oraymw's picture
By: oraymw, Matthew Watkins
Mar 27 2013 6:41am
5
Login or register to post comments
4217 views


Welcome back to Ars Arcanum, the MTGO stats-based column. Gatecrash has now been out for a month. I have already put up an in-depth format overview, as well as an analysis of over and under drafting. There have been a few high level events that feature the format, and people have been following the developing advice of pro players. Now is the time for a follow up article, where I go back into the format, and do an in-depth analysis of the current metagame. We’ll see what has changed and what has stayed the same. These follow up articles are especially interesting for seeing the ways that people respond to how a format is being drafted, so I’m excited to dig in.

I should note that I made a few key recommendations and predictions about the format in my GTC overview. Some of the main points in that article were that Dimir was the best deck in the format, that Boros was mediocre, and that Boros would lose some of its win rate in the weeks following the article. While some of the predictions were spot on, some of them were a little bit off.

As always, the data for these articles comes from watching hundreds of GTC matches on MTGO. For this study, I watched a little less than 600 decks in action over the course of about five days. The data comes from a random sampling of 8-4 drafts on MTGO. This is a little bit different from a 64-man draft in that the players are better able to respond to what is going on in their own draft, which often means that they will be able to adapt to the decks they will end up facing. This tends to flatten out win rates a little bit, as decks are better able to react to their neighbors. However, it also gives us a much more natural comparison with the way that most players will interact with the format. With all of that behind us, let’s dig in to the data.

Speed

Ending Turn of Gatecrash Draft Games: Study Two

Ending Turn of Gatecrash Draft Games: Comparing Study One and Two

Ending Turn Groupings Comparison

In these charts, we see the first of several surprises about the format. In the last study, I showed that GTC was much faster than RTR, but that it was a hair slower than AVR, by comparison. One of the main things that I recommended was to keep your curve as low as possible. However, my experience in the past has been that as people start to react against a fast format, it starts to slow down a little bit. Instead, we saw that Gatecrash has sped up significantly. Instead of just being a relatively fast format, we see that GTC is now the fastest format since triple Zendikar.

The comparison charts are especially telling in this regard. In our first GTC study, we saw that about 50% of games ended between turns four and eight. In this second study, we see that number increase to 57%. We saw a significant increase in the numbers for turns two through seven, and another big change in the numbers for turns twelve through twenty, which saw a 4.5 point drop. Whereas one in six games went past turn eleven in the first study, now only one in nine games gets that far. That means that a given player will often only see one game go past turn eleven in every other draft.

There are several reasons why the format ended up faster. The first and most important reason is that most people are correctly building their decks with a much more aggressive curve. In my last article, I recommended playing as few six drops as possible, while overloading on two drops. It seems that quite a few people came to the same conclusion. There were far fewer Ruination Wurms and Smog Elementals being played, which meant that more people were able to act in the early game. However, the second reason the format has sped up is for an opposite reason. While most people adjusted to the format by playing faster decks, a few people have started to “get bored” with the format, and they are building decks in order to experiment with more expensive cards. Those decks are light on early defense, and tend to get annihilated even more quickly than other decks in the format. Finally, players have started to figure out how to build each of the guilds correctly. The Orzhov, Dimir, and Simic decks have gotten much faster and they are focusing a lot more on having a good curve to enable early game action.

Having a faster format has a few big implications. First, it means that decks have to continue prioritizing cheap spells. Many people have talked about how GTC is a format where curve is more important than just about anything, and I definitely agree. There are a lot of people that are trying to make seven drops work, or they are trying to get away with greedy splashes. I can only say to resist that temptation. The best way to be successful in this format is to have a good curve.

With that said, I should note that this format often results in two decks with very fast curves that trade off creatures in the early game. Then, both decks will have mostly emptied their hands, and they will hit the midgame, from about turns 8 to 11, but they won’t have any more gas. Many of the winning decks that I saw were able to put out an efficient and resilient threat on these turns; something like a Zhur-Taa Swine, a Dinrova Horror, or Towering Thunderfist. You definitely don’t want many of these cards, but the winning decks tend to have something that is resilient and powerful that can take over the game after the initial flurry of inexpensive cards.

Guilds

Guild Popularity in Gatecrash

Guild Win Rate in Gatecrash

In my Gatecrash overview, I pointed out several factors which indicated that Dimir was a stronger guild than Orzhov. and that it would demonstrate that strength as people started to gravitate towards Orzhov, or figure out the best ways to beat the Church of Deals. The data for this study shows that I was correct in my assessment, but there were also a few interesting developments, and I’ll walk through each of them in this section.

First we should talk about how the popularity of decks changed in the past few weeks. Boros dropped several percentage points in popularity; apparently people started to figure out that it wasn’t a strong enough deck to be drafting at one quarter of the field. Meanwhile, Orzhov saw an increase by a few percentage points. It went from being the third most popular guild, to edging out Boros as the most popular guild. Simic saw a substantial drop in popularity, as did Gruul, while Dimir saw an increase by a couple of percentage points. In my last article, I introduced the idea of popularity ratios, and I think those numbers are particularly useful at analyzing where the format is at now:

GTC

Boros

Simic

Orzhov

Gruul

Dimir

Pop Ratio One

1.26

1.11

1.03

0.91

0.69

Pop Ratio Two

1.18

0.98

1.21

0.86

0.78

 

This side-by-side comparison helps us see that Boros and Simic saw a significant decrease in drafters, Orzhov and Dimir saw a significant increase, and Gruul saw a slight decrease. These numbers are actually exactly what we would predict to see after taking a look at the last article. The winning decks saw an increase in popularity, while the losing decks saw a decrease in popularity. I should also note that the color white became even more heavily overdrafted, making up nearly 50% of the field overall. Meanwhile, green dropped down to being the second least drafted color, just ahead of blue. The place where we see a big change is in the Win Rate chart.

The first thing that we should notice about the win rates is that Dimir actually managed to increase its win rate. In my first GTC overview, I predicted that Dimir would be strong as people started to draft it correctly, and as it faced an Orzhov heavy metagame, and that is exactly what happened. Meanwhile, Orzhov’s win rate dropped dramatically, by about 20 percentage points. Those twenty points were spread out between Boros, Gruul, and Simic, which all put up decent, albeit average results.

The question that we need to answer is “What happened to Orzhov?” There were a number of factors involved in its fall. The first was that part of this drop is a natural result of overdrafting, the second is that we saw the game plans of other decks start to react to Orzhov, and thirdly we’ll see that Orzhov suffered at the hands of poor drafting.

The first factor is overdrafting. As we saw, Orzhov saw a .18 increase in its popularity ratio. It went from being drafted at about 20% of the field, to being drafted at almost 25%. Meanwhile, Boros saw only a small decrease, while Dimir saw a substantial increase. All of this leads to Orzhov seeing a significantly lower number of its cards. On top of this, nearly 1/3rd of all the Gruul decks were splashing white. Furthermore, players finally recognized the power of extort, and they were making use of it in Boros’s neighbors, which meant that Orzhov could no longer depend on picking up all of the extort cards. It would have been a tremendous surprise not to see Orzhov lose some of its win rate. However, overdrafting certainly doesn’t account for a 20 point decrease. We would expect to see maybe a five percent drop, but these numbers suggest a cocktail of factors.

The second factor is that the other decks in the format correctly adjusted their game plans to fight against Orzhov. The Orzhov deck centers around playing many inexpensive cards with extort. It tries to hit its opponent for a substantial amount of damage in the early game, and then use the time it has earned to drain the life out of its opponent. In a format dominated by Boros and Gruul weenie rush strategies, this was the perfect strategy to dominate the format. The problem is that once people saw how good Orzhov was, they started to adjust their game plans. Bigger and more resilient creatures became more important. As soon as Orzhov finished fighting off the initial onslaught, the Boros deck would drop a Towering Thunderfist. Orzhov would be out of removal, and Basilica Guards doesn’t do much to block a 4/4. This meant that Orzhov was unable to get in enough extort before losing the game to one of these big creatures. Dimir and Simic both compounded the problem; the pilots of decks in both of these guilds demonstrated a much better ability to build a cohesive deck. Also, both of those decks are very good at taking over the midgame, which made them perfect for dominating Orzhov. Again, all by itself, a change in strategy would not signal such a dramatic change in win rate, but when we consider that Orzhov was also fighting tooth and nail for its extort cards, we see that it simply didn’t have enough extort to close out the midgame and stay alive against a larger threat.

Finally, and most importantly, Orzhov suffered from poor drafting. This expressed itself in two major points. First, Orzhov decks were full of greedy splashes. There were many Orzhov decks that featured Ember Beasts, Truefire Paladins, Sage’s Row Denizens, or even things like Mortus Strider. I saw four Orzhov decks that were splashing for Aurelia, and a couple that were splashing for Lazav. The worst was that they were often splashing for mediocre cards. Because of this, they would often just fail to play things out in the right order for the first few turns, and they were punished dramatically by a fast format. Secondly, people seemed to think that Orzhov was meant to be a slow control deck. These decks featured high curves, a lot keyrunes, and several gates. They leaned heavily on Dutiful Thrull, and really hoped to make the game go as long as possible. Orzhov decks are at their best when they are streamlined and when they are not splashing. They need a very low curve, lower than any other deck in the format, because every bit of mana is essential for Orzhov. Again, by itself, this factor wouldn’t have killed Orzhov so badly; it is definitely a powerful enough strategy to withstand a handful of bad drafters. The problem is that all three of these factors combined together to make Orzhov a terrible choice against the metagame.

#SkillGame

This leads me to my final point about GTC drafting; Gatecrash is one of the most skill-intensive draft formats in recent history. It is comparable to a set like Innistrad or Rise of the Eldrazi. For example, in the last GP we saw Eric Froehlich, Gerry Thompson, David Ochoa, Shahar Shenhar, Conley Woods, LSV, Josh Utter-Leyton, Zvi Moshowitz, Gaudenis Vidugris, Shuhei Nakamura, Brad Nelson, and Own Turtenwald all make the top 50, in a GP of more than 1600 people. We saw one of the most star studded top 8s in history at Pro Tour Gatecrash, but more importantly, we saw that pros with a track record of good limited performances surge ahead during the Gatecrash draft rounds. For my own experience, I’m currently at a 77% match win rate in the format, and I’ve won 50% of my GTC drafts, including a huge string of victories in the last few weeks. People who have proven themselves to be top tier limited players continue to put out results in this format, while many other people struggle to put together anything cohesive.

In this section, I’m going to step away from the statistics for just a moment to talk about the factors that make this format so skill intensive. I’ll talk about the speed of the format, the range of values for individual cards, and the complexity of game play.

First of all, the speed of the format has a huge impact on the skill level for the format. Speed affects a formats skill level in several ways. For one, it makes a format much less bomb-focused. Cards like Aurelia’s Fury, Clan Defiance, or Aurelia, the War Leader are still very strong in this format, but they are much less important in a format where you often don’t get to six mana, and where you see a much smaller percentage of your deck. In most formats, it would be entirely appropriate to take one of these cards for your first pick and then force your way into their colors. Normally, you would have enough time to draw this bomb and cast it, so it wouldn’t matter very much if you missed out on a little bit of card quality by forcing these colors. However, in Gatecrash, you can’t rely on always drawing these cards, let alone casting them. It is vital that you don’t sacrifice the consistency and power of the rest of your deck because of one early pick. For myself, I usually try to stay as open as possible throughout the entire first pack. I still usually end up in Dimir, but that is because the other guilds tend to be drafted a little bit too heavily. In Gatecrash, your first two picks don’t really matter, and your first six picks are expendable. If you find yourself struggling in this format, ask yourself whether you are approaching the format with enough discipline and flexibility to avoid getting sucked in by early bombs. Instead of focusing on the power level of individual cards, it is much more important to pay attention to the synergies between all of your cards.

The speed of the format is also vital in that it forces players to be more disciplined with their mana. Building a good mana base is the hardest skill in Magic, and this format really tests that skill. It is important to make sure that you can cast your cards early and often. Splashing is very bad in this format, especially greedy splashes. Often, they just dilute the synergy of your deck and make it so that you die to an aggressive start from any other the other guilds. Your deck should be two colors almost every time, and you should only be splashing for truly powerful cards, and it is often incorrect to splash even in those cases.

The second thing I want to discuss is the non-intuitiveness of the format. In most formats, you can figure out the approximate power level of all the cards, and when you find yourself having to make a choice, you just defer back to that knowledge. However, since Gatecrash is such a synergy based format, the picks are often not straightforward. Dimir is probably the very best example of this. Over the course of the format, I have faced the choice between Death’s Approach, Balustrade Spy, and Sage’s Row Denizen many times. Death’s Approach definitely has the highest abstract power level, but it is also the worst of the bunch when it is bad. All of these cards haven’t different strengths in different forms of the Dimir deck, and my choice always depends on what cards I have previously drafted. Likewise, Dimir often faces a choice about which cipher card to play. Do you play Hands of Binding, Midnight Recovery, Call of the Nightwing, Last Thoughts, or Shadow Slice? Each of these choices is entirely dependent on the cards in your deck. Midnight Recovery gets a lot better when you have lots of creatures you want to get back, but Last Thoughts is better when your deck is more controlling and filled with powerful removal spells. Call of the Nightwing is especially good when you have a healthy combination of Sage’s Row Denizens, Undercity Informers, or Corpse Blockades, but in other situations you would much prefer to have a Midnight Recovery. Shadow Slice is good in the more aggressive versions of Dimir, while Hands of Binding is very good in the more extort/evasion centered decks. The problem is that the power of all of these cards depends on the contents of your deck, so you need to be able to quickly understand how you plan to win the game and how the value of all your cards change from draft to draft.

The last way that Gatecrash is incredibly skill intensive is in the actual games. Part of this is a function of the speed of the format. Because the games end so quickly, you are able to make far fewer decisions over the course of a game. This means that any given decision that you make takes on additional weight. Missing one or two damage early in the game makes a huge difference when you only have eight turns, instead of twelve. Choosing the correct card is much more difficult if you are going to end up with three cards in your hand at the end of the game. Losing card advantage is a much bigger blow when you are in a tight race, and using your mana correctly is much more important when you will often not have more than five lands in play.

In the end, this is a format that requires players to have a lot of skill, and those skills are often not the kinds that are developed in other formats. If you find that you are not one of those players, don’t be too discouraged. There are plenty of opportunities for you to improve your game. Having this type of challenge is a great opportunity to improve your game. But more importantly, Dragon’s Maze isn’t very far away, and it will likely change the skills needed to succeed in many dramatic ways.

For those that need a little help improving their skills in Gatecrash, I have a few places that I can point you. First, you can check out my stream at twitch.tv/oraymw. I try to stream on weekends, and my goal for the stream is to supplement these articles in order to help people figure out how to apply the things that I’m demonstrating. Another resource is the new article series “The Slow Bleed” by Zach Orts; if anyone remembers the column “Waiting for Godot” by Ryan Spain, then Zach’s article will look very familiar. He talked to Ryan, who gave him permission to use the same article format. He walks through an entire draft, including all of the picks, and the play decisions that he makes in each game. He has done an incredible job with that, and it’s a fantastic resource for helping people get to the next level in their drafting. Finally, there is just no substitute for getting on MTGO and drafting as much as possible in a format. Take the opportunity to practice with these tools, and you’ll see a gradual improvement in your play.

Conclusion

Here are my final conclusions about the format:

1.       While Gatecrash seemed very fast in the first study, it has only gotten faster. Curve is essential, and you just don’t have time to mess around.

2.       As a result of the initial flurry of spells, having powerful and resilient midgame creatures becomes extra important.

3.       Orzhov saw a dramatic fall from its place at the top of the format, as a result of being overdrafted, seeing better competition, and being drafted poorly.

4.       Dimir proves its place as the best guild in the format.

5.       Gatecrash is one of the most skill intensive formats in recent memory.

6.       The secrets to winning in Gatecrash are discipline and flexibility.

7.       Individual cards in Gatecrash have ranges of values depending on the contents of a player’s deck. This means that players need to constantly reevaluate their cards.

8.       Gameplay is even more important in a fast format with a lot of synergy. If you have loose gameplay, it is likely a significant factor in lowering your win rate.

As always, you can follow me on twitter @oraymw for updates about articles. I’ve also put up a Tumblr account at http://oraymw.tumblr.com/ where I post links to my articles. You can go there and subscribe to the RSS feed, and then you’ll be able to get updates whenever a new article goes live.

Finally, I encourage you to check out the podcast that I do with my buddy Zach Orts, which is called All in the Telling. In it, we look at stories from a professional standpoint in order to get a better understanding of why they are important to the human experience. But mostly, we just talk about what makes awesome stories awesome. You can also follow Zach on twitter at @zvazda.

Ars Arcanum Archive


 

3 Comments

Another great article! by Lobster667 at Wed, 03/27/2013 - 07:03
Lobster667's picture
5

Not much else to say here, I think you're right in basically everything here. I think especially people figuring out Simic has sped the format up (that is one of the decks I didn't realize the power of early on personally, at least).

Also, am I a criminal for saying that I might like this format as much as triple Innistrad? I really enjoy all the factors you mentioned goes into the skill-intensity (and my MW% is at 75, just below yours).

For those that would like to by oraymw at Wed, 03/27/2013 - 12:53
oraymw's picture

For those that would like to check out Zach Orts article "The Slow Bleed" you can find the URL here: .

http://puremtgo.com/articles/slow-bleed-gtc-draft-1

I meant to link to this in my article, but I managed to be a dummyhead and forget to include the link :( So... it comes in the comments instead!

One little note that I didn't by oraymw at Thu, 03/28/2013 - 14:36
oraymw's picture

One little note that I didn't include in the article was that Dimir face an abnormally high number of mirror matches during this study compared to less popular decks from other formats. Because of this, I decided to run the numbers and see where Dimir ended up when it was just facing other guilds. The win rate comes out to 66.66%, which is a 4.5 percentage point boost (by comparison, Boros w/o the mirror only got a less than one point boost, while Orzhov lost 1.2 points).

There are a lot of interesting implications about this number, but the main point is that it suggests that Dimir is even stronger than the numbers we saw in the actual article.