fbpx

Innovations – A Karsten-Style Analysis of Extended


Grand Prix: Oakland!

Monday, January 18th – Extended PTQs are in full swing, and Patrick Chapin provides the latest metagame statistics! Channeling the great Frank Karsten, Patrick takes a detailed look at the tournament results so far, and creates a breakdown of the Winner’s Metagame. This is a fantastic article for anyone serious about cracking into Pro Tour play.

Frank Karsten was inducted into the Magic the Gathering Hall of Fame, last year. His three Pro Tour Top 8s (including Worlds 2008) and 292 Pro Points are obviously very fine accomplishments, but one of Frank’s largest contributions to the game, and the one for which he is most well known, is his writing.

Frank is known for his analytical mind and willingness to spend as much time as it takes to understand all that can be gleaned from tournament data. For years, his columns were among the most important to read to stay current with the metagame. One of the primary ways he did this was by calculating the “Winner’s Metagame.”

Frank cautioned against relying on who won too heavily, as well as relying too heavily on how many people showed up with a deck. The system he used for calculating his “Winner’s Metagame” was to compile all of the applicable tournament results he could, then award 6 points to every archetype for every time someone finished 1st or 2nd with it. Then he would award 4 points to every archetype for every 3rd or 4th place finish. Next he would add 3 points to every archetype for each 5th to 8th finish. This way, he would have information about what was making Top 8’s, but it would be weighted based on how well it did in the Top 8.

After tallying all these numbers, Frank would divide each archetype’s point total by the total number of points in the pool. This percentage would show how prevalent each archetype was at the winning tables, a.k.a. the “Winner’s Metagame” or popularity percentage.

Today, I will be using Frank’s method on the Extended information available from the first two weeks of PTQs. This is not an exhaustive list of every PTQ that took place, but these are all of the published results that I know of.

1/12 Magic Online PTQ
Big Zoo 1st, Fast Zoo 2nd
Faeries 3rd, Martyr 4th
Scapeshift 5th, Misc. Aggro 6th, Faeries 7th, Fast Zoo 8th

1/10 Magic Online PTQ
Thopter 1st, Burn 2nd
Scapeshift 3rd, Dredge 4th
Scapeshift 5th, Dark Depths 6th, Scapeshift 7th, Scapeshift 8th

1/10 Milan PTQ
AIR 1st, Rock 2nd
Thopter 3rd
Bant 5th, Death Cloud 6th, Scapeshift 7th, Dark Depths 8th

1/10 Rome PTQ
Scapeshift 1st, Thopter 2nd
Death Cloud 3rd, Thopter 4th
Hive Mind 5th, Scapeshift 6th, Fast Zoo 7th, Dredge 8th

1/9 Indianapolis PTQ
Big Zoo 1st, Fast Zoo 2nd
Fast Zoo 3rd, Burn 4th
Scapeshift 5th, Hypergenesis 6th, Thopter 7th, Fast Zoo 8th

1/9 Montreal PTQ
Faeries 1st, Martyr 2nd
Thopter 3rd, Big Zoo 4th
Fast Zoo 5th, Fast Zoo 6th, Thopter 7th, Fast Zoo 8th

1/9 Atlanta PTQ
Dark Depths 1st, Fast Zoo 2nd
Hive Mind 3rd, Affinity 4th
Big Zoo 5th, Fast Zoo 6th, Faeries 7th, AIR 8th

1/03 Magic Online PTQ
Affinity 1st, Bant 2nd
Fast Zoo 3rd, Death Cloud 4th
Thopter 5th, Dark Depths 6th, BRW Aggro 7th, Dark Depths 8th

1/02 Seattle PTQ
AIR 1st, Fast Zoo 2nd
Scapeshift 3rd, Burn 4th
Bant 5th, Teachings 6th, Thopter 7th, Thopter 8th

1/02 Magic Online PTQ
Doran 1st, Fast Zoo 2nd
Burn 3rd, Faeries 4th
Burn 5th, Fast Zoo 6th, AIR 7th, Dark Depths 8th

1/02 Minneapolis PTQ
Fast Zoo 1st, Faeries 2nd
Sullivan 3rd, Affinity 4th
Big Zoo 5th, Dredge 6th, Scapeshift 7th, Thopter 8th

Now, that is already probably more data than you want, but I list it here as a reference in case anyone is interested in which Top 8s were won by which deck. It should be noted that the Milan PTQ did not report the 4th place finisher. That is fine; we should calculate the results we do have and divide by the number of points possible with these results.

Winner’s Metagame:

Winner's Metagame

The total of 99% is just a result of rounding. I have taken the liberty of organizing the results among Aggro, Combo, and Control, arriving at the following:

46% (159) Aggro – Fast Zoo, Big Zoo, Burn, Affinity, Rock/Doran, Bant, Misc Aggro,
28% (97) Combo – Scapeshift, Dark Depths, AIR, Dredge, Hive Mind, Hypergenesis
26% (92) Control – Thopter, Faeries, Death Cloud, Martyr, Misc Control

In addition to breaking down what the metagame looks like today, Frank was also fond of examining the change in the metagame every two weeks. These are the first two weeks of the season, and it is highly speculative to base the PTQ metagame on Magic League results or some other previous Extended season metagame. As such, the reader is encouraged to try this method on their own once they have the results from between the 1/16 PTQs and 1/23 PTQs. When calculating these new percentages, do NOT add in the results from weeks 1 or 2. Once you have the results from weeks 3 and 4, compare them to the previous week to get an indication of what decks are falling out of popularity and which are gaining steam.

All of this information can be useful for figuring out the metagame as it is today and where you think it will go next week. Correctly anticipating the metagame is one of the secrets for which GP Masters like LSV and Saito are known. Still, you should play a deck with which you are comfortable. If you do decide to pick a deck to master, this metagame breakdown can give you an idea of which decks are worth testing against.

Let’s take a look at examples of each major deck in the metagame. The lists I present here are the most successful of each of these archetypes in PTQs, and as such are probably the best versions to test against.

Up first, the decks you NEED to have in your gauntlet. Let’s start with the big one, Zoo. While there are all sorts of flavors of Wild Nacatl decks, they basically fall into two categories: Fast Zoo and Big Zoo. Fast Zoo is certainly the more popular of the two, and is by far the most popular deck in the format, weighing in at a hefty 20% of the winner’s metagame. It is basically a collection of the fastest and most powerful aggro cards in the format. I have listed two variations here, though they really aren’t that different from one another. I recommend testing against whichever of the two you think will be harder for your deck.



As you can see, Goblin Guide has fallen out of favor and the primary question seems to be Bant Charm versus Steppe Lynx. Either way, just test against whichever you think is harder for you and test against it a lot. This is the deck to beat.

The buzz after Pro Tour: Austin was about the dominance of Big Zoo, with Rubin Zoo putting up incredible numbers. Today, the format has shifted a bit and Big Zoo decks have fallen in popularity. They are a lot like Fast Zoo decks, except they generally cut Tribal Flames and Kird Ape, as well as some other early creatures, for a better Stage 2 and Stage 3, with “Big” cards like Baneslayer Angel, Umezawa’s Jitte, Punishing Fire, Elspeth, Bloodbraid Elf, and Ranger of Eos, as well as Noble Hierarchs to speed things up. The theory is that by slowing down the Zoo strategy and going a little “Bigger,” you can gain an advantage in the common Zoo mirrors. The downside is that you are a little slower versus everyone, giving others a chance to do unfair things to you, such as combo off or assemble Thopter-Sword.



Again, I recommend testing against the build that you think will be harder for you. Your gauntlet should probably have two Wild Nacatl decks in it, one fast and one big, so as to make sure you are actually pushing your deck and forcing it to adapt to the hostile Extended format. The best way to beat decks like this is to do something really unfair like Scapeshift, Dredge, or Hypergenesis. If you are not playing a combo deck, however, you are going to need lots and lots of cards that are good against aggro, such as Kitchen Finks, Firespout, Path to Exile, and Engineered Explosives, in order to gain ground on them, since their cards are SO good.

Up next, we have the most popular Control deck in the format (if you want to call it control, since many consider it a combo deck). Thopter Foundry + Sword of the Meek is one of the absolute most powerful combinations in the format, and I have listed two takes of the strategy. One uses only a single copy, instead focusing on resolving a Gifts Ungiven that presumably fetches Thopter Foundry, Sword of the Meek, Academy Ruins, and Tezzeret the Seeker. At this point, no matter how the opponent divides the piles, you will be able to assemble the combo and begin making a 1/1 flier and gaining a life for every mana you spend.

If this doesn’t sound as broken as many things people are doing, you are right, but the key is that it is incredibly resilient and it doesn’t need to win immediately as long as it wins. Once you get the combo going, no aggro deck can race you (you make all the chump blockers you want and start getting ahead on the board and in life), Dark Depths can’t get through, Burn can’t kill you, and most control decks can’t get out of the unrelenting stream of tokens.

The other build I list is one that uses many copies of both parts of the combo and just tries to get it going as fast as possible. Whereas the first is a control deck with a combo kill, the second is a combo deck that is very controlling in nature. Again, test against which ever you think will be harder for you. The best way to beat decks like this is with Extirpate, though Cranial Extraction, Ancient Grudge, and Gaddock Teeg are all good.



Personally, in the current Extended season, I prefer U/W Control decks with a Baneslayer kill instead of a Thopter kill, such as my U/W Control here or Michael Flores’s U/W control here. I would also note that I have since added Glen Elendra Archmages to the main deck, among other changes. I will be including my current U/W Control list at the bottom of this article, but do not want to list it here as the focus of this section is on the decks from the first two weeks of PTQs.

Up next, we have the most popular true combo deck in the format, Scapeshift. This deck is somewhat similar in spirit to Tooth and Nail, in that it just sort of hangs out and then kills with One Big Spell. It is slower than most combo decks, but it is far more resilient. Its advocates claim that its weakness against control decks and some combo decks is not that big a deal, since it has such a good aggro match-up and beats some of the more fragile combo decks with just a token amount of permission.


The idea behind the deck is probably familiar to you: just cast Scapeshift when you have at least seven lands on the battlefield and you deal 18 to your opponent by getting Valakut and six lands that count as Mountains. An 8th land means dealing 36 damage, if need be.

While this deck is quite popular, I think it is very overrated and predict it will decline in popularity in the weeks to come. The only thing that may get in the way of this decline is if Kyle Boggemes and DJ Kastner’s innovation of Path to Exile is adopted. With Ponder plus Coiling Oracle and Wood Elves, it is common for this deck to find itself with a random dork on the battlefield and one mana. Right now, Scapeshift uses a few Firespouts (or occasionally Dead/Gone or Magma Jet) to deal with problem cards like Gaddock Teeg and Meddling Mage. Path to Exile accomplishes this goal, but rather than stranding the Scapeshift player with a dead card against combo or control, provides even more acceleration, which is often the perfect way to fight decks with Counterspells.

For instance: Turn 1 Ponder, Turn 2 Coiling Oracle plus Path to Exile, Turn 3 Wood Elves plus Sakura-Tribe Elder (or Path to Exile), Turn 4 Scapeshift with double Remand (or Cryptic Command) back-up.

This breakthrough is not enough to get me on board, but it has been the only impressive Scapeshift deck I have played against. Still, this is the most popular combo deck, so I recommend a copy of it in your gauntlet, and that copy should probably be built like this, rather than with Path, since this is the most well known list.

Up next, a surprise return of the Faerie tribe. Whereas Mono-Blue Faeries was a dominant deck last season, this season saw a huge decline in the tribe as people tried to make up for the loss of Riptide Laboratory. In the end, time has shown that the direction to take a dedicated Faerie deck now, is to go back to Black for Bitterblossom, Doom Blade, and sideboard cards in a style of deck that is much more tempo oriented and reminiscent of 2008 Standard Faeries.


The Faeries deck is an important one to have in your gauntlet, as it is a great example of a Blue tempo deck that wins with incremental advantage, as opposed to the combo-esque nature of Thopter decks. These types of decks are often a bit vulnerable to quick attacks, though if you are not playing an aggro deck, your best bet is to try to attack them from a variety of angles. This is one deck where you really can’t have dead cards and expect to beat them.

Up next, we have a deck that is chronically ignored in testing, presumably because of the frustration some players experience when playing against it. Burn is a deck that is mind-numbingly simple, but potentially lethal, as Goblin Guide and Lightning Bolt have gone a long way towards pushing it over the top. The deck is literally this close to be 40 Bolts and 20 Mountains. The best ways to fight this deck are to be faster, or life gain/damage prevention (COP: Red is legal!), or a quick clock backed up by permission (which are like Healing Salve). This deck’s biggest advantages are its Zoo matchup and that it is the most consistent deck in the format (all of its cards are the same and cheap).


If you have time to test against seven decks, these first seven are the ones to include in your gauntlet. If you only have time for five, I recommend Fast Zoo, Thopter, Scapeshift, Big Zoo or Burn, and Faeries or Dark Depths.

Speaking of Dark Depths, this strategy has been falling out of favor a little since PT: Austin (where it was one of the best performing archetypes). In Rome, at the World Championships, Dark Depths was among the worst performing archetypes. Since then, it has experienced a small rebound, but it is still not the threat it once was, now that Ghost Quarters, Path to Exiles, Repeals, Cryptic Commands, Bant Charms, and more are everywhere. Still, if you have time to test against 7 decks, this one might make the cut, since it can do some pretty busted things. Okay, just one busted thing, but a 20/20 Flying indestructible guy on turn 2 is pretty busted.


At this point, we are starting to get into the fringe decks, but I am including sample lists here as it is still important to mix up your testing. While most of your testing should be against Fast Zoo, Thopter, and Scapeshift, it is also good to get a few games in here and there against a wide variety of decks. A useful tactic is to find the extremes in the metagame and test against those, as detailed here.

All-In Red (AIR) is not always the most fun deck to playtest against since games are won or lost often with almost no relevant decisions on either player’s side, other than the absolute most basic uses of mana. On the upside, however, you can playtest 10 games against AIR and get a pretty good idea of the feel of the match-up in under an hour. The way to beat this deck is to not lose to Blood Moon. Beyond that, cards like Path to Exile, Mana Leak, and pretty much any fast relevant disruption/bounce/permission/removal at all are good.


Affinity is one of those decks that never seems to go away. The reason for this is that it is relatively easy to play at least halfway decently, and it is almost always underrated by a large chunk of people who shave Affinity hate (figuring few people play it). As such, Affinity has become a total “metagame deck” where savvy players try to predict the exact perfect time to pick the deck up (presumably when everyone has cut their Katakis and Grudges).

Aside from artifact hate, other ways to beat Affinity include fast combo and Wrath of God. Personally, I think this is a poor time to play Affinity, despite all of the people cutting Grudges from their lists, as too many people play decks that you can’t beat very well, and the people you do beat have cards that you often can’t beat if they draw them.


My man Christian Calcano is up next, with a Doran Rock that he brewed up on a car ride we shared en route to a PTQ in Baltimore (Thanks for driving, Sean!). Rock is always a tricky one to put in the gauntlet, since so many people play such a variety of cards, but the primary distinction is whether the deck is aggro or control. Either way, they usually use the classic combination of Green creatures, discard, and removal. Calcano’s list is a good example of an aggressive Rock deck (in real life this is a Junk deck, not a Rock deck, as it includes White and is an attacking deck). I have listed Death Cloud separately, as the more controlling B/G decks are a very different animal.



The next archetype to be discussed, Bant has been praised by some, ridiculed by others, but is rarely thought of as a Tier 1 deck. Instead, its pilots often say they selected it because it is fun, solid, and offers them a chance to play a real game. Personally, I think the strategy is a little underpowered, but at least you get to play a real game against everyone (no unwinnable matchups). The main problem with playtesting against such a deck is that it is the least extreme deck in the format. It is literally middle-of-the-road on everything, and thus testing against it is one of the least effective uses of your time.

First of all, just about every match-up 45-55, give or take. Second of all, the deck has some quality cards against everyone and some bad cards versus everyone. In true “bad midrange” fashion, the matchups are often all about if they draw the right cards to match up against you. This can be particularly bad if you playtest only a small number of games (sometimes even 10 is not enough), since if they just draw Meddling Mage and counter spells against the combo deck, or Tarmogoyfs and Rhox War Monks against the beatdown deck, you can get a very inaccurate feel of the match-up. Still, this is actually a deceptive strength of the archetype. Since it is so inefficient to playtest against it, many players in a PTQ will have almost no experience playing against it.


The next deck is very powerful metagame choice, but I don’t recommend playtesting much against it if you are playing Aggro, unless it gains some popularity. It can be very difficult to beat such a deck with Aggro, and playtest games are exceptionally slow. If you do test against it, skip to sideboarding relatively quickly so that you can find out if your graveyard hate helps you.

The fatal flaw to this archetype is two-fold. First, you can’t (or so the theory goes) build a Martyr deck that can be Aggro and non-Aggro. Your deck is either full of Wraths and Plows and lifegain (which is what they usually are), or it has disruption and card draw (and loses to aggro). The B/W lists of late try to straddle this line, but they are just as vulnerable as the rest to problem #2 – Extirpate or Cranial Extraction. Those cards are pretty rough, and leave you pretty much just a Baneslayer Angel Deck with no permission.

Something a few people have been asking me about is adding a Wild Cantor and a Soul Warden as well as a few Enduring Renewals. Obviously the primary advantage is that it gives you another infinite combo that beats most opponents, and it can be fetched by Ranger of Eos; however, there is also the hidden upside that if you have a Martyr of Sands and Enduring Renewal, you can just gain 12-21 life for every two mana you spend, which is almost impossible to race.

It should be noted that most lists no longer run Proclamation of Rebirth, instead using Phyrexian Arena, Ranger of Eos, Emeria, and sometimes even Moonlight Bargain to draw cards.

Test against this deck at your own peril. I highly recommend playing at least a couple preboard and a couple postboard, but a long set can be very time consuming. It is important to at least have a feel for the pace of the match.


The next fringe deck is actually hated on and talked about far more than the position in the metagame it occupies. Dredge has left people so scared from previous seasons that there is an incredible movement among people to play 4-7 graveyard hate cards, especially when you factor in other graveyard decks like Thopter and Martyr. Maindeck Relic of Progenitus, Extirpate, and Cranial Extraction are everywhere. On top of this, Meddling Mage and Gaddock Teeg can be a problem, and you literally have to deal with so many types of hate, from Muddle the Mixture finding Yixlid Jailer, to Tolaria West finding Crypt, to games starting with Leyline, to Ravenous Traps out of non-Black decks.

There is just too much hate to play this deck right now, even though it was one of the most hyped decks in Austin and Worlds. Test against it if you have time, but don’t be afraid to jump to sideboards if game 1 seems uninteresting. If you don’t have time, no problem; just remember to sideboard in cheap cards and cards that are relevant, such as sweepers, permission, and of course graveyard hate. Sideboard out slow cards. Try to anticipate what the Dredge player will look like after boarding, not just what he looked like game 1.


Extended has always been a hot bed of random combo decks, and this season is no different. Hive Mind has been making some minor waves, as combo players are trying to figure out what the next big thing is going to be. So often, the key to combo is to play the combo people aren’t expecting. LSV is one of the best in the world at this. When they expect TEPS, show ’em the Elves. Then, when they expect Elves, hit ‘em with the Swans. Then, when they expect the Swans, go back to TEPS!

This format has already gone from Hypergenesis to Dark Depths to Dredge and now to Scapeshift. Is Hive Mind next? Personally, I think it is the wrong way to go, since it is actually vulnerable to a lot of the same hate as Scapeshift. The classic way to beat this deck is to attack with cheap creatures. It may have a slightly better combo and control matchup than Scapeshift, but where Scapeshift is good against Zoo, Hive Mind is pretty weak, often relying on the desperation sideboard of Blood Moons and Firespouts (remember this if you are playing a deck that loses to Blood Moon!)


Oh, Hypergenesis, how far you’ve fallen. What was once the most feared deck in the format has now been almost entirely obsoleted. They can’t beat the Blue decks since they are so vulnerable to Countermagic. They can’t beat all the hate (like Meddling Mage, Chalice, Cranial Extraction and so on), they can’t beat Blood Moon. Their backup game is awful (7, 8, 9, 10, and 11 drops? Nice! Oh, I forgot… they also have Grey Ogres!).

Avoid this deck – it was one of the worst performing decks at Worlds and Austin – but playtest a few games against it. It won’t take long, as this is another deck that you can get a 10 game set under your belt in an hour or so, if you are fast. Again, don’t be afraid to go to sideboard matches early, as playtesting is about using your time effectively (not doing what is “fair” to the decks, hence test against versions of Zoo that are hard for you, and test mostly sideboard games against decks like Dredge, Hypergenesis, and Affinity).


I may have listed over 20 decks here, but that doesn’t mean you need 20 decks in the Gauntlet. Figure out how much time you have to test, and devise a plan for how to spend it effectively. You can only spend the time you can, so rather than just wish you had more, make the most of it. Fast Zoo, Thopter, and Scapeshift need to be the first three decks in the gauntlet. The next two are Big Zoo or Burn (whichever is harder for you) and Faeries or Dark Depths (whichever is harder for you).

You usually don’t need to get to the fringe decks until you have a pretty good idea of what sort of deck you are interested in battling with. However, once you do, by all means test against a wide range of enemies. This will help you learn the ins and outs of your choice. I know a lot of what I am saying you surely already know, but I am trying to make sure there is useful information here for all experience levels.

Before I call it a day, I am going to share the U/W Control deck that I am planning on piloting in the Detroit PTQ this weekend.


I was talking to Adrian Sullivan just before a PTQ began, nine days ago began. I mentioned that I wanted to play 25.5 land, and asked if I should just add another Relic. He suggested instead adding a Ponder, and on a whim I did. It was totally awesome. It is hard to describe just how good it was (considering how little it cost me), but suffice to say I am going to experiment with a 2nd or 3rd. It is a little slow, so I am not just moving straight to 4. I should also note that I was really happy with the Relics as well, and would probably add a third, except that there is a good chance I will end up with 2-3 Ponders which will help me find a Relic when I need it.

I want to emphasize that developing a plan for smart play testing time is much, much more effective than just sort of hanging out and playing whatever deck someone happens to have on them. You only have so much time to test, use it wisely. Whatever you end up playing this weekend, I wish you the best of luck!

Patrick Chapin
“The Innovator”