a Mishra's Factory against Balance I found out [have to equalize lands BEFORE checking creatures]). To resolve a spell, you just go through it line-by-line, so the card you draw is not one of the cards you Scry.

Note: If I happen to be wrong about this - something I would not give more than a 1% chance of happening - Visions is still better because you end up putting one of the cards in your hand (same as Sleight) but have the extra option of putting the other card back on top instead of being forced to plant it way down below.

Actually you have a third option of putting both cards on the bottom, though there's very little call for that - it might have been an out in old Soldevi Digger/Balance decks that used Demonic Consultation to empty their library (i.e what if you draw your Balance AND Digger off the Sleight?) but that's so exceptionally rare that it's not really worth thinking about.

[Bram]

Brilliant work. I had been wondering what a decent structure for a 'benchmark' test of this kind would be when I first saw the rumored card on mtgnews. You've provided a catch-all analysis in looking at the individual cases. I've read it thouroughly and am once more forced to agree with you

[CF]

I'll chime in with Bram. Thanks for the work, Matt. This cards was news to me, but I'll be sure to get 4 foils asap :oP

--
Chris

[DavidHernandez]

Since both of these are Sorcery speed, I'm starting to take a longer look at Impulse.  

"Dig down 4, get what you need, and pile the rest on the bottom for reshuffling with a Fetchland"

Why is Impulse seeing so little play?

--Dave.

[Rico Suave]

I believe it's because Impulse is 2 mana.

[Gabethebabe]

When I saw the card I was immediately convinced it would be better than Sleight of Hand.

It digs one deeper.

[DavidHernandez]

I believe it's because Impulse is 2 mana.

Yes, but the sorcery speed is like costing you a land drop.  Don't get me wrong, i see the power in these cards, but I wonder if cards like Opt and Impulse make up for their "lack of digging" or "1cc more" by allowing you to cast on the opponents turn.

This is also one of the reasons i recently dropped DA's from GAT.

Dave.

[Tristal]

What do you do first turn that matters as an instant in GAT?  Assuming you don't run Force Spike or Disrupt...

Yes, if you lead with Tropical + Sapphire, Sleight is probably worse than Impulse.  The major point of Sleight for me is that it practically guarantees you are getting a second turn land drop when cast first turn.  (Even more so with Serum!)  You can keep hands like Tropical, Sleight, Sleight, Dryad, Force, Force, Misdirection - where the same hand replacing Sleight with Impulse would be a mulligan.

Maybe you wouldn't play a hand like that either way, you've played the deck more than I have, so I'll bow to your judgement on it.  The one mana difference makes a tremendous impact on why Sleight is good, and I don't feel Impulse's instant-ness makes it that much better.

[Rico Suave]

Yes, I meant more along the lines of Sleight is a first turn play, while Impulse is not.  Due to the high number of set-up spells in these decks, it also means you may still be setting up on turn 3, which slows you down a lot.

It also lets you do such things as first turn Land, Mox, Dryad, second turn Sleight (pump), Brainstorm (pump), Duress (pump) for a 4/4 beater, as opposed to Impulse making it a 3/3 (and EOT Impulse would leave it at 2/2).  Being cheaper to cast is definitely a benefit.

[Eastman]

Trust me. I've won a tournament running 4 Impulse probably more recently than anyone on the site, and I don't hesitate in saying it is terrible. It's the reason I can't play chain's keeper, it's just too awful.