Friday, 18 July 2014

Can You Draft Obelisk of Urd?

In Magic 2014 there was Door of Destinies which was a pretty odd card to include in a set with no strong tribal theme. In Magic 2015 there is Obelisk of Urd which also requires you to have some creature types matching to be good, but which demands a lot less of you.


Sure the Obelisk costs six to the Door's four, but it has convoke, and you don't have to get it into play before you start playing your creatures.

So let's take a look at creature types in Magic 2015 to see if we can get anything done with the Obelisk.

Commonality Weighting
When I reviewed Door of Destinies I did some math on how common creatures of each type were. I counted the number of creatures of each type that were in each colour and commonality. Then I weighted the commonalities. There are 10 commons and only 3 uncommons in a pack, but there are also fewer uncommons in the set than there are commons, so any individual uncommon actually shows up about half as much as any given common, not less than a third as much.

This time I improved my approach. Instead of weighting based on how many different ones there are, I weighted based on how many would likely see play in a draft. I went through the list of Mythics, Rares and Uncommons counted the number that probably wouldn't be played in limited. I considered all lands "unplayable" not because they wouldn't be played but because the wouldn't count as one of the 23 non-land cards you put in your deck - that is, they wouldn't take up a slot that would otherwise be occupied by a common. I assumed that all playable Rares and Uncommons in a draft would be played.

I counted "unplayable" mythic rares in a special way. Since these often get taken for cash value, I counted all mythics would a double colour requirement as being 60% unplayable. I also counted big Garruk as 60% unplayable assuming you would splash him in a black or green deck.

So we know on average you get 30 commons, 9 uncommons, 2.525 rares and 0.375 mythic rares per draft. On average you aren't going to play 0.175 of those mythics, 0.764 of those rares, and 1.469 of those uncommons. When you put the rest into your 23, you will have 13.4 spots left for commons. So each card in your deck has abour a 58% chance of being common, 33% of being uncommon, 8% of being rare and a little under 1% of being mythic.

Divide that by the number of playable commons, uncommons, rares and mythics in each color and you get the approximate chance that a card is a given card of that type. We can multiply that by the number of times a creature type appears at that rarity and get the weighted value of that creature type in creatures per draft deck. I validated these numbers by adding them up to estimate how many total creatures you'd get in your draft deck based on them. They look a little lowish - estimating 12 to 15 creatures in a deck based on colour, so I increased all the numbers by about 20% to account for a bias towards creatures over spells.

One thing I noticed doing this analysis is the Walking Corpse isn't actually in packs. Its collector number is 278/269. So much for Necromancer's Stockpile.

Common Creature Types
Unlike Magic 2014, which had slivers, there are no stand-out common creature types in Magic 2015. The most common creature type by far is humans. A deck of 23 white cards would have 6.3 humans, blue would have 2.9 and red 3.6. Thus a two colour white/red and white/blue deck would average 4.6 and 5.0 humans. Mono red gets 5 goblins, mono black 4.8 zombies and mono white 3.1 soldiers and clerics a piece.

None of these are very high numbers. Four creatures of a type aren't going to make the obelisk very appealing. So let's look instead at combined creature types. Unfortunately there is very little intersection of creature types between colors. Not a single creature type other than human averages more than two appearances in a mono-color deck for two different colors. That means when you combine two colours you don't get any significant build up of creature types. A black/red deck will have two or three zombies and two or three goblins, but you also won't get much overlap of other types that might have been better shared like warrior.

So let's look at the case of a mono- or almost-mono- white deck. Such a deck might have 6 humans, 4 clerics and 5 soldiers. Is Obelisk good for that deck?

I think we'll guess that Obelisk is solid if it is hitting two of your creatures with the promise of possibly getting more. Since you can choose whichever type you drew the most of to Obelisk for, the question is how likely you are to draw a pair from that deck.

Well let's say you are hoping to put down the Obelisk on turn five with an assist from a creature. You might be able to put it down on turn three or four but tapping all your creatures to slam a card that buffs your creatures is typically a weak tempo play. On turn five you are looking at 11 or 12 cards.

This probability is very hard to calculate because in reality many of those clerics and soldiers are actually going to be your humans too. But let's pretend that wasn't the case. That gives at 87% chance to hit a pair. That's actually really high.

So lets imagine you get a deck that's a little bit more reasonable. You get 5 humans, goblins or zombies, and 3 of each of two other types. That still gives a 67% chance to buff two creatures.

Plus, unlike the Door of Destinies, Obelisk does something with only one creature. Six mana convoke may be a pretty high price for a +2/+2 aura for one of your creatures, but for a card that can have a high upside, it's nice that it isn't useless when the upside doesn't happen.

Token Generators
Something I've been ignoring up to those point is that Magic 2015 has token generators. I think we can rule out Sliver Hive/Obelisk as an untenable situation, given that it is two rares that then require you to find another combo piece to get going. But there are some other rares that if you happen to combine with Obelisk it is very nearly lights out.

Hornet Nest and Hornet Queen are both quite the dreams if you can ever live them. Chasm Skulker would do the trick and you might even have some other squids from Coral Barriers. Spirit Bonds on turn two with an Obelisk in hand is pretty much a game won. Goblin Rabblemaster is going to present a real challenge when the Obelisk is on his side.

But you aren't going to take Obelisk in your first three picks out of the first pack in the hopes of finding another rare. At uncommon we have Feral Incarnation, but that costs a lot and since there are virtually no other beasts in the set I don't see that as very realistic. First Response is potentially very powerful with an Obelisk, and has the advantage of being white so naming Soldier for Obelisk could be doing other work.

The real story, though, is the commons. One of those soldiers in white is actually two soldiers in Raise the Alarm. Much more shockingly, white has Triplicate Spirits. A first turn drop into raise the alarm would allow a third turn spirits and a fourth turn Obelisk giving three attacking 3/3 fliers. That's a very specific draw, but one that relies on a bunch of commons and the Obelisk. And if you don't pull it off until turn five, six, or even later, those three phantom monsters are still going to be good.

Conclusion
Obelisk of Urd looks is somewhere between fine and absurd, but the secret is that for the most part it is a white card. If I were going to draft Magic 2015 I would be very tempted to first pick an Obelisk if there were no other real stand-out cards. After that I'd be heavy into white hoping to get raise the alarms and spirits. I think the creature counts in white can totally support this card, particularly when you know that's what you are trying to do.

Triplicate Spirits is presumably just a great card anyway and I doubt it will be coming around late. Even so, as a common you can probably get one or two, and you can name human or soldier with the obelisk when the spirits don't show.

No comments:

Post a Comment