oraymw's picture
By: oraymw, Oraymw
Mar 20 2015 12:00pm
0
Login or register to post comments
7799 views


 

Welcome to another installment of Ars Arcanum, the MTGO numbers-based limited series. This installment is the Dragons of Tarkir (DTK) spoiler analysis. The goal of this article is to use numerical processes to analyze the new set, and to try to get a solid foothold for where to approach the draft format in the first couple of weeks. Hopefully I can establish a framework that will be useful for understanding and evaluating our experiences in the new format.

In past articles, I’ve explained most of the major concepts that I use to drive my analyses. I’ll link you back to this article where you can find a more thorough explanation, but basically I follow three principles. First, I analyze the creature stats because they are the foundation of limited. Second, I like to focus more on analysis and exploration. I find it more useful to think about the format and raise questions, then to try and make accurate predictions about a format that no one has yet played. Third, the most important aspect of these articles is the concept of rarity weight, which means that I count cards depending on how often they will appear in a draft on average.

This article specifically will focus on Dragons of Tarkir, Dragons of Tarkir, Fate Reforged draft (DDF). Unlike most big set analyses, it’s important that we take a look at Dragons of Tarkir as it relates to Fate Reforged. We already know quite a bit about FKK draft, and that information can certainly help guide our understanding of the format, Dragons of Tarkir packs will also be replacing two-thirds of those packs. This is a significant change in the format, and it will change the way that we understand Fate Reforged. One thing I won’t cover in much detail is the DTK/FRF sealed format; in that format, you’ll open 3 DTK boosters and 3 Fate Reforged boosters. Since there are fewer cards in small sets, the cards show up more frequently, so it’s important to remember that DTK/FRF sealed will be much more FRF centric than DTK.

Since understanding the context of DTK’s relationship with FRF is so essential to analyzing the set, I’m going to start off by taking a look back at the FRF spoiler analysis, and giving some accounting for the conclusions that I came to in that analysis.

Looking to the Past

Since my KTK Spoiler Analysis, I’ve started adding a section where I go back and review the things I covered in my previous spoiler analysis. This is a self-reflective process that will hopefully help me hone my craft so that I can produce better analyses and predictions. The goal is to understand why I made the predictions, what happened in the format afterward, and how to improve the analyses in the future. I think that it’s much too common for Magic writers to simply make bold predictions or extreme comments without ever having to account for those predictions, which makes this kind of look back vital.

1.         I predicted that Fate Reforged would speed up the format a little bit. I wasn’t sure to what degree it would speed up; I predicted that it would be noticeable and make a change on the way people drafted, but I also acknowledged that it could be a smaller acceleration. This is probably the most debatable of all the predictions that I made. In my FKK draft overview I collected a fairly large sample of data that demonstrated a mild acceleration in the format, specifically in turns seven through nine. Other MTG statisticians noticed a similar acceleration in the hard data.

2.         However, many professional players came to wildly different conclusions about the format. In fact, there were many pros that claimed that the format had actually slowed down significantly. This was very unsettling for me. There was plenty of data supporting the idea that the format had sped up, but players that I respect were reporting different interpretations of the format. I started out writing a full explanation of my thoughts about this issue, but it ran over 1500 words, and there just wasn’t room in this article. For that reason, I’ve included a link to my blog, which will often include Ars Arcanum appendix content. For this article, I’ve included a more detailed discussion about whether the format has sped up. For the purposes of this reflection, I’m going to cite the results as a mixed success, since the prediction matched the data, though it didn’t match the opinions of certain professional players.

3.         I predicted a shift towards two-color decks. I think it would be hard to argue against this prediction. I’ve seen a much higher concentration of two-color decks.

4.         I predicted that while 18 lands would still be the norm, that there would be a shift towards fewer lands. Again, I’ve seen this borne out in many drafts, and I think it was a very accurate prediction.

5.         I predicted that the format would be a play format. I think I was still right about this, but those that believe the format slowed down will tend to disagree.

6.         I predicted that blue would be the best color, and red would be the worst. I know that I certain basically all of my drafts in blue, and I’ve had remarkable success. Blue decks have also had consistently good results in limited format across the season. However, it still seems like blue is undervalued, and I think a lot of people would disagree with this assessment. I did think that white was the next best color, and I think that most reasonable players would list blue and white as the best colors in FRF, though black certainly makes a run for its money. Red is clearly the worst color. So, I’ll say this is a win, but a debatable one.

7.         I predicted that allied color combinations would become more popular as the base for people’s decks, and this was another very accurate prediction.

 

8.         I predicted that the prerelease format would be fast and punishing, and that White based aggro decks would probably be the best. On this one, there just isn’t enough data to make an accurate assessment. At my prerelease, 6 of the top 8 decks were white based aggro decks, but that’s a pretty small sample size. However, I also don’t worry too much about predictions for the prerelease since it is a one-off format.

Overall, I’d describe the Fate Reforged spoiler analysis as a mixed success. I think I did a great job at analyzing the data, but there were several outside factors that affected public perception of the format, and my analysis disagrees with that public perception. I do believe that FKK is faster than KTK, and that belief has led to FKK being my most successful personal draft format in history, but I do recognize that other people view the format differently.

Let me know what you think about the predictions in the comments. Obviously this reflection is subjective, so it’s important to get feedback from you the readers in order to improve my processes.

Modeling Megamorph

The first thing I realized when doing the Khans of Tarkir analysis was that morph was going to be the dominant mechanic of the set. The second thing I realized was that it would be incredibly difficult to model in this sort of article. In that article, I produced a complex, but effective model of what I thought the format would look like because of morph. I feel like it was a very effective analysis, and I decided that I would basically just reproduce that model here. To that end, the following few paragraphs is a description of that model as taken from that article.

“However, Morph is an incredibly difficult mechanic to model. Essentially, every morph gets played in some combination of three modes. Sometimes it stays a face down 2/2 for the entire game.  Sometimes you just play it face up for its normal CMC. And sometimes you play it face down, but then unmorph it later. All of these modes affect the format in drastically different ways, and make a dramatic change on what kinds of creatures and spells are going to be played in the format. It was an incredibly difficult job to figure out a way to model Morph in a manner that allows players to understand the format more deeply. As I mentioned earlier, I ended up with about 40,000 calculations for just this mechanic. I did find a model that I think will help us understand the format, but it requires a lot of mental effort to process, and it’s a little bit complicated to explain.

First, I made three models as the base, and a fourth model that looks at all three of those models at the same time.

The first model I call Morph. This model looks at every morph card as if it were a 3 mana 2/2 that would never turn face up in the course of the game. Obviously this isn’t how cards will be played in the format, but it is an important foundation for understanding what the format will look like. It gives us a much better picture of how the game will play out in the first five turns of the game, while people are still struggling to put together the mana to unmorph their creatures.

The second model I call Not Morph. This model looks at every morph card as if it were always cast face up for its full mana cost. Again, this model doesn’t accurately reflect what will happen in the game, but what it does do well is give us a sense of the size of the creatures in the format and what the format will look like in the later stages of the game.

The third model I call Unmorph. This model looks at every morph card as if it were played face down as a 2/2, but then turned face up for its morph cost later in the game. Again, this model doesn’t accurately reflect what will happen in the game, since many morphs will die, and in the late game you’ll often want to just play your morph face up. However, it gives us a very good idea of how the tempo of the game can shift dramatically because of morph in the midgame.

None of these three models are accurate, but by looking at all three of them at the same time, we can start to get an approximation of how the format will actually play out.

The final model attempts to do this. I call it DTK Model because it takes all three of the other models and balances them out. The purpose of this model is to give us an approximation of what the format will look when games transition between all three of these modes. While this is probably the most accurate of the three models, it is still not a very good representation by itself, but only gets its strongest meaning when we look at it in conjunction with the other models.

Finally, I encourage you to use these simply as tools to think about the format on your own. Drafting in general and morph in particular are dynamic systems, and slight differences and changes in behavior can have dramatic impacts on final results, so it is imperative that you use this information as merely a jumping off point for making your own thoughts on the format.

In the rest of the article, I’m going to be looking at specific aspects of the format, but I’m going to do so through the lenses of all four models at the same time. For example, the first section will be converted mana cost, and I’ll be showing you four graphs, which represent all four of these models. By necessity, this article will feature an even greater number of charts, since we’ll be comparing four models each time. Then I’ll be discussing what those graphs tell us. I’ll also talk about the other mechanics, especially the multicolor aspect, as we go through this information about Morph.”

This article has started off with so much talking. Let’s get into the data!

Converted Mana Cost

Creature Converted Mana Cost Comparison with KTK – MORPH

Creature Converted Mana Cost Comparison with KTK – NOT MORPH

Creature Converted Mana Cost Comparison with KTK – UNMORPH

Creature Converted Mana Cost Comparison with KTK – DTK MODEL

Four Models Comparison

I apologize once again for the sheer number of charts to convey this data, but I want to be able to provide you with a detailed picture of how I’m modelling the format. Each of these charts is compared with the same data from KTK, which gives us a pretty good idea of how the two sets are going to be different.

In the Morph chart, we really don’t see much of a difference. DTK has a similar number of morphs, and the creatures without morph have similar converted mana costs. There are a few more two and three drops in DTK, but that is because there are simply more creatures in DTK. If we were to only look at that chart, we wouldn’t see much of a difference. However, it’s the other charts where we start to get to all the juicy data.

 In the Not Morph chart, we see an immediate and stark difference from KTK. One of the biggest features of KTK was that there were very few creatures at the two and three mana slots, but there were a lot of creatures at the five and six slots, because the format expected you to load up your three drops with morphs. Essentially you were playing with five and six drops that you could just play on turn three. But DTK does something that’s quite a bit different. One of the most common things in the set is to see a 2/1 megamorph that flips up for 2 to 4 mana. There are also a lot of megamorph creatures that are small on their face up side. This is an interesting design peculiarity for the Megamorph mechanic. If the creatures are all just huge six drops, then you aren’t really that motivated to play them face down in the late game. But if the creatures are smaller, then it can be very important to play them face down in the late game so that your creature is big enough to tussle. In the late game, a 2/1 isn’t going to have much impact, but a 3/2 will have much more impact. Because of this, there are a lot of morphs that are small on their face down side. This means that the format will play out quite differently from Khans of Tarkir, because you’ll be able to play something on turn 2 a much higher percentage of the time.

In the Unmorph chart, we see the costs to flip up creatures that are face down. In KTK, the biggest number was five. The multicolor common cycle, as well as several other Morph creatures, would just sit on the board until turn five, and then the game would suddenly shift into something very different, because a person could just start flipping up their facedown creatures. Getting to five mana in KTK was essential, and one of the most important reasons to play 18 lands in the format. In DTK, we see something very different. In this set, we see a lot of creatures that turn face up for 2 or 3 mana, and we see am much lower number at five mana. This has a dramatic effect on the format. Creatures will be flipping up earlier, but it will be harder for them to eat a 2/2. Instead, they’ll be capable of turning into 3/2s in the middle of the game. This changes the pace of the game dramatically. Now, the format isn’t just about getting down morphs and hitting five mana; it’s going to be more about tactical decisions about when to flip up your individual morphs. I also expect a much larger number of games to involve morphs being played out on T5 or T6, and then flipped up immediately.

I’m not sure the full impact that this has on the set, but I do expect it to speed things up a little bit. The creatures cost less and they are going to trade more often, so the board is going to clear up much more often. There will still be board stalls, but I expect them to be a little bit less frequent in this set.

Power and Toughness

Power and Toughness Comparison – MORPH

Power and Toughness Comparison – NOT MORPH

Power and Toughness Comparison – MEGAMORPH

Power and Toughness Comparison – DTK MODEL

DTK Creatures

Morph

Not Morph

Megamorph

DTK Model

 

CMC

3.2166

3.4172

3.627

3.4202

Power

2.3004

2.5296

2.7357

2.5219

Toughness

2.6577

2.7868

2.9884

2.811

P/T Differential

-0.357

-0.257

-0.253

-0.289

 

 

Creatures define any limited format, more than any other card type. They take up about 50% of the set, and they determine which removal spells and strategies will be the strongest. The pace of play is also almost always determined by the creatures, and theses analyses have been consistently accurate in predicting the shape of the format. The most useful of these tools is the  creature P/T differential. Over the past couple of years it has been the most reliable method for predicting the speed of the format. P/T Differential represents the difference between the power and the toughness of the creatures in the set. Toughness is usually larger than power, so numbers that are between -0.1 and .1 tend to be the fastest formats. Formats with a differential between -0.3 and -0.1 tend to be medium speed formats, and formats with larger negative differentials tend to be slow. There are many other factors, and Theros wound up being the biggest exception for this rule, but it’s still a great indicator for format speed.

Like Khans of Tarkir, it’s hard to pinpoint exactly what these numbers will feel like in DTK. We’re looking at a format that has creatures played in essentially three different modes, and each of those modes has drastically different numbers. The goal of this analysis is to help you get an idea of what those numbers look like, but you’ll have to figure out for yourself where you think the format will actually fall.

The most important part of these numbers is that we see the Not Morph and Megamorph categories falling firmly in the medium speed format category, while the Morph category falls in the slower range. Khans of Tarkir was the exact opposite of this, and I’m curious to see what effect this has on the format. I think that it is pretty clear that triple KTK and FKK drafts were generally very slow formats. But Dragons of Tarkir looks like it is trending more towards a medium speed format.

The charts also give us some fascinating data. The most interesting of the four from my perspective is the Megamorph chart. This basically shows us what the format looks like with all of the creatures in the set, and with all the Megamorphs being played face down and then being flipped up for their mega morph value. That chart shows Toughness hanging a little bit above power, but not by a huge margin. It seems like Dragons of Tarkir is going to balanced much more closely between aggressive and control strategies.

Another important part of this chart that isn’t immediately obvious is that there’s an atypical bulge at 4 power and 4 toughness. In most formats, the numbers drop more sharply from 3 to 4, but in this format we have nearly as many creatures with 4 power and toughness as we have at 3. This is obviously influence by the tremendous amount of Dragons in the set. On that note, I actually put together a companion piece to this article about a week ago when the full spoiler came out. I decided that I wanted to know how many flyers were in the set as compared with past Magic sets, so I ran the numbers on that. I’ll link over to that post on my blog, but the TL;DR version is that DTK has a higher number of flyers per draft than any regular season format as far back as Lorwyn.

Tempo

Average P/T of Creatures in DTK per CMC – DTK Model

Average Greatest P/T of Creatures in DTK – DTK Model

I decided to limit this section to just two charts, mostly because the different version of these charts were too close to each other to be worth dumping 8 charts on here. In the first chart, we see the average size of creatures at each CMC slot. The second chart shows us the average size of creatures for each land drop of the game.

The idea behind the second chart is to get an idea of the size of creatures that you’ll have to attack into for each stage of the game. As you can see, toughness exceeds toughness on every land drop. This means that typically the creatures you are attacking into are going to have more toughness than the power of the creatures you are attacking with, which usually suggests unfavorable attacks. The gap between these two is less pronounced than it was in KTK, but there is still a significant gap. With that said, I find it fascinating that these two lines approach each other as the game progress. That isn’t necessarily true in other formats; in KTK, for example, that gap got wider as the game progressed. One culprit here is Megamorph. There are several cheap megamorphs that can flip up and often attack in the mid stages of the game, but they aren’t significantly larger than everything else on the field. However, when the late game Megamorphs get online, they are going to be able to smash through opposing defenses. Basically, several colors get some small Megamorphs and some large ones. The large ones play like Woolly Loxodons; they typically cost 6 or 7 to unmorph, and then they are tremendous and drastically change the state of the game.

The other major contributor to this effect are the dragons. Once a player gets to six or more mana in a game of Dragons of Tarkir, things start to change drastically. From that point on, they are going to be able to slam down large evasive creatures with powerful effects that are hard to kill. We can see that there are ways built into the set to defend in the early game, but the late game is going to be characterized by dragons that come down and reshape the entire nature of the set.

I think that a games of Dragons of Tarkir will follow a typical pattern. One deck will be trying to put their opponent off balance by coming out fast out of the gates, establishing an early lead on board, and then using dragons to punch through the last bits of damage in the late game. On the other side there will be decks that try to wrest control of the board away from their opponents in the early game. Their goal is to stabilize and then drop dragon after dragon until their opponent’s just run out of answers. It seems like there’s a lot of tension built in between these types of decks, and it’s definitely not clear which of those strategies will be more dominant.

Typically at this point I’ll dive into a discussion of the mechanics in the set, but this article is already above 4000 words, so I decided not to go into that here. However, if you want to find where I discussed those mechanics in more detail, you can find that here.

Conclusions and Recommendations

Thanks again for reading! Here are my conclusions and recommendations:

1.      While Megamorph doesn’t seem very different from Morph on first blush, it turns out that the cards with Megamorph are built very differently. Many of the Megamorphs are early game cards that can be played face down for a bonus in the late game. This is a large fundamental shift in the environment, and will lead to significantly different play patterns. I suspect that this will make the format a little bit faster than Khans of Tarkir.

2.      Dragons of Tarkir has shifted the tempo of the format into earlier stages of the game. The creatures have less toughness in the midgame than in Khans of Tarkir. The set also has a significant tipping point when players hit six mana because they will be able to start playing lots of dragons. These shifts probably mean that the format will be more aggressive than Khans of Tarkir, but it probably won’t be a really fast format either. Also, it means that the play patterns will likely feel like two different games; the early game struggle for board control and the late game dominance of dragons.

3.      Dragons of Tarkir has a lot of dragons. It has more flyers per pack than any regular season draft format going back to at least Lorwyn. Being able to deal with flyers will be essential, and there will be a lot of evasion to end board stalls in the late game.

As always, thanks for sticking through to the end of the article. Remember to check out my blog at oraymw.wordpress.com, which will have additional information about the set, as well as various other topics throughout the week. Follow me on twitter @oraymw for updates and discussions about Magic, and to see when new articles have gone up. Finally, leave any thoughts in the comments below. See you next time!

Ars Arcanum Archive