Big Data from Barcelona: Modern Metagame and Archetype Performance

This article is about the most recent Modern Grand Prix, not about the Mythic Championship. There are three good reasons to focus on the less prestigious of the two event. One is that all the information for Mythic Championship IV is already out there. Unlike in previous years, Wizards themselves have been very frank about revealing exactly how popular and how successful the various archetypes were at the tournament. For Grand Prix Barcelona, in contrast, there’s exclusive, never-before-published data to look at.

The second reason is comparability. Say you’re wondering what to expect from the main event at the upcoming MagicFest in Minneapolis. The Mythic Championship field composition and results only provide marginal insight here. GPs feature a different crowd, a different metagame, and even a different set of rules. For instance, it is possible to maindeck Leyline of the Void when you know ahead of mulligans what your opponent is playing, and it isn’t when you don’t. At the same time, the GP was very much infused with MC technology. Three quarters of Grand Prix Barcelona’s players updated their deck submission on Friday evening or later.

Finally, there’s a reason why “big data” has become a buzzword in recent years. It is better to draw on larger samples when you’re looking for connections—for example between deck choice and success—that include an element of uncertainty. Good fortune and bad luck only evens out over the long run. You might want to argue for quality over quantity when it comes to game results. But if we assume a random distribution of mistakes, they too become just another element of chance that gets filtered out in sufficiently large samples. With more than three times as many players who completed about four times as many Modern matches, Barcelona’s GP thus can offer more reliable data than Barcelona’s MC.

Grand Prix Barcelona’s Modern Metagame

  • 161 W/U Control (10.6%)
  • 157 Izzet Phoenix (10.4%)
  • 143 Eldrazi Tron (9.5%)

Blue-White Control had won the previous Modern GP, where Eldrazi Tron had had the best overall record, and Izzet Phoenix had been among the top decks for months prior. These three led the charge into Day 1, which leaves the question, where was Hogaak?

  • 121 Jund (8.0%)
  • 120 Hogaakvine (7.9%)
  • 111 Humans (7.3%)

Regarding Hogaak’s continued apex-level power, the cat had been well out of the bag by Friday, but player numbers on Saturday morning didn’t quite reach a level to match. Expect the metagame share to grow, but don’t underestimate the role Hogaak already played at this event. In absolute numbers, the GP had 22 more Hogaak players than the MC. The deck also had more total followers at this GP than the old Bridgevine deck had assembled at Grand Prix Dallas-Fort Worth before the banning of Bridge from Below.

  • 77 Burn (5.1%)
  • 64 Tron (4.2%)

Rounding out the Top 8 of most popular archetypes were Burn and Green Tron. Paradoxically, the story here is both about how little the Modern GP metagame changes over time as well as about how much it changed already. These two had been among the three most-represented decks at the majority of GPs for the better part of a year. On one hand, it’s impressive how far their metagame share has dipped. On the other hand, you still can’t leave home without a plan to beat either.

  • 40 Red Prowess (2.6%)
  • 38 Dredge (2.5%)
  • 35 Thopter Foundry (2.3%)

Note that 33 of 38 Dredge decks incorporated Hogaak, Arisen Necropolis as well.

Hogaak, Arisen Necropolis

Also note how much more fractured this metagame looks compared to the Mythic Championship. At the latter, 10 decks accounted for 82% of all players, whereas the top 10 decks only made up 68% of Barcelona’s Grand Prix field. This difference exists across all formats, but Modern remains special. Nowhere else can you find this many archetypes that see play.

  • 25 Devoted Vizier (1.7%)
  • 24 Affinity (1.6%)
  • 23 R/G Valakut (1.5%)
  • 20 Infect (1.3%)
  • 20 Spirits (1.3%)
  • 19 Hatebears (1.3%)
  • 18 Scales (1.2%)
  • 16 Merfolk (1.1%)
  • 15 Amulet (1.0%)
  • 15 Neobrand (1.0%)
  • 14 Bogles (0.9%)
  • 13 Jeskai (0.9%)
  • 12 Red Prison (0.8%)
  • 11 Rock (0.7%)
  • 9 Blue Valakut (0.6%)
  • 9 Esper Shadow (0.6%)
  • 8 Ad Nauseam (0.5%)
  • 8 Esper Control (0.5%)
  • 8 Goblins (0.5%)
  • 8 Mardu Pyromancers (0.5%)
  • 7 Mill (0.5%)
  • 7 Ponza (0.5%)
  • 7 Storm (0.5%)
  • 6 Grixis Control (0.4%)
  • 6 Snow Control (0.4%)
  • 5 Blue Tron (0.3%)
  • 5 Elves (0.3%)
  • 5 U/R Thing (0.3%)
  • 5 X Rack (0.3%)
  • 4 U/R Breach (0.3%)
  • 4 Aggro Eldrazi (0.3%)
  • 4 Grixis Shadow (0.3%)
  • 4 Shadow Zoo (0.3%)
  • 4 Taking Turns (0.3%)
  • 4 Valuetown (0.3%)

Performance of Major Archetypes

Usually, this is where I would post win rates. But I’m no longer sure that win rate constitutes a good measure of success. Winning five of eight matches equals a fantastic win rate of 62.5%. Only the most dominant decks—think Scapeshift at Grand Prix Denver—ever maintain such a percentage over a large number of matches. On the flip side, there’s no doubt about what it means to be eliminated with a record of 5-3. It doesn’t mean to be successful.

Magic tournaments themselves measure performance in points, and that may be the best yardstick after all. Positive outliers affect overall win rate and average points per player equally, whether you count the outlier as 12-3 or as 36 points. The advantage of the latter metric lies in its ability not to get distracted by Day 1 records that fall short of securing a Day 2 berth. Qualifying for the second day is an important goalpost, and a good measure of success should account for that.

The average number of points per Grand Prix Barcelona player, across all archetypes, was 13.5. The pilots of the most successful archetype won almost two matches more on average:

  • Hogaakvine players averaged 19.4 match points

There’s no big surprise here, although there is a big difference in points. No other archetype with a double-digit number of players came close to such a result.

  • Hatebears players averaged 16.1 match points
  • Thopter Foundry players averaged 16.0 match points
  • Humans players averaged 15.1 match points
  • Amulet players averaged 15.1 match points

The biggest surprise here concerns white Hatebears, most of whom featured a splash of colorlessness. The deck earned the second highest point average. Seven of 19 players qualified for the second day, and the following version finished best:

Axel Signargout, 76th Place (11-4)

1 Blast Zone
1 Eiganjo Castle
4 Eldrazi Temple
1 Flagstones of Trokair
4 Ghost Quarter
1 Horizon Canopy
3 Shefet Dunes
1 Silent Clearing
7 Snow-Covered Plains
1 Blade Splicer
4 Eldrazi Displacer
4 Flickerwisp
4 Giver of Runes
4 Leonin Arbiter
2 Restoration Angel
4 Thalia, Guardian of Thraben
4 Thought-Knot Seer
2 Wall of Omens
4 AEther Vial
4 Path to Exile

2 Gideon, Ally of Zendikar
1 Grafdigger's Cage
1 Leonin Relic-Warder
2 Oust
1 Pithing Needle
3 Rest in Peace
2 Sorcerous Spyglass
2 Stony Silence
1 Winds of Abandon

Many others splashed for Tidehollow Sculler and Wasteland Strangler. Some ran either Soulherder and Spell Queller or Scavenging Ooze and Renegade Rallier. But they all shared the base of Giver of Runes, Leonin Arbiter, Thalia, Flickerwisp, Aether Vial, Path to Exile and a full playset of Ghost Quarter.

  • Red Prowess players averaged 14.9 match points
  • Eldrazi Tron players averaged 14.7 match points
  • Burn players averaged 14.6 match points
  • R/G Valakut players averaged 14.5 match points
  • Tron players averaged 13.9 match points

The decks above all performed somewhat above average, while the subsequent all performed a little subpar. Most notably, Jund, despite putting two players into the finals, didn’t do well overall. Only two other Jund decks finished inside the Top 100.

  • Izzet Phoenix players averaged 13.2 match points
  • Jund players averaged 12.6 match points
  • Dredge players averaged 12.6 match points
  • Scales players averaged 12.4 match points
  • W/U Control players averaged 11.7 match points

The most represented archetype, U/W Control, didn’t do well either. Only decks that were a fraction as popular performed worse. To wit:

  • Affinity players averaged 11.6 match points
  • Infect players averaged 11.4 match points
  • Spirits players averaged 10.9 match points
  • Merfolk players averaged 10.9 match points
  • Devoted Vizier players averaged 10.4 match points
  • Neobrand players averaged 9.9 match points

Performance of Minor Archetypes

The fewer players enter a tournament with a certain deck the fewer outliers does it take to sway the average. As such, much of the following might be meaningless fluctuation. Specifically, it probably was a fluke that the four players who picked Blue-Red Through the Breach averaged more match points than the players on Hogaak.

Just in case, here’s the highest-finishing U/R Breach list:

Nicolas Baldy, 194th Place (9-6)

2 Desolate Lighthouse
1 Flooded Strand
7 Island (335)
1 Misty Rainforest
1 Mountain (343)
1 Polluted Delta
1 Prismatic Vista
4 Scalding Tarn
2 Steam Vents
2 Sulfur Falls
4 Emrakul, the Aeons Torn
3 Snapcaster Mage
1 Jace, the Mind Sculptor
2 Blood Moon
3 Cryptic Command
1 Flame Slash
2 Force of Negation
1 Izzet Charm
4 Lightning Bolt
1 Magmatic Sinkhole
4 Panoptic Mirror
4 Remand
3 Serum Visions
1 Spell Snare
4 Through the Breach

1 Abrade
3 Anger of the Gods
2 Ceremonious Rejection
1 Dispel
3 Dragon's Claw
1 Engineered Explosives
2 Force of Negation
2 Surgical Extraction

The other three all looked very similar. Aside from Magmatic Sinkhole and Force of Negation, there’s nothing new here, and previous results never gave any indication that the archetype was on the upswing. The corresponding record of 26-19 isn’t statistically significant either. Of course, that doesn’t prove it was a fluke. It seems unlikely that opposing countermagic was what was holding the deck back, but maybe it just took Force to force through Through the Breach for a breakthrough.

Here are all the decks with a playerbase smaller than 1% of the field whose positive performance looks dubious because of it:

  • U/R Breach players averaged 19.5 match points
  • Esper Control players averaged 18.4 match points
  • Esper Shadow players averaged 18.1 match points
  • Rock players averaged 16.6 match points
  • Mardu Pyromancers players averaged 15.8 match points
  • Blue Valakut players averaged 15.3 match points
  • Snow Control players averaged 13.7 match points

It’s much easier to believe in the underperformance of fringe archetypes. After all, they must be fringe archetypes for a reason. Indeed, way more of them did badly than did well.

  • Blue Tron players averaged 13.2 match points
  • Storm players averaged 12.4 match points
  • Aggro Eldrazi players averaged 12.3 match points
  • Goblins players averaged 11.6 match points
  • Mill players averaged 11.3 match points
  • U/R Thing players averaged 10.2 match points
  • Bogles players averaged 10.1 match points

If you saw Grand Prix Barcelona’s Top 16 lists, you might not have expected to find Bogles this low. Mats Törnros—who just earned his second Grand Prix Top 8 in June—took the deck to ninth place and 39 points. But the other 13 Bogles players in the tournament all failed to qualify for the second day. None of them even managed to win five matches.

This, by the way, is the advantage of average points over win rate in action. Bogles’s point totals correspond to a positive overall record, although the deck had a very negative weekend.

  • Ad Nauseam players averaged 9.8 match points
  • Ponza players averaged 9.4 match points
  • Jeskai players averaged 9.1 match points
  • Grixis Shadow players averaged 9.0 match points
  • Shadow Zoo players averaged 9.0 match points
  • Valuetown players averaged 8.3 match points
  • Red Prison players averaged 7.8 match points
  • Taking Turns players averaged 7.3 match points
  • Elves players averaged 7.2 match points
  • Grixis Control players averaged 7.2 match points
  • X Rack players averaged 4.2 match points

Last and least, it’s almost poetically appropriate for The Rack and company to bring up the very back of the rear. It’s a strategy that strives to make others discard in a world where many actively want to dump cards in their graveyard. This is no accident waiting to happen. It’s more akin to shooting yourself in the knee in an attempt to hit the foot of the person behind you.


GP Barcelona Modern Metagame


Scroll to Top