Sunday, 12 July 2015

Magic Origins Draft Set Analysis Part I: Overall Numbers

For a similar look at things which focuses a bit more on historical comparisons, I recommend Matthew Watkins's excellent Ars Arcanum

In this post, I want to look at the overall numbers which define this draft format. Questions like "How many creatures are there?" "How big are they?" and the like will be answered.

As for methodology, all of the count numbers I am giving here are based on a per draft basis. What this means is, the number I show will be the number you will expect to be opened in the entire draft for one 8-player draft (on average). For example, the number of 2-drops per draft is 38.6, which means that in the average draft, 38.6 creatures with a Converted Mana Cost (CMC) of exactly 2 will be opened by all the players at the table across the total 24 packs.

There were a number of judgment calls I had to make along the way. Is Harbinger of Tides a 2 CMC spell or 4? (I counted it as 2). What about X spells? (I counted them with X = 1). Obviously the rules view these a particular way, but here I am more concerned about how they actually play out - I tend to try to put everything as cheaper here, as you can cast the cards there, so that's where they start to make an impact. Furthermore, I tended to count all the variable-sized things as their smallest version (so renown is off), with the one noticeable exception of Revenant, which I labelled a 3/3. In all of this, there was some guesswork, but I tried to do my best to account for variability. I make this note here, though, so everyone is aware that this stuff will change things, though because these cards aren't very plentiful, it won't have a big effect on the overall format.

With that out of the way, let's start looking at things. We'll start with the mana curve of creatures. Here's a chart of the CMC of all creatures broken out by colour:


Here's the same chart, but without the overall count (so as to be easier to see color-by-color):


So we see a plurality of creatures in the set cost 3. G and W have the most cheap guys. U somehow has a glut of 4-drops.

Next, we'll take another pass at the creatures, but this time by power and toughness instead of by CMC:


Here we see that a big percentage of cards in the format have 2 power and/or toughness. The toughness curve is just to the right of the power curve, so the back-ends tend to be just a little bit higher than the fronts. We can see, though, that there are very few creatures with either power or toughness greater than 4. Given that there are 8 players in the draft, this works out to around 2 such creatures per player per draft, and given that some of these will go unplayed (due to ending up out-of-color or actually unplayable, as several of them actually are), most games you won't see one. If we extend this down to include 4 p/t critters as well, it comes to closer to 3 creatures per player by power, and 5 by toughness. This likely means just under one per game in terms of power, and a bit over one by toughness.

Now, let's break this down by color, with separate graphs for power and toughness:




Black has a higher-than-average percentage of the 3-power creatures, but the big news here is that blue is blowing the 3-toughness out-of-the-water, at 38% of the total.


Next, let's combine these results to look at the typical size of a creature for its cost:



These averages are, once again, weighted by their frequency in packs (i.e. commons count for more than rares). We can see that the slopes are generally pretty gradual (and most of the big spikes are because there are very few different cards and that color and CMC combination), and the biggest upward spike happens at 6 CMC - although you can definitely argue that 2-3 CMC creatures are more efficient in some absolute sense.

Last but not least for creatures, let's look at my favorite little image for checking out the vanilla stats in a format, the Eat/Bounce/Trade Chart:


The way to read this is that, if you have a creature of the size of one of the top two rows, it tells you how many creatures at a given CMC either eat that creature, bounce with it, or trade with it, from each of the three sections, respectively. For instance, if I have a 2/2, at a CMC of 2: 2.37 creatures per draft will eat it (from the top section), 4.752 creatures will bounce with it, and 22.525 creatures per draft will trade with it; at a CMC of 3, that shifts to 9.44 eating it, 10.59 bouncing with it, and 30.30 trading with it.

I do want to note that none of these charts take abilities like first strike or double strike into account, relying instead on pure size. In this format, you need to account for a shift of two commons, two uncommons, and a rare (all in white and red).

Anyway, the big takeaway here, to my eyes, is that there's very little that interacts favorably with a Bear. If you're on the play, the only card that would eat it at 2 mana or less actually ETB tapped (Shambling Ghoul), so you're actually down to Consul's Lieutenant and Knight of the White Orchid. Even moving up to CMC 3, you're only at about 2 cards per player per draft, and again, some of these are going to be stuck in sideboards as their drafters will shift to other colors. So it looks like the real way to deal with these things is to trade, which leads me to think it's going to be fairly important to have a pretty reasonable number of 2-drops in the format. I don't think you necessarily need to have a deck chock full of them, but you are really going to want at least a few - and there aren't all that many to go around.


Finally, let's take a quick look at the interactive spells en masse:


I don't know that this tells you all that much really, but the big glut at 4 CMC is interesting - basically every color but blue has a common removal spell, so there you go.












No comments:

Post a Comment