Thanks!I was pleasantly surprised at how well it seemed to go together on my first test run of it period. I think 3 JVP is probably fine for the moment. I originally had 2 selfless spirit in, but I ended up cutting one. What do you think I should cut for 1? I don't think I need as many as 3.
Tough call.
My initial thought is to cut a Spell Queller, because although the card is awesome, opening hands with double Queller seem pretty rough. I imagine there would just be a lot of turns where you're holding mana up, and you end up doing nothing.
The obvious choice is probably that Vapor Snag, but it seems like a waste of JTU's -3 ability to cut it.
skaab ruinator might be the U beater. It doesn't have hexproof like geist of saint traft does, but it can block too while only coming in for 1 less damage and gains recursions plus the added ability of pitching it to JVP as you mentioned. It really Ruined some of my opponents days already. Eh? eh?
Something to note with this list if anyone wants to try it; consider holding lands to pitch to JVP if you have enough mana to do what you need to. LD takes precedent, but this deck has a low curve (no 4 drops) and we don't usually need 7 mana up with a vial.
This is a pretty good tip. I imagine that's part of the trick of playing with filtering effects: knowing when to stop playing land. Holding up lands also probably baits out late game discard spells.
Playing millions of cards every turn... Slowly and systematically obliterating any chance my opponent has of winning... Clicking the multitude of locking mechanisms into place... Not even trying to win myself until turn 10+ once I have nigh absolute control... Watching my opponent desperately trying to navigate the labyrinthine prison that I've constructed... Seeing the light of hope fade and ultimately extinguished in an excruciatingly slow manner... THAT'S fun Magic.
We have 2-3 users that are dramatically making this thread incomprehensible and non-productive for anyone else to possibly join in the discussion. This needs to change.
Every time I see [ktkenshinx] post in here, I get the impression of a stern dad walking in on a bunch of kids trying to do something dumb and just shaking his head in disappointment.
Near Mint: The same as Slightly Played, but we threw some Altoids in the box we stored it in to cover up the scent of dead mice. Slightly Played: The base condition for all MTG cards. This card looks OK, but there’s one minor annoying ding in it that will always irritate and distract you whenever you draw it. Moderately Played: This card looks like it survived the Tet Offensive tucked inside the waistband of GI underwear. It may smell like it, too. Heavily Played: This card looks like the remains of Mohammed Atta’s passport after 9/11. It may be playable if you double-sleeve it to stop the chunks from falling out. The condition formerly known as "Washing Machine Grade" Damaged: This card is the unfortunate victim of a Mirrorweave/March of the Machines/Chaos Confetti/Mindslaver combo.
[M]aking counterfeit cards is the absolute height of dishonesty. Ask yourself this question: Since most people...are totally cool with the use of proxies...what purpose do [high] quality counterfeit cards serve?
I'm glad that my simulation was accurate. The verification is really nice; my math aptitude isn't good enough for me to do it myself, unless I simplify the problem like I did here:
EDIT: **For the purposes of checking my program, I just ran a Monte Carlo simulation calculating the probability of finding at least one copy of a 4-of in your opening hand of 7 cards, with a 60-card deck. The analytical solution is 39.95%. My simulation returned 39.98%. The only other possible problem is that there is a bug in my code, but only for that particular case that tiago_oldman was concerned with.
From that case, I knew that a million trials wasn't enough to get the hundredths digit accurately, so I just shaved it from my ultimate answer of 4.3%.
For what it's worth, I just ran the same simulation with ten million trials, and the result was 4.31338%. In essence, moving up an order of magnitude in trials bought me an extra digit for my answer. There's clearly diminishing marginal returns on CPU time though, so as long as the Monte Carlo simulation returns a "good enough" answer, that's really all I can hope for.
The advantage of a Monte Carlo simulation rather than analytical solutions, is that it allows me to check complex probabilities involving Æther Vial. For example, let's say tiago_oldman wanted a more accurate number, where we actually needed the resources to put Arbiter into play by T3. That answer involves either finding a white mana source by turn 2 (the number of which varies from deck to deck) or an Æther Vial on T1. While it's not impossible to calculate that analytically, each extra dimension of complexity added makes it harder and harder to calculate. At some point, it just becomes easier to simulate.
The problem with monte carlo simulations is nto the simulation, is the usage of a comuter pseudo number generator that repeats itself (the whole chain) in a much shorter run than most people expect.
For example if you are running a system on a linux 2. X kernel its RNG has a cycle of only 65K numbers ( i know sicne I had to write apatch to the kernel to solve that in a system the company I worked on relied). ( I am not sure about modern windows) That means running millions of interations are near worthless since they start to repeat themselves (and can accumulate even more error). That is one reason there is still so much research in making good RNG. Even most simulation software sell plugins for their really good RNG and these are "usually" not what people have at home. Last time I tried to show that to someone at my computer I made a simple code that kept generating Radom numbers between 1 and 1000 until at least one occurrence of EACH number had happened.... it ran for 2 days until the other guy was convinced that 731 was never going to appear even at infinite time with that random seed.
ps. Revised my calculations side by side with his, and got the line where I made a mistake. Evne so the whole point was that I made a GUESStimate and people somehow get offended that a guesstimate misses by 5%... Probably the innerent bias on the deck distribution derived from the fact that you shuffle in patterns causes more deviance than that!
I don't think anyone was offended by your 1-in-10 God-hand GUESStimate. To me the tone seemed more like "Hoho, I'd kill for 1-in-10 odds!" rather than "Haha, look at this fool!".
I may have been slightly offended* by the later post when you pooh-poohed a reasonably good Monte Carlo result while holding up an erroneous analytical result, though. (Glass houses, ya know?)
*I wasn't actually offended. I was truely quite amused by the irony of it all; sorry for the schadenfreude.
This is my new deck, it diverges a lot from your typical D&T in that it is not based around search-blocking but it is an option to sideboard those cards in. It is a very controlling creature based deck.
@Sabinfrost this looks like the same kind of WU deck that got on camera during the last GP. While I like the idea of playing probe to dig for threats I don't think the deck wants to be playing it. Taking hits from the mana base or from spells really hurts when you aren't packing a ton of removal like other decks. We got to see that in the feature match when the D&T player died to his own probes.
Alright guys, I've spent most of the morning flipping through the pages of this primer trying to figure out what I should build I have the majority of the creature base for mono white, uw, and eldrazi and taxes. Any advice or suggestions towards what seems to be the best death and taxes list in the current format would be greatly appreciated.
So I've been brewing some Wr lists and as i was doing so I realized that the Nahiri Emrakul combo might be something worth exploring in Death and taxes. From my experience the deck often has trouble closing out a game or just needing some general utility and with our disruption and tempo/control elements Nahiri could be very reasonable in our deck. Now I could be completely and totally crazy as it definitely is a little bit of a nonbo with vial/thalia GoT/Arbiter but i'm definitely going to start exploring a Nahiri and Taxes shell and i'd love any help or opinions people on here would be willing to offer. I'm still in the VERY early stages of deck design and it definitely needs some tuning but this is the list I'm currently working from.
The first few things i'm noticing is whether or not Simian Spirit Guide is worthwhile. It allows me to pump out early Magus's, Blood Moons, Nahiri's, Arbiters, or Either Thalia, but it does lower my creature count which already feels a little low.
I'm going to start my testing here but I figured I'd share this with all you lovely people and see if we can come out with a powerful new variant of Death and Taxes for Modern!
Alright guys, I've spent most of the morning flipping through the pages of this primer trying to figure out what I should build I have the majority of the creature base for mono white, uw, and eldrazi and taxes. Any advice or suggestions towards what seems to be the best death and taxes list in the current format would be greatly appreciated.
I would go with this when it comes to the guts of the deck:
These are the cards you're going to find in virtually every D&T list. Deviations from this core are strictly experimental and frankly have not put up the requisite results to be taken seriously in a competitive aspect. In addition to this core, you should consider one of the following packages:
After that, you can add some odds and ends and you should be good to go in the maindeck. Sideboarding is a longer, more difficult conversation, so we'll approach that in another post.
Thankk you for that. You are my etarnal senpai, my mystic, my mangara, my geist.
I laughed out very loud hahaha. He's primer in Merfolk! I don't know how he got accurate in DnT, but he's got a lot of good ideas to share and spare.
@ASteelersMidget I thought about Nahiri'mrakul once before, but forgot it when our WR fastland came up. But isn't Kiki-Resto better? Since you don't have any tax to worry about and in a world where dredge is a threat and jeskai decks are tier 1 decks, you may not want to be screwed by grafdigger's cage imo... Just thoughts tho. Your tests might contradict my suspicions
I have pretty well all of this, and the mono white has been the version I've been working towards I'll type out my list that I own when I get home. Thank you
The problem with monte carlo simulations is nto the simulation, is the usage of a comuter pseudo number generator that repeats itself (the whole chain) in a much shorter run than most people expect.
For example if you are running a system on a linux 2. X kernel its RNG has a cycle of only 65K numbers ( i know sicne I had to write apatch to the kernel to solve that in a system the company I worked on relied). ( I am not sure about modern windows) That means running millions of interations are near worthless since they start to repeat themselves (and can accumulate even more error). That is one reason there is still so much research in making good RNG. Even most simulation software sell plugins for their really good RNG and these are "usually" not what people have at home. Last time I tried to show that to someone at my computer I made a simple code that kept generating Radom numbers between 1 and 1000 until at least one occurrence of EACH number had happened.... it ran for 2 days until the other guy was convinced that 731 was never going to appear even at infinite time with that random seed.
While what you're saying is true to a certain extent, it fails an important common sense test. Namely, that if RNGs were so error-prone that MC simulations didn't work, no one would use them. Instead, MC simulations are used all the time for a wide variety of applications, so the RNGs must not be so bad as to preclude their utility.
I'm not going to claim to be a randomness expert, but if my MC simulation were off by a full percent at a million trials, the simulation would be essentially useless. I won't deny that the output value bounces around a little bit (as it should), but it seems to change about +/- 0.1%. To make it better, I could simply include more trials, but that small of a difference in this particular context doesn't matter.
Evne so the whole point was that I made a GUESStimate and people somehow get offended that a guesstimate misses by 5%... Probably the innerent bias on the deck distribution derived from the fact that you shuffle in patterns causes more deviance than that!
This thread has an obligation to provide relevant strategy information to a large number of people, including the lurkers who don't represent themselves in the thread. Your mistake wasn't just a 5% difference; you were off of the observable value by a factor between 2 and 3, and it was right in the regime where you don't want to make that kind of error. Sure, you didn't take a 20% chance and represent it as a 50% chance, but you also didn't take a 1% chance and represent it as a 2.5% chance. You took a reasonably rare event (fewer than 1 in 20 games) and you represented it as a far more common event (more than 1 in 10 games).
If your intuition had been similar to agua_benta's intuition...
...we wouldn't even be having this conversation. Again, no one is offended that you took a guess. It's just that veteran players know your guess was bad, without even having to do the math. And because part of our job in this thread is to communicate common strategy options with the deck, we really shouldn't let your error slide, because it's going to lead someone else astray.
As for the "inherent bias [in] the deck distribution derived from the fact that you shuffle in patterns", this is the textbook definition of insufficient randomization. To be clear, judges are typically only going to call you on this if you are aiming to get an unfair advantage by not randomizing your deck. But if your shuffling is so bad to cause an observed error of a relatively rare event to jump by 5%, then there is probably a problem.
Alright guys, I've spent most of the morning flipping through the pages of this primer trying to figure out what I should build I have the majority of the creature base for mono white, uw, and eldrazi and taxes. Any advice or suggestions towards what seems to be the best death and taxes list in the current format would be greatly appreciated.
The short answer is "WB Eldrazi and Taxes".
The long answer is that all of the builds are reasonably competitive, and from your reading, you probably know that we don't have a stock list to give you. The primer post is a good place to start with reference decklists, but other than that, you'll have to tune something to your metagame.
The first few things i'm noticing is whether or not Simian Spirit Guide is worthwhile. It allows me to pump out early Magus's, Blood Moons, Nahiri's, Arbiters, or Either Thalia, but it does lower my creature count which already feels a little low.
I know what you mean. I'm currently brewing a WR list, and I'm (slowly) typing up a mini-primer for WR D&T. I may end up running SSG out of the sideboard, or alternatively, I may end up running CotV out of the sideboard. T1 Chalice on 1 is rough against a lot of decks, and may be worth considering as a possible line of play.
I'm not personally a fan of Nahiri (because it's expensive through a Thalia tax, and I tend to like having early lines of play), but do some testing and let us know how it goes. I'm sure it will occasionally win games, but I just envision it being more useful as something else.
For what it's worth, Lt. Glitter also does testing on the WR version, so maybe he has better insights than I do.
Playing millions of cards every turn... Slowly and systematically obliterating any chance my opponent has of winning... Clicking the multitude of locking mechanisms into place... Not even trying to win myself until turn 10+ once I have nigh absolute control... Watching my opponent desperately trying to navigate the labyrinthine prison that I've constructed... Seeing the light of hope fade and ultimately extinguished in an excruciatingly slow manner... THAT'S fun Magic.
We have 2-3 users that are dramatically making this thread incomprehensible and non-productive for anyone else to possibly join in the discussion. This needs to change.
Every time I see [ktkenshinx] post in here, I get the impression of a stern dad walking in on a bunch of kids trying to do something dumb and just shaking his head in disappointment.
Near Mint: The same as Slightly Played, but we threw some Altoids in the box we stored it in to cover up the scent of dead mice. Slightly Played: The base condition for all MTG cards. This card looks OK, but there’s one minor annoying ding in it that will always irritate and distract you whenever you draw it. Moderately Played: This card looks like it survived the Tet Offensive tucked inside the waistband of GI underwear. It may smell like it, too. Heavily Played: This card looks like the remains of Mohammed Atta’s passport after 9/11. It may be playable if you double-sleeve it to stop the chunks from falling out. The condition formerly known as "Washing Machine Grade" Damaged: This card is the unfortunate victim of a Mirrorweave/March of the Machines/Chaos Confetti/Mindslaver combo.
[M]aking counterfeit cards is the absolute height of dishonesty. Ask yourself this question: Since most people...are totally cool with the use of proxies...what purpose do [high] quality counterfeit cards serve?
kambal kills! this card is a clear winner in eldrazi and taxes. always gets value from removal and completely shuts down storm. the life loss and life gain is relevant in almost every match up. i think he is a clear winner over thalia, heretic cathar. she comes down a little late to be effective, where kambal is a bomb turn three or turn 10. here is my current list.
As for the "inherent bias [in] the deck distribution derived from the fact that you shuffle in patterns", this is the textbook definition of insufficient randomization. To be clear, judges are typically only going to call you on this if you are aiming to get an unfair advantage by not randomizing your deck. But if your shuffling is so bad to cause an observed error of a relatively rare event to jump by 5%, then there is probably a problem..
All the ways people use to shuffle are heavily pattern biased. If you pile shuffle.. then well more obvious then that? Rifle shuffle also mostly is a split->spread-> combine classical list operation. It is not fully reversible just because you (and anyone) fail at doing a perfect split the a perfect spread then a perfect merge. But MOST of the cards are spread and merged almost perfectly. That keeps a close approximation of the perfect function that is fully reversible (and by definition is it is reversible it is absolutely not random), whjere the only deviation comes from " shufling error".
All these shuffle methods are WORSE randomness generators than the already bad computer ones. Just because the pattern of redistribution is too complex for a human to follow it does not mean it is really random. It is hard to simulate the real relevance of this aspect because it is hard to measure how perfectly one rifle shuffle is made.. the better it is made the less random it is.
The first few things i'm noticing is whether or not Simian Spirit Guide is worthwhile. It allows me to pump out early Magus's, Blood Moons, Nahiri's, Arbiters, or Either Thalia, but it does lower my creature count which already feels a little low.
I know what you mean. I'm currently brewing a WR list, and I'm (slowly) typing up a mini-primer for WR D&T. I may end up running SSG out of the sideboard, or alternatively, I may end up running CotV out of the sideboard. T1 Chalice on 1 is rough against a lot of decks, and may be worth considering as a possible line of play.
I'm not personally a fan of Nahiri (because it's expensive through a Thalia tax, and I tend to like having early lines of play), but do some testing and let us know how it goes. I'm sure it will occasionally win games, but I just envision it being more useful as something else.
For what it's worth, Lt. Glitter also does testing on the WR version, so maybe he has better insights than I do. [/quote]
I've been playing a lot of Gideon Ally of zendikar in my Wg list I've been playing and while it sometimes comes up, thalia is hoenstly usually perished by the time turn 4 rolls around.
Also with my meta currently full of Control, Bant Eldrazi, and Valakut decks this feels like it'll give me an edge greater than any of the other lists. As for Grafdigger's cage while it is an issue, between sideboard options like wear // tear, Chalice, Engineered Explosives, and Smash to smithereens i'm not too concerned about a cage. Worst comes to worst i nahiri until i find one or i board out emrakul and a few nahiri's. I'll be sure to post updates to this as I get more testing.
@Sabinfrost this looks like the same kind of WU deck that got on camera during the last GP. While I like the idea of playing probe to dig for threats I don't think the deck wants to be playing it. Taking hits from the mana base or from spells really hurts when you aren't packing a ton of removal like other decks. We got to see that in the feature match when the D&T player died to his own probes.
I didn't watch the GP. Probe helps a lot for maximising Meddling Mage. It makes the deck a bit more consistent as well. There is a risk involved but there are also sideboard options to help in matches where life is an absolute premium.
My initial thought is to cut a Spell Queller, because although the card is awesome, opening hands with double Queller seem pretty rough. I imagine there would just be a lot of turns where you're holding mana up, and you end up doing nothing.
The obvious choice is probably that Vapor Snag, but it seems like a waste of JTU's -3 ability to cut it.
I see what you did there.
I know we play creature spells for a reason, but there are some spicy flashback options too: Artful Dodge, Saving Grasp, Feeling of Dread, Momentary Blink, and Think Twice.
This is a pretty good tip. I imagine that's part of the trick of playing with filtering effects: knowing when to stop playing land. Holding up lands also probably baits out late game discard spells.
WUDeath&TaxesWG
Legacy
UBRGDredgeUBRG
UHigh TideU
URGLandsURG
WR Card Choice List
WUR American D&T
WUB Esper D&T
The Reserved List
Heat Maps
The problem with monte carlo simulations is nto the simulation, is the usage of a comuter pseudo number generator that repeats itself (the whole chain) in a much shorter run than most people expect.
For example if you are running a system on a linux 2. X kernel its RNG has a cycle of only 65K numbers ( i know sicne I had to write apatch to the kernel to solve that in a system the company I worked on relied). ( I am not sure about modern windows) That means running millions of interations are near worthless since they start to repeat themselves (and can accumulate even more error). That is one reason there is still so much research in making good RNG. Even most simulation software sell plugins for their really good RNG and these are "usually" not what people have at home. Last time I tried to show that to someone at my computer I made a simple code that kept generating Radom numbers between 1 and 1000 until at least one occurrence of EACH number had happened.... it ran for 2 days until the other guy was convinced that 731 was never going to appear even at infinite time with that random seed.
ps. Revised my calculations side by side with his, and got the line where I made a mistake. Evne so the whole point was that I made a GUESStimate and people somehow get offended that a guesstimate misses by 5%... Probably the innerent bias on the deck distribution derived from the fact that you shuffle in patterns causes more deviance than that!
I may have been slightly offended* by the later post when you pooh-poohed a reasonably good Monte Carlo result while holding up an erroneous analytical result, though. (Glass houses, ya know?)
*I wasn't actually offended. I was truely quite amused by the irony of it all; sorry for the schadenfreude.
4x Path to Exile
4x Gitaxian Probe
4x Judge's Familiar
4x Selfless Spirit
4x Thalia, Guardian of Thraben
4x Meddling Mage
2x Kira, Great Glass-Spinner
2x Vendilion Clique
2x Magus of the Moat
4x Flooded Strand
4x Hallowed Fountain
1x Eiganjo Castle
1x Minamo, School at Water's Edge
5x Plains
1x Island
2x Ghost Quarter
4x Celestial Colonnade
There is no sideboard currently posted as it's still in early development.
WB Ayli, Eternal Pilgrim | WBR Queen Marchesa | WUBR Breya, Etherium Shaper
Modern:
/ Death and Taxes | W/ Soul Sisters | Spirits | U Faeries | Tempered Steel Affinity
2x Blade Splicer
1x Emrakul, the Aeons Torn
4x Flickerwisp
1x Kitchen Finks
4x Leonin Arbiter
4x Magus of the Moon
3x Simian Spirit Guide
3x Thalia, Guardian of Thraben
2x Thalia, Heretic Cathar
4x Inspiring Vantage
2x Mountain
7x Plains
4x Sacred Foundry
1x Slayers' Stronghold
2x Blood Moon
4x AEther Vial
4x Nahiri, the Harbinger
The first few things i'm noticing is whether or not Simian Spirit Guide is worthwhile. It allows me to pump out early Magus's, Blood Moons, Nahiri's, Arbiters, or Either Thalia, but it does lower my creature count which already feels a little low.
I'm going to start my testing here but I figured I'd share this with all you lovely people and see if we can come out with a powerful new variant of Death and Taxes for Modern!
I would go with this when it comes to the guts of the deck:
4 Æther Vial
4 Flickerwisp
4 Ghost Quarter
4 Leonin Arbiter
4 Path to Exile
3-4 Thalia, Guardian of Thraben
These are the cards you're going to find in virtually every D&T list. Deviations from this core are strictly experimental and frankly have not put up the requisite results to be taken seriously in a competitive aspect. In addition to this core, you should consider one of the following packages:
1. Eldrazi package
- 4 Caves of Koilos
- 3-4 Eldrazi Displacer
- 4 Eldrazi Temple
- 4 Godless Shrine
- 0-2 Reality Smasher
- 0-2 Relic of Progenitus
- 2-3 Shambling Vent (note that this card will probably be made obsolete by Concealed Courtyard soon)
- 4 Thought-Knot Seer
- 3-4 Tidehollow Sculler
- 4 Wasteland Strangler
2. Mono-W land hate/value package
- 0-2 Aven Mindcensor
- 3-4 Blade Splicer
- 0-2 Brimaz, King of Oreskos
- 1 Eiganjo Castle
- 2-3 Horizon Canopy
- 0-2 Kitchen Finks
- 0-2 Restoration Angel
- 2-4 Selfless Spirit
- 4 Tectonic Edge
- 0-2 Thalia, Heretic Cathar
- 0-2 Wall of Resurgence (you might want one of Blinkmoth Nexus, Darksteel Citadel, or Mutavault if you choose to run this)
3. WG ramp package:
- 2-4 Blade Splicer
- 3-4 Horizon Canopy
- 0-2 Kitchen Finks
- 4 Noble Hierarch
- 4 Razorverge Thicket
- 2-4 Restoration Angel
- 2-3 Scavenging Ooze
- 2-3 Selfless Spirit
- 2 Stirring Wildwood
- 2-3 Tectonic Edge
- 4 Temple Garden
- 2-3 Thalia, Heretic Cathar
- 2-4 Voice of Resurgence
After that, you can add some odds and ends and you should be good to go in the maindeck. Sideboarding is a longer, more difficult conversation, so we'll approach that in another post.
Legacy: Merfolk U; Shadow UB; Eldrazi Stompy C
Pauper: Delver U
Vintage: Merfolk U
Primers:
@ASteelersMidget I thought about Nahiri'mrakul once before, but forgot it when our WR fastland came up. But isn't Kiki-Resto better? Since you don't have any tax to worry about and in a world where dredge is a threat and jeskai decks are tier 1 decks, you may not want to be screwed by grafdigger's cage imo... Just thoughts tho. Your tests might contradict my suspicions
Edit: double post
2. Mono-W land hate/value package
- 0-2 Aven Mindcensor
- 3-4 Blade Splicer
- 0-2 Brimaz, King of Oreskos
- 1 Eiganjo Castle
- 2-3 Horizon Canopy
- 0-2 Kitchen Finks
- 0-2 Restoration Angel
- 2-4 Selfless Spirit
- 4 Tectonic Edge
- 0-2 Thalia, Heretic Cathar
- 0-2 Wall of Resurgence (you might want one of Blinkmoth Nexus, Darksteel Citadel, or Mutavault if you choose to run this)
I have pretty well all of this, and the mono white has been the version I've been working towards I'll type out my list that I own when I get home. Thank you
I'm not going to claim to be a randomness expert, but if my MC simulation were off by a full percent at a million trials, the simulation would be essentially useless. I won't deny that the output value bounces around a little bit (as it should), but it seems to change about +/- 0.1%. To make it better, I could simply include more trials, but that small of a difference in this particular context doesn't matter.
This thread has an obligation to provide relevant strategy information to a large number of people, including the lurkers who don't represent themselves in the thread. Your mistake wasn't just a 5% difference; you were off of the observable value by a factor between 2 and 3, and it was right in the regime where you don't want to make that kind of error. Sure, you didn't take a 20% chance and represent it as a 50% chance, but you also didn't take a 1% chance and represent it as a 2.5% chance. You took a reasonably rare event (fewer than 1 in 20 games) and you represented it as a far more common event (more than 1 in 10 games).
If your intuition had been similar to agua_benta's intuition...
...we wouldn't even be having this conversation. Again, no one is offended that you took a guess. It's just that veteran players know your guess was bad, without even having to do the math. And because part of our job in this thread is to communicate common strategy options with the deck, we really shouldn't let your error slide, because it's going to lead someone else astray.
As for the "inherent bias [in] the deck distribution derived from the fact that you shuffle in patterns", this is the textbook definition of insufficient randomization. To be clear, judges are typically only going to call you on this if you are aiming to get an unfair advantage by not randomizing your deck. But if your shuffling is so bad to cause an observed error of a relatively rare event to jump by 5%, then there is probably a problem.
The short answer is "WB Eldrazi and Taxes".
The long answer is that all of the builds are reasonably competitive, and from your reading, you probably know that we don't have a stock list to give you. The primer post is a good place to start with reference decklists, but other than that, you'll have to tune something to your metagame.
I know what you mean. I'm currently brewing a WR list, and I'm (slowly) typing up a mini-primer for WR D&T. I may end up running SSG out of the sideboard, or alternatively, I may end up running CotV out of the sideboard. T1 Chalice on 1 is rough against a lot of decks, and may be worth considering as a possible line of play.
I'm not personally a fan of Nahiri (because it's expensive through a Thalia tax, and I tend to like having early lines of play), but do some testing and let us know how it goes. I'm sure it will occasionally win games, but I just envision it being more useful as something else.
For what it's worth, Lt. Glitter also does testing on the WR version, so maybe he has better insights than I do.
WUDeath&TaxesWG
Legacy
UBRGDredgeUBRG
UHigh TideU
URGLandsURG
WR Card Choice List
WUR American D&T
WUB Esper D&T
The Reserved List
Heat Maps
artifacts: 4
4 aether vial
creatures: 30
2 dryad militant
3 eldrazi displacer
4 flickerwisp
2 kambal, consul of allocation
4 leonin arbiter
3 thalia, guardian of thraben
4 thought-knot seer
4 tidehollow sculler
4 wasteland strangler
instants: 4
4 path to exile
lands: 22
4 caves of koilos
4 concealed courtyard
4 eldrazi temple
4 ghost quarter
2 shambling vent
2 snow-covered plains
2 snow-covered swamp
creatures: 4
1 kataki, war's wage
2 kor firewalker
1 reality smasher
2 rest in peace
2 stony silence
instants: 3
1 anguished unmaking
2 zealous persecution
sorceries: 4
2 sunlance
1 thoughtseize
All the ways people use to shuffle are heavily pattern biased. If you pile shuffle.. then well more obvious then that? Rifle shuffle also mostly is a split->spread-> combine classical list operation. It is not fully reversible just because you (and anyone) fail at doing a perfect split the a perfect spread then a perfect merge. But MOST of the cards are spread and merged almost perfectly. That keeps a close approximation of the perfect function that is fully reversible (and by definition is it is reversible it is absolutely not random), whjere the only deviation comes from " shufling error".
All these shuffle methods are WORSE randomness generators than the already bad computer ones. Just because the pattern of redistribution is too complex for a human to follow it does not mean it is really random. It is hard to simulate the real relevance of this aspect because it is hard to measure how perfectly one rifle shuffle is made.. the better it is made the less random it is.
Basically a worse kor skyfisher, and that card isn't used very often (even though i like it). No flash makes it very bad.
I'm not personally a fan of Nahiri (because it's expensive through a Thalia tax, and I tend to like having early lines of play), but do some testing and let us know how it goes. I'm sure it will occasionally win games, but I just envision it being more useful as something else.
For what it's worth, Lt. Glitter also does testing on the WR version, so maybe he has better insights than I do. [/quote]
I've been playing a lot of Gideon Ally of zendikar in my Wg list I've been playing and while it sometimes comes up, thalia is hoenstly usually perished by the time turn 4 rolls around.
Also with my meta currently full of Control, Bant Eldrazi, and Valakut decks this feels like it'll give me an edge greater than any of the other lists. As for Grafdigger's cage while it is an issue, between sideboard options like wear // tear, Chalice, Engineered Explosives, and Smash to smithereens i'm not too concerned about a cage. Worst comes to worst i nahiri until i find one or i board out emrakul and a few nahiri's. I'll be sure to post updates to this as I get more testing.
I didn't watch the GP. Probe helps a lot for maximising Meddling Mage. It makes the deck a bit more consistent as well. There is a risk involved but there are also sideboard options to help in matches where life is an absolute premium.