[Article] Metagaming Legacy Before GP NJ and Eternal Weekend

[Lemnear]

Hey Lemnear, could you explain more what you're talking about in the last paragraph?from what I can tell, you are misunderstanding the chart, but it's difficult to tell exactly what you are talking about.

Sure. My issue is NOT that you've considered CoM to actually hit 60% of decks as statistically ~65% of decks run at least Brainstorm in their 60 atm. These ~65% include combo decks as well, which can kill you before CoM comes down and therefore should not be part of the statistic of decks CoM is effective against. If they are not included in these estimated 60% you picked, the number of decks, which go into the midgame with their cantrips (aka past turn 3) and are the target for CoM, is too high, but whatever :). I asume we agree that we count mostly Delver and Blade variants in that category and those don't make up 60% of the metagame.

The problem with that table is this part:

In this table, we are assuming that 60% of the field is playing one of these Xerox style decks or a draw focused deck. You can go through the table yourself, but something huge to look at is that if you are only able to win 20% of your matches where the mainboard Chains of Mephistopheles is ineffective, but have a 70% win rate with the Chains which shuts off the cantrip of the decks, you break even on your expected win percentage at 50%

You can't tie your estimated, average win-% (here 50%) to a card and make a statement with the graph which boils down to (correct me if I'm wrong) "I won 70% of those matches just because I have CoM in the deck and managed to win 20% of matches against decks which CoM is a blank against" as neither the table nor the formula takes the waxing/waning impact of CoM into account if drawn at certain stages of the game. I hope we agree that the value of CoM varies depending on your opponents plays and the turn when you find & resolve CoM rather than your opponents deck constellation and therefore I think the table is very misleading.

[Richard Cheese]

Don't know if being in the Northeast matters but what I saw from the SCG this weekend I want to do two things:

• Play Chalice on 1
• Be ready to beat miracles

This goes against what I usually play but these decks were very prevelant at the open. The only problem with chalice decks is in a big tournament the variance gods are going to smite you. But I have been eyeballing the Kronberger list from BOM.

Yeah, I'm thinking the field will probably be mostly Delver (UR, BUG, RUG, in that order) and Miracles, but with a good amount of D&T, Elves, and Sneak/Show present.

I would love to play Aggro Loam, but it's slow and inconsistent in the early game.

Fast combo like Belcher or TinFins is also high on the list, but you might as well just assume that ever round you're sitting down across from at least 4 Force, 4 Daze, 4 Brainstorm.

Could get in on the Delver circlejerk and play BUG, but picking a DTB for a major event seems like something you should only do if you're good at this game. I need that left-field factor to buy me free wins and make the games interesting enough to not just drop after one loss.

For what it's worth, I don't go to these big things expecting to make day 2, I just want to have more wins than losses at the end of day 1. Plus I'm there for a week with the wife to visit friends and family, so there's not going to be any late-night practice sessions or last-minute changes. I guess what I'm saying is, if you're bored and want to grab a drink after round ~4, you know who to PM.

[iGrok]

Sure. My issue is NOT that you've considered CoM to actually hit 60% of decks as statistically ~65% of decks run at least Brainstorm in their 60 atm. These ~65% include combo decks as well, which can kill you before CoM comes down and therefore should not be part of the statistic of decks CoM is effective against. If they are not included in these estimated 60% you picked, the number of decks, which go into the midgame with their cantrips (aka past turn 3) and are the target for CoM, is too high, but whatever :). I asume we agree that we count mostly Delver and Blade variants in that category and those don't make up 60% of the metagame.

The problem with that table is this part:



You can't tie your estimated, average win-% (here 50%) to a card and make a statement with the graph which boils down to (correct me if I'm wrong) "I won 70% of those matches just because I have CoM in the deck and managed to win 20% of matches against decks which CoM is a blank against" as neither the table nor the formula takes the waxing/waning impact of CoM into account if drawn at certain stages of the game. I hope we agree that the value of CoM varies depending on your opponents plays and the turn when you find & resolve CoM rather than your opponents deck constellation and therefore I think the table is very misleading.

Ok, I think I've found our communication issue! The graph does not say "I won 70% of those matches becausechains was in the deck.". What it says is, "I won 70% of those matches when chains was in the deck." How or if you played the card isn't at question here. Maybe you played it, maybe you discarded it to Lilianna to pump goyf, or maybe you never saw the card, and won anyways becausethat can happen. I'm not looking at what specific impact it has on any game in particular.

The chart gives a lot of example cells at various winrates so that whatever you think the winrates might be, you can get a rough feeling for if its correct. A more exact method would be to use the formula and exact numbers after testing, but we can make rough guesses.

As for the numbers we picked:
"60% of the meta is affected by chains" sounds reasonable to me, although I suppose it depends if you count Elves or storm (i count elves but not storm). I also don't know the European meta very well. I also don't know the Japanese meta, but I assume that at an American GP, the American metagame will be the most represented.
"70% against affected decks" sounds in the right neighborhood. Again, this includes all the games you don't see chains and win anyways. But at the end of this post I'll throw you some lower comparison numbers.
"20% against non-affected decks" was picked mostly for the shock value. If you're playing Jund, you are going to win more than 20% of matches against decks that can't out-card advantage you, even if you are playing a dead card or two. You won't see them every game, and under some situations you can discard it to grow goyf, and eek some marginal value from it. This number also doesn't care how you win, just that you do.

So to give an example that might seem a little more tame, if 60% of the meta is affected by a hatecard, and you win 61% of the affected games and 35% of the rest, you're still better off playing that hatecard main, regardless of what that card is.

He's not saying that 70% and 20% the actual numbers, just that 70% is in the right ballpark, and 20% is around the flip point if 70% is accurate. Actual testing is of course needed to confirm, but it's a strong theory. From my end, I'm just fascinated by the math behind it, because it challenges something that I, and probably most people, thought. If all you care about is winning, playing simetimes-dead cards might actually be the right choice.

[JamieW89]

"70% against affected decks" sounds in the right neighborhood. Again, this includes all the games you don't see chains and win anyways. But at the end of this post I'll throw you some lower comparison numbers.

Winning 70% of the matches against decks affacted by draw-hate with this list? Seems way too high imo. The graph doesn't include 40-50% numbers, which, to me, seem more likely to be true than 70%.

So to give an example that might seem a little more tame, if 60% of the meta is affected by a hatecard, and you win 61% of the affected games and 35% of the rest, you're still better off playing that hatecard main, regardless of what that card is.

No. What you should look at is the changes in MW% with/without said hatecard in those matchups. (Preferably the effect on each individual matchup compared to that decks metagame share, but making 2 or 3 subgroups is fine to get an overview of the effects)
Using your method you could decide that an already well positioned deck requires a narrow hate card while in reality that card has a negative effect on your overall win%.

From my end, I'm just fascinated by the math behind it, because it challenges something that I, and probably most people, thought. If all you care about is winning, playing simetimes-dead cards might actually be the right choice.

Obviously playing a dead card is worth it if that card alone would get you to a 70% matchup vs 60% of the field. I doubt anyone would find that shocking.

[thecrav]

Obviously playing a dead card is worth it if that card alone would get you to a 70% matchup vs 60% of the field. I doubt anyone would find that shocking.

I believe the people who've been maindecking Red Blasts can attest to that!

[iGrok]

Winning 70% of the matches against decks affacted by draw-hate with this list? Seems way too high imo. The graph doesn't include 40-50% numbers, which, to me, seem more likely to be true than 70%.

If your win% with hatecards against affected decks is 40-50%, you've got a bad hatecard (or a bad deck), unless you're winning more than 65% of your other games even with two dead cards in your deck. I'm not a Jund player, but my understanding is that Chains is a good card to bring in against decks that like to draw cards. If your winrate with the card is <50%, your non-affected winrate needs to be between 50-70%. Jund is having trouble because it's not having that kind of winrate against these decks (again, thats my understanding anyway).

If I knew how well Jund performed against the field with chains in the deck, the chart wouldn't be needed :). As is, it helps to figure out where the line is, using estimations based on current sideboarded game winrates against the metadecks and non-meta unsided mulligan games winrate. These are rough approximations, but they can give a general sense of net positive or negative.

No. What you should look at is the changes in MW% with/without said hatecard in those matchups. (Preferably the effect on each individual matchup compared to that decks metagame share, but making 2 or 3 subgroups is fine to get an overview of the effects)
Using your method you could decide that an already well positioned deck requires a narrow hate card while in reality that card has a negative effect on your overall win%.

I agree that looking at changes in MW% would be great to see, especially if it was split deck-by-deck. But, that's significantly more in-depth than the scope of one article.

I don't agree that this method could make net-negative cards seem positive. How do you figure that?

Obviously playing a dead card is worth it if that card alone would get you to a 70% matchup vs 60% of the field. I doubt anyone would find that shocking.

Apparently, a lot of people do. :)

[Ellomdian]

Don't know if being in the Northeast matters but what I saw from the SCG this weekend I want to do two things:

• Play Chalice on 1
• Be ready to beat miracles

So, maybe Tezz could be your friend for a few weeks? :p

[tescrin]

So on my way to get food, I thought about Eidolon taking the spot of Chains as it also punishes the whole cantripping-stuff, while also presenting a clock if topdecked and a threat to combo. If you rather run Thoughtseize than Hymn, I cannot see serious restrains for the manabase.

Eidolon
(Taken with the bag of salt that I am obviously not a significant name in Legacy, ahem) I've been saying for a few months that any non-blue deck running red is basically obligated to run Eidolon. He's the only thing that makes up for the lack of White in such a deck (by giving the obligatory hatebear + discard synergy.) Jam 4 and don't look back. The only reason that bandwagon isn't a thing is because red is bad and double-red is worse.

Chains
To the article: Library > Chains. As someone who uses library and ran Chains for a good 6-months in the sideboard; Chains is basically really-cool, really-expensive, hot-garbage. I love the card, I would love for it's value to skyrocket so I could really stick it to some poor sucker on ebay; but that card is balls IME. It's not worth the card slot a lot of times against Delver (it just gets countered, it's win more, or it's straight tempo loss) and against S&T variants you still lose because they combo through it. It also has good potential to simply fuel PiF. It is hit or miss against elves, and they carry 8-10 outs post-board (3-4 decay + 3-4 gsz + 1-2 sage) meaning if you DID get the right combo with it, they can probably just pop it.

Library
Library is always live. I say that even if SotL is flashed in. What..? I just don't draw and nothing bad happens? Dang. And if you're even slightly ahead? Screw it, punch the turbo-fuck button and jizz all over your poor opponent.

[Lemnear]

Eidolon
(Taken with the bag of salt that I am obviously not a significant name in Legacy, ahem) I've been saying for a few months that any non-blue deck running red is basically obligated to run Eidolon. He's the only thing that makes up for the lack of White in such a deck (by giving the obligatory hatebear + discard synergy.) Jam 4 and don't look back. The only reason that bandwagon isn't a thing is because red is bad and double-red is worse.

I don't care for names, but for good ideas :)

I'm with you here and would have to ponder over double red for Eidolon vs. double black for Hymn and if they can even played in the same deck or if Hymn need to be replaced with Thoughtseize. Guess people with more experience should think about those options.

Chains
To the article: Library > Chains. As someone who uses library and ran Chains for a good 6-months in the sideboard; Chains is basically really-cool, really-expensive, hot-garbage. I love the card, I would love for it's value to skyrocket so I could really stick it to some poor sucker on ebay; but that card is balls IME. It's not worth the card slot a lot of times against Delver (it just gets countered, it's win more, or it's straight tempo loss) and against S&T variants you still lose because they combo through it. It also has good potential to simply fuel PiF. It is hit or miss against elves, and they carry 8-10 outs post-board (3-4 decay + 3-4 gsz + 1-2 sage) meaning if you DID get the right combo with it, they can probably just pop it.

That section is spot on. Grinselbrands draw is indeed pointless to hate out as you'll lose to the 7/7 lifelinking, flying body. Elves have indeed plenty of options to get rid of CoM, but I guess Elves was a sidenote only anyways. I haven't thought of cantrips used to fuel PIF with a CoM on the field. Nice idea.

Library
Library is always live. I say that even if SotL is flashed in. What..? I just don't draw and nothing bad happens? Dang. And if you're even slightly ahead? Screw it, punch the turbo-fuck button and jizz all over your poor opponent.

Library at least always advances your gameplan unlike CoM. We could ponder over Miiri's Guile or SDT instead, but that's not the place. Thanks for your input

[Gunz24]

Interesting that I posted something very similar to this in the Jund thread prior to your blog and your membership on The Source.

Jund bros (and gals?),

Everyone in this thread is losing their damn mind. Treasure Cruise has everyone in fits and understandably but some of these suggesstions are just a touch overboard. Cutting Hymn for example? We run Hymn because our combo match-up is terrible. More often than not I'm siding this card out against any delver deck which is where Treasure Cruise is the most threatening so the idea of "feeding their graveyard" is overblown. Not to mention if your concern is Treasure Cruise how does Inquisition help? The simple answer is it doesn't. Keep the Hymns and sideboard smarter.

Lose Punishing Fire? This isn't the first time or the last we'll talk about this possibility but removing our Fire package seriously undermines our longevitiy. Not to mention it may be our strongest card against delver variants where, again, Treasure Cruise is the most damaging to us.

I've given this Treasure Cruise problem a lot of thought and I have good news and bad news. Good news is we can beat it, the bad news is I'm nearly certain the way we can most efficiently do so is a $400 card on the reserve list. Whomp whomp. Chains of Mephistopheles is the answer to our Jund woes.

Think about how we are able to interact with Treasure Cruise. One way is to attack the graveyard and prevent them from delving. This may seem relavtively easy because traditionally our sideboards are full of graveyard hate but when you get to the true reason for that its because our Reanimator match-up is less than ideal. We bring in graveyard hate that wins us the game not that helps us durdle. If we take this route and main deck our hate we're affecting our deck's bottom line: Tarmogoyf. And for what in exchange? So they can't cast one spell? That doesn't win us the game in the same way eating Griselbrands with Scavenging Ooze or cracking a Spellbomb on a Dredge deck does. All this strategy does is waterdown our power in match-ups where we need to rely on forcing poor trades. The same match-ups where a 4/5 or a 5/6 Tarmogoyf is a serious issue for our opponent.

Another way is to just outright counter the Cruise. Incliment weather, sorry dude. Except we can't do this. Sure we could bring in Red Blast and Pyroblast from the board but that isn't what our deck wants to be doing. How often have you taken this approach in a match and found yourself sitting with a Red Blast in hand wishing it were something relevant? "But it can kill a Delver!" Yea, it can. So does Punishing Fire. The closest we come to this is Thoughtseizing, Hymning or Duressing away a Cruise before they can cast it and we are capable of doing that aready without drastic deck composition changes.

So this brings me to the solution. Anyone who's resolved a Chains of Mephistopheles against a blue player knows how devistating it is. I have never lost a match where a copy of Chains stayed in play. Never. With Chains, Treasure Cruise becomes "Exile your graveyard and discard 3 cards" for U. Great! Traditionally this was a sideboard card because in the match-ups it was good in it was really good and the match-ups its bad against its hot garbage. The current metagame has changed that. The amount of Treasure Cruises running around in my opinion justify main decking Chains of Mephistopheles. Best case scenario you crush your opponents card advantage and then grind them out the way only Jund can. Worst case scenario you pitch it to a Liliana activation and grow your goyf with the ever elusive enchantment card type.

My propsed list for the current metagame would be as follows:

3 Bloodbraid Elf
4 Dark Confidant
4 Deathrite Shaman
4 Tarmogoyf

4 Liliana of the Veil

2 Lightning Bolt
3 Abrupt Decay
3 Punishing Fire

4 Thoughtseize
3 Hymn to Tourach
1 Maelstrom Pulse

1 Sylvan Library
1 Chains of Mephistopheles

1 Forest
1 Swamp
2 Wooded Foothills
4 Verdant Catacombs
3 Bloodstained Mire
3 Wasteland
4 Grove of the Burnwillows
3 Bayou
2 Badlands

Sideboard:

1 Sylvan Library
1 Chains of Mephistopheles
1 Umezawa’s Jitte
1 Toxic Deluge
1 Grafdigger’s Cage
1 Scavenging Ooze
1 Nihil Spellbomb
2 Golgari Charm
1 Massacre
2 Duress
2 Ancient Grudge
1 Engineered Plague

I really love the 2 Library configurations and in some match-ups we want to be in that mode but against other decks we need to turn on 2Chainz. A note about the interaction of Sylvan Library and Chains of Mephistopheles if you choose to take this route. Library is a YOU MAY ability meaning if Chains is out you will need to defer from "looking" at the other 2 cards in your draw step. Frequently Library use is played in the manner Scry 3 would be except you're technically drawing these cards. If you do this with Chains in play you will need to discard cards and then put some back or pay life. Tragic. Don't get caught screwing this up, it will end you.

Finally, if the meta shifts or your opponent decided to move to Dig Through Time over Treasure Cruise your Chains of Mephistopheles will do nothing. Dig puts the cards in hand not draw so in this way Dig gets around Chains just like Bob does for us. If you encounter this I'd advise switching to double Library mode and heavier discard, Duress would be a must.

I'll be putting some testing into this leading up to Legacy Champs in Philly and will follow up with you.

Stay strong Jund Bros. Don't panic.

[Vicar in a tutu]

Interesting that I posted something very similar to this in the Jund thread prior to your blog and your membership on The Source.

I play Jund as my main deck in legacy. I have the cards to build anything, but I prefer Jund (my brain can't handle storms, if you get what I'm saying). The problem with your solution to Treasure Cruise is that a miser Chains, in a deck with 0 cantrips, 1 Library (which is a nonbo with Chains) and 4 Dark Confidants, is very difficult to find. Is this supposed to fight a U/R Delver deck with a million ways of finding Treasure Cruise? Over any given tournament, you might get to cast Chains once while a typical Treasure deck would cruise away at warp-speed.

[Lemnear]

I play Jund as my main deck in legacy. I have the cards to build anything, but I prefer Jund (my brain can't handle storms, if you get what I'm saying). The problem with your solution to Treasure Cruise is that a miser Chains, in a deck with 0 cantrips, 1 Library (which is a nonbo with Chains) and 4 Dark Confidants, is very difficult to find. Is this supposed to fight a U/R Delver deck with a million ways of finding Treasure Cruise? Over any given tournament, you might get to cast Chains once while a typical Treasure deck would cruise away at warp-speed.

This. It's a 300$+ too-Late-to-be-effective random draw/topdeck which does nothing to adress Junds problems (combo, no card selection, Terminus) nor brings damage to the table.

[FoolofaTook]

Rather than just beating a dead horse (as fun as that can be), can we turn this thread into a meaningful metagame/deck choice discussion for Eternal Weekend and the GP?

More specifically, can anyone tell me why I shouldn't just pound 3 Bloodies at breakfast and play Burn?

That sounds like a great choice. You can win or lose fast and then watch people gawking at TC fueled mayhem for the rest of the round. Watching a guy pull just what he needs off of his TC is fun but watching him pull 2 lands and a cantrip is priceless.

[Richard Cheese]

That sounds like a great choice. You can win or lose fast and then watch people gawking at TC fueled mayhem for the rest of the round. Watching a guy pull just what he needs off of his TC is fun but watching him pull 2 lands and a cantrip is priceless.

Maybe I'll just maindeck Misinformation and keep putting fetches on top of their library before they get to 7 cards! Where is your blue god now?!?!

[Ellomdian]

Maybe I'll just maindeck Misinformation and keep putting fetches on top of their library before they get to 7 cards! Where is your blue god now?!?!

You fool! Don't make ANOTHER Alliances Uncommon an $80 card overnight!

[FoolofaTook]

Maybe I'll just maindeck Misinformation and keep putting fetches on top of their library before they get to 7 cards! Where is your blue god now?!?!

Works against Elves, Miracles, Viking Funeral, Reanimator, Dredge, Treasure Cruise, etc.

Think we could talk WotC into a strictly better upgrade with Draw 1 Card added in?

Heck, I might even test out a Scry 1.

[Julian23]

Somehow I feel putting Fetchlands on top of their library kind of defeats the purpose ;-)

[Ellomdian]

Somehow I feel putting Fetchlands on top of their library kind of defeats the purpose ;-)

The purpose being to Dilute their draws with low-impact cards (since they don't run many 'real' lands) while making their Cruises more expensive and forcing them to pop the first fetch unless they want to draw blanks?

I'm all for that