Standard: Composition of Jund decks

Here's a table of successful Jund decks over the last couple months. The four decks on the left are Reid Duke from GP Miami, Paul Cheon and Jose Francisco Silva from this PE, and Owen Turtenwald from his last Jund article. (The other decks in the table are pre-Voice of Resurgence.)

Here are the common elements of our four "up-to-date" Jund maindecks:

   25 lands
    (20 duals + 2 Kessig Wolf Run + 3 other)
   7 "small spells" (4 Farseek + Arbor Elf / Rakdos Keyrune / Ground Seal)
   14 creatures
    (2 Olivia + 4 Huntmaster + 4 Thragtusk + 2 Garruk Primal Hunter + 2 other)
   7 removal spells
    (1 Tragic Slip + 2 Putrefy + 4 others)
   3 Bonfire of the Damned
   1 Rakdos' Return

This formula sums to 57 slots; most of the decks spend the last 3 wildcard slots on "big spells" -- the last Bonfire, extra Olivia or Rakdos' Return, or the potentially-big Mizzium Mortars.

We can say 25 lands / 7 small spells / 7 removal / 21 big spells is the usual formula for standard Jund.

Duke's Miami deck follows this formula the closest:
- His 2 "other" creatures were Vampire Nighthawks.
- His 4 "other" removal were 2 Pillar of Flame, 1 Abrupt Decay, and a 2nd Tragic Slip.
- His 3 wildcard spells were Bonfire #4, Olivia #3, and Rakdos' Return #2.
- His manabase was 22 duals, 2 Wolf Run, and 1 Cavern of Souls.
- His "small spells" were the 4 Farseek, 2 Ground Seal, and 1 Rakdos Keyrune.

Of the four decks we're looking at, Cheon's is the outlier -- it plays 2 Liliana, 2 Arbor Elf, and 3 Forest, which are all "throwbacks" to pre-Voice of Resurgence.

The patterns are a bit different if we look at players' 75s instead of their 60s:

Counting sideboards, our Jund decks all need access to 3 Ground Seal, 2 Vampire Nighthawk, 2 Liliana, and 3 Pillar of Flame, plus some mixture of 5 discard spells. The average sideboard is mostly taken up by filling out those quotas, with a few slots left for big creatures or extra removal or Underworld Connections. Our formula of 25 / 7 / 7 / 21 expands to something like 25 / 10 / 11 / 24, leaving only 5 flexible spaces for sweet sideboard tech.

Reid Duke is really running a "modest" Jund deck here -- he skips out on stuff like Sire of Insanity and Mizzium Mortars, but he includes all 4 copies of Ground Seal and Tragic Slip. The most glamourous thing about the deck was probably that it had 2 Ruric Thar instead of Vraska or whatever, but the Ruric Thars never even made it into the text coverage.

Probably the "big mana" engine makes sideboarding more important with Jund than with other standard decks -- if only 28/60 of the deck is action spells, swapping out 6 cards can make a big difference in what the deck does (unlike in a deck with 40 creatures).

Standard: Jund is the slowest

Some opening hands
From M3G1 of Michael Hetrick's Jund videos -- an opening hand on the play:

t1 nothing
t2 nothing
t3 Putrefy
t4 possibly nothing

Hetrick:
"Even if we draw a red source quickly enough, this hand doesn't do anything, still, for quite a while. The best that we could draw would be Farseek, obviously, but we'd still need to draw the correct card after that. This being an unknown matchup, I'm gonna have to mulligan. This hand just wouldn't be very good against aggro."

Here's the six we find:

t1 nothing
t2 Ground Seal
t3 probably a Keyrune
t4 probably a Huntmaster

Hetrick:
"Alright. Sure."

The second hand casts some spells by turn three, but against aggro, it wouldn't offer any resistance until the turn-four Huntmaster.

Though they're both better than a mull to five, neither of these hands is very likely to win against aggro.

Does Jund not run early drops?
Not really.

For three mana or less, a normal Jund deck has something like 8-10 removal, 4 Farseek, 2 Keyrune, and 2 Ground Seal, plus possibly maindeck Lilianas. This represents 16-20 of the 35 or so spell slots in the deck -- but only the removal and the Lilanas will affect the board.

Discounting the "set-up" cards (mana acceleration and cycling Ground Seals), we can expect about 10 spells from a Jund deck that will affect the board for three mana or less. A lot of aggro decks run 30+ spells that affect the board for three mana or less.

Does Jund need early drops?
Maybe not.

Jund is capable of fighting aggro with stuff like t3 Huntmaster or t4 Thragtusk. On the play, the Farseek plan can actually work against Champions of the Parish and Flinthoof Boars.

But probably yes.

Jund decks have a ton of Pillar of Flame, Tragic Slip, Vampire Nighthawk, and other cheap anti-aggro spells in the sideboard. The transformative aspect of this implies that the optimal plan for beating fast creatures is the opposite plan from how the deck is built. We could say that Jund decks are pre-boarded for Reanimator and other big-spell matchups.

Owen Turtenwald's Jund article this week mentioned that Farseek is the best card in the deck and basically showed that no hand with Farseek should ever be mulliganed. Turtenwald plays Keyrunes to supplement Farseek; traditionally, turn three mana acceleration in constructed is a sign of a slow format. The giant importance of a Rampant Growth that doesn't even block should illustrate how slow the Jund deck is.

How can such a slow deck be good?
It is good when the other decks are slow.

As far as I know, Jund does poorly in aggro matchups and really well in matchups where it can win with Rakdos' Return and Sire of Insanity. This means that it's a good deck in top 8s and day 2s, but it mostly loses in online daily events, where grinders play 19-land mono-red decks (for budget reasons and clock reasons).

Do the blue decks have early drops?
Yes -- they have Augur of Bolas and Azorius Charm.

The UWR decks floating around Magic Online these days are all big spell decks, with counterspells, Planeswalkers, and Sphinx's Revelations. A typical blue deck doesn't use many slots for pure removal (burn spells or Detention Sphere), but Augur of Bolas and Azorius Charm mean UW can play 8 two-drops that both draw cards and slow an aggro rush.