Happy new year, Moderners! It’s good to be back to the Nexus, even if I expect I’ll do far fewer articles in 2017 than I did last year. Between personal and professional developments, it’s been a very busy few months for me, but I’m excited to return and plunge into Magic’s newest set. No promises about how many articles I’ll do or when they’ll hit the Nexus feed! That said, I can always promise the articles I can do will offer the same rigorous, data-driven analysis readers have come to expect.
I’m bringing that spirit to a new Aether Revolt staple that needs no introductions. You’ve dreamed of playing with it since the Masters of Modern preview. You’ve read about it all week. Soon, your poor creatures will be feeling the “THIS IS KALADESH” kick at a tournament near you. Now all that’s left is to break out those metagame statistics and, in true Nexus fashion, quantify just how good Fatal Push really is.
It’s been a while since a card had this much hype. Eldritch Evolution, the Birthing Pod that never was, wasn’t even close. Authors and players across the Modern content-sphere haven’t been able to contain their Push praise: see the cheers of Paulo Vitor Damo da Rosa, Sam Black’s (premium) writeup, or the measured analysis of our own Ryan Overturf. In anticipation of a dozen more articles on the card’s power, and to avoid fitting the Reddit madlibs template, I’m staying away from these theoretical and high-level surveys of Fatal Push’s potency. Instead, I’m taking a deeper, data-driven dive into Fatal Push’s contextual relevance in Modern.
Today’s article continues the Lightning Bolt index tradition I wrote about almost two years ago. In that article, I used metagame prevalences and deck compositions to quantify Bolt’s effectiveness in any given period of time. This article will do the same for Modern’s newest removal icon through the introduction of the “Push Index.” Think of it as the quantified measure behind Jordan’s “Push Test” discussed last Friday. By analyzing the top decks in current Modern and cataloging their creature rosters, I’ll suggest an objective and quantifiable tool to determine if Fatal Push is pure heat or nothing but hype.
Creature Hits and Creature Misses
We already know why Push is Wizards’ glittering holiday prize for our format. It’s a one-drop removal spell in a color that has conspicuously lacked quality one-drop removal. Its revolt trigger, a real hoop in Standard, is a fetchland formality in Modern. Furthermore, Modern is dominated by low-cost creatures that die to even an unrevolted Push, with almost everything else getting Pushed overboard after revolt is met. It even kills creature lands, an honor not even venerable Abrupt Decay boasts. You can check out Ryan’s, PVDDR’s, Jordan’s, or Sam’s articles for more content on these and other Push merits. Here, I want to shift our attention away from Push itself and over to Push’s targets.
Unlike Path to Exile, both the old Bolt and the new Push are conditional removal spells that can’t dispatch everything. Their strength in Modern is directly related to how many creatures they can reliably remove. Consider the iconic Lightning Bolt. For Bolt-slingers everywhere, the card is at its strongest in a format where 100% of the decks are playing creatures with exclusively 1, 2, or 3 toughness, and no cards to boost their toughness. Also, no Slippery Bogles. Bogles need not apply to this article. In that theoretical metagame, Bolt would be the optimal removal spell, killing every creature with no drawback for only one mana. By a similar token, Bolt would be horrible in a format where every deck is playing only creatures with 4+ toughness. This spectrum was at the core of my Bolt Index article, which attempted to quantify Bolt utility based on creature weakness or resilience to Bolt.
Push operates in a similar fashion. In a format where every deck is playing only creatures that cost 0-2 mana, Fatal Push is the new Bolt. Especially if some of those 0-2 mana creatures are Bolt-resistant beefcakes named Tarmogoyf or Death’s Shadow. By contrast, a format that’s all Tron and Breach Valakut all the time is too heavy to Push around. At least it’s a flavor win: no one kicks, let alone pushes, Primeval Titan anywhere. As with the Bolt Index accounting for toughness, a so-called “Push Index” would need to account for mana cost. Specifically, for unconditional hits (i.e. creatures costing 0-2 mana), conditional hits (i.e. creatures costing 3-4 mana), and misses (i.e. creatures costing 5+ mana).
The following sections divide the common Tier 1 and Tier 2 creatures into unconditional hits, conditional hits, and misses based on their mana costs. This will give you a comprehensive list of all the relevant targets in (or out) of Push’s crosshairs. It’s also the foundation for the quantitative analysis later: we can’t talk about the metagame showings of various creatures without identifying those creatures first. All sections are nested under spoiler boxes to save you from too much scrolling. Also, hit me up in the comments if you think I missed a major entry.
If you’re already a Push believer, skip the “unconditional” and “conditional” sections. Note the distinction between unconditional and conditional hits. Even though Push removes both, it only removes the conditional hits if revolt is online. As this doesn’t happen reliably in every deck, we’ll need to weigh those conditional hits less than the unconditional hits, which is why we split them up now. Jump ahead to the “misses,” which will be the primary hurdles to Push’s future Modern success.
Unconditional hits: 54
I’ve omitted Sakura-Tribe Elder from the count because its purpose is to die and I’ve never seen a game lost to Steve beatdown. The same is not true of Insolent Neonate, which is why the menacing looter still makes the cut.
Conditional hits: 21
Kitchen Finks is conspicuously absent because its current metagame prevalence has taken a real dive. Abzan Company’s decline is a big player here, although Company is certainly not the only deck to use Finks.
I’ve included Etched Champion here even though Push can theoretically kill it while metalcraft is inactive. This is a relevant distinction because removal spells like Spatial Contortion and Wasteland Strangler can kill a Champion where Push cannot, so we should acknowledge the opportunity cost of playing Push over something that does hit Champs.
Hits and misses in the metagame context
The purpose of these lists isn’t to dazzle readers with their lengths. Yes, between these three lists you’ll find basically every mainboarded creature in top-tier Modern. True, Push unconditionally removes 56% of those creatures, and conditionally takes out 15% more. Sure, only 28% of Modern creatures effectively dodge Push on casting cost alone. Despite those impressive-sounding stats, none of this is a true measure of Push’s effectiveness because none of those creatures exist in a vacuum. They only matter insofar as a certain metagame is playing certain creatures.
If a disproportionate share of top-tier decks are playing creatures in the “misses” category, who cares if those creatures only make up 28% of Modern’s collective creature roster? They might make up 60% of creatures played in Tier 1 decks! This is why our Push analysis can’t stop at a list of what it can and cannot remove. We must extend that analysis to how those creatures appear in real metagames.
Calculating a Push Index
As with the old Bolt Index, the new Push Index is just one method of rating Push’s effectiveness in any given metagame. We’ll calculate it by listing the Tier 1 and Tier 2 decks, determining the mana-cost distribution of their creatures, finding what percentage of creatures in tier decks succumb to Aether Revolt’s new toy, and weighting the results based on metagame share. In the end, this will help us decide when Push is good, when Push is better in one metagame than in another, and how heavily to invest valuable removal slots in the card.
Defining the metagame
To calculate the Index, we first need a metagame. At the risk of stealing Jason’s metagame update thunder, which I’m told is roaring to a Modern Nexus site near you shortly, here are some rough numbers on Tier 1 and Tier 2 decks from December 2016. These stats might change when Jason does the final metagame update push, but it’s a solid starting point for the New Year and for our Index.
Preliminary December 2016 Metagame
|Deck Name||Metagame %|
|Death's Shadow Zoo||2.4%|
(Metagame side note: reactive decks didn’t really end the year on a good note…)
Now that we know what decks make up our current metagame, we need to go a level deeper and see what creatures make up those decks. The following spoiler boxes list the copies, counts, and costs of the average creatures in representative lists. To capture the essential top-tier creatures, I looked at the highest-finishing lists from high-attendance events. The results are recorded below. Feel free to skip ahead to the end if you don’t need the full deck-by-deck breakdown.
With 55 unique creatures and 205 total copies, Modern’s Tier 1 and Tier 2 decks cover a lot of creature real estate. This is the current environment Fatal Push is stepping into, and it will be our starting point for calculating Push’s initial Index.
Our next step is to add these creatures, and their mana-cost bins, to our metagame prevalence table. We’ll add four columns: one for the total count of creatures in a given deck, and one each for the number of unconditional hits (0-2 CMC), conditional hits (3-4 CMC), and misses (5+ CMC) in each deck. This will establish the creature landscape across Modern’s current top-tier.
Creature Distribution in the 12/2016 Metagame
|Deck Name||Metagame %||Total Creatures||Unconditional Hits|
|Death's Shadow Zoo||2.4%||15||15||0||0|
Even before “scoring” this table with any formal measure, we can eyeball its vulnerability to Fatal Push. With 205 creatures played across the top-tier decks, a full 148 (72%) are unconditional Push hits. Another 15% are conditional, with only 13% falling in the miss category. From a macro-perspective, that’s good news for Push, and we might be tempted to stop there by defining the Push Index as the percentage of creatures in a format which die to the spell (72% or, adding the conditionals, 87%).
Push Index complications
Unfortunately, this misses two key elements of an effective Push Index. First, it’s an inaccurate measure of how many creatures Push actually hits in a metagame. Those 205 creatures aren’t distributed between 12 decks evenly split with 8.3% shares. Their host decks have shares ranging from 9.4% to 2%. Imagine a world where the 2% deck (pretend it’s Gruul Zoo) played around 40 creatures and the 9.4% deck (pretend it’s Jeskai Control) played about five. We’d obviously need to weight the 2% deck less than the 9.4% deck. This means we’ll need to apply some metagame share-based weighting to these creature counts. Bookmark this: we’ll call it the Absolute Push Index later in the article.
Problem number two with this simplified Push Index: it doesn’t help us compare Push effectiveness between metagames with more or fewer total creatures throughout the top-tiers. Consider two metagames. In one, it’s an aggro-heavy format with 10 top-tier decks, each playing 40 creatures with CMC 0-2. In the second, it’s a control-heavy format also with 10 top-tier decks but eight of them are totally creatureless. The remaining two decks are aggro holdouts still running 40 creatures with CMC 0-2. Our rough Push Index would give both metagames a 100% unconditional hit rate, which isn’t quite accurate: Fatal Push doesn’t do anything against 80% of the decks in that second metagame! Put another bookmark here: this will become our Relative Push Index later on.
To address these problems and get our Absolute/Relative Index calculations, we’ll need to dig deeper than just adding the percentages on our table above.
Applying metagame weights
Our first job is to convert the count of creatures in each bin (conditional hits, unconditional hits, misses) into a weighted score. We’ll do this by expressing each count as a percentage of the total creatures in its deck, and then multiplying that percentage by the deck’s metagame share. For example, Infect has 17 creatures, all of which are unconditional hits. We convert that count of 17 to a score of 1 (i.e. 100% of Infect’s 17 total creatures are unconditional hits). We then multiply that score by the deck’s share to get its weighted score (i.e. we multiply Infect’s score of 1 by its share of 9.4% to get a weighted score of .094). To keep the numbers manageable, we’ll also multiply that final number by 100 so we’re not stuck in small decimals.
The table below applies this formula to all our decks above, calculating percentages in columns 3, 4, and 5, and then calculating the weighted score in columns 6, 7, and 8.
Push Index Creature Percentages and Weighted Scores
|Deck Name||Metagame %||Unc. Hits|
% of Total
% of Total
% of Total
|Death's Shadow Zoo||2.4%||100.0%||0.0%||0.0%||2.4||0.0||0.0|
As a final calculation step, we need to apply an additional modifier to the unconditional hits category. Even though conditional Fatal Push hits should certainly count towards Push’s overall effectiveness, we shouldn’t weigh them as heavily as an unconditional hit: you don’t need any setup to kill a Tarmogoyf. You need at least a fetchland crack to remove Thought-Knot Seer. This begs the question: in what percentage of turns or games can you reliably trigger revolt and use Fatal Push to remove a 3-4 cost creature?
As a rough measure of revolt consistency, I looked at the average number of fetchlands played in black-based, top-finishing lists of top-tier decks: 8.79, rounded to 9 (because you can’t really have fractions of cards). I then used the hypergeometric distribution to determine the chances of having at least two fetchlands by turn three, which is often the earliest you need to kill a CMC 3-4 creature. That chance is about 41%. This is the multiplier we’ll apply to the conditional hits to weigh them less heavily than the unconditional ones: in essence, a conditional hit becomes an unconditional hit by turn three in about 41% of games. In reality, it’s probably a bit more if we include hands with two non-fetches and one fetch, dead creatures we use to trigger revolt, and games where you don’t need to remove the creature until after turn three. But 41% is a good low-end estimate for the share of games where you can voluntarily trigger revolt and kill a 3-4 mana creature.
Push Index Final Weighted Scores
|Death's Shadow Zoo||2.4||0.0||0.0|
If you think the 41% chance under- or overestimates the scenarios in which you can voluntarily trigger revolt, feel free to use a different multiplier for the unconditional hits column! Also, let me know in the comments what alternative you used.
From scores to a finalized Index
Scores? Check. Weights? Check. Now, let’s set up the final Index.
Let’s start with the bookmarked Absolute Push Index from earlier (the measure of Push effectiveness in a vacuum). We’ll express this as the sum of our unconditional and conditional weighted scores, divided by the total possible score in the metagame. In this case, the total possible score is 61.2, i.e. the sum of all the top-tier metagame shares. As another example, if the top-tier decks collectively made up 65% of the format, the total possible score would be 65.
Adding up our scores we get:
Absolute Push Index (score 1): 48.2 (relative to the entire metagame)
Absolute Push Index (score 2): 77.6 (relative to the top-tier metagame)
Stated another way, in a metagame where these top-tier decks make up 61.2% of the format, Push is going to unconditionally kill creatures in at least 48.2% of your total matchups. More importantly, it will unconditionally remove creatures in 77.6% of your top-tier matchups (48.2/61.2).
This is an excellent hit-rate. For reference, the hit-rate in a metagame where every single top-tier deck played exclusively 0-2 mana creatures would be 100%, which is a mere 23% higher than our current rating of 77.6%. A metagame where every single top-tier deck played exclusively 3-4 mana creatures would only have a 41% hit rating (or whatever other multiplier you used for the conditional hits). Our current metagame has a Push Index almost twice that. Push sure looks hot going into 2017.
But what about our Relative Push Index? You’ll remember from our earlier bookmarks that the Absolute Push Index is great at telling us how effective Push is in a vacuum, but isn’t a strong measure when comparing two distinct metagames. Enter the Relative Push Index. This will be expressed as the Absolute Push Index multiplied by the total number of creatures in the top-tier metagame. For this particular sample, that was 205 total creatures.
Relative Push Index: 161.4
This Index will help us compare Push’s effectiveness in any given metagame to Push’s historic effectiveness in any other metagame. Here’s an example.
Imagine it’s September 2017 and you are deciding how many Pushes to run in your Abzan deck. You look back to February 2017 and you see the Relative Push Index was, say, 165. You remember that Push was spectacular in February 2017, and you ran the full playset. Next, you look back to June 2017 where the Relative Push Index (which you would calculate) was only 135. You remember Push was pretty bad in June because Bant Eldrazi, Grixis Control/Midrange, and Gx Tron were everywhere; indeed, you shaved your Push count to just 1 in the maindeck and 1 in the board. Finally, you calculate the Relative Push Index for September 2017 and find that it’s a solid 160. Based on this, you would probably include as many Pushes as you used during the February 2017 metagame when the Index was at 165. Or maybe include 3 in the main and 1 in the board, if you’re worried about the slight drop. Either way, this would help you guide your Push decision when building decks.
Using and adapting the Push Index
Ultimately, you can use the Push Index as a data-driven tool to determine when Push is good and how many you should (or shouldn’t) run. I suggest you calculate the Push Index in every metagame update for 2017, comparing those numbers to your own experience with Push and using that data to make informed deckbuilding and sideboarding decisions. I suspect that a 161.4 Relative Push Index means that most decks that want Push should be running 2-3 copies in the maindeck, but I could easily see that number pushing up to 3-4. I’ll keep checking in on the Index as decks adapt to the new card (more Bant Eldrazi and Gx Tron?) and players run more creatures to dodge the spell (Tasigurs and Anglers are looking hotter than ever).
Banlist update thoughts
Our January 2017 Modern banlist update came a week early and it’s a stunner: Golgari Grave-Troll and Gitaxian Probe are banned in Modern. To say nothing of Standard’s upheaval (farewell, Emrakul, the Promised End, Smuggler’s Copter, and Reflector Mage), this update promises to make a sizable impact on competitive Magic. My original draft of this article had some banlist thoughts including a Troll ban prediction, but I’m changing gears to offer quick roundup thoughts on Modern’s latest seismic event.
- Good riddance, Golgari Grave-Troll
Dredge is not a turn four rule violator. It also doesn’t occupy an objectively oppressive share. Unfortunately, Dredge has reduced Modern diversity by forcing players to run 4-5 or even more graveyard hate cards in their sideboards. In turn, this has had the disgusting effect of pushing out all other graveyard decks from the top-tier. Remember Abzan Company? Living End? Grishoalbrand? All these decks are in Tier 3 or lower, and that’s a major loss for format diversity. Given these circumstances, Troll’s banning was unsurprising and warranted. I applaud Wizards for depowering the deck without killing it: Dredge should remain at least a Tier 2 deck, with intermittent Tier 1 appearances. Meanwhile, other graveyard decks should return to the format, and the Dredge sideboard subgame should diminish. This ban is an overall win for metagame diversity and Modern.
- Mixed outcomes after Gitaxian Probe‘s ban
I’m less happy about Probe’s demise. On the one hand, it’s a slick limited-ends ban that depowers a variety of fast, linear, non-interactive decks without gutting any of them: see Infect, Death’s Shadow Zoo, and UR Prowess in the top tiers alone. Looking only at that impact, it should slow down Modern while allowing those faster decks to remain somewhat viable. On the other hand, Probe’s removal is a big hit for reactive, fair, interactive Delver of Secrets decks, which I imagine is an undesired but foreseen consequence of the ban. Perhaps Wizards is banking on Fatal Push picking up the Delver slack? More likely, they felt it was a necessary cost to push back top-tier fast decks. As a whole, this is a clear issue of addressing the symptom of a disease, not its cause. The cause is a lack of generic answers and policing cards/strategies in Modern. The symptom is fast, top-tier linear decks which fair decks can’t restrain. As such, the Probe ban will go down as yet another victim of Wizards’ inability to print viable Modern answers and interaction. Push is a good step in the right direction. This kind of ban, however, is not—particularly with all the ensuing ban mania it will promote.
- Looking ahead, unban Preordain or Stoneforge Mystic by April 2017
As hinted at above, Modern needs built-in answers for fast, linear, non-interactive decks. Those answers cannot be bans. This format cannot keep enduring bans; it is becoming increasingly difficult to tell new players it’s a safe investment when we average one to two bans per year. Wizards must introduce new cards to empower internal answers. Yes, Fatal Push is an excellent step in this direction), but we need more Pushes, as well as reprints (Prohibit, please?) and unbans. The only two unbans which address this standing need are Preordain, for interactive decks to dig for answers, and Stoneforge Mystic, to deploy a Batterskull roadblock against decks that refuse interaction. Jordan has already written extensively on the cantrip, and I agree with most of what he said. I’ll also emphasize that Preordain was initially banned because it powered up blue-red combo decks, all of which are now gone. Indeed, the card is safer still with Probe out of the format. As for Mystic, David has been testing the card for months and it seems much safer than many have alleged. Wizards should take a look at both cards and unban one by April.
I’m giving these thoughts at the tail-end of an article, taking the risk of provoking controversy without adequate room to explain these ideas. I’m somewhat bummed that the banlist update came early and will likely divert attention from my Push Index analysis to the ever-polarizing topic of the banlist, but that’s Modern as you know it. Welcome back, I guess?
Thanks for reading and, again, it’s good to be back writing about Magic’s best format. Although I won’t write as regularly as in 2015 or most of 2016, I’ll do my best to keep some data-driven articles coming as we enjoy what looks to be an exciting 2017 for Modern. Hit me up in the comments with Push questions and ideas, or the inevitable ban and unban conversation, and I’ll talk to you all soon!