The first results of Ravnica Allegiance Standard events are in, and we’re finally living in a world with all 10 shock duals. In this new world, it’s natural to ask how far you can stretch your mana bases. Today, I’ll offer some insight by analyzing the mana bases of the Top 4 decks from last weekend’s SCG Open in Indianapolis—the largest Ravnica Allegiance Standard tournament to date.

That tournament had excellent live coverage, both video and text. This is in stark contrast to the Grand Prix at MagicFest New Jersey, which had neither. To make matters worse, the lack of coverage was “announced” in a reply to a tweet. I find these developments and the way they are communicated very disappointing. I was asked to do GP coverage for many years, and the last time I did text coverage, I proudly provided deck breakdowns, a Top 8 Twitter thread, and lots of other timely updates that added to the prestige of the event. For GPs that I did not attend, I could always trust other coverage reporters to give me a good and timely overview. I will miss all these things going forward. It’s the end of an era.

But the SCG Open did have proper coverage, so let’s honor their event. In the Top 8, there were surprisingly no copies of Skewer the Critics or Wilderness Reclamation, even though Mono-Red Aggro and Nexus decks were heavily played in Day 2 and on MTG Arena. Clearly, best-of-3 is a completely different format from best-of-1. Let’s check out the highest-placing deck lists.

Sultai Midrange

Anthony Devarti, 1st place at SCG Open Indianapolis

In the previous Standard, the number one deck was arguably Golgari Midrange. In Ravnica Allegiance Standard, this hasn’t changed. The only difference is that the deck now splashes blue. Hydroid Krasis is the real deal, and Breeding Pool makes it all possible.

In his trophy-winning deck, Anthony Devarti chose to run eight blue lands: four Breeding Pool, two Watery Grave, one Drowned Catacomb, and one Island. In contrast, the Sultai Midrange build of 8th-place finisher Abe Corrigan and 11th-place finisher Dylan Donegan ran 10 blue lands: four Breeding Pool, three Drowned Catacomb, three Watery Grave, and zero Island. All lists ran four Merfolk Branchwalker and four Jadelight Ranger, whose explore triggers as a whole roughly count as one extra blue source. So it’s effectively nine blue sources versus 11 blue sources.

Now, how many blue sources do we need to consistently cast Hydroid Krasis?

Assuming that green mana is never going to be an issue, a deck with nine blue sources in its mana base would be 91.4% to cast a 2GU spell on-curve, which is acceptable but maybe a tad low. It would be 95.9% for the version with 11 blue sources, which seems like overkill. Here, “cast on curve” means having at least one blue mana on turn 4 on the play after mulligans, conditional on drawing at least four lands, in accordance with the assumptions from my classic article.

In my evaluation, 10 blue sources hits the sweet spot. So that’s nine lands, plus the explore creatures. This yields 94.0% to cast Hydroid Krasis on curve as a 2GU spell and 98.4% as a 4GU spell. While a mini-Krasis may not be plan A—you’d hope to cast it for X=4 or more if possible—it’s important that you’re able to do it when your draw calls for it.

My suggestion would be to adjust Anthony Devarti’s mana base by cutting one Memorial to Folly and one Island, and adding one Watery Grave and one Drowned Catacomb. This increases the likelihood of casting Hydroid Krasis and Ravenous Chupacabra on turn 4 while simultaneously ensuring that you won’t draw basic Island in your opening hand. Given that the deck contains 1GG and 2BB spells, I don’t see a lot of upside to basic Island. There aren’t many Field of Ruin or Assassin’s Trophy around.

Bant Flash

Jonathan Hobbs, 2nd Place at SCG Open Indianapolis

History of Benalia and Frilled Mystic in the same deck? Is this for real?

Well, in his 26 land mana base, Jonathan has 17 white sources, 16 blue sources, and 16 green sources. Ignoring the potential card draw from the two copies of Depose // Deploy, which together count for half a source at best, this means that he is 88.2% to cast History of Benalia on-curve and only 81.9% to cast Frilled Mystic on-curve. (These probabilities take into account that there are 26 lands total and that there’s a lot of overlap in the blue and green sources.)

If you would replace the Plains by a Hinterland Harbor, then it would turn into 84.7% for History of Benalia and 89.1% for Frilled Mystic. That would represent an improvement overall, but it still doesn’t reach the 90% or higher consistency that I like to see. My conclusion is that Jonathan’s mana base is too ambitious, and that it would be better to cut the History of Benalia from the deck, perhaps replacing it with Deputy of Detention or Knight of Autumn as alternative 3-drops.

That said, I like how his archetype contains a variety of instant-speed effects that keep the opponent guessing. If they play around Frilled Mystic, they might walk into Angel of Grace. If they play around Settle the Wreckage, then you get them with Seal Away. And if they don’t cast anything, then you just evolve Growth-Chamber Guardian. That strategic element can be very powerful.

White Weenie Splash Blue

Max Magnuson, 3rd place at SCG Open Indianapolis

In the previous Standard, White Weenie decks often splashed red for Heroic Reinforcements and sideboard cards. With the introduction of Deputy of Detention and Hallowed Fountain in Ravnica Allegiance, Benalish Marshal decks have the option of splashing blue.

But is eight blue sources enough for Deputy of Detention?

One thing to keep in mind here is the land count. Since my definition of “cast on curve” conditions on drawing at least three lands by turn 3 on the play, you can improve your consistency by playing fewer lands. Eight blue sources in a 24 land deck means that 33.3% of your lands produce blue, but the ratio is 38.1% in a 21 land deck. It’s weird, but that’s how it works. You could argue that I should count Legion’s Landing as a land, but a timely transformation requires a lot of 1-drops and no interaction from the opponent, so I’d rather not rely on it.

In any case, eight blue sources is not enough. In this 21 land deck, you’re only 85.8% to cast Deputy of Detention on-curve. If you would replace a Plains by an Island, then this would increase to 89.6%, but at the cost of reducing the probability of casting Benalish Marshal on-curve from 100% to 94.1%. I don’t think that’s worth it. Also, Island doesn’t help cast two 1 mana creatures on turn 2. Azorius Guildgate is an option, but a tapland is really awkward in a deck with such a low curve.

Ultimately, splashing a 3-mana card with eight sources is substantially more difficult than splashing a 4 mana card with eight sources. I would recommend against it. My view on splashing Theater of Horrors in Goblin Chainwhirler decks is similar.

If you want to have access to Negate and Spell Pierce in the sideboard of your White Weenie deck and believe that an 86% castable Deputy of Detention is better than a 100% castable Conclave Tribunal or Adanto Vanguard, then I would limit the number of main deck copies of Deputy of Detention to two or three. This would at least reduce the risk of drawing multiples without a blue source. Besides, I am not convinced the deck even wanted twelve 3-drops in the first place.

Esper Control

Nick Cowden, 4th place at SCG Open Indianapolis

Kaya’s Wrath and Absorb are substantial boons to control players, but running both in the same deck is challenging.

Nick Cowden’s 27-land mana base has 16 white sources, 17 black sources, and 17 blue sources. Ignoring the potential card draw from the two copies of Search for Azcanta, which might count for half a source or so, he is only 82.6% to cast Absorb on-curve and only 82.2% to cast Kaya’s Wrath on-curve. (These probabilities take into account the exact make-up of the dual land mana base and that there are 27 lands total.)

These probabilities are far below where I would like to be. You could increase them by adding Guildgates, but that would only exacerbate the tapland problem. Instead, I would move the three Thought Erasure from the sideboard to the main deck for their surveil card selection, perhaps shaving an Island, a Syncopate, and a Cast Down. This would be largely in line with the Esper Control build of 7th-place finisher Andrew Davis. The resulting mana base still won’t be able to cast Absorb consistently on-curve, so I would limit Absorb to three copies, but at least things would improve for Kaya’s Wrath.

General take-aways

  • To splash Hydroid Krasis, I suggest 10 sources. This way, you can consistently cast it as a 2GU spell if need be.
  • History of Benalia and Frilled Mystic in the same deck is still very ambitious. Even with up to 12 shock lands and 12 check lands, you will run into colored mana problems.
  • I don’t like splashing a 3 mana spell with only eight sources of your splash color.
  • For CCDD cards, you really need 17 or 18 sources of each color (which is more than what you’d need for a 2CC spell) and you can’t afford to run any off-color lands. Get that basic Plains out of your Frilled Mystic deck and that Island out of your Kaya’s Wrath deck.
  • I didn’t really cover this, but if you’re building a 3-color midrange deck for the new Standard and wonder about the best mix of shocklands and checklands, then an average guideline would be to run 11 shocklands and 9-10 checklands.

While I computed the percentages in this article with my own simulations, you can get them yourself by using the website mtgoncurve.com. It’s an awesome tool that is based on my definitions. You simply input a deck list, and then a simulation runs to calculate the probabilities for each card in the deck, all automatically recognized by the website. It considers the exact makeup of your mana base, although card draw spells are not taken into account.

I found out about this website in a Reddit post by its creator probableluck. Due to slightly different assumptions in the mulligan and scry strategy, the results from the column P(mana | cmc) won’t match up completely with my numbers, but the difference is generally no more than a fraction of a percentage point. I believe the underlying calculations are sound, so I encourage you to check it out.