Throne of Eldraine contains a new cycle of mono-color lands that will surely become Standard staples. The opportunity cost of putting a Castle into your deck is relatively low, and the first few copies are close to basics with upside.
Their abilities are all useful as late-game mana sinks:
- Castle Ardenvale reminds me of Vitu-Ghazi, the City Tree, which won the 2005 World Championship. It will be a great fit for Selesnya Tokens or any deck based around Divine Visitation.
- Castle Vantress provides blue players with a free way to sift towards the answers or win conditions they need. Scrying is useful for any deck, but it synergizes particularly nicely with God-Eternal Kefnet.
- Castle Locthwain reminds me of Sea Gate Wreckage, a land I have played in Modern. It’s not a great fit for control decks due to the life loss, but aggro decks that empty their hand quickly can easily exploit the free card draw.
- Castle Embereth will fit extremely well in Mono-Red Cavalcade. You could view it as a Contested War Zone without the control switching, a Shefet Dunes without the sacrifice, or a Ramunap Ruins without the ban. Okay, that last one is a stretch, but this land is still amazing in the right deck.
- Castle Garenbrig could ramp out an enormous Voracious Hydra in Standard or an early Primeval Titan in Modern. Based on how powerful Temple of the False God was in Onslaught Block Constructed, I wouldn’t underestimate conditional ramp.
How often will they enter untapped?
To provide some deck building insight, I will focus on the following question: If you have drawn at least one Castle by Turn X, what is the probability that you have also drawn a corresponding basic-type land by that turn?
To determine this probability, I assume that play/draw is decided randomly, that you play 36 spells and 24 lands with 4 Castles, and that you make reasonable London mulligan decisions. I implemented this in a Python simulation; the file starts with additional details on why I chose this conditional probability, how London mulligans are handled exactly, and the impact of different land counts.
The results for turns 2 and 3 are shown in the following table.
|Number of basic-type lands in deck||Turn 2 probability||Turn 3 probability|
What Does That All Mean?
Suppose you run the following mana base:
- 8 Mountain (say, 4 Stomping Ground and 4 Steam Vents)
- 4 Castle Embereth
- 12 other lands (say, 4 Temple of Mystery, 4 Forest, and 4 Island)
The above table states that if you have drawn at least one Castle Embereth by turn 2 (respectively, turn 3), you are 72.4% (respectively, 77.1%) to have drawn a Mountain by that turn as well. This assumes that play/draw is randomized and that you make reasonable London mulligan decisions.
How many basic lands do you need for Castles?
For most decks, the consistency numbers with 8 basic lands already seem decent to me. This means that multicolor decks with playsets of all its shock duals could easily run a few Castles without requiring substantial mana base sacrifices. With 8 basic-type lands, Castles would enter tapped only about 1/4 of the time in the early game.
Low-curve aggro decks without catchup mechanics may have to be a bit more careful with tapped lands. These decks really need to curve out and want a reliable source of untapped colored mana on turns 2 or 3. For these decks, I would like to support Castles with at least 10 and preferably 12 basic-type lands. With 12, Castles would enter tapped only about 1/8 of the time in the early game.
How many Castles should you play in your deck?
Castles aren’t legendary, so nothing is stopping you from playing the full 4. Heck, some land-heavy multicolor decks may even run more than 4 total Castles. But I expect the most common number of total Castles will be 3 (or maybe 2) because of diminishing returns.
Due to the high activation cost, additional Castles beyond the first are unlikely to be activated, and drawing multiples early on can be awkward. Cards with these properties are often played as 3-ofs (and sometimes as 2-ofs). Yet the correct number depends on a lot of factors.
You should play more Castles if:
- Your deck has few one-drops or other tap-lands. In that case, dropping a tapped Castle on turn 1 is not a big problem.
- Your deck has a low mana curve but lacks mana sinks. In that case, the activated ability will be particularly valuable in the late game.
- The activated ability matches your deck’s strategy well. For example, Castle Embereth seems awesome for an aggro deck.
- Your deck has many copies of the right type of basic (say, 12 or more) for each Castle.
You should play fewer Castles if:
- Your deck already has a lot of tap-lands (such as Temples) and lacks good catchup mechanics (such as Wraths) to make up for missing your curve.
- You need more basics to consistently cast your one-drops on turn one in aggro decks.
- Your deck already has a strong late-game with multiple mana sinks. In that case, you’re unlikely to spend mana on Castle activations.
- The activated ability doesn’t match your deck’s strategy at all. For example, Castle Embereth seems questionable in a control deck.
- Your deck has relatively few copies of the right type of basic (say, 8 or fewer) for each Castle.
- Ravager Wurm sees a lot of play. I can already imagine a Temur Ramp deck with Ravager Wurm and Quasiduplicate tearing down Castles in the new Standard.
I will add 4 Castle Embereths to 20-land Mono-Red
Mono-color aggro players generally need untapped mana on turn 1 and will often run 20 total lands in a mix between basics and Castles. A natural question for these players is whether they should run 3 or 4 Castles.
The table at the start of this article doesn’t provide a full answer to that question, so I adjusted my simulation’s deck composition while retaining the other assumptions, including mulligan strategy. This yielded the following numbers:
3 Castle Embereth and 17 Mountains:
- On turn 1, if you have drawn at least one Castle Embereth, you are 98.3% to have drawn a Mountain as well (and thus 1.7% to have the nightmare mono-Castle land draw).
- On turn 5, if you have drawn at least four lands, you are 58.4% to have drawn at least one Castle Embereth.
4 Castle Embereth and 16 Mountains:
- On turn 1, if you have drawn at least one Castle Embereth, you are 97.3% to have drawn a Mountain as well (and thus 2.7% to have the nightmare mono-Castle land draw).
- On turn 5, if you have drawn at least four lands, you are 69.9% to have drawn at least one Castle Embereth.
Mana base construction is always a trade-off where you must place subjective weight on the quantifiable upsides and downsides. But I’ll take a 1 percentage point increase of the nightmare mono-Castle draw in exchange for a 11.5 percentage point increase of being able to give my creatures +1/+0 in the mid-game, so my first Mono-Red builds will contain 4 Castle Embereths.
Likewise, I’ll start with 4 Castles in other mono-color aggro decks as well. I’m looking forward to seeing them in action!