Probability Theory of Drawing Cards

[Crackdown]

Although I like to make up my own weird decks, I also like reading about tournament decks and sideboards. When reviewing the lists, I notice quite often that a sideboard contains just one or two cards to counter a potential threat. What is the probability of drawing those cards when needed?

Suppose for instance that Kor Haven could shut down an early beatdown, so you decide to include an appropriate card to take it out before your opponent gains control. How many response cards should you have? Should you assume that they will have 4 Kor Haven and intend to play it soon? If so, do you include 4 Port or sowing salt?

What I'm trying to figure out is how much probability theory goes into this analysis. Does the player calculate how many turns he can go without a response and then put in enough response cards so that the probability of drawing the appropriate response is x% within x turns?

Or is the whole process a lot less formal than that and just includes a lot of playtesting?

 

[galtwish]

This is hardly a casual topic , but here's a quick reply.
As a long time deck builder (que music), I have often asked the same question. My deckbuilding aquaintences all have different methods on how to build. Here are some basic tenets:
1>Make sure you don't die to beatdown on there 4th turn if you are going second
2>If a card wins you the game you put in
4 if you are beatdown
3-4 if you are aggro/control
2-3 if you are control
combo is a case by case basis
Morphling is a good example of this. You will see 2-3 Morphling in Draw-Go, if they play it, 3-4 in ConTroll, and 4 in Accelerated Blue. 4 Masticores in Accelerated Blue, 3-4 in Ponza, and 1-2 in either Squirrel Deck, if at all.

The appropriate resonse cards in the deck are generally part of the sideboard, or pre-sideboard strategy incorporated in the main deck to give a higher winning percentage against a certain deck. Scott Johns has a great series of Articles running on magic.mindripper.com right now about sideboarding, so I'll defer to his explination. I tend to go for results that smash the largets crossection of the field. Here's my current Ponza Sideboard as an example:

3 Cursed Totem-Replenish, Accelerated Blue, Rebels, misc stuff
4 Scald-Islands really suck against this deck
2 Boil-Ditto
4 Keldon Vandals-Smashes Monolith agaisnt AccelBlue and Bagain, while providing more beatdown (desperately needed against both), a bad Treachery target, iosn't Misdirectionable, and also helps against Ponza, Tinker, Wildfire and any random artifact decks.
2 Blood Oath-Against Replenish and all the random decks at my local tournament, because I see a lot of Exploration/Sneak Attack/Bargain/mid-range creature decks. Really helps steal a win, like Final Fortune in old Deadguy Red (Worlds 98 deck).

I also bring the following in case I want to change my sideboard before the match:
4 Cave-In-Squirrel, Elf, Suicide Black, Sligh
1 Flowstone Slide-Rebels, Control Green
3 Flashfires
1 Boil
2 Crater Hellion-Mid-range decks, surprise factor
4 Rack & Ruin-Tinker, Wildfire

I really suggest that the best way to be ready for a tournament is to go an hour early and bring 30 extra cards for your sideboard. Scout what other people are regestering and playing and make your sideboard (and maybe even change up your maindeck a bit) right before you have to turn in you deck list. remember the character of your deck and build appropriately.

Erik