As Smmenen says above, there's no such thing as a matchup percentage because skill matters so much. But even taking your skill and the skill of your opponent as a constant, unless a matchup is very lopsided 10 games isn't enough to tell who is favored based on the results alone. For example, if you win 6 of 10 games, the 95% confidence interval for your win percentage is 31% to 83% (for stats nerds, this is using the Wilson method: http://vassarstats.net/prop1.html).  If you win 8 of 10 games the 95% confidence interval is (49% to 94%). This ignores sideboarded and unsideboarded games being different, which increases the number of games needed. To get a sense of which deck is favored from a small number of games, you need a lot of intuition or knowledge separate from the actual results.

[on Tezz 2]

Exactly. Of the cards listed, only using Ponder to bottom cards (or Fatesealing yourself with Jace) makes the *stronger* non-artifact cards more likely to be at the top of your library. The others all make the *weaker* cards more likely to be at the top of your library and provide more reason to use the +1.

Two ideas for Auras that enchant players.

Mind Lock

Enchantment- Aura
Enchant Player
Enchanted player can’t search his or her library.

Curse of Stupefaction

Enchantment- Aura Curse
Enchant Player
Enchanted player can’t play spells with a converted mana cost 1 or less.

This is not a drawback of Tezz 2's +1 ability. In general, you're just as likely to move your weaker (non-artifact) cards to the bottom of your library as your stronger cards. Your stronger cards aren't any more likely to be in the top 5 cards than at any other place in your library, so this ability has no effect on the "strength" of your library's top 5 cards, for example (assuming you had 10 or more cards left to start with!).

Adding "Sacrifice ~: add to your mana pool" would make it a bit more interesting.


I highly doubt they'd print anything this strong.

Evil Wild Growth

Enchantment- Aura
Enchant land
Whenever enchanted land is tapped for mana, its controller adds to his or her mana pool (in addition to the mana the land produces) and loses 1 life.

It has some limited offensive use compared to Wild Growth. Just tossing this out there as an Aura example that isn't the usual black reanimation/discard/simple creature buff or debuff.

What you've calculated is a measure importance of winning the coin flip to winning the match (based on your assumptions, the calculation is correct. I go through more details here: http://www.themanadrain.com/index.php?topic=25791.0).

Setting aside issues of mixing sideboarded and unsideboarded games, these game results would seem to indicate pretty clearly that you win a higher percentage of games on the play than on the draw. But if you test statistically whether the win percentages in the games on the play are different than the games on the draw, you can't say these percentages of games won on the play versus on the draw are different with a high degree of confidence (the p-value is about 0.18). That is, even if the win percentage on the play equaled the win percentage on the draw, when you played this many games you'd see an average win percentage at least as different as what you observed here 18% of the time.

It would be extremely hard to "play the same game twice" in any reasonable way and there's also no need to do so to get at this question. You can simply alternate which deck goes first, play each to the best of your ability, and see the fraction of games won by the deck going first (unsideboarded for simplicity). The trouble is that unless the effect (amount by which going first is better than going second, or vice versa) is quite large, say 20 percentage points, you won't reasonably be able to test enough games to learn anything with a high degree of statistical confidence. And even if you do, that's only one matchup.

Even completely obvious conclusions, such as that replacing Vampiric Tutor with Imperial Seal would make a deck worse, can't reliably be confirmed without lots and lots of statistical evidence that you are unlikely to have. To quote a discussion along these lines.

This article didn't quite articulate why knowing how to win a post-sideboard game on the draw can be so important in some matchups. If you are a serious underdog in game 1, you are very likely to lose game 1 and need to win a post-sideboarded game on the draw to win the match. By comparison, if you are highly favored in game 1, you are very likely to win game 1 and only to need to win a post-sideboarded game on the play to win the match.

For game 1, improving your chance to win by 5 percentage points on the draw and 5 percentage points on the play has the same effect on your chance to win the match as improving your chance to win game 1 by 10 percentage points on the draw and 0 percentage points on the play, or vice versa.

The same is not true in general of post-sideboarded games. Improving your chance to win an on-the-draw post-sideboarded game by 10 percentage points could be worth more, less, or the same in terms of your chance to win the match as improving your chance to win an on-the-play post-sideboarded game by 10 percentage points, depending on the particulars of the matchup.

This chain of reasoning is subtle, and historically I haven't been that good at explaining it without resorting to hard to follow mathematics.

If your G1 percentage is good, you win more matches when you have skewed postboard odds (of winning based on being on the play/draw) than with even ones, even given the same average of (your chance to win on the play post-sideboard and your chance to win on the draw post-sideboard).

I just want to comment that testing 8-10 (or even 20) games shows very little statistically unless the results are extremely mismatched. And if you claim that one deck is extremely favored over another one, people naturally will doubt whether your testing was an accurate guide to the matchup. For much more detail, see http://www.themanadrain.com/index.php?topic=20129.0.

What do you do if you draw Omniscience?

Clearly Dragon Breath is on average much worse to draw than the cards that would replace it in your deck. But my point was just that it has some use if stuck in your hand.

Not completely useless, you can cast it on Blightsteel Colossus (likely after Oathing it).

Manipulate Fate is superior to Foresight in that it gives you a card right away. Sorcery, 1U: Search your library for three cards, exile them, then shuffle your library. Draw a card.

Thalia, Guardian of Thraben
1W
Legendary Creature - Human Soldier   
First Strike
Noncreature spells cost 1 more to cast.
2/1

Card image: http://forums.mtgsalvation.com/attachment.php?attachmentid=127171&stc=1&d=1326344441

Other than being Legendary, this is a strict upgrade over Glowrider.  Would seemingly have a place in G/W hatebears (http://www.starcitygames.com/magic/vintage/23257_The_Guide_To_Vintages_Landscape_Attacking_The_Red_Zone.html for a list).  

Ponder is restricted.  I assume you meant something else instead...

Very nice article.  Re: proxies, high-quality proxies are clearly desired but are quite constrained by WotC's legal actions.  ELD even had to stop giving away his proxies as door prizes, for example.  I'm sure others know more details.  

Good instructions on creating high quality proxies may help.  However, even if you make instructions were easily accessible, most people wouldn't find it worth the effort if they're only going to make 10-15.  The "printing out slips of paper" solution works great among friends but not for tournament play.

How would you sideboard against another Oath deck (and how it would differ by Oath variants, if applicable)?

Quick math related to the bolded sentence: if you start with Time Walk in graveyard and Will, Blessing, and 2 Demons in your deck (in random order- doesn't matter what other cards are there as long as none are creatures), there's a 1/6 chance that you mill Will but not Blessing.  Unless Will is the first card (of these four), milling Will but not Blessing can't occur.  If Will is the first card, then this occurs if one of the Demons is second, which conditional on Will being first happens 2/3 of the time.  So 2/3 * 1/4= 1/6.  Sound correct?

I just stated some conditions where milling for bombs does matter.  Under what conditions does milling bombs not matter?  Assuming the opponent's deck has no tutors (which would increase the value of milling for bombs), if you can't mill enough cards over the course of the game that the opponent is at risk of running out of cards, milling won't have any value.  That is, if the game is certain to end before the opponent would run out of cards in library (I'm abstracting away from the fact that your opponent's strategy might change when you mill them), all the milling you do has no value. 

The reasoning: without loss of generality, milled cards in this example could come off the bottom of the deck (since each ordering is equally likely).  When the conditions in the previous paragraph are met, milling is just removing cards that would never be accessed anyway.

The value of milling an opponent (or the harm of milling yourself) can be seen in the following hypothetical situation.

The board is at a stalemate.  Either player who attempts to break it will lose.  However, your opponent has a bomb that will win the game the turn she draws it.  You have more cards in your library; thus, you have an inevitable victory without needing to mill the opponent, were the bomb not present in her deck. 

If you successfully mill the bomb, you will win.  Otherwise, you will lose.  Flip this scenario and you can see when milling yourself could be harmful. 

That isn't to say this situation, or something resembling it, comes up in actual Innistrad drafting.  But the strategy could hold water. 

John Jones came up with quite the innovation, playing 2 Black Lotuses and 2 Mox Emeralds.

I'm the farthest thing from a Dredge expert, but Black Lotus, Ancestral Recall, and Bazaar of Baghdad in opening seven probably has a good chance to win on turn 1 in a Fatestitcher list like the 1st place one here: http://www.themanadrain.com/index.php?topic=43240.msg594113

Now, it's extremely unlikely to get such a starting hand, but play a few hundred games and you're likely to see it (Lotus Petal replacing Black Lotus would also have a chance).

If there are n eligible cards in the graveyard, get the smallest die available with >=n faces.  On a roll of i from 1 to n, pick the ith eligible card from the top.  Otherwise, reroll until you roll 1-n.  Anyone who plays one of these cards can pick up a 20 sided die which should generally be sufficient.

The winning list is missing Mox Sapphire (and if I counted right, is not coincidentally 59 cards).

Having not read this article, I'd say that if adding an extra Ponder/Preordain to a 59 card deck while turning into a 60 card deck improves it, the correct thing to do is to add that Ponder/Preordain in the place of a different card, so your deck remains 59 cards.  Similarly, if your 60 card deck can be improved by making it 61 with an extra Ponder/Preordain, it can be improved more by removing another card at the same time.

Tutors and the ability to fit in specialized bullets is about the only reason I can think of why this might not hold (no deck will have enough cantrips that exact mana ratios are that important), but that seems implausible. 

Fissure is another tournament legal instant land removal spell.

And Lavaball Trap, Rith's Charm (nonbasic only), and Wrecking Ball.

Interesting, thanks!

This list has 0 Ancient Tombs but has 60 cards anyway.  I'd guess there was a mistake somewhere.

I found the lack of Trinisphere in the 3rd and 5th place MUD lists interesting.

Either pilot (Nam Tran and Jason Pare, respectively) want to comment on their reasoning in leaving it out?