MTG Salvation

Quote from herkamurjones »

Quote from timewalkinonsunshine »

I posted a GP Toronto tournament report on reddit for your reading pleasure.

Lucksacked the Gemstone Caverns FOUR TIMES in 13 rounds played (32 games, not sure how many were on the draw). That is a hell of a lot of luck dude! Good finish, solid report. Lets keep taking those turns

Since I was a little bit bored, I did a hypergeometric distribution calculation on the chance to see Gemstone Caverns, when you have 2 on your deck. The calculation does not consider mulligans.

So, the chance to see exactly one Caverns in your starting hand is a little bit shy of 21%. The chance to see one or two is a little bit over 22%.
So if we plug this data in a binomial distribution and use a Student's t distribution to calculate the confidence interval, in average after 32 games you would have seen (7 +- 1) Gemstone Caverns for 95% accuracy. Looks like our friend is either VERY unlucky to only have seen that many Gemstone Caverns or he is not shuffling his deck enough times to reach a sufficiently random state.

OTOH: 2 Gemstone Caverns really seems the sweet spot. Running just 1 makes the probability of seeing one on your starting hand go down to 11.6% and going to three, soars it up to 28%, with the caveat of having a risk of seeing doubles on the starting hand to 31% and multiples across the game to be very likely to happens (something we do not want).

This post has been sponsored by my procrastination to finish a presentation. I hope you liked it.

Edit3: The calculation assumes a normal distribution approximation. To correctly calculate the confidence interval of a binomial distribution, it is a more complicated process. I also was pretty lax on the confidence intervals, as I just did a z=2 for 95%. That is a rough approximation.
Edit4: The chance to only see 4 Gemstone Caverns in 32 games is funnily, one in 33. This means that what happened on the GP for Mr. Wong happens only about 3% of the time. Talk about unlucky!

That was a cringey monologue, but MTGSalvation got a shoutout!

Storm lists also use one Snow-Covered islands for the fringe Gifts Piles where you want lands.

Unexpected things happening on PT Rivals of Ixalan: Lantern with the turn 4 kill.

This is generally the argument brought by bad players - actually, this is the argument brought by people who make excuses for their own failures.
The first thing you have to put in your head, the same as poker players is: bad beats happen. There are situations where you have to shrug off and realize that this game has probability involved.
What you can't do, though, is using this way of thinking to every loss. There's hardly a game where you don't make mistakes or that you can't evolve your play. Sometimes it's about not maindecking a certain card, not taking a right mulligan or even not choosing right wether to play or to draw.
You also can't generalize and say that limited is all about luck - you just chose a ****ty format to play, There are limited formats that are really skill intensive - RoE draft, for instance is one of the most skill-testing environments in all magic. Decent limited formats test your drafting, deck-building and playing skills.
Then again, AVR is a ****ty format.

Confirmed via twitter, Exhaustion should now be fixed.

Quote from herkamurjones »

Quote from necrosed »

Also, I would like make something clear: don't be hellbent in the whole turn4 Jace thing. No one would ever do that in a game if it means you instantly lose the game.
Please don't make a bad play and put the blame on the card.

You make a great point. Shaheen Soorani was on the First Strike podcast last night and said basically that. When this is played in a U/W control shell, you're not looking to slam him down on turn 4; more likely turn 8+ after you've taken control of the game.
HOWEVER, we are a different style control deck and usually don't have 2UU lying around. I do feel that if we have the ability to drop Jace on turn 4 when we can brainstorm a Temporal Mastery back to the top to miracle next turn, we kind-of don't care if he dies. It's almost like our plan becomes:

1. Set up miracle
2. Absorb some damage
3. Proceed to "go off"

I agree with this completely. I think it's also pretty reasonable to do something like turn 5 Jace, bounce a threat, Gigadrowse a second threat, or turn 7 Jace and Exhaustion. Or sometimes you just let him set up your turn 5 miracle + mine and let him die. He doesn't have to be a turn 4 play, you don't have to 0 him every turn. There is power in that flexibility!

I posted a GP Toronto tournament report on reddit for your reading pleasure.

Yes, thats a risk if you dont know what you are doing certainly. I've cast a lot of Nahiri's in my life, and they dont simply lose you the game.