Perfecting and then running that script took a little while.
I have a python script that crawls MTGGoldfish and saves decklists from there, along with deck names, players, tournament name, and some other data, and dumps it to a python dictionary that I can update easily with new data. The annoying part with that is that MTGGoldfish will throttle you if you access too many pages too quickly, so I had to put in a few hundred millisecond sleep between page calls. It took several days to grab it all.
I then found out that MTGGoldfish doesn't have any GP or higher level tournaments farther back than November or December 2014, so I went through all of mtgtop8 manually and found the tournament IDs for all tournaments at 2 stars or higher from that time back to August 2011 and then wrote another script to crawl those tournament IDs. Then I merged the two datasets together. I don't keep updating with new mtgtop8 stuff, though. I know I'm missing such-and-such PPTQ in Kansas because of it, but I'd also have to figure out a way to strip out double entries. I worry that it wouldn't be as simple as checking date, tournament name, and player. The full dataset that I have is about 68000 decks.
In order to do the card text searches, I downloaded mtgjson and can merge card text from there into individual decklists. I do that on the fly, even though it's (kind of) slow, because if I merged it all and saved it into a file then my dataset would go from 68MB to 600MB and I don't feel like doing that.
This might be a long shot, but is there anyone fluent in javascript that can get Karsten's manabase code working for them? I've been fiddling around with it since it was brought up here, but I keep getting a syntax error.
Wow. This thread is getting pretty serious. All this is nice and all, but that's a lot of info to be swallowing in one sitting. Not sure if I'm ready for this kind of data.
Oh well, guess it's time to stop-a complainin' and start-a readin'
I tried to condense it to a tl;dr in that second post (did it manually by looking at the top cards). I'll update it with only September's days in at few days, and try to distill it into something easier to digest.
I tried to condense it to a tl;dr in that second post (did it manually by looking at the top cards). I'll update it with only September's days in at few days, and try to distill it into something easier to digest.
Don't get me wrong, all this info deserve praise and I can only wish all primers coulde have such dedication put throught them. However, burn is my secondary deck and I feel like absorbing all of this is overwhelming and maybe a bit much. But It's definitely welcome informations.
I brought in the Cage and the Paths. Do I need the Paths?
Does the sideboard look alright in general?
I would bring in the Paths because you want to keep them off of broken stuff like Heritage Druid, Elvish Archdruid, and Ezuri. Cage is fine as well, because it keeps them from digging and putting broken stuff on the battlefield. I assume you're playing 4 Blazes, right?
I tried to condense it to a tl;dr in that second post (did it manually by looking at the top cards). I'll update it with only September's days in at few days, and try to distill it into something easier to digest.
Don't get me wrong, all this info deserve praise and I can only wish all primers coulde have such dedication put throught them. However, burn is my secondary deck and I feel like absorbing all of this is overwhelming and maybe a bit much. But It's definitely welcome informations.
It's definitely hard to convey information about a lot of decks at once without just throwing a lot up there. I'm trying to keep it from being a straight up core dump. It will probably be cleaner when it's only 1 month worth of data.
So, I played yersterday FNM with my BW tokens deck against a boros burn deck. He won 2-0, I just couldnt keep up.
He told me he sided out the goblin guides on the draw. I was really impressed to see someone sided them out. Is that correct? His argument was that I had many chumpblockers. Insight?
Thanks. Yes I'm on a fairly standard 4-of everything list so 4xSearing Blaze did good work. I've found T1 on the draw bolting their Llanowar Elves or Elvish Mystic to be a good way of slowing them down. My opponent said the Cage really did work when I played it as he had CoCo in hand and couldn't cast it.
So, I played yersterday FNM with my BW tokens deck against a boros burn deck. He won 2-0, I just couldnt keep up.
He told me he sided out the goblin guides on the draw. I was really impressed to see someone sided them out. Is that correct? His argument was that I had many chumpblockers. Insight?
If you're going to side out a creature, Swiftspear should come before Goblin Guide because T1 Guide is far more powerful. He's worried about it being nullified by blockers and should therefore cut the one that's worse on the turns before blockers show up.
Did he say what he sided in? Probably just enchantment hate? I suppose it's reasonable to cut Swiftspears there so that you maintain burn spells that just go around all of the 1/1s.
he brought 2 wear/tear, 1 searing blood, and 1 of something I dont remember. I actually think that the swiftspears are more relevant against tokens. Say I am making 2/2's, the GG just gets chumped and gives me some lands. The swiftspear, however, has a bigger ass and survives those chumps. I usually get rid of them before the goblin, to be honest, she is more pesky to deal with.
That said, I would have removed maybe the 4 rift bolts (since thez are a bit slower in case I land a Sorin or an active shambling vent).
I said I'd wait until the end of the month to put up stats for September 2017, and I'll keep doing this after each month. The following is a lot of data that's tough to digest, but it might be helpful to you to know what kind of lifegain to expect at various times and what counterspells people are playing. This analysis comes from MTGGoldfish data with some cleaning procedure applied to it to try to clean up any bad deck names. I then go through these deck lists and find instances of "counter target", "lifelink", and "gain N life"/"gains N life" in the card text. On top of that, I search for Leyline of Sanctity, Oketra's Last Mercy (no one plays it anyway), Chalice of the Void, and Phyrexian Unlife. The September 2017 deck population is 529 decks.
Card fractions: This is averaged over the whole meta and is the fraction of decks that play the card in the maindeck/sideboard without taking into account how many copies they play. These give a general idea of what cards to expect from a randomly chosen deck.
Maindeck: The top cards are Cryptic Command at 11.9%, Scavenging Ooze at 11.2%, Shambling Vent at 10%, and Lightning Helix at 9.45%. Top other counters are Mana Leak and Remand at 7%ish. Chalice of the Void is 7.4% and Collective Brutality is 6.8%.
Sideboard: The top sideboard hate cards are Collective Brutality at 21%, Dispel at 15.5%, Leyline of Sanctity at 10%, and Negate at 9.26%. Ceremonious Rejection and Disdainful Stroke don't matter for us, but satisfy my search criteria.
The first column is the average copies per deck of this archetype if the deck plays this card, so that this number is not affected by a deck that doesn't play the card. The second column is the fraction of decks of this archetype playing this card, and this number is affected by decks that don't play the card.
This is just a core dump of the fraction of decks out of the total for this month with each name.
6.43% Eldrazi Tron
6.05% RG Titanshift
5.29% Burn
5.29% UBR Death's Shadow
4.16% Abzan
4.16% UR Gifts Storm
4.16% WU Control
4.16% Affinity
3.21% Elves
3.21% Lantern Control
3.21% Eldrazi and Taxes
3.02% WUR Control
2.84% Vizier Company
2.46% Ad Nauseam
2.27% BGx Midrange
1.70% Death and Taxes
1.70% Dredge
1.70% BG Tron
1.70% Bogles
1.51% Living End
1.51% Amulet Titan
1.32% WB Smallpox
1.13% Merfolk
1.13% Jund
0.95% URG Gifts Storm
0.95% Jeskai Tempo
0.95% Mono U Tron
0.95% WBR
0.95% WURG Knightfall
0.95% Kiki Chord
0.76% WUG Knightfall
0.76% Bant Eldrazi
0.57% W
0.57% WUB
0.57% UBR Delver
0.57% WUBRG Death's Shadow
0.57% Bant Company
0.57% BRG
0.57% WUBG Humans Company
0.57% Gifts 3c
0.57% Goodstuff 3c
0.57% BRG Death's Shadow
0.57% R/B Hollow Call
0.38% Tooth and Nail
0.38% Unknown
0.38% UBRG Scapeshift
0.38% WBRG
0.38% BRG Titanshift
0.38% Counter Cat
0.38% Faeries
0.38% W/B Eldrazi
0.38% W/R Prison
0.38% UR
0.38% UB
0.38% WRG Titanshift
0.38% WBRG Death's Shadow
0.38% RG Tron
0.38% B/R Pack Rat
0.38% Skred Red
0.38% UR Control
0.38% WUBG Death's Shadow
0.38% UG Infect
0.19% WUB Control
0.19% WU
0.19% WURG Gifts Storm
0.19% Grishoalbrand
0.19% G Tron
0.19% RG
0.19% U/R Through the Breach
0.19% U Control
0.19% UBR
0.19% WUBR Death's Shadow
0.19% 8-Rack
0.19% WURG Copycat
0.19% WURG Hatebears
0.19% Abzan Eldrazi
0.19% Grixis Delver
0.19% Gruul Zoo
0.19% Melira Company
0.19% WUB Death's Shadow
0.19% G/R Land Destruction
0.19% G/R Ponza
0.19% R
0.19% WUBR Copycat
0.19% WUBRG Knightfall
0.19% Hulk Breach
0.19% URG Control
0.19% WBG
0.19% WUBG
0.19% UBRG Death's Shadow
0.19% BG Infect
0.19% Martyr Proc
0.19% WUBRG
0.19% WURG
Edit: Addendum with data sorted by card rather than by deck. Specifically, this is being posted because of Leyline. Leyline sees some maindeck play in Lantern and Bogles, and is in 10% of sideboards. 80%+ of Lantern, Ad Nauseam, and Bogles decks play it in the sideboard and are the primary decks you'll see it from.
Columns are average number of copies in decks that play it, fraction of decks of each type that play it, and fraction of the meta that each deck comprises.
The Mardu one is trying to go all out on lowering their CMC, so I'm not too surprised that it would cut Eidolon. I don't necessarily agree that lowering your CMC by playing Burst Lightning and Shard Volley and cutting Eidolon and shaving a Boros Charm is really something that you want to do. I suppose the idea is to really just treat Burn as a 7 card combo deck and fly by the seat of your pants. If that's the goal, I think the build is a reasonable conclusion of having that goal, but I'm not sure the goal is worth having.
As for the RWg one, they're definitely not looking to lower their CMC and I don't agree with cutting Eidolon here. Eiolon is still good even if it's not that great against Eldrazi Tron.
At least they're choosing to play burn spells over Eidolon and not playing Shrine of Burning Rage, so they have that going for them, which is nice.
To me, Eidolon is part of the reason the deck is as good as it is, which means cutting Eidolon is not something I'd do. It's almost (but not quite) like saying "I want to play Burn without Lightning Bolt".
To me, Eidolon is part of the reason the deck is as good as it is, which means cutting Eidolon is not something I'd do. It's almost (but not quite) like saying "I want to play Burn without Lightning Bolt".
I'd agree that Eidolon is the most individually powerful card in the deck, and a clear reason why Burn rose to Tier 1 in the first place. However, I also think Eidolon isn't as good as it used to be.
This isn't to say that I think it should be cut - I don't - but I do see why people might want to try experimenting without it.
I don't think it's accurate to say that Eidolon "isn't as good as it used to be".
The two decks people talk about in regards to Eidolon are GDS and ETron. Eidolon is strong against GDS. If they remove it, you made them spend resources to deal with it and that disrupts the flow of their game to some degree. If they don't, it actively applies pressure to them and allows you to sandbag burn spells until you're able to finish them. I'll side out Eidolon against ETron all day, but it's not actually a new thing to have a tier 1 Tron deck that you side Eidolon out against. I sided it out against Gx Tron for the last 2 years. If Eidolon was great 2 years ago (and it was), it's still great today.
I also don't agree that if you're "looking for pro points" you cut Eidolon. Consistency is actually king for those people.
1. Eldrazi Tron. What are our priority targets with Revelry? If they have a map on the battlefield and are tapped out--do you just go ahead and blow it up to prevent Tron? Or do you wait in case they get Chalice?
How does Ballista play into it? Do you kill it on sight?
Deflecting Palm? It just seems awkward. They have to have an attacking creature to use it--a lot of times they cast a TKS before that point, so it gets eaten. A lot of times, I feel like I'd be better off having any burn spell.
2. Elves--WTF are we doing for elves? They're everywhere.
I've got 1 lavamancer in the board, and he obviously does work.
I'm changing my 3 Relic to 2 Relic and 1 Grafdigger's cage to keep them off coco/Chord.
I'm putting 1 copy of searing Blood in the board.
I've seen some people say Anger of the Gods--I just don't like it for Burn. Ensnaring Bridge buys us time but they can chord up a rec sage.
Do you "Bolt the bird?"
Private Mod Note
():
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
BW BW Tokens
RG Dredgeplendid Reclamation
RW Boros Burn
GB Elves
RB Dark Goblins
WU Azorius' Relic
I have a python script that crawls MTGGoldfish and saves decklists from there, along with deck names, players, tournament name, and some other data, and dumps it to a python dictionary that I can update easily with new data. The annoying part with that is that MTGGoldfish will throttle you if you access too many pages too quickly, so I had to put in a few hundred millisecond sleep between page calls. It took several days to grab it all.
I then found out that MTGGoldfish doesn't have any GP or higher level tournaments farther back than November or December 2014, so I went through all of mtgtop8 manually and found the tournament IDs for all tournaments at 2 stars or higher from that time back to August 2011 and then wrote another script to crawl those tournament IDs. Then I merged the two datasets together. I don't keep updating with new mtgtop8 stuff, though. I know I'm missing such-and-such PPTQ in Kansas because of it, but I'd also have to figure out a way to strip out double entries. I worry that it wouldn't be as simple as checking date, tournament name, and player. The full dataset that I have is about 68000 decks.
In order to do the card text searches, I downloaded mtgjson and can merge card text from there into individual decklists. I do that on the fly, even though it's (kind of) slow, because if I merged it all and saved it into a file then my dataset would go from 68MB to 600MB and I don't feel like doing that.
Mardu Burn
Monogreen Stompy
Legacy
Burn
Pauper
Dimir Flicker
Monowhite Tokens
Oh well, guess it's time to stop-a complainin' and start-a readin'
(W/B)BW Tokens(W/B) | (B/R)Rakdos Burn(B/R) | (U/R)Gift Storm(U/R)
How do I sideboard against Elves? A few people play it at my store, mostly Abzan or GB and usually without the Vizier/Druid combo.
My sideboard is:
1xDeflecting Palm
3xKor Firewalker - we have a LOT of Burn at my store! 17 players tonight and at least 4 on Burn. 2 weeks ago there was 26 players and 6-8 on Burn...
3xPath to Exile
2xRest in Peace
1xGrafdigger's Cage
3xDestructive Revelry
2xShattering Spree
I brought in the Cage and the Paths. Do I need the Paths?
Does the sideboard look alright in general?
(W/B)BW Tokens(W/B) | (B/R)Rakdos Burn(B/R) | (U/R)Gift Storm(U/R)
I would bring in the Paths because you want to keep them off of broken stuff like Heritage Druid, Elvish Archdruid, and Ezuri. Cage is fine as well, because it keeps them from digging and putting broken stuff on the battlefield. I assume you're playing 4 Blazes, right?
The sideboard looks reasonable to me.
It's definitely hard to convey information about a lot of decks at once without just throwing a lot up there. I'm trying to keep it from being a straight up core dump. It will probably be cleaner when it's only 1 month worth of data.
He told me he sided out the goblin guides on the draw. I was really impressed to see someone sided them out. Is that correct? His argument was that I had many chumpblockers. Insight?
BW BW Tokens
RG Dredgeplendid Reclamation
RW Boros Burn
GB Elves
RB Dark Goblins
WU Azorius' Relic
If you're going to side out a creature, Swiftspear should come before Goblin Guide because T1 Guide is far more powerful. He's worried about it being nullified by blockers and should therefore cut the one that's worse on the turns before blockers show up.
Did he say what he sided in? Probably just enchantment hate? I suppose it's reasonable to cut Swiftspears there so that you maintain burn spells that just go around all of the 1/1s.
That said, I would have removed maybe the 4 rift bolts (since thez are a bit slower in case I land a Sorin or an active shambling vent).
BW BW Tokens
RG Dredgeplendid Reclamation
RW Boros Burn
GB Elves
RB Dark Goblins
WU Azorius' Relic
I said I'd wait until the end of the month to put up stats for September 2017, and I'll keep doing this after each month. The following is a lot of data that's tough to digest, but it might be helpful to you to know what kind of lifegain to expect at various times and what counterspells people are playing. This analysis comes from MTGGoldfish data with some cleaning procedure applied to it to try to clean up any bad deck names. I then go through these deck lists and find instances of "counter target", "lifelink", and "gain N life"/"gains N life" in the card text. On top of that, I search for Leyline of Sanctity, Oketra's Last Mercy (no one plays it anyway), Chalice of the Void, and Phyrexian Unlife. The September 2017 deck population is 529 decks.
Card fractions: This is averaged over the whole meta and is the fraction of decks that play the card in the maindeck/sideboard without taking into account how many copies they play. These give a general idea of what cards to expect from a randomly chosen deck.
Maindeck: The top cards are Cryptic Command at 11.9%, Scavenging Ooze at 11.2%, Shambling Vent at 10%, and Lightning Helix at 9.45%. Top other counters are Mana Leak and Remand at 7%ish. Chalice of the Void is 7.4% and Collective Brutality is 6.8%.
Sideboard: The top sideboard hate cards are Collective Brutality at 21%, Dispel at 15.5%, Leyline of Sanctity at 10%, and Negate at 9.26%. Ceremonious Rejection and Disdainful Stroke don't matter for us, but satisfy my search criteria.
The first column is the average copies per deck of this archetype if the deck plays this card, so that this number is not affected by a deck that doesn't play the card. The second column is the fraction of decks of this archetype playing this card, and this number is affected by decks that don't play the card.
This is just a core dump of the fraction of decks out of the total for this month with each name.
Edit: Addendum with data sorted by card rather than by deck. Specifically, this is being posted because of Leyline. Leyline sees some maindeck play in Lantern and Bogles, and is in 10% of sideboards. 80%+ of Lantern, Ad Nauseam, and Bogles decks play it in the sideboard and are the primary decks you'll see it from.
Columns are average number of copies in decks that play it, fraction of decks of each type that play it, and fraction of the meta that each deck comprises.
http://sales.starcitygames.com//deckdatabase/displaydeck.php?DeckID=116403
http://sales.starcitygames.com//deckdatabase/displaydeck.php?DeckID=116454
Some interesting points:
Any thoughts? Personally, I still can't understand the latest dissatisfaction with Eidolon and the reasons behind people moving away from it.
As for the RWg one, they're definitely not looking to lower their CMC and I don't agree with cutting Eidolon here. Eiolon is still good even if it's not that great against Eldrazi Tron.
At least they're choosing to play burn spells over Eidolon and not playing Shrine of Burning Rage, so they have that going for them, which is nice.
I'd agree that Eidolon is the most individually powerful card in the deck, and a clear reason why Burn rose to Tier 1 in the first place. However, I also think Eidolon isn't as good as it used to be.
This isn't to say that I think it should be cut - I don't - but I do see why people might want to try experimenting without it.
The two decks people talk about in regards to Eidolon are GDS and ETron. Eidolon is strong against GDS. If they remove it, you made them spend resources to deal with it and that disrupts the flow of their game to some degree. If they don't, it actively applies pressure to them and allows you to sandbag burn spells until you're able to finish them. I'll side out Eidolon against ETron all day, but it's not actually a new thing to have a tier 1 Tron deck that you side Eidolon out against. I sided it out against Gx Tron for the last 2 years. If Eidolon was great 2 years ago (and it was), it's still great today.
I also don't agree that if you're "looking for pro points" you cut Eidolon. Consistency is actually king for those people.
1. Eldrazi Tron. What are our priority targets with Revelry? If they have a map on the battlefield and are tapped out--do you just go ahead and blow it up to prevent Tron? Or do you wait in case they get Chalice?
How does Ballista play into it? Do you kill it on sight?
Deflecting Palm? It just seems awkward. They have to have an attacking creature to use it--a lot of times they cast a TKS before that point, so it gets eaten. A lot of times, I feel like I'd be better off having any burn spell.
2. Elves--WTF are we doing for elves? They're everywhere.
I've got 1 lavamancer in the board, and he obviously does work.
I'm changing my 3 Relic to 2 Relic and 1 Grafdigger's cage to keep them off coco/Chord.
I'm putting 1 copy of searing Blood in the board.
I've seen some people say Anger of the Gods--I just don't like it for Burn. Ensnaring Bridge buys us time but they can chord up a rec sage.
Do you "Bolt the bird?"