Magic@Tapfor1

About Play-testing

Play-testing is a method for assessing the "playability" of a newly made deck.

Prior to this step, the pool of cards was evaluated to determine which cards/colors the deck would use, a mana curve was calculated and the quantity and type of lands were selected to complete the initial deckbuilding step.

But how well will this deck play?

Can a land drop be made each turn or will it become mana and/or color hosed?

Are the low, mid and high CMC spells balanced so there's appropriate plays available each turn, within the expected number of turns until a win?

Are the number of disruption or fetch cards balanced?

These are the types of questions best answered before the match starts.

Playtesting involves card shuffle, deal and draw steps, along with an analysis of the resulting hand.

A few minutes of playtesting can uncover flaws in the deck build, and allow for tuning the deck prior to a match.

The sections below describe this method in more detail.

Shuffling the Deck

Most likely the newly built deck is fully sorted by color and/or by converted mana cost and/or by spell type. These cards need to be randomized rather quickly.

The steps below will quickly produce deck where the mana is woven with the spells:

Sort the cards into three piles by type (creature | spell | land)
Mix the cards within each pile, but don't mix the piles
Set the spell pile on top of the land pile
Set the creature pile on top of the spell pile
Pick up the whole deck and deal as if to five players
Pick up piles in the order of 1, 3, 5, 2, 4.
Deal the cards as if to five players
Pick the piles up in 1, 3, 5, 2, 4, order
Perform several standard shuffles

The Pile shuffle does not randomize the cards; it just distributes the three types of spells more evenly within the deck.
Performing additional pile shuffles will eventually restore the original order of the cards.
There should be an even number of cards in each pile. If not, the deck contains too few or too many cards. Count and modify as necessary to maintain exactly 40 (or 60) cards.

At this point, the deck will now be arranged in a repeating pattern of two spells, then a land. The deck is considered "stacked" and must be further randomized.

Several overhand or rifle shuffles will now produce a more randomly distributed deck.

This type of shuffle can also be done after a long match where many lands and spells have been played to prevent card-type clumps.

As a matter of game etiquette, be sure to perform a few overhand and/or rifle shuffles in front of your opponent prior to the start of a match.

Your opponent does have the right to shuffle your cards, should they choose!

Deal a Few Hands

Now comes the challenging part of building a deck; evaluating if it does what was intended.

The idea here is to see what your options would be over several turns. It's likely that a sealed deck match will go for about seven turns before a win. Therefore, it's good to evaluate the initial hand, and the options available over six or seven draws.

Before dealing decide if you will play first or draw first, then deal a seven card hand, face up.

Count the number of Lands, Creatures and Spells

Will you make three consecutive a land drops?
Can you play any creatures on turn one?
Do you have other cards you would want to play on turn one?
Do you have cards to play on turn two?
Should you mulligan?

Draw a card.
If you played first, does this card fit the mana curve?

If this is your first turn, did it fill a void in the initial hand?
Can you make plays on the next few turns?

Continue drawing a card and evaluating options and plays.

Anticipate the types of plays your opponent might make:

playing a large or evasive creature
dealing non-combat or directed damage
casting counter or removal spells

Do you have cards in hand to deal with these?

After drawing your seventh card, consider your status against an ideal opponent (one that has no blockers in play or counter spells in hand):

determine if it's possible to establish a win condition.
how many turns are needed to deal 20 damage?
can all spells in hand be cast with the available mana?
are there any unplayable cards in the hand?

After a few playtest hands, evaluate the answers to these questions to help decide if a deck modification is needed.

Making Modifications

Math Problems

Probability and statistics should influence deck modifications.

Consider that there is only a 25% probability of getting a particular card in a hand of ten cards from a forty-card deck.
Ten cards is a fourth (one quarter) or 25% of the deck. (10/40 = .25) Or said conversly, there is a 75% chance of not getting a particular card after three draws. The odds only improve to a 35% liklihood of getting a particular card after seven draws.

So if the deck's win condition is to draw and play that one Worldspine Wurm card, chance (the Hypergeometric Distribution Function) says it will happen about once in three games.

It's important to know the anticipated probability of seeing a given card in the deck in order to evalute the observed frequency.

If winning the game hinges on playing one particular spell by turn seven, statistics says the deck needs to have four copies of that spell to make it likely to happen each game.

Mana Problems

Before changing the number of lands in the deck, determine if the amount of each color is mathmatically appropriate.

Is there a mana issue related to the deck's primary color(s) or splash color(s)?
With four forests in the deck, the odds of getting two of them to play a green