Meadbert's testing results with some new lists.

[meadbert]

I just finished testing results with 128 decks and I am going to briefly summarize my results here.  I am also going to go into three lists that did fairly well and that are somewhat unique.

First of all I did not test TPS.  The reason is that I had played about 200 games straight with it before I stated testing and thought I had a good grasp of how good it was. It won about 56% of those games which were against what I thought were the 16 best decks in the format.  This suggests that TPS is about a top 5 deck although maybe not the best.  It then turned out that those "best" decks faired poorly in this tournament and thus a stastical analysis indicates that TPS is merely middle of the road.  I am already off topic.  This is not about TPS.  Instead it is about what I did test.

I also inexplicably left out Intuition Tendrils despite the fact that is has done well for me in the past.

Overall the meta was about a quarter Drains.  Then it was about an 8th each of Rituals, Shops, Bazaars and Rods.  The final quarter would have been Jank decks.

The way I played the tournament was as follows.
Only a deck's records against non eliminated decks matter.  This is because I do not want top decks to be judged based on their ability to beat up on terrible decks.  Instead they should be judged against each other.

I played rounds just like in a tournament including the coin flip.  I played assuming that each player had a pretty good idea what he was playing against.  This makes cards like Cabal Therapy, Pithing Needle, Meddling Mage and Chalice of the Void stronger.  The impact of Meddling Mage was probably the biggest.  It also tends to help control decks since they know which hate to mulligan into or whether they need to mulligan to hate.

Once a deck was 2 rounds below 500 it was eliminated.
When the number of decks dropped to 32 you had to fall 3 rounds below 500.
At 8 you had to drop 4 rounds below 500.
The final two battle out till one deck is 5 rounds behind the other.  (This gets REALLY boring after a while!)

In general Shop decks did terribly.  The last 3 Shop decks alive were the new Transmuter list, Vroman's Uba Stax list (from 2006, not from recently), and then Roland Chang's World Championship list from 2005 with Thirsts.

Tez decks and Vault decks did okay.  In early rounds they had strong records, but then none made the top 16.  The last three were B/U Vault combo with Reconstructions, Meditate Remora and a Arsenal Bulllshit which is a Mono Blue list that uses Commandeers and Misdirections as extra Forces and Cunning Wish->Recall as Fact or Fictions 2-5.

Several Drain decks made the top 16 including Xerox, Becker Drain Tendrils and Bob Tendrils, but only 1 made top 8.  I will get to that later.  I am not sure that Xerox is that good.  I have tested a similar list in the past and it never did great.  I did add Misdirection to the Wish board which seemed to have been an improvement.

Bazaar decks faired poorly.  No Dredge decks made the final 8, but a statistical analysis of the results suggests that Wiley's Dredge list with Wastelands should have finished about 5th.  Basically it lost in the top 16 to a bunch of decks that did VERY well later on.
Fatestitcher Dredge, which I still believe is one of the top decks in the format if not the best lost early.  It was unfortunate in that its first two matchups were against Goblines (with maindeck Leyline of the Void) and Arcane Oath (4 Maindeck Tormod's Crypts).  Note that I was playing so opponents know to mulligan to Leyline/Crypt in game 1.  Wiley Dredge was the only deck with Multiple Bazaars to make the top 16.  I may have hated out Dredge too aggresively hurting Bazaar in general.

Two decks that were eliminated somewhat early but had really good records were Worse than GAT and Dark Drain.

For Worse than GAT I dropped 3 Scroll and 3 Brainstorm and then added Rods till I got to 4 and also added Goyfs.  The resulting list is probably better than the original Worse than GAT and wrecks the meta game I think.  GAT's record was insane but all of the decks it beat kept getting eliminated so Worst than GAT did not make top 16.

The other deck worth mentioning I called Dark Drain which is basically bastardized IT.  I ran 2 Deep Analysis along with 4 Accumulated Knowledge to go with 4 Intuitions.  I also had Necro, Bargain and Desire and 4 Drains and 4 Rituals.  The results was a deck with a silly good draw engine and a fast clock.  It did not have Duress and thus was a bit slow with disruption.  Also Necro, Bargain and DAs do not always play well together.
Anyway this list had a really good record like Worse than GAT, but did not make the top 16 because the decks it beat were eliminated fairly early.

The final deck which I will mention that did not make top 16 is basically Goth Slaver, but with Vault/Key instead:
2 Polluted Delta
3 Flooded Strand
2 Volcanic Island
2 Underground Sea
4 Island
1 Tolarian Academy
1 Black Lotus
1 Lotus Petal
1 Mox Ruby
1 Mox Jet
1 Mox Sapphire
1 Mox Emerald
1 Mox Pearl
1 Mana Vault
1 Sol Ring
1 Mana Crypt
1 Time Vault
1 Voltaic Key
1 Sundering Titan
3 Goblin Welder
1 Yawgmoth's Will
1 Demonic Tutor
1 Tinker
1 Time Walk
4 Force Of Will
1 Fact Or Fiction
1 Gifts Ungiven
4 Intuition
4 Thirst For Knowledge
4 Accumulated Knowledge
1 Hurkyl's Recall
4 Mana Drain
1 Ancestral Recall
1 Brainstorm
1 Mystical Tutor
sideboard
1 Trinisphere
4 Sphere Of Resistance
4 Tormod's Crypt
1 Empty The Warrens
1 Lava Dart
3 Yixlid Jailer
1 Echoing Truth

The board addresses Long and Dredge which are tough matchups.  I think Oath was finally Goth Slaver's undoing, but its overall record (7-2) was really good and indicates that it may be a top deck.  The trouble is that the decks it beat were eliminated while the Oath decks hung around.
Intuition is a huge bomb in the above list since with very little mana and Welder out you can Intuition for Vault/Key/Walk and just win.  With Walk you Weld in both combo pieces and with either combo piece you Weld in the other.
I suspect Welder will be weaker now that there are new Tinker targets so this deck's heyday may have already come and gone, but I thought it worth mentioning here.



The final 8 decks were interesting.  There were 5 Oath decks, 2 Long decks and 1 Non Oath Drain Deck.

I am not going to go into all of the Oath decks since 2 were janky homebrews.  One had Argivian Find and Rootmaze to go with Tyrants.  Those cards have a ton of synergy because you can Oath up Tyrant and then Find out Rootmaze and play it.  That bounces two of your opponent's permanents that will not come back into play tapped.  I might rework the list to use Raven's Crime and Life from the Loam in the main because Dredge into Oath is nice with Argivian Find in hand and both seem like they would be useful afer Oathing.

Old ICBM Oath was one of the top 8 decks.  This was a list back when ICBM ran 2xRod to go with Chalice, Force, Drain and maybe 2 Duress.  It seemed very strong right now.  Crucially it had a strong anti-Oath strategy post board involving Mishra's Factory and Life from the Loam.

James ICBM Oath was also a top 8 deck.  This is probably the least surprising deck to top 8 since it has been doing well at tournaments recently.

One of the two Long decks was Pitch Long.  Basically I took an ancient Pitch Long list and dropped the 3 Brainstorms for Ponder, Scroll and Winfall and it was great.  I actually think the list with Ponder, Scroll and Winfall might be better than the original.


The final 3 decks I am going to provide lists for and go into details.

3rd place was Turbo Oath:
4 City Of Brass
4 Forbidden Orchard
4 Gemstone Mine
1 Barbarian Ring
1 Nomad Stadium
4 Ancient Tomb
1 Strip Mine
4 Wasteland
1 Bazaar Of Baghdad
1 Glacial Chasm
1 Black Lotus
4 Crucible Of Worlds
4 Null Rod
2 Hellkite Overloard
1 Yawgmoth's Will
1 Demonic Tutor
1 Imperial Seal
1 Raven's Crime
1 Vampiric Tutor
1 Time Walk
4 Intuition
1 Ancestral Recall
4 Oath Of Druids
1 Fastbond
1 Life From The Loam
1 Regrowth
1 Krosan Reclamation
4 Argivian Find
1 Enlightened Tutor
sideboard
1 Trinisphere
4 Leyline Of The Void
1 Wheel Of Sun And Moon
1 Flame Jab
1 Ancient Grudge
4 Duress
1 Thoughtseize
1 Choke
1 Ray Of Revelation

There is a combo win with Fastbond, Crucible and then Wasteland, Nomad Stadium and Barbarian Ring.  Without Ring you still have infinite mana.
This deck is very strong.  Null Rod is wrecking ball right now and along with the 4 Rods, Turbo Oath runs 8 Tutors to find them and then 6 ways to get them back if countered (4 Argivian Find, Regrowth, Yawg)
Also, Rod is basically a one sided bomb since it only shuts off your Black Lotus.  It may seem like there are not enough ways to get Rod mana up in time, but running 4 Ancient Tomb and 1 Lotus leaves just 1 fewer accelerant than ICBM used to run and it is far more than most Fish decks run so Null Rod hits more reliably, more quickly and with more protection(Argivian Find) in this list than in any other list.

Oath is obviously a bomb.  Fastbond is a must counter and Crucible is usually dangerous as well.

It is not uncommon to hardcast Overlords.  I run two because sometimes one is in hand in which case I really want to just pay RR to swing for 10 twice with the other to win in 2 turns instead of 4 turns with Akroma.

Turbo Oath can do a number of things with Intuition.  You can find Oath or Orchard.  You can find Rod if you want Oath.  You can find Fastbond with 2xArgivian Find and Fastbond.  You can find any card with Yawg, Regrowth and card.  You can get 3 of (Bazaar of Baghdad, Strip Mine, Life from the Loam, Raven's Crime)  Note that Orchard can be added to that list as well if you already have Oath.

While Turbo Oath can abuse the yard, it also functions just fine without the yard by Intuitioning for Oaths/Rods/Orchards to win with Overlords.

The board has been adjusted to help beat Long.  Having 5 Duress effects to go with Trinisphere is nice.  Note that Trinisphere can be easily tutored up.
Generally board out 4 Intuitions, 1 Crucible and 1 Argiviand Find against Long.


The token Choke in the board is to mess up Basic Island.dec which is frequently a plan used to defeat you.

Leylines and Wheels are there for Dredge.
Rods and Raven's Crime come out against Dredge.

Ray of Revelation and Ancient Grudge are there because they are great with Intuition.

Flame Jab is the only solution to Magus of the Moon and there is basically no solution to Blood Moon.

Turbo Oath's weakness is that it has weak turn 1 plays.  Rod can come down on turn 1, but not reliably.  Instead turn 1 is usually used to set up a turn 2 Bomb.  Once a bomb is dropped on turn 2 the bombs can keep dropping 1 a turn for several turns.  The exception is of course Wasteland and Strip Mine which do hit turn 1 against Dredge all of the time.  Also, post board you have 5 Duress effects against Long to reliably have a turn 1 play.

2nd Place was Duressless Grim Long:
4 City Of Brass
4 Gemstone Mine
1 Forbidden Orchard
1 Underground Sea
1 Tolarian Academy
1 Black Lotus
1 Lion's Eye Diamond
1 Lotus Petal
1 Mox Ruby
1 Mox Jet
1 Mox Sapphire
1 Mox Emerald
1 Mox Pearl
1 Mana Vault
1 Sol Ring
1 Mana Crypt
1 Memory Jar
2 Simian Spirit Guide
2 Empty The Warrens
1 Wheel Of Fortune
1 Yawgmoth's Bargain
1 Necropotence
1 Tendrils Of Agony
3 Grim Tutor
1 Yawgmoth's Will
1 Demonic Tutor
1 Imperial Seal
1 Ad Nauseam
4 Cabal Ritual
4 Dark Ritual
1 Vampiric Tutor
1 Mind's Desire
1 Timetwister
1 Tinker
1 Windfall
1 Time Walk
1 Ponder
1 Hurykl's Recall
1 Ancestral Recall
1 Brainstorm
1 Chain Of Vapor
1 Mystical Tutor
1 Regrowth
1 Enlightened Tutor
sideboard
1 Darksteel Colossus
2 Simian Spirit Guide
2 Empty The Warrens
1 Rebuild
1 Hurkyl's Recall
4 Xantid Swarm
4 Orim's Chant

Mostly this is a traditional Grim Long list similar to what Stephen Menendian originally suggested.  There are a few huge differences.  The biggest change is that I dropped Duresses.  I sat down and was trying to figure out what was best to run for disruption in Long.  Duresses are strong against Combo and decent against Control, but they are pretty weak against Shops and terrible against Dredge.

Then there were Xantid Swarms which rip apart control, but are bad against the rest.  Orim's Chant is great against Dredge and Combo and decent against Control, but pretty weak against Stax (although possibly still better than Duress!)

I was temped to just replace Duresses with Orim's Chant and call it a day since it seemed to me that Chant was just better against the current meta, but then I decided that it was a combo deck so instead of figuring out disruption why not just try to win faster.  I could instead drop the 4 Duress for 2 Simian Spirit Guide and 2 Empty the Warrrens.  Warrens crushes Stax and Fish.  Warrens and Spirit Guide are annoying for Control decks because they never know when Ritual will be followed by SSG and Warrens.  Warrens can also be used to either race or buy time against Dredge.
I then put Xantids and Chants in the board to go along with more Spirit Guides and Warrens.

The other change I had to make was of course to replace the Brainstorms that were restricted.  I added Enlightened Tutor because it was broken.  Being able to ET for Lotus on upkeep makes Desire so much better.  Also you can find Necro, Jar, Bargain or occationally Mana Crypt.

I replaced the other Brainstorms with Ponder and then Ad Nauseam.  A token Ad Nauseam may seem strange, but it is just what this list needed.  The question was do I run Infernal Contract or Ad Nauseam and I ran Ad Nauseam which was huge all the time.

I may have also modified the original list to get the 4th Cabal Ritual.  I am pretty sure that 4 Cabal Rituals should be played.  They crush Shop decks since even with Chalice@0, Chalice@1, Resistor, Thorn and Trinisphere all out at once you can still generate mana with Cabal Ritual and threshold.

The above list nearly beats Dredge preboard and then wins barely post board.  It crushes Control post board thanks to 4 Swarm and does well enough pre board thanks to the threat of Warrens causing opponents to blow Forces on Dark Ritual occationally.

There was actually not much in the way of weaknesses for this deck and although it finished second, the statistical analysis indicates that it was the best deck.

The winning deck was Arcane Denial.dec.
Last year I suggested that Arcane Denial might be the best replacement for Brainstorm/Scroll and at the time proposed a list that used the Plagiarize/Cephalid Colliseam combo.  I then built a deck around assembling that combo along with 6 mana and Threshold.  It turns out that as long as you are assembling Threshold and 6 mana by turn 3 what you really want is Skeletal Scrying!  Scrying acts as a 1 card combo in that situation and is a nice replacement for Yawg in case it is Duressed.  I dropped the Plagiarizes entirely and scaled back Cephalid Colliseam to add other cards.  Here is the list:
1 Polluted Delta
2 Flooded Strand
1 Volcanic Island
2 Underground Sea
2 Cephalid Colliseum
3 Island
1 Tolarian Academy
1 Black Lotus
1 Mox Ruby
1 Mox Jet
1 Mox Sapphire
1 Mox Emerald
1 Mox Pearl
1 Sol Ring
1 Mana Crypt
4 Mishra's Bauble
4 Urza's Bauble
1 Darksteel Colossus
1 Empty The Warrens
1 Yawgmoth's Will
1 Demonic Tutor
2 Skeletal Scrying
1 Vampiric Tutor
1 Tinker
1 Merchant Scroll
1 Time Walk
4 Force Of Will
2 Misdirection
4 Thirst For Knowledge
1 Chain Of Vapor
4 Arcane Denial
4 Mana Drain
1 Ancestral Recall
1 Brainstorm
1 Mystical Tutor
sideboard
4 Pithing Needle
4 Engineered Explosives
4 Tormod's Crypt
1 Rebuild
2 Hurkyl's Recall

Perhaps an Island should be added to the board in place of a Needle.

Arcane Denial is an accelent card in conjunction with Baubles.  The reason is that you can counter your own Bauble and turn 2 cards into 3.  This also loads the yard like crazy and is what gets you threshold so fast.
Baubles sort of replace Brainstorm in that they show you a few extra cards on turn 1 so you find Force with more regularity against Long.

Denial replaced Scroll in that you can reliably have counter back up on turn 1 to counter the Crucible Necro or other bomb that drops quickly.

With 4xThirst and 4xDenial there is a really strong draw engine.  Denial's flexibility is also remenicient of Merchant Scroll.  You can use it with a Bauble or Mox to draw 3 cards or you can lose card advantage and pay 1u to counter a bomb.

As a general rule of thumb if I already have Force in hand or if I do not yet have my land drop for next turn then I go for the draw.  Otherwise I hold Denial in case I need to counter a bomb.

Playing through Chalice and Null Rod is easy.  With 10 ways to pitch artifacts (including Colliseums) you rarely get stuck with artifacts in hand.

Arcane Denial ends up being must faster than one might think.  The Baubles cycle through your "bad" cards so that all you end up seeing are good one.  This means you see far more Forces, Thirsts, Drains and Power than you should. 

The board is well suited for crushing Dredge since you can bring in Crypts, Explosives and Needles for Forces, MisDs and Drains (possibly a Bauble or 2).

Explosives are huge in a few matchups.  Most importantly they address Oath of Druids and Sphere of Resistance.  While Rod can be played through, Resistor is a huge pain because you must now pay 4 to Denial yoru own Bauble.  Also just cycling Baubles is slow.  Explosives can remove Resistors easily since you can play it for 0 and then pay BU to get 2 sunburst counters.
Having 4 bounce and 4 Resistors against Stax seems to be enough, but I did find that I wanted a 4th basic post board so perhaps that should be added.
Explosives was also strong against Duressless Grim Long since it answers Xantid.  It is actually not THAT great because if Long has turn 1 Xantid and turn 2 win then you might not have the 3 mana needed to use Explosives, but starting with turn 1 Explosives for 1 on the play is not terrible.

Thirst is way more busted in this list than in other lists since you run 17 artifacts pre board and then usually more post board.

This deck's Yawgs are absolutely insane thanks to Baubles and Denials.  It is very common to play a turn 2 or 3 Yawg to draw 7 cards off Baubles and Denials.  What is great is that you can actually draw up to 10 cards on your opponent's upkeep and maintain massive card advantage while your deck plays like it has a virtual Library of Leng out.
Also the information that is gained by Baubles is nice for a control deck.

Empty the Warrens is strong backup win.  Baubles help generate storm and Warrens also for nutty Yawg, Warrens, Walk plays which are frequently needed to race combo.

The final game that won the tournament for Denial went as follows:
Turn 1 Long opens with Xantid which resolves.
Denial goes Land, Mox, Denial on Baubles, Bauble.
Long Imperial Seals for turn 3 win.
Denial goes Lotus, Mana Crypt, Tinker, Walk.
Turn 2.b, Swing with DSC, YAwg, Lotus, Walk.
Turn 2.c win.

Baubles help you keep drawing aginst Mana Denial so you can keep hitting land drops against decks with Wasteland.  This was crucial in the Turbo Oath matchup.  I thought that Turbo Oath would surely have the advantage, but instead Arcane Denial was able to keep countering, bouncing and just ignoring Null Rod to finally win with a huge Yawg.

For those of you who love Drain decks I strongly recommend trying the above list.

[feyd]

You are a mad man to create a deck which utilizes arcane denial so effortlessly.  I commend you for that.  Your turbo oath list and arcane denial decks look fantastic; pretty cutting edge if you ask me.  I'd personally like to know if you tested any bomberman lists recently and, if so, how did they work out for you.  Obviously you are the master of dredge and enjoy beating people to death with an army of small zombies; I can see you like aggro oriented decks like GAT, dredge, and oath. 
Off topic: in your opinion do you think the new ink leviathan is worth a damn?

[zeus-online]

it'd be nice knowing which decks where included....That is so easy to discover that it suggests conspicuousness or little need for perspicacity in the observer. Long would win if everything was aggro or something.

You haven't mentioned fish decks, ad nauseam and a host of other decks which i find relevant.

Other then that, looks like an interesting experiment, and you must have too much time

/Zeus

[A.-1.]

I'm certainly not going to question the thoroughness of your tournament.  Your playtesting results in the past have proven very in-depth.  I do, however, wonder if your results may have been skewed.

Do you feel like your familiarity with certain decks, like Arcane Denial.dec, helped them to fare better? Would you expect the same results if you had not spent so much time with these decks?  I'm not doubting that these decks could have made the top 8 with another pilot.  I have tested the decks you have created in the past with favorable results.  I just feel like they may have placed higher than they normally would.

I played assuming that each player had a pretty good idea what he was playing against.  This makes cards like Cabal Therapy, Pithing Needle, Meddling Mage and Chalice of the Void stronger.  The impact of Meddling Mage was probably the biggest.  It also tends to help control decks since they know which hate to mulligan into or whether they need to mulligan to hate.

Even though a quarter of the decks were what you called jank, did you play like the opponent would know the entire decklist of the jank deck?  This would take away one of the few benefits that many homebrew decks have over more established decks.

Tez decks and Vault decks did okay.  In early rounds they had strong records, but then none made the top 16.  The last three were B/U Vault combo with Reconstructions, Meditate Remora and a Arsenal Bulllshit which is a Mono Blue list that uses Commandeers and Misdirections as extra Forces and Cunning Wish->Recall as Fact or Fictions 2-5.

Did these decks do poorly after the early rounds or were they eliminated because their wins were against decks that were terrible?

Note that I was playing so opponents know to mulligan to Leyline/Crypt in game 1.

Doing this against an opponent who always plays Dredge seems perfectly reasonable.  But you will not know every person who plays Dredge. 

Did any Fish-type deck do well?  What build do you feel is the best?

[meadbert]

Off topic: in your opinion do you think the new ink leviathan is worth a damn?

I actually generated most of these lists back in December so other than R/U Trans Stax I did not test either Leviathan or Progentius much.
I will say that dodging Welder is huge and will cause Uba Stax trouble.

it'd be nice knowing which decks where included....That is so easy to discover that it suggests conspicuousness or little need for perspicacity in the observer. Long would win if everything was aggro or something.

You haven't mentioned fish decks, ad nauseam and a host of other decks which i find relevant.

Other then that, looks like an interesting experiment, and you must have too much time

/Zeus

it'd be nice knowing which decks where included.
/Zeus

Shop Decks: Red Shop Aggro, Mono Blue Stax, Mud, Blue Serum Powder Stax, Mud Aggro, Gilded Claw, Vroman Uba Stax, Chang Stax, Gamble Turboland, Blue Shop Aggro, Serum Powder Uba Stax, Mad Man Mike Aggro/Combo, Uba Cap, Blax, Uba Land, R/U Trans Stax

Oath Decks:  Tez Oath, Oath Intuitive, Old ICBM Oath, Turbo Oath, Rector Tendrils, King James Oath, Rootmaze Oath, Mana Oath, Arcane Oath.

Skullclamp Decks:  Suicide Virus, R/G Xmas Beats, Jank, Kobolds, Elves.

Goblin Welder Decks:  Master Slaver, Capsule Slaver, Traditional Slaver, R/G Fishy Stax, Survival, Welder Combo, 5c Belcher, R/G Warrens Belcher, Painter Combo with Hatching Plans, Empty Slaver, Key/Top.

Null Rod Decks:  Vinelasher Kudzu.dec, BUG Fish, Worse than GAT, R/G Hate, Turboland, Shop Land, Squee Land, U/W Fish, Countryside Crusher.

Landstill decks:  White Landstill, Landstill with Bazaar/Squee, Red Landstill, Mystic Remora Mono Blue Landstill, Green Landstill

Bazaar of Baghdad Decks:  Madness, 42 Land.dec (I actually ported to vintage ... went 0-2), Strategic Dragon, Wiley/Waste Dredge, Field Dredge, Turbo Nauseam, Colorless Dredge, Winds Dredge, WGDX, Mana Dredge, Fatestitcher Dredge, B/G Combo Dragon, Dawn of the Dead.

Dark Confidant Decks:  Dark Illusions, Bob Nauseam, Erayo SS, Deez Goyfs, Magus of the Future.dec, Mask Not, Suicide Black, Vroman Painter, Super Fish, Mono Black Tendrils, Dark Gifts.

Drains:  Hulk Smash, Vinci Drain Tendrils, Vault Combo with AKs, Restricted List.dec, Plagiarize.dec, Crappy Turbo Gush, Replenish, Tinker Drain Tendrils, Tez Bomberman, Strategic Slaver, U/W Bomberman, B/U/W Bomberman, Goth Slaver, Arsenal Bullshit, Xerox, Vault Combo with Argivian Finds, Mono Blue Control with Tez, Gree Tez, Modified Meandeck Gifts, GAT, Drain Nauseam, Dark Drain, Composite Tez, Turtenwald Painter, Meditate Remora, Old Drain Tendrils, B/U Vault Combo, Becker Drain Tendrils, Beaumont Painter, Arcane Denial.

Rituals:  Long with Vault/Key, Meandeck Tendrils, Thirst Long, GWS Long, Duressless Grim Long, Pitch Long, Meandeck Ad Nauseam.

Jank:  Mono Red Goblins, Goblines, R/B Goblins, Meandeck Parfait, Dave Parfait, Witness combo, Time Spiral Turboland, Mono Red Burn, Janky Jank, Turn 2 Tinker.


That is the list of decks.  Keep in mind that as the tournament goes on your performance against decks that have been eliminated is ignored.  I do that specifically because I do not want to select based on which deck beats up on bad decks the best.  Instead I want to know which good decks beat other good decks.

You haven't mentioned fish decks, ad nauseam and a host of other decks which i find relevant.

/Zeus

Fish did fine untill towards the end.  Basically a meta with 5 Oath decks in the top 8 is a nightmare for Fish.  This is actually heavily mitigated by Meddling Mage.

Worse than GAT I consider to be a Fish deck and that did well.
The last two Fish decks to be eliminated were Vinelasher Kudzu.dec which is strong, but suffers against Oath and then U/W Fish.
The U/W list was actually a Feinstein list from before Gifts Restriction.  I then dropped 3 Brainstorms and added 3 Mindscensors
BUG Fish did not do well.  It did not have Meddling Mage to deal with Oath and it is in no position to race Oath.

There were five decks with Ad Nauseam.  First was Duressless Grim Long which only ran a token Ad Nauseam and finished in second.
Next was Meandeck Ad Nauseam which did fairly well finishing in the top 32.  Its losses were to Duressless Grim Long, Dark Drain(IT), and Bob Gifts which are all strong decks.

Third was Turbo Nauseam:
4 CityOfBrass
4 GemstoneMine
1 BarbarianRing
1 CephalidColliseum
1 TolarianAcademy
1 StripMine
4 Wasteland
3 BazaarOfBaghdad
3 NomadStadium
1 BlackLotus
1 LotusPetal
4 MoxDiamond
1 MoxRuby
1 MoxJet
1 MoxSapphire
1 MoxEmerald
1 MoxPearl
1 ManaVault
1 SolRing
1 ManaCrypt
4 CrucibleOfWorlds
1 EngineeredExplosives
1 DemonicTutor
1 ImperialSeal
4 Mindtwist
4 AdNauseam
1 VampiricTutor
1 TimeWalk
1 AncestralRecall
1 Fastbond
1 EnlightenedTutor
4 ArgivianFind
sideboard
4 GhostQuarter
4 PithingNeedle
3 Tormod'sCrypt
4 XantidSwarm

I have had some poor responses to this list even though I think it is obviously good.  Folks will compain about "jank" like Mindtwist, Mox Diamond, Bazaar of Baghdad and Argivian Find.  Mindtwist and Mox Diamond were restricted until fairly recently, Bazaar has been mentioned as a candidate for restriction and my love of Argivian Find has been well documented.  The total mana cost of the deck is far less than Meandeck Ad Nauseam so Ad Nauseam basically always wins.  Mindtwisting for 5 on turn 2 happens frequently too.
The weakness is the vulnerability to Leyline which is what held this list down the last time I tested it.  This time there were fewer Leylines and more Tormod's Crypt and Relics which can be answered with Engineered Explosives or Pithing Needle.


The only list to with Ad Nauseam to do poorly was Bob Nauseam.  You only had to lose your first two rounds to eliminated and this is what happened to Bob Nauseam.  It does not mean that it is not a good list.  It may have been unlucky like Fatestitcher Dredge.

EDIT:  I left out Drain Nauseam.  Drain Nauseam was basically Meandeck Ad Nauseam with Drains instead of Thoughtseizes.  I may have slightly reworked the manabase to play nice with Drain.  Probably I swapped a Swamp for an Island or something like that.  It compiled a 4-2 record in rounds which is pretty good.  Its losses were to Elves and Worse than GAT.

Do you feel like your familiarity with certain decks, like Arcane Denial.dec, helped them to fare better? Would you expect the same results if you had not spent so much time with these decks?

So the weakness of 2 fisted testing is that the pilot does heavily impact the results.  I am not sure to what degree familiarity with a deck helps or hurts it.  As an example, I would say that I play Turboland decks better than most, but I also play against Turboland decks better than most.  I have been to tournaments where I played Turboland and my opponents dropped Chalice@0 against me 7 times and never once dropped Chalice@1.  I definitely agree that I impact how testing goes, but it is tough to measure how and whether experience with a deck helps or hurts it.  I will say that Arcane Denial.dec does not seem to be THAT skill intensive to me.  Basically the toughest decision is when to use Denial to draw 3 cards on turn 1 and when to hold it as a counter and I already gave my two rules.

Even though a quarter of the decks were what you called jank, did you play like the opponent would know the entire decklist of the jank deck?  This would take away one of the few benefits that many homebrew decks have over more established decks.

I used the wrong word when I said Jank.  What I meant by Jank was lists that have neither Shops, Drains, Rits, Rods nor Bazaars.  By this measure ICBM James Oath is considered Jank when in reality nobody would consider it to be Jank.
I did not play as I knew the entire decklist.  Specicially I played as though you knew your opponents' 3ofs and 4ofs including the sideboard.  This meant that you knew that Strategic Dragon can transform into Oath.  It also meant that Dredge would know an opponent had 4xLeyline.  Dredge would not know if an opponent had 2xPlanar Void though.

Tez decks and Vault decks did okay.  In early rounds they had strong records, but then none made the top 16.  The last three were B/U Vault combo with Reconstructions, Meditate Remora and a Arsenal Bulllshit which is a Mono Blue list that uses Commandeers and Misdirections as extra Forces and Cunning Wish->Recall as Fact or Fictions 2-5.

Did these decks do poorly after the early rounds or were they eliminated because their wins were against decks that were terrible?

B/U Vault combo was eliminated by Goblins and Xerox.  Xerox has an insane Draw engine for the Drain mirror.  I do not know how Goblins did well there.

Arsenal Bullshits final 3 losses were losing twice to Arcane Denial and once to Duressless Grim Long.  I cannot say it was a bad deck.  It may have just been unlucky in that it was put up against the top two decks so early in the tournament.

Meditate Remora's last three losses were two against Oath decks and one against Elves.  Elves is naturally a terrible deck if you are running Mystica Remora.

Note that I was playing so opponents know to mulligan to Leyline/Crypt in game 1.

Doing this against an opponent who always plays Dredge seems perfectly reasonable.  But you will not know every person who plays Dredge. 

In round one you will not know who is playing Dredge.  After round one any team will have scouted enough to know at least each and every Dredge player.  That information is extremely valuable since you want to know if you should mull to Wasteland in game 1 or Vamp for Tormod's Crypt instead of Recall and it is easily gathering since dredging 40 cards is difficult to do discreetly.

[Grand Inquisitor]

WTF
 

Madman, indeed.

This level of precision in work is always appreciated when shared, so thanks.  However, what are we supposed to take away here?

The methods you employed for your testing are obviously complex, but are they useful or illustrative?

I just finished testing results with 128 decks

But many of these lists only played two sample matches, right?...

Once a deck was 2 rounds below 500 it was eliminated.

You even give an example of a well known metagame force falling to this small sample size:

Fatestitcher Dredge, which I still believe is one of the top decks in the format if not the best lost early.  It was unfortunate in that its first two matchups were against Goblines (with maindeck Leyline of the Void) and Arcane Oath (4 Maindeck Tormod's Crypts).

Given the strength of this contender, it's ability to beat a large part of the field, and the unlikelyhood of meeting any deck with that much MD hate (nevermind two such decks in a row), doesn't this largely invalidate the process as a meaningful metagame evaluation?

I'm not saying this is useless information, that your post lacks quality or that this isn't good stuff to have.  It just seems to be that (especially relative to the effort), there's not a strong link between your system of evaluation and your findings about three (innovate, but unpreedented) decks.

For example, it would be very hard for me (an experienced T1 player) to look at something like turbo oath and imagine how it would perform against something like uba stax or even Tezzeret.  If you had said something like, "Turbo Oath won 45 of 60 preboard matches against Tezzeret; the common critical plays were..." then it's much easier for this audience to look at it and give it a smell test.  I especially don't want basic users looking at this and running out to pick up playsets of arcane denial

Why did you pick the system you did?  Or, if I'm missing the point, what is the purpose of this post?

[FlyFlySideOfFry]

Isn't Arcane Denial a Sorcery speed Thirst for Knowledge that requires you to run bad cards to make it reasonable to run? Why not just dedicate 7 slots to ak-int.

Edit: To elaborate isn't it just that out of all the Drain decks this is the one that just happened to get lucky draws/pairings? Between all the current available engines it just seems that Denial+Baubles is just weak and I'm surprised that it did so well. More importantly it seems like Oath decks did so well because over 50% of the field was either aggro or shops, which doesn't seem to be the case in real tournament settings. It would probably have worked out better to test a smaller pool of decks that are actually being run in tournaments with a percentage correlating to actual tournament data of T8s and gotten more rounds per deck than testing out a bunch of decks most people would agree are bad (or at least not the most effective list they could be) and get the result that "ya, Oath slams a janky metagame".

[benthetenor]

Alright, I'll bite.

What did Arcane Denial.dec lose to? Apparently, it's quite good in the Oath-heavy metagame that comprised the Top 8, but what about against the standard decks? It does look really interesting, and I'm always a fan of decks that draw an insane number of cards.

I think that the fact that "good" decks lost in early rounds is a good sign that the system works, because statistically speaking, good decks will lose  early in a large tournament. Luck happens. The only information that this analysis didn't give us was about the mirror matches, but it was specifically designed so that that would be impossible. The basic thing that you can take away from this is, in a wide-open metagame, the top decks might shake out this way.

I do want to commend you on what was obviously a lot of testing. And if the result is a sweet new deck, then all the better!

[Anusien]

Did you consider running multiple copies or tweaks on the same deck rather than adding new archetypes?  Your archetype list in many respects reads like the worst cross-section of 2006.

[wiley]

The methods you employed for your testing are obviously complex, but are they useful or illustrative?

This would be similar to having a 128 man tournament with some (not all) of the luck factor reduced.  I would say that is useful.  As far as illustrative, well some of those decks would seldom be played by players with skill, but that doesn't mean they are bad decks.  This tournament gives probably your best look at what those decks are capable of when piloted competently.

The biggest problem is that, while there are multiple types of decks, none actually appear in multiples.  This ignores the popularity factor, which is certianly somthing that you should realize before taking this (or any tournament results) as canon.

As far as questioning familiarity with decks in two fisted testing; 1) Bert has a huge amount of experience with long type decks as well as drains, oath and all of his project decks, he plays more than anyone I know.  2) His knowledge of all of his pet decks helps the established archetypes far more than the reverse.  He explained this when he talked about his own experience with turboland.

Isn't Arcane Denial a Sorcery speed Thirst for Knowledge that requires you to run bad cards to make it reasonable to run?

Short answer: no.  Long answer: you're thinking of Mystic Denial, Arcane denial is a much better card.

Fake EDIT: Sarnathed for some of the post.

[FlyFlySideOfFry]

Isn't Arcane Denial a Sorcery speed Thirst for Knowledge that requires you to run bad cards to make it reasonable to run?

Short answer: no.  Long answer: you're thinking of Mystic Denial, Arcane denial is a much better card.

Fake EDIT: Sarnathed for some of the post.

Mystic denial is a 1UU instant: counter target creature or sorcery spell. Arcane Denial is a 1U sorcery: discard a 0cc artifact draw 3 cards.

[Grand Inquisitor]

This would be similar to having a 128 man tournament with some (not all) of the luck factor reduced.

I don't think this means what you think it means.

As for Bert's competence, I've seen nothing to question it in his posting.  What I'm curious about is the consequences of his choice of testing methods has on what he's telling us (apparently that these three decks are pretty good).

When a series of multi-person tournaments are culled for trends, top decks show (some composite) of what is effective, popular and/or well-played.  Having one person play a single tournament with perfect information, a single playstyle in a synthetic metagame, the result is not that you reduce the luck factor.  As for what the result is, I'm still trying to figure that out.

[meadbert]

There are definitely huge problems with the methodology.  Letting good decks go 0-2 drop is probably one of them.  Two fisted testing is very problematic.
For instance with Meddling Mage or Cabal Therapy one of the toughest plays it figuring out what might beat you in a game you are already winning.  In real game situations I might forget various plays that can still beat me.  Might Pyroclasm or Warrens still beat me.  The trouble with two fisted testing is you always know your opponent's hand "and what they are thinking" thus it is "impossible" to forget about the card they are about to beat you with.

Definitely there are huge problems in the system.  This is not mean to be some sort of authoritative declaration of what top 8s should look like,  Under no circumstance should this be taken more seriously than normal tournament results.  Where this is useful is regarding decks where there is very little tournament data.  How many players have take 8 Baubles and 4 Arcane Denials into a type 1 tournament recently?  If 5 players tell me they did and they fared poorly then their experience means more.  Since we do not have the tournament data, I would argue that this testing is the best we have to go on.

Because of the testing methodology decks that finished poorly played few games and decks that finished with good records played many games.
What this means is that comments regarding poor decks mean much less than comments regarding decks that finished off quite well.
If some deck goes 2-4 in games but is eliminated it hardly is convincing data to show it is a bad deck.

The decks that made it far were played a lot.  Arcane Denial went 70-46 in games.  That is a lot of games and a lot of testing.  Such a record is unlikely unless either Arcane Denial is good or their is a severe flaw in my testing and I am either playing other decks very poorly or somehow causing bias for Arcane Denial.  Even if there is some bias, the record is strong enough to suggest that Arcane Denial is good.  When I say that Arcane Denial is good it is with high confidence.

Anyway, my point is not to supplant any commonly held beliefs.  There is no discussion of Tez vs TPS or Tez vs Dredge here because folks play those matchups enough and know how they turn out.  These results are intended merely to provide a little light for the corners of the vintage meta game that normally stay in the dark.  If anyone wondered how Worse than GAT might do, we now have a little bit of information.

What I encourage is for those with interest to try Arcane Denial or Duressless Grim Long and ideally improve upon them.

[pierce]

there are many ways to do research. Explain how this approach is benificial, and not just riddled with inbreed thought and extremely subjective play.

Otherwise, I'm pretty sure this is a tragic example of something cool that doesn't benefit anyone.

[The Atog Lord]

OK, Pierce, I'll take a stab at explaining the value of what Meadbert has done here. I use two-fisted testing to prepare for tournaments when other options are not available. It is often far more beneficial than attempting to gain insight by testing against the denizens of MWS. On the one hand, there are no doubt problems with the methodology. It is going to prevent more tricky plays from working properly, and you can't pretend that you don't know what both decks have in hand. Moreover, it is going to favor the deck the player knows better. On the other hand, there are problems with standard playtesting too. One player may be more skilled than the other, or one player may be better able to play his deck than the other. Nothing is perfect.

Two fisted testing can give some valuable insights. It often reveals key cards in a match, and highlights some salient features of the how different cards interact. It can help reveal, for example, whether the disruption in one's sideboard is sufficient to halt Ichorid.

In this case, did Meadbert's result prove that Arcane Denial is the best deck in the format? No, of course not. But he did provide evidence that it may be more viable than previously thought. That itself, as well as other other results he provided, are quite worth reading.

Thank you, Meadbert, for sharing your results with us.

[meadbert]

Isn't Arcane Denial a Sorcery speed Thirst for Knowledge that requires you to run bad cards to make it reasonable to run? Why not just dedicate 7 slots to ak-int.

Actually this is a great question and I want to comment more on it.  What has already been mentioned is that Arcane Denial costs 2 rather than 3 which makes it less expensive than Thirst for Knowledge.  This mitigates its "Sorcery Speed" limitation since you can drop it on turn 1 and not mess up your Drain.

It is illustrative to compare Arcane Denial to Thirst for Knowledge and Merchant Scroll.

In the Thirst comparison note that both cards want an artifact to be good.  Most thirst decks run around 13 artifacts.  This list run 17 artifacts, 16 of which can be Denialed. 
First lets compare the early game.  Thirst is rarely a turn 1 play.  Meanwhile Denial is.  It can either counter a bomb or gain card advantage with any of your 16 0cc cards.
By turn 2 Thirst is arguable the better card.  There is a reason this list runs a full set of Thirsts, but by turn 4 the tables start to turn in Denial's favor again.
Denial can be played in response to any of your own spells.  For example if I cast Thirst EOT and it is Drained, then I can Denial my own Thirst in response and draw 3 cards with Denial.  For that reason Denial is actually more powerful than Thirst once I have 5 mana up.
The main argument against Denial comparing it to Thirst is that it is not useful without an artifact to pitch.  Thirst is also worse in such a situation and Denial can still counter a bomb so it is not like Denial sits dead in your hand.

The second card is the more accurate comparison.  Although Denial seems to function like Thirst because it interacts with Baubles, it really is more of a replacement for Merchant Scroll.  I covered this earlier but Scrolls most important uses used to be drawing 3 cards with Ancestral Recall and grabbing Force on turn 1 against Combo.  Scroll could also find bounce as a solution or Mystical for Yawg in the late game so its flexibility went beyond the first two examples, but those were two of the most common uses.

Arcane Denial can be used for either use.  Denial can be used in conjunction with a Bauble or Mox to draw 3 cards similarly to Scroll.  It is true that the card advantage is +1 instead of +2, but you dig just as deep and for 1 less mana and a turn earlier.  This makes it more likely that you find counter magic.  Generally an extra card of card advantage is better than an extra mana thus Scroll for Recall is better than Denial, but Denialing a Bauble is still pretty good even if not as good as Scroll->Recall.

The second comparison is the Scroll for Force comparison.  Denial can be used to counter a spell with -1 card advantage for  just like Scroll could by grabbing Force.  In this manner Denial is actually an improvement since you can delay your decision.  For instance you can keep Denial up on turn 1, but if you opponent mearly Brainstorms or Ponders then you have Drain mana up by turn 2 and you can potentially use Denial to draw 3 cards on turn 3 with Drain mana up.  If you Scroll for Force on turn 1, then you are stuck with Force.  You cannot change your mind once you have Drain up and decide you want Recall instead.  By delaying the choice Denial offers more flexibility and in at least this respect is an improvement over Scroll.

The final argument against Arcane Denial is that it forces you to run "bad" cards like the Baubles.  How bad are the Baubles?  As I mentioned before drawing a card is slightly better than producing a mana.  Baubles draw a card.  Lotus Petal produces a mana.  Lotus Petal is restricted.  Petal can produce any color mana, but then Baubles give you information regarding your opponent's hand or library.  Each is broken post Yawg.  Each pitches to Thirst.  Each can be Tinkered out.  What I am claiming here is that Baubles are not much worse than Petal.  In fact I clearly find them superior to Petal in this list since I run 8 Baubles but no Petal.  Petal gets Drain up on turn 1.  Baubles help draw into Force on turn 1.  Baubles end up having synergy with the deck in the following ways:

Force/Drain: Baubles occationally draw Force on turn 1 and give you crucial information regarding your opponent's hand making decisions about countering spells easier.
Denial:  Draw 3 cards.
Thirst: Draw 3 cards.
Skeletal Scrying:  Draw 5 cards on turn 3 thanks to Threshold.
Yawg:  Allows insane Yawgs.
Mystical/Vamp:  Allows for Mysticall/VAmp for Force on turn 1 and then draw into Force before your opponent's turn.
Tinker:  Can be Tinkered out.
Empty the Warrens:  Easy storm enabler.
Lands:  By cycling having your 8 "worst" cards cycle for free you can get away with running fewer lands.
Tolarian Academy:  Generate an insane amount of mana.

The result is that Baubles end up having synergy with nearly the entire deck.  With so many uses for Bauble, it must be totally dead in some situation in order to be a bad card.  Instead Bauble cycles for 0 which is not bad at all.

Arcane Denial's power is comparable with if not as good as Thirst for Knowledge and Merchant Scroll and Mishra's Baubles are not much worse than Lotus Petal.

[FlyFlySideOfFry]

Isn't Arcane Denial a Sorcery speed Thirst for Knowledge that requires you to run bad cards to make it reasonable to run? Why not just dedicate 7 slots to ak-int.

Actually this is a great question and I want to comment more on it.  What has already been mentioned is that Arcane Denial costs 2 rather than 3 which makes it less expensive than Thirst for Knowledge.  This mitigates its "Sorcery Speed" limitation since you can drop it on turn 1 and not mess up your Drain.

It is illustrative to compare Arcane Denial to Thirst for Knowledge and Merchant Scroll.

In the Thirst comparison note that both cards want an artifact to be good.  Most thirst decks run around 13 artifacts.  This list run 17 artifacts, 16 of which can be Denialed. 
First lets compare the early game.  Thirst is rarely a turn 1 play.  Meanwhile Denial is.  It can either counter a bomb or gain card advantage with any of your 16 0cc cards.
By turn 2 Thirst is arguable the better card.  There is a reason this list runs a full set of Thirsts, but by turn 4 the tables start to turn in Denial's favor again.
Denial can be played in response to any of your own spells.  For example if I cast Thirst EOT and it is Drained, then I can Denial my own Thirst in response and draw 3 cards with Denial.  For that reason Denial is actually more powerful than Thirst once I have 5 mana up.
The main argument against Denial comparing it to Thirst is that it is not useful without an artifact to pitch.  Thirst is also worse in such a situation and Denial can still counter a bomb so it is not like Denial sits dead in your hand.

The second card is the more accurate comparison.  Although Denial seems to function like Thirst because it interacts with Baubles, it really is more of a replacement for Merchant Scroll.  I covered this earlier but Scrolls most important uses used to be drawing 3 cards with Ancestral Recall and grabbing Force on turn 1 against Combo.  Scroll could also find bounce as a solution or Mystical for Yawg in the late game so its flexibility went beyond the first two examples, but those were two of the most common uses.

Arcane Denial can be used for either use.  Denial can be used in conjunction with a Bauble or Mox to draw 3 cards similarly to Scroll.  It is true that the card advantage is +1 instead of +2, but you dig just as deep and for 1 less mana and a turn earlier.  This makes it more likely that you find counter magic.  Generally an extra card of card advantage is better than an extra mana thus Scroll for Recall is better than Denial, but Denialing a Bauble is still pretty good even if not as good as Scroll->Recall.

The second comparison is the Scroll for Force comparison.  Denial can be used to counter a spell with -1 card advantage for  just like Scroll could by grabbing Force.  In this manner Denial is actually an improvement since you can delay your decision.  For instance you can keep Denial up on turn 1, but if you opponent mearly Brainstorms or Ponders then you have Drain mana up by turn 2 and you can potentially use Denial to draw 3 cards on turn 3 with Drain mana up.  If you Scroll for Force on turn 1, then you are stuck with Force.  You cannot change your mind once you have Drain up and decide you want Recall instead.  By delaying the choice Denial offers more flexibility and in at least this respect is an improvement over Scroll.

The final argument against Arcane Denial is that it forces you to run "bad" cards like the Baubles.  How bad are the Baubles?  As I mentioned before drawing a card is slightly better than producing a mana.  Baubles draw a card.  Lotus Petal produces a mana.  Lotus Petal is restricted.  Petal can produce any color mana, but then Baubles give you information regarding your opponent's hand or library.  Each is broken post Yawg.  Each pitches to Thirst.  Each can be Tinkered out.  What I am claiming here is that Baubles are not much worse than Petal.  In fact I clearly find them superior to Petal in this list since I run 8 Baubles but no Petal.  Petal gets Drain up on turn 1.  Baubles help draw into Force on turn 1.  Baubles end up having synergy with the deck in the following ways:

Force/Drain: Baubles occationally draw Force on turn 1 and give you crucial information regarding your opponent's hand making decisions about countering spells easier.
Denial:  Draw 3 cards.
Thirst: Draw 3 cards.
Skeletal Scrying:  Draw 5 cards on turn 3 thanks to Threshold.
Yawg:  Allows insane Yawgs.
Mystical/Vamp:  Allows for Mysticall/VAmp for Force on turn 1 and then draw into Force before your opponent's turn.
Tinker:  Can be Tinkered out.
Empty the Warrens:  Easy storm enabler.
Lands:  By cycling having your 8 "worst" cards cycle for free you can get away with running fewer lands.
Tolarian Academy:  Generate an insane amount of mana.

The result is that Baubles end up having synergy with nearly the entire deck.  With so many uses for Bauble, it must be totally dead in some situation in order to be a bad card.  Instead Bauble cycles for 0 which is not bad at all.

Arcane Denial's power is comparable with if not as good as Thirst for Knowledge and Merchant Scroll and Mishra's Baubles are not much worse than Lotus Petal.

That is an excellent description of Arcane Denial it looks like an amazing play countering your own spell like TfK in response to their counter to draw 3 cards. Even though it is a highly situational card it seems like these situations where it would be on par with Merchant->Ancestral arise very often. You can even use it in a desperate situation to try and force through Tinker/Will. I'd love to see this deck run in a tournament and I would read the report with a thirst for knowledge on this deck. (sorry about the pun )

[Kiriyuu]

Hi Meadbert!

Could you share your Dark Drain list, I've tried to construct how I felt it should go, but can't fit in everything from your description, so I'd like to see the differences

Thanks.

[Implacable]

Well, as a note on early testing: it shellacks Drains and gets shellacked by heavy mana denial.  More to follow.

[Bongo]

Amazing work, thank you very much! Such an effort is greatly appreciated.


You said the following about Duressless Long:

There was actually not much in the way of weaknesses for this deck and although it finished second, the statistical analysis indicates that it was the best deck.

Could you explain this in more detail?

[meadbert]

You said the following about Duressless Long:

There was actually not much in the way of weaknesses for this deck and although it finished second, the statistical analysis indicates that it was the best deck.

Could you explain this in more detail?

Basically once a deck was eliminated all games against it were completely ignored.  This meant that once the final two decks were Arcane Denial and Duressless Grim Long only that matchup mattered.  Arcane Denial won that matchup 14-9 in rounds thus it was the "champion," but Duressless Grim Long actually had the better record against other decks including other top 8 decks so it might be the better deck overall.

For the statistical analysis I basically plugged in all of the results from all 128 decks into a rating system similar to the RPI or Pomeroy or Massey ratings used for College Football and Basketball and used that.  According to the ratings Duressless Grim Long was actually the best deck and Wiley Dredge was actually number 5 although it missed out on the top 8.

[A_Outcast]

I have a question about your Arcane Denial.dec matchups.

When you had Arcane Denialm in hand, did "knowing" your opponent had a bomb in their hand (since you were playing both sides) influence you to save the denial to counter their bomb? Or was it a based on some other factors as to whether or not to save your Arcane Denial?

The reason I ask is because if you "know" to save it for their bomb it really does alter how the matches play out because they could drop a bomb on you right after you Arcane Denial yourself for the 3 draw instead of you saving it for them. 

[feyd]

He did give some examples of when he would keep it.   Hell...even if he did use it as a cheaper thirst for knowledge the guy is running  such a biblical amount of counter magic that there's a good shot he'd draw into it during your upkeep.  There's also a second scenario which could happen: he could just arcane denial his own bauble or whatever and then be caught with his pants down during your turn (e.g no hard counters or pitch counters in hand).  Obviously he knows the deck better than anyone else because he invented it but the deck still did not have a pristine record against everything; it had quite a few loses to the second best deck.  In either case the deck may have just gotten very lucky...that is to say it's worst matchups may have been avoided during the tournament.  Like Meadbert said in his post, one deck which he thought could have done fantastic in the tournament was eliminated very early on due to a few decks having exactly the sideboard strategy necessary to beat it...and the luck. 

[pierce]

OK, Pierce, I'll take a stab at explaining the value of what Meadbert has done here. I use two-fisted testing to prepare for tournaments when other options are not available. It is often far more beneficial than attempting to gain insight by testing against the denizens of MWS. On the one hand, there are no doubt problems with the methodology. It is going to prevent more tricky plays from working properly, and you can't pretend that you don't know what both decks have in hand. Moreover, it is going to favor the deck the player knows better. On the other hand, there are problems with standard playtesting too. One player may be more skilled than the other, or one player may be better able to play his deck than the other. Nothing is perfect.

Two fisted testing can give some valuable insights. It often reveals key cards in a match, and highlights some salient features of the how different cards interact. It can help reveal, for example, whether the disruption in one's sideboard is sufficient to halt Ichorid.

In this case, did Meadbert's result prove that Arcane Denial is the best deck in the format? No, of course not. But he did provide evidence that it may be more viable than previously thought. That itself, as well as other other results he provided, are quite worth reading.

Thank you, Meadbert, for sharing your results with us.

my question has nothing to do with personal preference or pawning noobs on mws.

I'm not letting it go as 'nothing is perfect'. That argument is arbitrary at best.

At the point which one person did all the testing, as well as deck tweaking, these results are unusable. His own play style and skill level are fundamental aspects to this approach, and I don't think anyone can argue that. But is is his play style and skill level that will operate each deck. Simply put, I doubt his competancy with all the decks he used. It's no surprise one of his homebrews won due to this.

Combine this with the largely outdated and janky gauntlet (Did I see junky junk as an archetype?wtf?) and you get a pile of worthless data. That much should be obvious by the very unreal winner, which I speculate will have very little, if any impact on the current metagame.

[meadbert]

I have a question about your Arcane Denial.dec matchups.

When you had Arcane Denialm in hand, did "knowing" your opponent had a bomb in their hand (since you were playing both sides) influence you to save the denial to counter their bomb? Or was it a based on some other factors as to whether or not to save your Arcane Denial?

The reason I ask is because if you "know" to save it for their bomb it really does alter how the matches play out because they could drop a bomb on you right after you Arcane Denial yourself for the 3 draw instead of you saving it for them. 

The heuristic that I used very consistently was that I played Arcane Denial on turn 1 if I either had Force in my hand or if I was missing my turn 2 land drop.

[Implacable]

Preliminarily: the deck has very strong matchups against Drains, it seems.  The Denial + Bauble engine, as a secondary draw mechanic, is very effective against them, and your Yawgmoth Wills can be cast very early for a very powerful effect, as the Baubles often add a free quasi-Ancestral to the Will.  Against Shops, the deck suffers greatly: Wasteland is bad enough, but cards like Chalice@0 and Sphere are extremely difficult to play against.  The deck is reliant almost as much as a Storm deck on casting tons of spells (or even more so, if not to win then at least to get going), and therefore, while Wire and Stack and Crucible are mediocre, Spheres, Thorns, and Chalices are lethal.  Against Fish, the deck is fairly effective: Denials are very useful for countering the only thing you care a lot about (Null Rod).  But again, there's still a vulnerability to mana denial.

With that in mind, here are the changes I've made:

-2 Skeletal Scrying
-1 Mana Drain

+1 Time Vault
+1 Voltaic Key
+1 Misdirection

-2 Cephalid Colisseum
-1 Island

+2 Polluted Delta
+1 Volcanic Island

These are relatively minor changes, obviously.  In my opinion, while Scrying is sure great, if you're playing five-mana spells you're winning anyway, because mana's this deck's chokepoint, and Drains shouldn't be out of proportion to MisD, given how much of the deck's game takes place before having two blue active.  In the mana base, the Volc and the two Deltas both give me flexibility for the new sideboard and, in the Delta's case, give me a little more flexibility against Wastes.

New Sideboard:

4 Annul
2 Ingot Chewer
2 Leyline of the Void
2 Pithing Needle
2 Pyroclasm
2 Yixlid Jailer
1 Rack and Ruin

I don't like the old sideboard at all: to my mind, it neither addresses the deck's weaknesses at all nor addresses what it does target very efficiently.  It would be like if somebody built a Shaymora board packed with Boils, so that they could beat decks with Islands, which are of course their worst matchup.  This board has, in testing, done a pretty good job of beating up Stax, dealing with Fish, and addressing the Dredge issue, the less of which is said the better (really, who cares beyond statistics?  I'm not going to delve into the fine art of hoping you draw enough hate).

Thoughts?

[meadbert]

Preliminarily: the deck has very strong matchups against Drains, it seems.  The Denial + Bauble engine, as a secondary draw mechanic, is very effective against them, and your Yawgmoth Wills can be cast very early for a very powerful effect, as the Baubles often add a free quasi-Ancestral to the Will.  Against Shops, the deck suffers greatly: Wasteland is bad enough, but cards like Chalice@0 and Sphere are extremely difficult to play against.  The deck is reliant almost as much as a Storm deck on casting tons of spells (or even more so, if not to win then at least to get going), and therefore, while Wire and Stack and Crucible are mediocre, Spheres, Thorns, and Chalices are lethal.  Against Fish, the deck is fairly effective: Denials are very useful for countering the only thing you care a lot about (Null Rod).  But again, there's still a vulnerability to mana denial.

I agree with most of what you said there.  Resistor is a huge problem.  Chalice and Null Rod are not friendly, but they are not THAT bad either since you have so many ways to get rid of artifacts in hand.  These included:
4 Thirst for Knowledge
4 Arcane Denial
1 Brainstorm
2 Cephalid Colliseum
1 Tinker (Null Rod only)

Definitely the Stax game preboard is tough.  Uba Stax is even tougher since Uba Mask is a beating.

With that in mind, here are the changes I've made:

-2 Skeletal Scrying
-1 Mana Drain

+1 Time Vault
+1 Voltaic Key
+1 Misdirection

Scrying is amazing in many matchups, but against Shops it can be relatively weak since you may not have  available, or you cannot afford a significant  and furthermore it does not pitch to Force.  In the past a main reason I used Scrying was to doge Welder on DSC by RFGing my Tinkered out Mox.  That can now be addressed by a new Tinker target thus it is less important.

Dropping Drain seems wrong to me.  It is still a mana source which is important against Shops.  Drain is what allows drawing 6 cards off Thirst + Denial the following turn.

I actually I have considred adding Painter/Grinstone more seriously than Vault Key.  Painter at least allows for pitching Baubles to Force and MisD so it helps against Null Rod.
I do not see how Vault/Key helps against Null Rod.

-2 Cephalid Colisseum
-1 Island

+2 Polluted Delta
+1 Volcanic Island

Dropping Colliseums for fetches could definitely be an improvement against Shops.  Colliseums can be really broken in the late game when they load the yard for Yawg, but in the early game they are terrible.
I would not drop an island.  In fact I am considering adding an Island to the board.  The only need for red is Empty the Warrens in the maindeck.

These are relatively minor changes, obviously.  In my opinion, while Scrying is sure great, if you're playing five-mana spells you're winning anyway, because mana's this deck's chokepoint, and Drains shouldn't be out of proportion to MisD, given how much of the deck's game takes place before having two blue active.  In the mana base, the Volc and the two Deltas both give me flexibility for the new sideboard and, in the Delta's case, give me a little more flexibility against Wastes.

New Sideboard:

4 Annul
2 Ingot Chewer
2 Leyline of the Void
2 Pithing Needle
2 Pyroclasm
2 Yixlid Jailer
1 Rack and Ruin

I don't like the old sideboard at all: to my mind, it neither addresses the deck's weaknesses at all nor addresses what it does target very efficiently.  It would be like if somebody built a Shaymora board packed with Boils, so that they could beat decks with Islands, which are of course their worst matchup.  This board has, in testing, done a pretty good job of beating up Stax, dealing with Fish, and addressing the Dredge issue, the less of which is said the better (really, who cares beyond statistics?  I'm not going to delve into the fine art of hoping you draw enough hate).

I actually considered Denial's sideboard its strength.
The main weakness is Sphere of Resistance.  This is addressed by 3 bounce and 4 Engineered Explosives.  The Explosives in particular are very powerful.

4 Tormod's Crypt and 4 Explosives have synergy with the Thirsts and Baubles.
I could definitely see adding an Island or two to the board or adding a Volcanic Island for anti Stax Red cards.

I am not a big fan of adding Annul to combat Resistor and Thorn since it only works on the play and it competes with Denial.

[Implacable]

I agree with most of what you said there.  Resistor is a huge problem.  Chalice and Null Rod are not friendly, but they are not THAT bad either since you have so many ways to get rid of artifacts in hand.  These included:
4 Thirst for Knowledge
4 Arcane Denial
1 Brainstorm
2 Cephalid Colliseum
1 Tinker (Null Rod only)

Definitely the Stax game preboard is tough.  Uba Stax is even tougher since Uba Mask is a beating.

I disagree very strongly with your assessment of Null Rod's impact on the deck.  You run, what, 12 non-artifact mana sources?  It's almost impossible to cast Thirst with Rod on the table - if they have even one more piece of mana denial, you're usually locked out of the game. 

As a side note, I've not yet run up against Uba Stax.  There'd be plusses and minuses: while Mask is a beating, it means that they have it instead of mana denial, and my deck runs enough artifact removal post-board to often nail it quickly.

Scrying is amazing in many matchups, but against Shops it can be relatively weak since you may not have  available, or you cannot afford a significant  and furthermore it does not pitch to Force.  In the past a main reason I used Scrying was to doge Welder on DSC by RFGing my Tinkered out Mox.  That can now be addressed by a new Tinker target thus it is less important.

Dropping Drain seems wrong to me.  It is still a mana source which is important against Shops.  Drain is what allows drawing 6 cards off Thirst + Denial the following turn.

I actually I have considred adding Painter/Grinstone more seriously than Vault Key.  Painter at least allows for pitching Baubles to Force and MisD so it helps against Null Rod.
I do not see how Vault/Key helps against Null Rod.

First, Vault/Key doesn't help against Rod: it helps against the field at large, because it's simply the best kill condition out there.  The random victories you get while running it with the black tutor suite - and this deck's extraordinary draw engine - are easily worth it. 

Second, it seems like you're evaluating Drain on specious grounds.  Sure, if you get the mana open to Drain, then it'll be great against Shops!  But the point of Shop decks is that you don't get that mana available, and if you do, you're likely to be ahead.

The same is true of Scrying: you're analyzing it on the grounds that you're already in a position to win (I won't bother with the Tinker argument, as you set it aside).  It's cool: but so's Braingeyser.  Basically, I cut the Scryings for more win conditions, because in the context of how you're evaluating them, they're really the same thing.

Dropping Colliseums for fetches could definitely be an improvement against Shops.  Colliseums can be really broken in the late game when they load the yard for Yawg, but in the early game they are terrible.
I would not drop an island.  In fact I am considering adding an Island to the board.  The only need for red is Empty the Warrens in the maindeck.

Coliseums are also great if you're going to win.  Resolving a Yawgmoth's Will in this deck, more than in any other, is almost always game: and in fact, casting Will as soon as you've got the mana to make it resolve is often the right play, given the incremental advantage you get via Baubles.  If you're abusing Coliseum, you've won.

I actually considered Denial's sideboard its strength.
The main weakness is Sphere of Resistance.  This is addressed by 3 bounce and 4 Engineered Explosives.  The Explosives in particular are very powerful.

4 Tormod's Crypt and 4 Explosives have synergy with the Thirsts and Baubles.
I could definitely see adding an Island or two to the board or adding a Volcanic Island for anti Stax Red cards.

I am not a big fan of adding Annul to combat Resistor and Thorn since it only works on the play and it competes with Denial.

Essentially, my post-board plan is to completely transform the deck against Stax: I was actually MWSing against Soly and, post, he commented on how much the deck seemed to change.  I cut back on Denials, cut the MisDs, and cut some Drains on the draw or random things on the play to bring in heavy, varied artifact hate that dodges Rod, Sphere, and Chalice in varying degrees (by being cheap, non-artifact, and of varied CMC respectively).  Because I believe Rod, not Sphere, is the main threat against this deck, I've focused on that; but even if you believe otherwise, Annul is a lot more effective than Pithing Needle or Explosives or what have you.  The idea is that it replaces the more expensive countermagic, and then artifact destruction mops up what the new counter-suite of 4x Annul, 4x Force, and 2x Denial doesn't hit.  What it comes down to is this: pre-board, Denial doesn't really beat Stax so much.  It's 40/60ish against a good Shop player with heavy mana denial.  So it's not enough to shore up holes; you have to make radical changes.

[Eastman]

Really awesome work, this is one of the most useful things I've read in awhile.  In particular, I'm not at all surprised to see shops fare so poorly.  I think shop decks as an archetype are a lot weaker than they are often thought, and this reinforces that. 

[FlyFlySideOfFry]

I think the weakness in Shop decks comes from the fact that they require more luck than any other deck. You need the right set of lock pieces for the deck you're facing in your opening hand or you lose. This is because the high artifact concentration doesn't usually allow for tutors or effective draw spells. Then post-board when you finally know what you're facing and what lock pieces you need there is a large variety of powerful hate available that destroys Shop decks. This is why it is usually only very good players with almost perfectly tuned decks and strategies who top 8 with Shops.