The Standard(T2) Metagame in March

Written by derflippi on March 25, 2010




Players always want to figure out a good way to win. One important choice on the way there is what deck you chose. The deckchoice however cannot be done in a vacuum. The optimal deckchoice depends on the metagame. It's a bad idea to chose Scissors, when everyone else choses Rock.

I don't claim myself to be a great player; I make "play mistakes" like incorrect blocks or attacks all the time, but I believe I have a good understanding of greater Magic strategy.

With GP Brussels within reach, Standard (T2) strategy is the big topic at the moment. For this article, I collected the data of over 3000 decks played on magic-league this month in March. When useful, I put the extracts of the table firectly to this article. However, If you wish to look at the comprehensive table, you can download it here. This is especially useful to verify or re-evaluate my following points.

First, it will be interesting to see breakdown of the metagame:

580    19,1    Jund
317    10,44    UW Control
180    5,93    Naya Mystic
176    5,8    Bant
161    5,3    Vampires
154    5,07    MonoRed
153    5,04    Rb Bligthning
127    4,18    MW Aggro
115    3,79    Open The Vaults Control
113    3,72    Eldrazi Green
79    2,6    UWR Control
76    2,5    Boros
71    2,34    GW Rock
70    2,3    Cruel Control
66    2,17    Valakut Ramp
64    2,11    Allys
62    2,04    Summoning Trap
       
509    16,6    Other

It is not surprising to see Jund on Top. In comparison to my last Standard (T2) metagame analysis, it became even more popular.

Metagame in December:
14,93     Jund
12,92     Mono Red
7,39     Gwb Aggro
4,67     Naya
4,41     Boros Bushwhacker
4,38     Summoning Trap


MonoRed did not really lose players. It's just that I now treated Monored and Blightning Burn as different decks. GWb Aggro almost completely disappeared. One problem could be that it plays Fetchlands, Lotus Cobra, Maelstrom Pulse and Baneslayer Angels. Due to its budget restrictions, it did not see much play in DCI tournaments, so was nowhere to be seen in winning decklists there. When a deck doesn't prove itself anymore, people stop playing it. Although it was the best deck in December, people now don't play it. Instead, they use their Knight of Reliquarys in a new deck: Naya Mystic. The Zoodeck uses the most efficient creatures of Naya colors, and its namesake Stoneforge Mystic, among with a small toolbox of Equipments. Here's how Tom Ross played it at PT San Diego:

Boss Naya played by Tom Ross
Main Deck Sideboard
1 Scute Mob
2 Birds of Paradise
4 Wild Nacatl
4 Noble Hierarch
2 Stoneforge Mystic
4 Knight of the Reliquary
1 Dauntless Escort
4 Bloodbraid Elf
4 Ranger of Eos
3 Lightning Bolt
1 Path to Exile
1 Basilisk Collar
2 Oblivion Ring
1 Behemoth Sledge
2 Ajani Vengeant
3 Misty Rainforest
4 Arid Mesa
2 Stirring Wildwood
1 Raging Ravine
2 Terramorphic Expanse
1 Tectonic Edge
1 Rootbound Crag
5 Forest
1 Sejiri Steppe
2 Plains
2 Mountain
1 Goblin Guide
4 Cunning Sparkmage
1 Basilisk Collar
1 Behemoth Sledge
1 Stoneforge Mystic
2 Qasali Pridemage
3 Dauntless Escort
2 Manabarbs


Another completely new deck is the UWb Open the Filigree deck. It is a deck all around Open the Vaults, with several Esper-Cycle cards and Sphinx of Lost Truth:

Open the Filigree played by ykpon
Main Deck Sideboard
1 Swamp
3 Plains
3 Marsh Flats
2 Kabira Crossroads
5 Island
4 Glacial Fortress
2 Celestial Colonnade
4 Sphinx of Lost Truths
1 Sharuum the Hegemon
4 Glassdust Hulk
4 Filigree Angel
4 Architects of Will
4 Spreading Seas
4 Open the Vaults
1 Mistvein Borderpost
2 Journey to Nowhere
1 Jace, the Mind Sculptor
3 Courier's Capsule
2 Fieldmist Borderpost
3 Day of Judgment
3 Oblivion Ring
1 Day of Judgment
4 Negate
1 Sanguine Bond
1 Telemin Performance
4 Celestial Purge
4 Kor Firewalker



A general note about the current metagame is that the top decks can't be said in one sentence anymore. It's not just Jund, MonoRed and GWb Junk. Today UW Control, Bant, Vampires and Naya Mystic can also be considered as top decks, at least by a look at how often these decks are played.

So, Jund is the most played deck. This implies that decks beating Jund have better chances of winning a tournament than decks losing to it. One of these decks that are said to beat Jund is the almost monored Blightning Burn deck:

Rb Blightning played by Ding Yuan Leong
Main Deck Sideboard
4 Arid Mesa
4 Dragonskull Summit
4 Lavaclaw Reaches
8 Mountain
4 Scalding Tarn
4 Ball Lightning
4 Goblin Guide
4 Hell's Thunder
4 Hellspark Elemental
4 Blightning
3 Burst Lightning
3 Earthquake
4 Lightning Bolt
2 Quenchable Fire
4 Searing Blaze
4 Deathmark
2 Doom Blade
4 Goblin Ruinblaster
3 Manabarbs
2 Quenchable Fire


But I'll not keep the analysis with just numbers showing how often a deck was played. To extract useful information out of the datatable, another important value has to be looked at: How often a deck won. With the data on relative won tournaments and relative proportion in the complete field, I calculate a so called conversion rate for the decks. A comprehensive explanation of why this conversion rate is meaningful, and how it's formed can be found in my December analysis.
Short, the conversion rate tells if a deck has good or bad matchups against the metagame on average. A deck with a conversion rate of exactly "One" won exactly 50% of its matches.

Looking at the Top12 played decks (more than 2,5% of the metagame), here's their conversion rates:

1,68    Boros
1,4    Jund
1,22    Vampires
1,21    Bant
1,06    Eldrazi Green
1    Naya Mystic
0,97    UWR Control
0,89    MonoRed
0,84    Rb Bligthning
0,67    UW Control
0,67    MW Aggro
0,67    Open The Filigree


Direct comparisons however are difficult. That is because a deck with smaller sample size shows a  high variance in its conversion rate. Additionally, a deck that is played by more than 12,5% (tournaments are usually 8-man queues) is more likely to be treated incorrect. Theirs conversion rate appears to be worse than it is, normed with a quotient to nullify mirrormatches. However, what appears as a problem affects Jund only, and as I'll further explain, does not make a statement about the current metagame more difficult. Surprisingly, it will even stress my point.

In December, Jund had a conversion rate of 0,94. I already analyzed there why this does not say Jund was a bad deck. Since Jund was played alot, even if Jund wins a tournament, several decks of the same category don't win. More mirrormatches decrease the conversion rate. That's why Jund's conversion rate was below 1.

In March, the amount of UW Control decks being played is roughly double the amount of Bant decks. At the same time however, the two different decks won the same number of tournaments. The only conclusion is that UW Control must be much worse than Bant. Winning less than 50% matches on average is also true for the Open the Filigree deck, Mono White Aggro decks and the red ultra aggressive deckbuilding approaches.

Naya Mystic, Eldrazi Green and UWR Control play on the same level. They're not bad decks, yet don't have dominating winning percentages at the moment.

The three best decks right now are Jund, Vampires and Bant. I do not want to make a definite statement about Boros yet, since its sample size is relatively small. It's possible that variance creates the 1,68 rate of Boros, or that it's not played by enough players to reflect "the average player". Being played by only 2,5% is not a big enough quantity to make me judge Boros.

Vampires, played by a constant number of followers is simply good deck on average.

The strength of Bant can be valued higher than Vampires however. While the latter had constant appearances at tournaments, Bant became much more popular only recently. Therefore, its numbers describe the current situation better. Played only rarely at start of this Month, it became the third most popular deck in the end. This didn't hurt its conversion rate, so the metagame matrix does not make it a good deck because of some random factor.

In December, I said Jund is a good deck. Now, I say it's the best deck.

Many writers said that in the past. But finally, I am able to actually prove it; with empiric data.
Jund has has the highest quantity in the metagame. This means the statistic informative value is not blurred by variance too much. We know a deck that popular has a smaller nominal than real conversion rate. The nominal rate of Jund is already at 1,4. But the real value is even higher!  The step from nominal to real values makes Jund better, while other decks stay equally good. Going by the nominal values, Jund is already the best deck, so it's not needed to get fixed. So the nominal values will suffice for now, so I won't go more into detail on the real rates.

One last tip the statistic provides is a hint to a correlation of Bant being played and Junds winning percentages. The more players played Bant, the less did Jund decks win relatively. (Instead, Bant won). Unfortunately, the sample size to prove that interesting relationship is not big enough yet. I'm sure it's possible to get a better idea on that after Brussels.

The way I analyzed the metagame doesn't tell anything about exact cardchoices, instead it's all about strategy. Just like I lost no word on how to mulligan in a specific matchup, I cannot give an answer for the number of Putrid Leeches  one should play in his Jund deck.

I trust in the conversion rates. Since I looked at over 3000 decks, they reflect the average outcome for a deck. A high expected average outcome suggests the deck is good. Therefore, I suggest to play Bant or Jund in Brussels, as those are clearly the best decks in the current metagame. If I had to put a bet, I'd say Bant wins Brussels.

Here's two recent example decklists:

Bant played by Ludz
Main Deck Sideboard
6 Forest
2 Island
1 Plains
2 Sunpetal Grove
4 Celestial Colonnade
1 Sejiri Steppe
3 Stirring Wildwood
4 Misty Rainforest
4 Verdant Catacombs
4 Birds of Paradise
4 Knight of the Reliquary
4 Noble Hierarch
2 Thornling
4 Baneslayer Angel
4 Rhox War Monk
3 Rafiq of the Many
4 Lotus Cobra
1 Sphinx of Jwar Isle
3 Finest Hour
1 Sphinx of Jwar Isle
3 Mind Control
3 Negate
3 Bant Charm
3 Jace, the Mind Sculptor
2 Harm's Way


-

Jund
Main Deck Sideboard
4 Dragonskull Summit
3 Forest
3 Mountain
4 Raging Ravine
4 Savage Lands
4 Swamp
4 Verdant Catacombs
4 Bloodbraid Elf
3 Broodmate Dragon
4 Putrid Leech
4 Sprouting Thrinax
2 Siege-Gang Commander
2 Bituminous Blast
4 Blightning
2 Garruk Wildspeaker
4 Lightning Bolt
3 Maelstrom Pulse
2 Terminate
3 Deathmark
3 Goblin Ruinblaster
3 Great Sable Stag
3 Malakir Bloodwitch
3 Mind Rot


I'm always happy to work with feedback I get on my articles. Send an email to cmaflippi@magic-league.com with any suggestions or questions on my analysis.

Due to time constraints(army), I can't continue this articles series anymore. So if you're an skilled player, with some basic maths knowledge, why not share your experiences and write a strategy article? The articles deparment of magic-league.com is always looking for more authors. Contact Weedmonkey if you're interested.

Until next time, there will be a next time, become a better player, by playing on magic-league.com, the home of tomorrow's mtg-stars.

Back to Magic: the Gathering Articles

Comments:
by Ludz on 2010-03-25 08:54 CET

Nice article!

How do you think play skill and play testing effects these stats? I think that might be the reason why decks like UW control and Open The Filigree aren't doing any well, since the players who invented those decks and played them at the PT, they sure did playtest alot, because the decks seems harder to play. Compare to a deck like Mythic or Burn, and they are kind of straight forward.

Also, what's the Scapeshift doing there? ;) (just ctrl+F for scapeshift)

Anyway, nice article. Hope someone will continue this while you're away!


by derflippi on 2010-03-25 12:19 CET

The more players play a deck, the more realistic are the results for the deck. When 700 players play Jund, there will be bad players and good players.
However, when only 85 players played Boros, it's possible that it was played by more good than bad players. So I believe as long as sample size is big enough, playskill can be ignored.


by nopenopenope on 2010-03-25 13:01 CET

Burn in t2 is much more complicated that burn you know from past and Ext. The manacurve isn't ituitive and the sideboard is one of the most complicated. Well, all decks can be hard to masterize, but stop saying burn is easy to play.

For fun, take the Rb Blightning played by Ding Yuan Leong and sideboard it for Jund.


by darkgoyfidnt on 2010-03-25 13:30 CET

Your analysis in saying UW Control is worse than mythic bant is very, very wrong. The first problem is your sample data is magic-league.

UW Control is a deck that simply does better in the hands of pros. Noobs such as myself playing the deck without realizing all the nuances really drive this deck down more than any other.

Average playskill of players has a huge amount of relevance, regardless of sample size. The UW Control/Bant numbers would be much, much closer among pro players.


by derflippi on 2010-03-25 14:15 CET

The data reflects the average player. This makes it possible to say something about deck strengths empirically. Trying to argue with "it's just wrong" doesn't work here. Please prove your point.


by coboney on 2010-03-25 16:19 CET

I think what he's trying to say is that some decks require a higher level of skill and practice.

What this is, though is only 1 part of understanding the format. This gives us play data and bulk data about the metagame here on Magic-League, which can be used to help understand it better. We see that UW for example is having some trouble, and we have to reconcile that with what else we've seen.

Is UW a good deck? Is it a deck forced by 'pros' who really want to play blue? Is it a deck that is unforgiving of mistakes that hurts it more at lower levels?

Thats the type of questions you need to ask and think about to understand the format. Numbers are only one part, drawing conclusions from numbers is big.


by Djinn on 2010-03-25 18:58 CET

Great job, definitely helpful for the upcoming GP. I agree on the fact that playskills can be ignored when it comes to Jund or another deck that is heavily played and scores well.


by DonDiggy on 2010-03-26 07:34 CET

Nice Statistics, thx Flippi.

Imo you made a great job. But i think the Results have to be used with attention. Simply due to the fact that good players tend to play Jund or UW Control while almost every bad player gets his hands on Naya, WW and Vamps.

I think Naya will will win Brüssel, because its many options reward good play and punish bad play all the time. In the Hands of an awesome player its supposed to be the current showstopper in T2 (no bad matchup exept for UW maybe).


by tchiseen on 2010-03-27 11:03 CET

I think I might agree with DonDiggy that Naya has a chance of winning at the top flight.

Great article by the way. Keep up the good work.


by neosystems on 2010-03-28 18:35 CET

I'd just like to point out that Flippi is brilliant. :o

"Therefore, I suggest to play Bant or Jund in Brussels, as those are clearly the best decks in the current metagame. If I had to put a bet, I'd say Bant wins Brussels."

5 Jund, 2 Bant, 1 Naya

Jund vs. Bant in the finals (although it was Jund that won).


All content on this page may not be reproduced without written consent of Magic-League Directors.
Magic the Gathering is TM and copyright Wizards of the Coast, Inc, a subsidiary of Hasbro, Inc. All rights reserved.


Contact Us | Privacy Policy
Join Swagbucks!