xger's picture
By: xger, Xger
Apr 21 2017 12:00pm
Login or register to post comments

For those unfamiliar, oraymw for a while wrote a series called Ars Arcanum that provided a statistical approach to new sets and tried to divine how the new set would play out compared to the old sets. He is unable to prepare one for Amonkhet, and as I always found those articles useful, I decided to make my own. Let's get to my take on this approach to Amonkhet!

Amonkhet Spoiler Analysis

This article is going to present a statistical look at Amonkhet and attempt to draw comparisons to older environments and give possible ways the format will evolve. This is largely based on creatures, as creatures make up the majority of almost every limited deck. It also matters what rarity a card is--while Rhonas the Indomitable is undoubtedly strong, you simply won't see it very often in draft (or sealed). Accordingly, I've weighted the cards according to rarity (very similar to how the Ars Arcanum's articles approached the analysis). Here are the weights and calculations, for those that are curious:

The main starting point is 24 packs--the number opened in a draft. So, then the question becomes how often you will see certain cards in a draft pod. Amonkhet has the standard:

  • 15 Mythics
  • 53 Rares
  • 80 Uncommons
  • 101 Commons

Of course, the booster pack is the same as normal--10 commons, 3 uncommons, 1 rare/mythic, and 1 land. For purposes here, I exclude foils, as that would be much more difficult to model and has a fairly low impact on the analysis (considering that foils would appear with the same relative rarities as non-foils, just with only a handful for the draft). So, here are the weighted ratios for each rarity, presented as Number per Pack * Packs per Draft Pod / Number in the set:

  • Mythic - 15/121 * 24 / 15 = 0.1983
  • Rare - 106/121 * 24 / 53 = 0.3966
  • Uncommon - 3 * 24 / 80 = 0.9
  • Common - 10 * 24 / 101 = 2.3762

So, any given mythic will appear in about 1 in 5 draft pods, whereas any given common will appear between 2-3 times, on average. Now, if I just used those numbers straight up, it would skew the averages, so the next step is to normalize them, so that the averages can be meaningful:

  • Mythic: 5.1% of the averages
  • Rares: 10.2% of the averages
  • Uncommon: 23.2% of the averages
  • Common: 61.4% of the averages

The numbers above are rounded, but I did not round in the calculation steps, only the final numbers.

Converted Mana Cost

Here again, I must credit oraymw for the data he made available to the public following his Kaladesh entry. While anyone could cultivate it, it is extremely useful to have it readily available, and very kind of him to leave it open to the public.

Here, I'm going to only present the last few years of data as opposed to as far back as the data goes:

The converted mana cost is key to the flow of the format--if all of the creatures worth playing cost 4 or more, the format is necessarily going to be quite slow. The numbers above are taken from the public data, and then I added Aether Revolt and Amonkhet using the same methodology as the previous sets to ensure an accurate representation. An important note about the included trendline--because I shortened the number of sets presented, the trendline isn't the same as the one shown in the Kaladesh Spoiler Analysis. That trendline was still going upwards overall, but here we see that since Return to Ravnica, the sets on average have decreased in CMC. The exact starting point is relatively arbitrary--no matter what method I chose a starting point by, there are going to be reasons for and against it. Here, I chose Return to Ravnica for two reasons: it made the information easy enough to digest in a graph, and Return to Ravnica is considered the latest surge in player base and when Wizards began printing considerably larger quantities of cards.

Here, Amonkhet clocks in at an average creature CMC of 3.34, not taking into consideration Embalm. However, 70 of the 130 creatures are 3 CMC or less, so the average is a bit skewed by some of the large creatures such as Seraph of the Suns. Another note: my creature count is probably a bit different than if you strictly looked at cards that say "Creature" in the type line. When a card always (barring removal) makes a creature, I include it. So, Sandwurm Convergence is listed as a 5/5 in my data because you will get a 5/5 after casting it. Similarly, Cartouche of Solidarity lists as a 1/1 for 1 in my data. So, knowing that a decent amount of the creatures are 3 CMC or less, the format is likely to be similar to most, as that is how most skew. As for the average, I like to compare it to first sets since that is the best corollary. Here, the closest (since Shards) to Amonkhet is the original Innistrad, with Amonkhet being just a tiny hair faster. It is still slower than Shadows over Innistrad by a decent bit. 

When you compare the number of creatures at each casting cost, Amonkhet stands out a bit in that there is a sharper peak at 3 than normal. Compared to Kaladesh, there are fewer creatures overall, and they cluster more heavily at the 3 drop slot. However, the tail end is a bit lower than Kaladesh, but that is likely because of Embalm. Embalm--the new mechanic that is essentially creature flashback--gives players something to do with their mana in the later turns, so there is less need for 5+ CMC creatures in general. The average embalm cost is 4.76, with the most common embalm cost being 5 mana. 

Creature Power and Toughness

Following the format of the previous articles, here are the power and toughness charts:


First, a note: sorry that Aether Revolt is not in here. I did not quite plan out gathering this data and writing up the article as well as I should have, so I was unable to compile the data necessary to include Aether Revolt here. Fortunately, Aether Revolt is a small set, so the comparisons would already be a big strange, so it should not be a large loss. If I do more of these articles, I will include Aether Revolt, and any subsequent gap sets.

So, Amonkhet will have creatures a little bit smaller than Kaladesh, but a bit bigger than a number of previous sets. The closest sheer size is Dragon's Maze, and the closest first set is Khans of Tarkir. Efficiency-wise, the closest match is Journey into Nyx, with only a few decimals off. The biggest takeaway is that Amonkhet will be a bit powered down when compared to Kaladesh, but it is more powerful than both Battle for Zendikar and Shadows over Innistrad. There are some cards driving the size in Amonkhet, as we've seen a number of previously more expensive cards get shifted down (such as Harsh Mentor as a red bear, or Those Who Serve being cheaper than Pillarfield Ox). Next up, here are the charts showing where Amonkhet ranks when looking at both CMC ratios and overall size:

So, following Kaladesh was an interesting place for Amonkhet, but you can see that the set is much closer to the norm than Kaladesh was. While it is still on the outer ring, it is reasonably close to Battle for Zendikar, Khans of Tarkir, and Shadows over Innistrad. At this point, it seems that Kaladesh was an outlier, likely due to vehicles. If we could find a way to similarly model equipment, I would not be surprised for Mirrodin limited to have a similar quality, where a new card type is considerably stronger when it is introduced versus the later 'safe' levels. If vehicles return, it is unlikely they will be as powerful as those in Kaladesh. 

Back to Amonkhet though--as I was saying, Amonkhet doesn't stand out in any particularly strong fashion, it mostly seems to fall in line with other sets. This likely means that the set will have a similar level of power to previous sets, which is a swift change from triple Kaladesh where the game could be decided in the first few turns. 

oraymw used to use a metric he called Power/Toughness differential. He stopped it for Kaladesh (though he calculated it in the public data). However, I find it a useful metric in attempt to divine as much as the format as possible. So, here is the differential:

Here, a negative ratio means that the average toughness is higher than the average power. The more negative the number, the stronger the chance that the format will have creatures knocking off of each other or simply gumming up the board. Simply put, Devoted Crop-Mate cannot really get past Those Who Serve. For instance, Gatecrash had an even differential, and that format is often considered to be quick the fast format. There, the battalion mechanic encouraged attacking as well. Amonkhet does have exert--only usable on an attack--but it also has embalm. So, on one hand you are encourage to act, but on the other, trading is better, which generally makes attacking harder. 

The other metric presented invovled oraymw's personal take on the speed of formats, so I cannot continue that here. Nor do I have a good replacement, so that is part of the reason I brought back the differential. If I continue to do these articles, I may also attempt to tackle the question of how to measure the speed of a format.

One last note: these numbers do not assume that you would play a Plague Belcher and hit itself--the assumption is that it hits a lesser valued creature. It would be difficult to model the self -1/-1 counter theme in a genuine way. If you always assume that it hits itself, you miss out on considerations like if you have another creature under a pacifism, or a small creature that is no longer doing anything on the board. The reason I went with excluding the -1/-1 counters is simply--the player will always have control over how to play it best, so the safest assumption is that the player will always get the most power and toughness on the board (which of course discounts cards like Channeler Initiate). 

Evasion and Card Advantage

Amonkhet is not very high on the evasion scale:

So, what does this value mean? I continued the methodology previously used, and it assigns values to different types of evasion as follows:

  1. Slight combat ability (presumably first strike)
  2. Difficult to block (I placed menace here)
  3. Flying
  4. Unblockable

Each creature gets a point value as listed above, and a 0 if it has nothing. On that basis, Amonkhet is a fairly underpowered set, with the lowest value in the sample. The latest set with a lower value is Mirrodin Besieged, so that is probably the closest comparator. It is also important to note that this low evasion value comes with a low p/t differerntial and a mechanic that encourages trading. In other words, this could a board clog set until whichever player draws their out, moreso than normal.

As for card advantage, there are two mechanics that play strongly into that theme for Amonkhet: Cycling and Embalm. Not only is cycling coming back, but it is bringing many new powerful additions, such as Censor, Deem Worthy, and Pursue Glory. This allows decks to play otherwise borderline or situational cards as they are not going to way down a player's hand. There are two commons in particular that will be strong cards simply because of the potential advantage of cycling:

Both of these are the types of creatures that are typically underwhelming in limited because you may never see your 6th or 7th land. This time, however, you are able to cycle them away, so they are not dead early game, and they can break open the board in the late game.

Next up is Embalm, a mechanic that is inherently card advantage. Creature flashback is something that will always be good, and it even makes otherwise normally unplayable cards into contenders. Consider Sacred Cat, a 1/1 for 1 with lifelink. Normally, this creature would not be worth the effect, but when you can get in for a few points early and are able to chump twice, it is suddenly gaining enough life to be worthwhile. It also means when players are in top deck mode that the player who has embalm creatures is just going to get very far ahead. Then there is embalm creatures like Angel of Sanctions, which is just a card advantage beat stick.

Lastly, there is Aftermath. If you are able to get good value out of both halves of the card, you are effectively drawing another card, just the same as flashback. The real challenge in relying on Aftermath as a means of assessing card advantage overall in the set is that you will not get the advantage if you cannot play both halves. If you are solidly Green/Red, and then open Cut/Ribbons in the last pack, it is still premium removal, even if you will never cast the second half. However, when you are able to play both halves, you can get absurd value. Using Prepare/Fight to kill 2 creatures and gain double life in a turn will be a huge swing. Essentially, Amonkhet is filled with card advantage potential.

Unfortunately, I have no real way of statistically applying a measure that is equivalent to the other data available, as I have no record of how cards were scored. Comparing to Kaladesh offer some idea though: Embalm is less common than Fabricate was, but it is a far more powerful mechanic. Energy is probably slightly comparable to Exert, though Energy would come out ahead in that. Embalm will be very strong in comparison, as getting a copy of your creature is often worth a card, and sometimes would be worth more than a card.


So, first thing to note is that the removal percentage is not high, but it's not abysmal either, joining the cluster around 18% with sets like RTR, Khans, and Origins. It is however, a bit lower than recent sets, though only a slight be lower than Kaladesh. There will be 2-3 removal per pack, on average, so there should be some options. Next up, how much does the removal cost?

Amonkhet has reasonably priced removal. If you were to take out Eternal Masters due to its particular nature, Amonkhet would be completely in keep with recent sets. One thing to consider though is that Amonkhet's removal includes some more situational removal such as Blazing Volley and Trial of Ambition. The overall lower cost does indicate there is chance to gain some mana advantage, and Deem Worthy even presents a chance to get card advantage at the same time.

Next, here is the effectiveness of the removal in Amonkhet, compared to other sets:

Most of Amonkhet's removal ensures that a creature dies, instead of tricks that may or may not really remove the creature. Of course, as normal, blue has a lot of bounce, which brings down the effectiveness score. However, this time blue has Illusory Wrappings, which is a particularly strong blue removal. Just keep in mind that it only overwrites the power and toughness, so a creature can still be pumped, and it can still activate abilities. The trend overall seems to be in moving back toward more effective removal. For references, here is how the number is derived:

  • Exiled - 5
  • Dead (killed, countered, discarded) - 4
  • Incapacitated - 3
  • Weakened - 2
  • Temporary - 1

Of course, Amonkhet did have a decent number of Weakened ratings as well, because the -1/-1 theme fits in that space.

Previously, there was a restriction metric as well. However, applying the method in the notes, I came up with a number wildly different than any set in the data, so I am not presenting that here. I will note that most of the restrictions are in red, where it only deals a certain amount of damage. There are also a number of removal that need you to have a creature in play, either to fight or to sacrifice, so that does impose some restrictions. Otherwise, the only real restrictive card is Time to Reflect which is just rarely going to do anything.

Wrap Up

Well, I hope you were able to find this analysis useful, and that I did a satisfactory job of filling the void of Ars Arcanum. In general, it seems that Amonkhet will play out similarly to recent sets, other than Kaladesh--which now appears to be a lone outlier. Keep in mind that there seems a decent chance that the boards will get gummed up, so effects that break through stales might be very valuable (such as Open Into Wonder). It should also be nice that the removal is not as bad as it has been in some recent sets, so hopefully the games will be less bomb centric.

I do believe that a lot of games will be about grinding value. With Embalm, Cycling, and Aftermath, there is just so much value built into the set itself, that it will be easy to build a deck capable of grinding considerable value over time.

If you have any suggestions, I would love to hear them. If oraymw wants to and is able to come back to continue these articles, that would be great, but I will provide them in his stead as needed, so I would greatly appreciate any comments or suggestions!

Good luck at the prerelease and see you online!


xger21 on MTGO.