starcraft 2 and magic

Well, here's your standard apology-blogpost for not updating. I moved across the country and haven't been playing much Magic- there are a few guys here that are interested in learning, so we'll see what that brings. I'm hopeful.

Before I moved, I was playing the Starcraft 2 beta. Obviously amazing. I've been watching a bunch of videos about high-level strategy and play in it, and it's really fascinating to see all the perspectives and ideas on an entirely different competitive game, with its own jargon, metagame shifts, and general ideas.

What's really interesting to me is the lack of theory there is about the game- in video games, it's just not needed. You don't need to come up with elaborate concepts; things happen in real time, units take X time and Y resources, which are harvested at Z rate. If a player kills someone else, you can watch the replay and see exactly why. Pretty straightforward. There is one broad concept that's interesting to me for its application to Magic, though: the idea of a "timing push" ('push' basically just means an attack). The idea is to do your attacks at a timing that is either very convenient for you- such as when you're throwing down an expansion, or a key upgrade just finished as you're waltzing across the map- or extremely inconvenient for your opponent, such as when they're teching to something incredibly powerful, but either don't have it yet, or don't have them in any significant number, or when they just finished an expansion and haven't broken even on it yet.

How does this apply to Magic? When you're thinking about how matchups work, especially sideboard cards, don't just think of what your cards are doing, but when. Okay, that card is a fantastic hoser for your opponent's entire strategy... but it comes out when you're two thirds of the way to dead. Instead, you should try to fill in the gaps in your own tempo, and look for points where your opponent isn't doing that much that's very powerful, or they have to do a bunch of awkward things and can't react to what you're doing. This means you want to look for cards that come right at the moment when your opponent has to tap out for some spell that's usually safe to cast, and you want to play three-drops against decks with Cryptic Command or Thirst for Knowledge (so that they have to decide between countering it and drawing cards). Say you're playing an aggressive deck or Elf combo against scapeshift. Their two most important "timings" are turn three for Firespout (post-board, sometimes pre) and Cryptic, and usually winning on turn five. This means that you'll want to cast small guys (hopefully disruptive ones) after their turn three to put them in a maximally awkward position, and they can't stop your entire turn with Cryptic.

This is drifting toward very traditional theory, but I'll write more as I think more.

ptq post-mortem

Short version: 0-2 drop, to Brozek Boros both times.

Slightly longer version: I played the 75 in the last post, except I thought that Burrenton Forge-Tender would be better than Loaming Shaman.

Round one game one, he gets going with one, then two, then three goblin guides, but doesn't have the removal to stop me from comboing on turn four. I sideboard:

+4 Jitte
+3 Ranger
+1 Warcaller
+1 Viridian Shaman
+1 Burrenton Forge-Tender
-3 Cloudstone Curio
-1 Viridian Zealot
-1 Boreal Druid
-1 Arbor Elf
-1 Elvish Visionary
-1 Elvish Archdruid
-1 Summoner's Pact
-1 Primal Command

Pretty messy what I took out. Games two and three, he mulls to five and still beats me. Mostly I blame Jitte for this; I thought it would be crazy against a deck of almost nothing but X/1s, but what I didn't realize is that in this matchup, Jitte does not have the text printed on it. Instead, it is errata'd to have the following ability instead:

2, sacrifice an untapped creature: target player discards an instant.

...which is obviously a pretty terrible card. There's absolutely no way I should have cut elves at random to put in that thing; it's just way too slow and expensive for a deck that can just go "kill you." The fact that he beat me on mulls to five shows that. Yes, sure, I can blame it on him having every removal spell he possibly could, but that's just what his deck does. I misunderstood the matchup (it's the only deck I haven't played against extensively on MODO) and got crushed, as well I should have.

Second round was about the same; he deals me fifteen on the third turn the first game, putting me at one, and I can't combo next turn. Game two he has nothing and I have turn one Forge-Tender (which he attacks with Guide into) so I slowly, slowly, slowly beat him down. Game three I don't do much and he has a bunch of removal.

I'm going to try sideboarding Refraction Trap and see how that does.

selfish elf report

Selfish Elf sucks.

Current spergelves decklist:

2 Arbor Elf
4 Heritage Druid
3 Boreal Druid
4 Llanowar Elves
4 Nettle Sentinel
1 Joraga Warcaller
1 Viridian Zealot
4 Elvish Visionary
4 Elvish Archdruid
1 Ranger of Eos
1 Regal Force

2 Primal Command
4 Summoner's Pact
4 Glimpse of Nature
3 Cloudstone Curio

4 Verdant Catacombs
4 Misty Rainforest
1 Temple Garden
1 Pendelhaven
5 Forest
3 Horizon Canopy

Sideboard
3 Ranger of Eos
4 Umezawa's Jitte
1 Loaming Shaman
4 Ghost Quarter
1 Joraga Warcaller
2 Viridian Shaman

Only major change is two Warcaller between main and side. It's the best card to tutor up with Ranger of Eos against basically any scary cards when the game goes a bit long; you just play one with all your mana and if they don't kill it, you probably win. If they do, play another one. Just ran a two-man against Zoo to try it out, and the game went long (read: turn five) and he had a Meddling Mage and Steppe Lynx on Heritage Druid against my board of a couple elves, so I Primal Commanded to gain seven and search up Ranger Of Eos. Next turn, Ranger for two Warcallers and block Steppe Lynx. Next turn, Warcaller meets Path to Exile. Next turn, Warcaller wins the game. Surprisingly, slow-seeming strategies like this are actually really good against Zoo as long as each card you play trades with at least one of theirs. Warcaller is a great card at drawing a removal spell kicked once, great at giving you a solid team kicked twice, and wins the game unless they kill it kicked three times... plus, it's another one-mana guy, so unlike most sideboard cards, it doesn't fuck up your Glimpse plan.

sideboarding with and against elves

I've played a couple DEs with a barely-modified version of Matt Nass's list (aka Spergelves), and I've been incredibly impressed with Ranger of Eos. Before he won, I was always under the assumption that Ranger would be too slow and mana-intensive to be useful; spending four mana to get a guy that trades with their one-drop can't be worth it, you'd rather just Weird Harvest, untap and win, right? Actually, no, the card practically instantly wins against them, since even a terrible removal spell is better than you had before, and it plays incredibly well with Curio and Glimpse. I honestly felt like I could play against Slow Zoo all day, it was that good.

Previously, I was running the red splash for Blood Moon, but it's just really bad in the current metagame: it was in there mostly to give you game against Tribal Zoo, which no one really plays any more (surprisingly). Against Zoo with Hierarch, though, it's basically incapable of stealing games like it did, leaving you jack shit against them. I'll miss it against Faeries, but that's faded away a bit too, because of the bad Zoo matchup.

Part of Elves's success, I feel, is that people's maindecks aren't prepared for it, and unless they're super on top of the latest MTGO daily events so they know what dual land corresponds to what sideboard card, they won't know what Elves is doing post-board. If you throw some Rangers in there, however, this advantage goes away... it's a tough situation, since Nass's sideboard really is that good. Against Zoo it doesn't matter too much since you can shout at your opponent "I AM GOING TO CAST RANGER OF EOS" and they'll still lose to it, but against control, it'll influence what cards they're bringing in- if a DDT (cheesy but catchy name for the deck, innit?) player thinks you're using Damping Matrix instead of Ghost Quarter, that will influence how they play and sideboard; only a week ago, people would have brought in Hurkyl's Recall, which is now an obvious blank.

Depending on the hate that people come up with, I'm excited to try a black splash for Prowess of the Fair. I've played all of one game with it, and it was a while ago, but it won! Huzzah. Anyway, it seems comparable to Fecundity last season for obvious reasons, and definitely helps your beatdown plan, as well as the fact that it's an enchantment you can tap to Heritage Druid, and get an extra mana from Archdruid.

If I was playing a PTQ today, I'd run this sideboard (assuming two maindeck Primal Command):

4 Ghost Quarter
4 Umezawa's Jitte
3 Ranger of Eos
1 Loaming Shaman
1 Burrenton Forge-Tender
1 Elvish Champion
1 Viridian Shaman

The Champion is for any adopters of the deck or Conley devotees trying out some Bant concoction. In those matchups against midrangey decks that pack a bunch of crap like Meddling Mage, you basically just want to make your guys bigger than theirs and beat down, and unblockability seems like it'll help there. I haven't tested the mirror more than a couple times, but Champion seems like a no-brainer one-of, especially with Jitte. Speaking of Jitte, still playing four of them seems questionable, since they only come in against the moderately-played Faeries and Elves, but I feel like there's room in the sideboard to make those that much better.

There's been a bunch of talk about what the right hate cards are against the deck, and I can confidently say that by far the best one is Darkblast. Barring Pendelhaven or ridiculous Nettle-Nettle-Heritage-Archdruid draws, the deck cannot beat it if the opponent can cast it turn one. This is very different from Explosives, which you can play around by just attacking a bunch, or slow shit like Night of Soul's Betrayal where they need other cards to stop you from turn three or fouring them. Chalice is reasonable, but most of the time you can get an elf or two down before they play it, and if their draw is otherwise slow you'll have time to find artifact removal. In summary: just play Darkblast. It is good, and not just against this deck; it's really strong against Brozek Boros and DDT as well.

Also depending on what the metagame looks like, I'm excited to start testing a list resembling this:

Selfish Elves
4 Heritage Druid
4 Nettle Sentinel
4 Llanowar Elves
4 Arbor Elf
4 Joraga Warcaller
4 Elvish Visionary
4 Elvish Archdruid
1 Regal Force
4 Summoner's Pact
4 Glimpse of Nature
3 Strength of the Tajuru (SELFISH ELF)
2 Primal Command
18 land

Obviously, it's a bit more beatdown-centric, and the general idea is to be much more resilient against Darkblast and Volcanic Fallout. The disadvantage is the loss of Curio, and only eight Llanowars instead of ten or eleven, but the combo is still strong since it plays more one-drops than most Elf decks, as well as having a beatdown plan worthy of Extended. It'll be a couple days before Worldwake is on Magic Online, but I'll be trying it out the instant I can get my virtual hands on the virtual cards. I will probably start doing the Spongebob the first time I win the game with Selfish Elf.

how to play elves

Since the GP win, I assume that there'll be a lot of discussion on Elves in the coming weeks- sideboarding strategies, different builds, numbers of Curios, and other such minutia. In my opinion, the variations in lists and sideboards are, combined, roughly 10% as important as knowing how to play the deck effectively. This is not a deck you can pick up and play; despite what Nass said, he didn't just have experience with the Standard version, he's played the Extended version pretty extensively (there's really no way to arrive at a list like that unless you do; a less experienced person would take LSV's word that Thoughtseize and even Night's Whisper is the way to go). Despite that, he played really, really horribly, but got lucky enough that it didn't matter. Don't try that at home: the majority of games lost by the deck could have been won with better play.

As I see it, there are three main aspects to playing it well: goldfishing, general strategy, and sideboarding. Goldfishing is probably the easiest to learn: just lock yourself in a room with either a paper or online copy of the deck (MWS works better here than MTGO) and combo until your hands bleed. Most of it gets pretty rote, but there are still some strategic and probabilistic decisions to be made, and most of them involve Summoner's Pact. If that card didn't exist, aside from the amazing level of consistency it would take away from the deck, it would be reasonably simple to play. Most of the time, you just need to know what to get in certain situations, like if you untap on turn two on the play with a Forest and Nettle Sentinel in play, and a hand of:

Verdant Catacombs, Summoner's Pact, Summoner's Pact, Nettle Sentinel, Nettle Sentinel, Glimpse of Nature

Obviously, it's ridiculous, and pretty much the best you can ask for. Assuming you want to show off the turn two kill to your friends, the play here is pretty simple: fetchland to Forest, Pact for Heritage and play it, play Nettle, tap them for GGG, Glimpse, Nettle (GG), tap Nettles and Pact (5*G) for Regal force, tap Nettles (8*G), cast Regal, do the happy dance as you draw six cards plus what you drew off Glimpse already and easily turn two. Obviously, this is a pretty basic goldfishing example with the absurd triple-Nettle-and-(pseudo)-Heritage in it, but it shows a few basic principles: first, don't be afraid of Pact. Pact is your friend, it's a nice card that would never kill you unless you've done something very very bad. Don't be afraid to Pact when you can Glimpse but can't fully combo with 100% certainty; the deck is just under a third one-cost guys, so if you can draw three or four cards off Glimpse and won't completely run out of mana, it's probably worth it. There are no hard and fast rules; just practice a bunch and see what works for you.

That example was using Pact in a situation where it was used early, for a very clear, obvious combo purpose. Sometimes, it's not that simple. For example, you could be trying to combo with one Glimpse, two Nettles in play, you've used your land drop already, you have seven mana floating, and you're down to lands and Pact in hand. What do you go for here? If your instant response was Regal Force, that is wrong if you absolutely must combo (but may be correct in some game situations). It'll use up all your mana, and you'll have to ship the turn with two untapped Nettles and a bunch of cards in hand. You might also think that the better play is to go for the third Nettle Sentinel, but this doesn't do anything for you- you're a mana short of going for Regal now, but if you draw a few more cards, you'll have essentially infinite. The correct play is to Pact for Elvish Visionary. When you play it, you'll draw two cards and have five floating. This means that if you draw into Pact or Regal, you'll win outright; if you draw Primal Command, you can use a bunch of mana but go for another Visionary to keep going; and you have twice as many chances to draw one-drops to keep things going.

This is why goldfishing is so crucial with the deck: when you're not even playing for two tickets in an online 2-man, you can try to combo on turn three when it looks like you shouldn't. You'll probably be surprised how often you'll draw the combination of cards you need if you assume that if you don't combo on turn three, you've lost. Then, when you need to do this in a tournament, you'll feel at least reasonably confident, and look like a true genius/total lucksack when you win with from a position that seems impossible.

As far as general strategy outside the combo goes, you'll need to always have a plan based on your hand of what your next few turns are going to look like, especially against decks that are beating you down. Against Zoo and the like, the tension for them is between removing your creatures and playing threats of their own: if they do nothing but play Goyfs and Knight of the Reliquary, you can goldfish them pretty easily. If they do nothing but kill all your creatures as they hit the table, you'll eventually draw into some way to combo, since if they only have one threat, it'll give you all the time in the world. Because of this, you'll need to think carefully about the creatures you're playing, and how your board position stacks up to the aggro player. Generally, if you do nothing but play dorky elves with no Heritage Druid, eventually they'll think you just don't have anything, and play more threats to kill you before you do. The most crucial thing is knowing when to chump block: if you do it randomly, you're not going to have enough mana to go off, and you'll die anyway; instead, you need to think about what, exactly, your plan is when you untap, and if you can combo, especially through a single removal spell. Generally how I win games against Zoo is to play out my Llanowars and sub-Llanowars, keeping Heritage in hand if I have just one along with a Glimpse, and waiting to go off until they either tap out for more creatures or I know that one removal spell isn't going to stop me. Minor note: it's often best to play Heritage as your third creature, giving them the choice between letting you tap your guys for mana, or using a removal spell on a suboptimal target. They'll misplay by killing non-Heritage when you can combo, or misplay by saving it when you just wanted to tap them to play out your hand.

It doesn't affect strategy too much, usually, but know that if you untap with Archdruid, you almost always win.

The Thopter-Depths matchup is, surprisingly, one of the more straightforward ones. They don't have much removal pre-board, so Archdruid will generally win the game by himself, and if they don't have the 20/20 really early, they'll have to spend their mana tutoring and casting Thirst for Knowledge. If you can't combo, just play out your entire hand and see if they have anything fast enough to stop it. There's really no use playing conservatively here.

Against slow blue control decks, just play a few 1/1s, keeping all your good spells in hand. Attack for three or four, turn after turn, until they feel like doing something about it. Then combo them. They have almost no clock whatsoever, so there's hardly ever a reason to play out more than four creatures or to try to force a combo into open mana. Try to bait them into countering superflous elves.

Against everything, just remember: what you want to present to your opponent is a difficult situation where they're facing threats on the table, but have no earthly idea whether or not you can combo. Keep this in mind at all times; sometimes a play will come up where you can trick your opponent one way or another.

As far as sideboarding goes, the important thing is to think about what you want your gameplan to be, and sideboard out cards accordingly- this is much more important than what you bring in. Often, people will over-sideboard cards that seem good, but don't help your game plan at all. Remember: if you are bringing in more than six cards, you are doing it wrong. This holds true against almost every deck. Generally, you want to take out a couple one-drops, unless hitting your 1/1s early and often is important (like against Faeries), and take out a couple Visionaries if your plan is to combo them as soon as possible, since they can be a bit slow. Faeries is the matchup that is worth mentioning specifically: they won't let you combo after turn three hardly ever. Instead, you want to make them trade their two-mana answers for your one-mana dorks. Take out everything that costs more than three, as well as Curio. On the play, take out a couple Pact; on the draw, Archdruid, since it's the only card you have that they can trade profitably with. The matchup basically comes down to Jitte, so hope you get one.

I hope this was helpful to everyone that wants to play the deck. Now get out there and... stay at home goldfishing.