Monday, July 26, 2010

Fine Tuning your Cube

My brother sent me an Excel spreadsheet with his propose cube on it. If you're unfamiliar with the concept, a "cube" is a home made assembly of magic cards, usually with the best cards ever printed. With enough cards you can separate them into new "booster packs" and play with them without actually having to pay for new product. Also, it's all kinds of fun. But back to the matter at hand. My brother sent me a list of magic cards that he's planning to put in his cube, and I'm going to try to forge those into an excellent drafting environment.

It's not going to be easy. Or short. y'see, I've never done this before, and I've never crafted a draft environment before, so I'll be figuring this out as I go along. Great, huh? Let's get down to it.

Let's talk numbers. Each drafter gets three packs of 15 cards. That's 45 cards per drafter, and fielding a table of eight requires 45*8=360 individual cards. You want more cards than that so there's some variance as to what you expect to see every draft. In this case my brother has designed his cube with twice that many cards, 720 so that he can make either two eight man drafts, or have just one draft with a significantly varying card pool. Ok? now how do you divide that into colors? Five colors, multicolor cards, artifacts and lands. If you just divide by eight we get 90 cards per type. Straightforward enough? Well, until you get to multicolored. You can adjust the numbers by having more artifacts and less land, for example, but one of the stated goals of the cube is that it ought to have interesting nonbasic lands to draft. But all this is worked out already in the spreadsheet; he's got the cards all listed, sorted by converted mana cost and creature/noncreature, and including alternates for all the slots.

What we need to do is figure out if this cube will make for an interesting drafting environment. To do that we'll try to identify archetypes and build decks with the cards. First up:

Red Deck Wins
The idea behind red deck wins (or sligh, or by another name) is to drop quick red creatures backed up by burn and kill your opponent before he gets a chance to get going. Usually hard to do in a draft since people are always splashing your best burn for removal. Still, It's always good to make the red deck first in a new format because that gives you a measure of the speed of the format. The cards:

(1 drops) Mogg Fanatic, Jackal Pup, Greater Gargadon, Magus of the Scroll, Grim lavamancer, Gorilla Shaman, Goblin Welder, Spark Elemental, Goblin Cadets, Goblin Cohorts, Jackal Familiar
(2 drops) Hellspark Elemental, Blood Knight, Keldon Marauders, Mogg War Marshal, Ashling the Pilgrim, Plated Geopede, Slith Firewalker, Stigma Lasher, Emberwilde Auger, Ember Hauler
(3 drops) Magus of the Moon, Ball Lightning, Hell's Thunder, Zo-zu the punisher, Taurean Mauler, Sulfur Elemental, Countryside Crusher
(4+ drops) Rakdos Pit-dragon, Avalance Riders, Anger, Blistering Firecat, Flametongue Kavu, Rakka Mar, Skizzik, Siege Gang Commander, Kiki-Jiki, Mirror Breaker, Bogardan Hellkite
(Tattermunge Maniac, Boggart Ram Gang, Figure of Destiny, Ashenmoor Gouger, Giant Solifuge and Rakdos Guildmage in multicolored, Juggernaut in Artifact Creatures)

That's a pretty solid mass of creatures, and while a number will get grabbed by anybody with a red splash (Flametongue Kavu, Grim Lavamancer, Gorilla Shaman for example), quite a number of those cards aren't any good outside a dedicated red deck wins. Jackal Pups right off the top. Let's look at the spells.

(1 mana) Red elemental blast, Reckless Charge, Chain Lighting, Genju of the spires, Lightning bolt, Dead/Gone
(2 mana) Ancient Grudge, Chain of Plasma, Magma Jet, Fork, Flame Rift, Shrapnel Blast, Lash Out
(3 mana) Price of Progress, Pyromancer's Swath, Blood Moon, Flamebreak, Char, Rift Bolt, Browbeat, Wheel of Fortune, Pyrostatic Pillar, Seismic Assault
(4+ mana) Manabarbs, Relentless Assault, Chandra Nalaar, Fireblast, Jokulhaups, Form of the Dragon, Banefire, Rolling Thunder, Earthquake, Beacon of Destruction)

Huh. There aren't as many spells, and a good number of them are splashable to other colors. Not as good of a sign.

Artifacts: (Note, plenty of mana acceleration, going to skip over it by and large)
(0 mana) Mox Ruby
(1 mana) Skullclamp, Aether Vial, Black Vice, Cursed Scroll)
(2 mana) Umezawa's Jitte, Lightning Greaves, Winter Orb)
(3 mana) Tangle Wire, Trinisphere, Sword of Fire and Ice, Sword of Light and Shadow, Loxodon Warhammer
(4+ mana) Goblin Charbelcher, Nevinyrral's Disc, Sigil of Distinction, Memory Jar

Again, people are going to be taking artifacts pretty highly, and there are more artifacts that are technically an option, especially if you want to splash other colors or accelerate faster.

Lands: The vast majority of the lands (70/90) are color fixing lands, which mostly a dedicated red deck can ignore. It might be worthwhile picking up some red fetches if you've already got a geopede, but by and large we'll ignore them. let's look at the playable utility lands:
Strip Mine, Wasteland, Mutavault, Mishra's Factory, Rishidan Port, Barbarian Ring, Forgotten Cave

Not much there, and almost anybody's going to grab Mutavault and the Factory. The land destruction works better in a red deck than otherwise; in a red deck wins you can expect to play your spells out more quickly than your opposition so trading one land of yours for one land of theirs is usually a good deal. Still, with that many decent utility lands and mana fixing lands the occasional bit of land destruction goes well with most decks.

Ok, so, let's say I'm building the deck. What cards do I include? I'm going to build the deck with only half the cards in the pool, and cut many of the splashable cards.

(1 mana) Jackal Pups, Tattermunge Maniac, Goblin Cohorts, Figure of Destiny, Chain Lighting, Red Elemental Blast, Cursed Scroll
(2 mana) Keldon Marauders, Stigma Lasher, Ember Hauler, Flame Rift
(3 mana) Magus of the Moon, Ball Lightning, Boggart Ram gang, Seismic Assault, Pyrostatic Pillar
(4+ mana) Siege Gang Commander, Giant Solifuge, Kiki Jiki, Avalanche Riders, Chandra Nalaar, Earthquake,
Lands: Wasteland, Barbarian Ring, Forgotten Cave, 14 mountain.

Eminently playable. Got a good number of one drops and a solid number of creatures to follow up with. Jackal pups into Keldon Marauders into Ball lightning into Flame rift or Solifuge or whatnot gives a potential turn four goldfish, even without shoving all the best red cards into the deck. Figure, Chandra, Earthquake and Siege gang give your deck late game power too.

So what does this thought experiment tell us? There are enough cards to make a red deck wins regardless of which half of the cube you pick up. If nobody at the table is drafting red you can make a deck that will goldfish on turn four. So if you're trying to draft a control deck it'll have to be reacting before turn four so that you can stabilize. Whether the deck is legitimate (In those times where nobody else uses red as a main color) really depends on the power level and availability of the rest of the decks in the metagame, which won't develop for some time. In the meantime though, we've got more decks to build.

Goblins in Standard

This Friday I took my Goblin deck out to the local Friday Night Magic, and battled it out over the course of four rounds. Here's the report.

First off, living where I am (Park Falls, WI), the local FNM is an hour away in Ashland, WI. It's a new store, opened up a couple months ago, granting a great improvement over the 2-3 hours I'd drive previously. Hopefully that means I'll be able to attend more often. But anyways, goblins. On to the list!

4 Goblin Guide
2 Quest for the Goblin lord
1 Forked bolt
4 Lightning Bolt
4 Goblin Bushwhacker
4 Dragon Fodder
4 Warren Instigator
1 Ember Hauler
3 Searing Blaze
4 Goblin Chieftain
2 Staggershock
1 Goblin Ruinblaster
2 Siege Gang Commander

2 Teetering Peaks
3 Smoldering Spires
1 Tectonic Edge
18 Mountain

It's a red deck wins, with goblin flavoring. It's very fast, aggressive and surprisingly powerful. But if you're making an aggressive deck like that, don't you want mostly four ofs? So why all the odd numbers? I'm a bit embarrassed to say this, but it's a card availability issue. I haven't been buying most of the cards I've wanted to in the past year, and so I don't have a lot of the commons and uncommons I'd like. For example, I'm pretty sure I could do with more taplands. Their effects are pretty powerful. Not sure I want the full eight but I might. The presence of playsets of the rares and mythics might give the lie to this statement, but I make sure to trade into new, good goblins when they're printed. But anyways, let's get into the cards one by one.

4 Goblin Guide
This guy is good. Really good. Dropping him turn one on the play is easily good for six damage when they're getting their taplands or mana elves online. Later in the game he's still decent with haste and the possibility of pump. His drawback isn't as terrible as I initially thought; you generate enough tempo by swinging with him that the card advantage is less relevant. And hey, if he draws them out of a mana screw you're officially playing a more fun game of magic because of it.

2 Quest for the Goblin Lord
Only own two of these. I'm still not sure about this card, I haven't played the deck nearly enough to decide if the effect is too powerful or too slow. Sometimes it just never seems to get above three counters, and other times it makes an already decent horde into a impressive strike force. It's one of the cards that makes Dragon Fodder playable, so perhaps it'll get better or worse as I acquire other cards as well.

4 Lightning Bolt
What's that? Some of the best burn in a deck that tries to win quickly? I don't think I've ever heard anyone propose that before. Genius!

1 Forked Bolt
Should be a Searing Blaze, I think. Still a decent spell in it's own right, especially if they're playing out mana elves.

4 Goblin Bushwhacker
Now this guy is amazing. Despite his converted mana cost, he's actually a two drop, regardless. Even in a hand with six mountains he does more damage if you kick him on turn two than if you just play him out turn one. He acts as one turn of a lord effect, he gives everything haste, so he's also great on turn four next to a dragon fodder or ember hauler or warren instigator. He goes a long way towards making you feel like you're playing with 30 haste creatures, which is a great feeling let me tell you.

4 Dragon fodder
Once I start getting more of the other playables, I'll probably end up cutting this card out. It's got it's high points; it interacts well with all the lord effects. one Bushwhacker or Chieftain is enough to make these guys into respectable damage, and it certainly helps to get quest for the goblin lord online, and yet every time I look at it I wish it was Mogg War Marshal. Sigh. Still, getting two guys for two mana makes him pretty resilient to removal spells; they've got a wrath or pyroclasm effect or nothing.

4 Warren Instigator
Now this guy is nuts. I mean, everybody and their mother looked at him and thought "Not as good as goblin lackey", but Goblin Lackey is pretty darn good. As it turns out, this guy can be pretty good while still not being Lackey. Moreso than just the whole drop goblins into play thing, double strike is a very powerful ability, especially on a two mana guy. Given the ten lord effects I'm running and the Teetering Peaks, he often gets in for more than just two damage.

1 Ember Hauler
Double mogg fanatic? Sort of. As good as two mogg fanatics? Maybe. Just being a bear with the goblin type on him is decent in the deck, and being able to shock creatures at will is a useful ability. I expect he'll also do a number on opposing planeswalkers, which is a great thing to look for in a creature. Unfortunately, as a one of he didn't come up very frequently, and none of my opponents ran much in the way of planeswalkers, so I didn't get much opportunity to evaluate him. Once I crack some more M11 I expect I'll get the other 3.

3 Searing Blaze
That's right, I haven't been doing enough to crack myself a playset of Jace, the Mind Sculpter. Searing blaze has a significant drawback in that you need to play a land in order to get the most use out of it. So it's mostly a sorcery. That said, it worked wonders for me all night, clearing blockers and letting my goblin guides get in another two damage. And three damage to the player? For only two mana? Are you kidding me?

4 Goblin Chieftain
This guy is also really, really awesome. 1) He gives all your goblins +1/+1. That dragon fodder looks a lot better next to it. So does instigator and his double strike. 2) he gives everybody haste. On turns four and five that's significant if you're dropping your dragon fodders or siege gang commanders. 3) He himself has haste. No matter how clear your board is you can topdeck him and clock your opponent for two, threatening more on the next turn if you topdeck literally any creature in the deck.

2 Staggershock
This card I'm on the fence about. On the one hand, it's 4 points of burn for only 3 mana. Char and Flame Javelin were both excellent cards, so why not this? Well, staggering the damage out over two turns is less effective when you're trying to burn down their creatures and get your guys through. It's fine when you need to clear out a Wall of Omens after it blocks, but if they've got anything with power as well you're trading a card and a half to kill their guy, which isn't a good place to be. It does go well to their face (or their Jace), as those don't lose damage with the simple passage of a turn.

1 Goblin Ruinblaster
Could've used three more. As a 2/1 haste goblin for 3, he isn't bad. At four mana though, what with killing one of their nonbasics he's got the potential to be much better. While I'm not expecting to actively mana or color screw my opponent with this, it does set their gameplan back by a turn, whether that's laying a Sarkhan the Mad or a Sphinx of Jwar Isle. Either way, I'd prefer to lay some cheap dudes early and kill them before the game gets that far.

2 Siege Gang Commander
Another insane character. So sad that he's rotating in three months. Anyways, beyond the fact that he gets you four bodies for five mana, and that he lets you fling your team at those who would oppose you, this guy also interacts well with a goblin chieftan the turn before or a quest for the goblin lord; finishing the quest and making an impressive strike force with the boost. Oh, and did I mention being able to drop him with Instigator on turn 3?

1 Tectonic Edge
Again with the slowing down the opponent, although I'd prefer more ruinblasters to more edges. The edge makes colorless, which leaves you more vulnerable to Spreading Seas (the deck is mono colored, but most of the spells cost RR. Bushwhacker, Chieftain, Searing Blaze, Siege gang, so on.). It also is worse if you get an on curve hand.

2 Teetering Peaks
Worked out very well for me. Yeah, it's sorcery speed pump, but that comes with some pretty good stuff. Stick it on an instigator for four damage. Stick it on a dragon fodder token and force them to use their removal on it instead of your Chieftain, or take two damage. Very playable on turn three since the deck has many more two drop creatures than three drops.

3 Smoldering Spires
Makes a creature unblockable? Nice. Again, very good on turn three, especially with an instigator on the table. Why yes I'd like free goblins. I don't like including taplands in aggressive builds, but in this case I think the extra damage they offer justifies it. Not sure about playing with the full eight though.

18 Mountains
Dave Price tells me this is the best card in Magic. Who are you to argue with Dave Price?

So those are the cards. Allow me to lay out a couple lines of play with the deck; hands that came up when I was playing it.

Turn 2: Warren instigator. Turn three, Smoldering Spires making his creature unable to block, Goblin Bushwhacker. Swing in. On first strike damage, drop a goblin chieftain, pumping both attacking creatures. You do a total of eight damage on turn three, you can easily make that lethal on turn four.

Or Turn 3 Goblin Chieftain. Turn Four Dragon Fodder, Dragon Fodder, swing in for ten. That's only three cards!

Or one game, off a mull to five, on the draw: Turn two Instigator, Turn three bolt his Hypnotic Specter, use Instigator's ability to drop for free Siege Gang commander and Siege Gang commander. That got a concession.

Or one game, turn one quest, turn two bushwhacker, burned out. Turn three bushwhacker, burned out. Turn four Goblin chieftain, turn five Goblin Chieftan, Goblin Bushwhacker, activating the Quest. Swing in with three 6/3 goblins.

Quite literally more than half the creatures in the deck have haste, and more than half of that more than half give your other creatures haste. The deck can drop a nasty hand even after a board sweeper with just one or two cards, and hit you with it before you get a chance to untap. It'll steamroll you before you get a chance to do anything, unless you come out fighting and fighting hard. It's a real joy to play.

Monday, July 5, 2010

Maximizing Damage

Today we're going to learn how to maximize damage. I developed this theory first with respect to World of Warcraft, but the assumptions required are surprisingly general, and apply to all sorts of games. This will require math. College level math. If that's not your cup o' tea, then thank you, please come back another time. Or read over it and hope that you can pick through the details. I'm also going to be going through the simplest version of the calculation, with the simplest assumptions to show how it works. You can add more assumptions to make it more complicated and get better results if you'd like. Anyway, on to the problem!

Suppose you're playing a RPG. You have the option of selling your Sandals of Water Skiing and acquiring some War Boots of Goose Kicking. Is the deal worth it? Which item will allow your character to deal more damage, survive more hits or what have you? Well, I'm just going to assume you want to deal more damage, that you'll worry about defenses and peripherals (the sandals let you freaking water ski!) elsewhere. So how do we measure how much damage you do? Well, let's take a look at a fight. Or better yet, one attack with your weapon.

You swing your sword, what happens? Well, it hits, or it misses. He might dodge, block, parry, evade or what have you. And, if you're lucky, you might get a critical hit. Now, how much damage do you do with each of those outcomes? Well, if you hit, you'll do H damage, let's say. If you miss, you'll do M damage, and if you crit, you'll do C damage. Got that? This part's important: We can figure out how much damage you do on average by multiplying the probability of a particular swing type by the amount of damage it does:

1) Avg. Damage = Hh+Mm+Dd+Bb+Pp+Ee+Cc

Where the upper case letter is the amount of damage and the lower case letter is the probability of getting that outcome. It's also important to note that the probabilities are stochastic, that is, they add up to one. Or rather, that there's a 100% chance of getting one of those outcomes, and that there aren't any other possible outcomes. If we swing, one of those things will happen. Mathematically:

2) h+m+d+b+p+e+c=1

Ok? Now, I said I was going to keep this simple, so we're going to cut the options down to two possibilities, that you hit, and that you get a critical hit. Rewriting those equations (respectively) we get:

3) Avg. Damage = Hh+Cc

4) 1=h+c

Ok? See, this is the problem with physics. We get a problem, and then we have to assume away all sorts of stuff to make it solvable, and we end up with massless frictionless pulleys dropping weights in a vacuum, and it doesn't correspond with the real world at all. Or real video game in this case. But yeah, I fully intend to add this stuff back in at the end, so bear with the simple version for a bit. Anyways, we're gonna want to stick those two equations together. Solving equation 4 for h we get

5) Avg. Damage = H(1-c)+Cc

Which makes sense, as the probability of getting a hit is exactly the probability of not getting a crit, since those are the only two options we allow. The next question we have to ask is "What happens when we get a crit?" Here I'm going back to explaining in terms of World of Warcraft, since the mechanic in WoW is extraordinarily simple: A crit deals twice as much damage as a hit. Simply:

6) C = 2H

Substituting that back into equation 5, we get

7) Avg. Damage = H(1-c)+2Hc


8) Avg. Damage = H(1+c)

And that's your average damage, in all it's glory. If you know your average hit damage, and your chance of getting a critical hit you can figure this out. Well, actually it's your average damage per swing. If we divide your average damage by your attack speed (how many swings you make in a second) we get your damage per second (DPS)

9) DPS = H(1+C)/attack speed = A(1+c)

Where I define "A" as your hit damage divided by your attack speed. Same basic equation, but now it also accounts for weapon speed. I'm going to assume your attack speed remains constant, but later I'll also discuss changes to your attack speed.

Here's the tricky part. We need to maximize this equation. I mean, I can see that increasing your basic damage (H) increases your DPS, and increasing your crit chance (c) increases your DPS, but which one increases it faster? Well, multivariable calculus gives us a function for maximizing functions of more than one varaible. We take the gradient:

10) grad DPS = <1+c, A>

Ok? That gives us a vector, which points in the direction that'll increase the function most. Or decrease; the canny reader will point out that it's possible that it'll show the direction the function will decrease the quickest, but in this case it's definitely increase. You can check it yourself. Basically, the vector points in the direction of increasing c and A, and we already stated that as they increase, so does your damage. Anyways, let's move along.

Now that I have the gradient vector, well, what am I supposed to do with it? That's an excellent question, I'm glad you asked. Well, the gradient points in the direction of the maximum increase in damage. So if you boost your crit chance and your damage in precisely the ratio that the gradient dictates, you'll be boosting your total damage the most with the same amount of stat points. If you're getting stats along a different vector, then we can compare the relative efficiency of the vectors by seeing how closely they align with the gradient.

Ok, so where are we getting these other vectors? We get them from items. Items? Yeah, at the start I mentioned that I was trying to compare items; this is where it comes in. If I've got some boots that give 42 attack power and 1% crit, and I've got some other boots that give 70 attack power, which one is better? Well, you can render them down into item vectors, as long as you're talking in the same units as in equation 9, that is A in damage per second and c in crit%. In this case we have to remember that

11) 1 dps = 14 attack power

So we can set up two item vectors like this:

12) Item 1 = <3 , .01>

13) Item 2 = <6 , 0 >

There's a mathematical quibble here; can we treat items as vectors? Yes, yes we can. So long as we stay away from some not-gonna-happen ranges of data we're fine. (negative attack power or crit chances, over 100% crit chance, sticking more than one item in the same slot, that sort of thing.)

But on to the real problem. How do we compare item vectors to the gradient vector? We take the dot product. I'm gonna switch in bold letters to represent the contribution from the items, so A is the dps the item gives you, and c is the item's bonus to crit chance.

14) Item (dot) grad dps = A(1+c) + cA

If you sum that up, you get a number. Take that number and compare it against the number you get from other items. The largest number wins; using that item most increases your dps. If you keep doing the calculation you can select your items to maximize your dps. Neat, huh?

Ok, now step back a bit. What else does this apply to? Buffs? If you're trying to decide between blessing of might and blessing of kings then you can work them like items, make vectors and add it up. Sure it's probably going to be kings but it'd be nice to be sure. Ok, so you can also measure buffs. What about talents? Some basic ones yeah; + attack power or + crit chance talents, although odds are you'll want to get both. More? Skills, we can evaluate the efficacy of skills. All those numbers compare "white" damage, that is the damage you do when you automatically attack in each fight. "Yellow" damage from skills can be calculated much the same way.

What about more complicated stuff? Talents that give you benefits from increased critical hits, that sort of thing. Yeah, it can be done. It's a lot more complicated, it's manageable though.

Now think for a minute about the original assumptions; that there are only a couple possible outcomes of a swing, hits, crits, misses and so forth; that we can calculate the amount of damage in a critical hit, and that we can vary either damage or critical hit chance. Those are very general conditions, they can be applied to a vast range of potential games. Or other calculations in this game; you can figure out damage mitigation for tanks or rage generation or such.

Ok, there's the calculation and the big picture. I'm going to go through it quickly with miss chance and changing your attack speed.

Now, in equation 1) up there I posted factors for other outcomes to the swing; dodges, blocks, parrys, etc. Here I'm going to combine them generally into a "miss" factor

15) Avg. Damage = hH + cC + mM

And again those are the only possible outcomes.

16) 1 = h + c + m

Combining those two:

17) Avg. Damage = (1-c-m)H+cC+mM

Ok? Again C=2H, but this time we also know that M=0, the definition of a miss being that it deals no damage. Substituting these in

18) avg. Damage = (1-c-m)H + 2cH+0 = (1+c-m)H

Ok? We're also going to deal with the time factor, which we can't hide with a change of variable. Giving "T" as your attack speed:

19) dps = (1+c-m)H/T

Ok/ Now we take the gradient again

20) grad dps= < (1+c-m)/T, H/T, -H/T, -(1+c-m)H/(T^2) >

Four dimensional vector now since we're dealing with four variables. It's a little harder to read, what with the negative numbers. The vector is in order of Since you're attempting to decrease your miss chance and attack speed, so an item that gave you +1% hit would read like -1% miss. Here, let me make an item vector for you, with 1% hit and 3% faster attack speed

21) <1, .01, -.01, -.03>

See? The negatives will multiply by negatives and you'll get a larger dot product that way. So yeah, here you take the dot product again.

22) H(1+c-m)/T + cH/T -mH/T - T(1+c-m)H/(T^2)

Friday, July 2, 2010

Rebuilding Raksha, Golden Cub

This is another EDH deck that could use some retuning. This one has Raksha Golden Cub as the general. If you don't remember it in "Decks that I once loved (I and II)" it's because it was properly put away, not cluttering up my desk. It's mono white. Let's go over what's already in it, this time sorted by mana cost.

Lands: 38
That's a bit low; I don't have the artifact acceleration like in my Progenitus deck, and this one still curves up to nine mana spells. If I recall correctly, the main complaint I had with the deck before shelving it was that I never actually got to cast Akroma. The fact that I was running a mana light deck might explain all that.
Darksteel Citadel: Perfectly fine in a monocolored deck. There are a lot of board sweepers played in EDH, a few of which will also kill off the occasional land. Being indestructible means I still have one mana after the meltdown. Nice, even if it doesn't matter that much.
Windbrisk Heights: Let's see, the engine that powered a standard deck for two years? Good if I've got a way to attack with three creatures. I have? Excellent.
Terrain Generator: I like this card in a deck with mostly basics.
Urza's Factory: For T,7 I can make a 2/2 token. Yeah, it was an important 1 of in the Teachings matchup in Time Spiral Block Constructed, but it is a slow kill condition.
Forbidding watchtower: And here's one that's even slower.
New Benalia: I guess scry 1 is nice?
Secluded Steppes, Drifting Meadows: Cycling is good. I like lands that cycle.
30 Plains: More interesting than your standard plains, actually. Contained within this stack of lands is one of the decks we prototyped for a playtest of one of our Awesome Games. So a number of these sleeves contain a basic plains as well as a bit of paper with other rules text on it:

Blue Card
Look Ma, I'm on TV! An appearance on a friendly news show does wonders for your organization.
Gain 3 media.

When it became apparent that I didn't have enough sleeves on hand to have both a playtest game and an EDH deck I went with both. Anyways, back to the matter at hand.

1 drops:
Skullclamp: What can I say about skullclamp that hasn't been said already? Quite a bit, actually. Skullclamp is good because it trades one sort of resource for another. It trades creatures in play for cards in hand. Cards like that are dangerous because you have to be very accurate with the rate. If you don't make the rate punishing enough then the card is broken and bad things happen. If you make it too punishing then nobody plays with it. Clamp was printed too much on the broken side of things. So if it's broken enough to be unfun, why am I playing it?

Two reasons: First off, this sort of card is extremely skilltesting. Yeah, you can get by off of power alone, but if you're really good you'll do a lot better with it than you would as a mediocre player. Randy Buehler famously said that Necropotence made him good; learning exactly how much life to trade for cards made Buehler the powerhouse that he was. The second reason was that I wasn't playing when the clamp was legal. Insanely broken card? Let me in on the action!

Viridian Longbow: Aside from the clamp this deck runs an equipment theme. Viridian Longbow is one of the weaker members. If I want to kill creatures I've got wrath effecs. If I want to kill players I'm usually better off attacking. (Yeah, those two excuses don't go together that well, I know.)

Bonesplitter: Equipment theme. You read the text box on Raksha Golden Cub? Vigilance, As long as Raksha is equipped, cats you control get +2/+2 and gain doublestrike. I want equipment around to stick on her so my other cats get huge. Not that I have many other cats in here as of this update.

Soul Warden: Classic Multiplayer one drop. Gains a lot of life. Expecially if you put it in a deck like this, which has a strong token theme. (This deck has a strong token theme).

Weathered Wayfarer: This guy is good; great at getting you more lands. He even gets nonbasics. Y'know, when I really need that Forbidding Watchtower. Anyways, I'm still leery about putting in too many search effects. This is one.

2 drops:
Jotun Grunt: hoses the occasional graveyard. Not nearly as effective as I'd like.
Cloak and Dagger: +2/0 and Shroud. No rogues though.
Knight of the White Orchid: Another good land search. That's two.
Auriok Steelshaper: Makes things cheaper to equip, pumps knights and soldiers of which there are plenty. (All three of my two drop creatures, for example.)
Veteran's Armaments: Gives a massive pump to one creature in a swarm. Also equips free to soldiers, like that steelshaper.
Lightning Greaves: A prince among equipment. More than once I had this and Armaments in play come turn seven. Play Raksha, Equip armaments for free, then equip Lightning greaves. Swing in with a 6/7 doublestrike vigilance shroud haste creature. That deals general damage. Good times.
Planar Collapse: A wrath effect for two mana? Sure I have to wait a turn and sure there have to be enough targets for it, but a deal is a deal!
Armillary Sphere. That's three search effects
Empyrial Plate: Another good equipment. Good if I've got a full hand, so I guess it just combos well with skullclamp. Like there's anything that doesn't.
Leonin Sun Standard: Multi pump effect for armies of creatures. Going to get to those armies of creatures sooner or later, I promise.

3 drops
Oblivion Ring: I could go either way on this card. On the one hand it's still just a 1 for 1, and people can kill it to get their offending permanent back. On the other hand, it answers most everything.
Devout Witness: Spellshapes Disenchant. (T,1W, Discard a card: Destroy target artifact or enchantment). Useful ability, especially when it repeats.
Order of Whiteclay: Gets creatures back from my yard, as a repeatable effect. Maybe he's why Viridian Longbow is in here.
Story Circle: Just hoses some decks. I remember one game where I had this keeping off a host of dragons from Dragon Roost. I lost that game when he got more dragons than I had mana to keep them away. Might cut this card, it can really annoy people.
Mirror Entity: What's that? Did I hear there was a token theme to this deck? Well you're in luck Mr. Douglas.
Prison Term: Targeted removal, but you can switch it to better and better creatures.
Aura of Silence. Talked about this card in my Reaper King articles. From now on it's officially Aura of Science!.
Mobilization: Finally a token horde producer. A very, very slow one. Still combos with skullclamp.
Nuisance Engine: Also produces tokens. Pest tokens. tokens that chump one by one while I'm waiting to find...
Loxodon Warhammer: +3/0, trample and lifelink? Also an excellent equipment.
Whispersilk Cloak: Shroud and unblockable. Neat.
Fireshrieker: Equipped creature has double strike. Great with the big creatures that are also in this deck.
Sculpting Steel: Copies clamp. also anything else that might be around. Will usually end up as an equipment of some sort.

4 drops:
Sigil of the new dawn: Repeatable Raise Dead if you have mana up at the right times. Not as good as I've hoped.
Auriok Windwalker: 2/3 flyer that shifts equipment around. Made it a wizard, in the same set as Raksha and two sets after Steelshaper. Really? Really? out of three relevant creature types you had to make it a wizard?
Breath of life: 4 mana to reanimate something. On the one hand, four mana. On the other hand, there are targets for it.
Kithkin rabble: Theoretically gets huge if you've got a token horde. But if you've got a token horde, what do you need a huge guy for? Better because it also counts whatever enchantments I might have, but it still encourages me to overextend into an opposing wrath effect.
Springjack Shepherd: Chroma goats. You know, so far the token producers I've gotten have been sub par.
Cradle of vitality: Makes creatures better if I've got the mad lifegain going on. Do I have lifegain, and in what condition is it's sanity? Soul warden (repeatable, but in 1 life chunks) and the warhammer so far. More coming up, like...
Faiths Fetters: An excellent control card. Kills somehting, stops it from activating it's abilities and even gets you +4 life when it enters play. Neat!
Nevinyrral's Disk: I've been talking and talkinga bout how this is a token deck with equipment in it, so it makes sense I've got a card in here that kills all of both. Probably gets shipped to a deck where I need answers more.
Ajani Goldmane: Supports tokens, supports Cradle of vitality and draws hate away from me. Planeswalkers are like that.
Galepowder mage: Negates blockers and reuses comes into play effects. Like springjack shepherd.
Endless Horizons: Number four in our shuffling series, but this one I don't mind since whenever it hits play, inevitably it's going to thin the deck down to where it's easy to shuffle. But Endless horizons is interesting for other reasons. If it works, it's a free card every turn, and all the mana you'll ever need. And it means you're drawing spells from here on out. But it's when it doesn't work that's the problem. I can see that jerk grinning from here, you know the one, the guy who just can't vindicate it fast enough because he knows you'll never play Akroma now. Still, I really like this card, even if I might only want to play it with seven or eight lands in play.
Taj-Nar Swordsmith: Kicker X, gets you equipment costing X or less. Probably should put him in the 5 mana colum since we all know he's getting clamp. Also a cat. Shuffle #5

5 drops:
Archon of Justice: He's big, he flies and he makes people really regret messing with my stuff. He's in.
Battlegrace Angel: Also big, also flies and works with cradle.
Boon Reflection: Also works with cradle. Still don't know that I have enough life gain effects to make either worthwhile.
Scourglass: Like Disk, only it leaves my equipment around.
Death or Glory: Half of a living death. The worse half. Haven't gotten a chance to play with it, it might be fun.
Knight Captain of Eos: Three guys for the price of one. Not as good as...
Cloudgoat Ranger: Even more guys for the price of one. Finally something that looks like token production.
Stonehewer Giant: Searches up equipment, and sticks it on something. Yeah, he's good in this deck. Search #6

Six drops:
Crovax, Ascendant Hero: Pumps tokens, shrinks opposing creatures, takes out the occasional Birds of Paradise and is even hard to kill.
Adarkar Valkyrie: Another angel, this one that raises things from the dead. Again, how can I say no? It's easier when I remember how many times exactly that didn't do a thing for me.
Austere Command: Customizeable board sweeping. Generally the deck has stuff that fits into all four categories on this card though.
Spirit of the Hearth: Another 4 power flyer. Was it two for one tuesday in the rare bin? Anyways, giving me shroud is neat, and he brings our cat count up to two.
Purity: 6/6 flyer for six, has the chance to give me life if someone's playing burn. Nobody's playing burn in multiplayer. Except this one deck... nevermind.
Jareth, Leonine Titan: Blocks anything, even most eldrazi. Hard to kill. Also our third and final cat.

Mass Calcify: For seven mana it destroys all nonwhite creatures. Wipes a number of creatures while leaving mine intact. Hopefully nobody else is playing mono white.
Sigil of Distinction: Ok, technically this is an X mana artifact, but it gets better the more land I've got in play so I've got it over in the X spells. Doesn't work well with the searchers since they both put it into play, but it can be huge.
Benalish Commander: Also an X spell don't be fooled by the 3W mana cost. Basically works like Kithkin Rabble once he hits play, but he has the advantage of giving you a soldier for every turn before that. Not the best guy in the cycle but servicable.
Myojin of Cleansing Fire: 4/6 creature, hard to kill until he kills everything else. Neat. Eight mana though. Eight mana buys you a lot, for example...
Akroma, Angel of Wrath: You know what? Let me just quote it's text box: Flying, First Strike, Trample, Haste, Vigilance, Protection from Black, Protection from Red. No rest, no mercy, no matter what. Akroma is just awesome.
Evangelize (With Buyback): It steals creatures and it's a repeatable effect! How awesome is that? Too awesome; it can really make people want to quit playing, which is exactly what I don't want them to do.

And, of course, Raksha, the Golden Cub. The cat lord who has all of three cats to buff. Whoo

Ok, that's the composition. Let's see what all we've got: 38 lands, 27 creatures (Plenty of soldiers, knights, not many cats. More wizards than I would like. Bunch of angels, spirits, avatars. Possibly enough clerics to support some tribal effects. Definetly enough to support steelshaper's buff.), 17 artifacts (11 equipment, 5 non, and Steelshaper's Gift which can be either), 11 enchantments, 6 sorceries and 1 planeswalker.

So the question is, what do I want to do with this deck? As I see it there are three directions I could take it. One is to go more in the way of soldiers and token production, building a massive army and attempting to swarm people. The second is to make it more of an angel theme deck; hopefully more midrangey. The third is to take out a lot of the worse creatures and try to make it a control deck. I've got more wrath effects that I can put in.

Currently, I've rebuilt it into more of an aggro deck. I've cut some of the worse, higher casting cost cards and replaced them with lower casting cost cards that are still bad. Err, anyway. It's doing better as an aggro deck, but could still stand some tweaking. I'll talk about that in a future post.

Short version: Needs more spectral procession.