Showing posts with label Probablity. Show all posts
Showing posts with label Probablity. Show all posts

Wednesday, December 30, 2015

1-ofs

Four weeks ago we discussed Hypergeometric Distrubtion and introduced this series. We also discovered that if you run four copies of a card in a 60-card deck you have a 40% chance of having at least one copy in your opening hand.

Here is the introduction: http://mtgolibrary.blogspot.se/2015/12/1-ofs-2-ofs-3-ofs-and-4-ofs-and.html

This week we are talking about 1-ofs. Single copies of a card in a deck is generally a mistake. The reason is that there will be many games of Magic where you never see the card. If you can win without ever seeing the card in a game, why is the card even in your deck? You should just increase the number of copies of some other card.

The Hypergeometric Distrubtion calculator from part 1 shows us that the probability of having your single card in your opening hand is only 11,6%.

So when do legitimately run a single copy of a card in your 60-card deck?




You can search for the card
A good example of this is the dreaded 1-land-spy deck in Pauper. It needs only one land because it can search for that lonely forest easily. Here is an example deck list: http://tappedout.net/mtg-decks/1-land-combo/

In the deck mentioned above, 1-land-spy in Pauper there is an example of this as well. You only need one kill card, Haunting Misery, because you will draw your entire deck and can then kill with it.

You can draw your entire deck
It is similar to being able to search for the card.

Your card is restricted
Congratulations, you are playing Vintage. You are awesome.

You really want to run 5 copies of a card
If you have a card that is a 4-of and you then want just one additional copy that means that the second card has to be a 1-of.

You need to run a single copy because of mana curve considerations
I think this is less likely to come up than it was for 3-ofs or 2-ofs. Stay away from 1-ofs if you can.




Wednesday, December 16, 2015

3-ofs

Two weeks ago we discussed Hypergeometric Distrubtion and introduced this series. We also discovered that if you run four copies of a card in a 60-card deck you have a 40% chance of having at least one copy in your opening hand.

Here is the introduction: http://mtgolibrary.blogspot.se/2015/12/1-ofs-2-ofs-3-ofs-and-4-ofs-and.html


This week we are talking about 3-ofs. As an example we will be using Spire Golem in Pauper Delver decks.

The Hypergeometric Distrubtion calculator from part 1 shows us that the probability of having at least one Spire Golem in your opening hand (if you run three Golems) of 7 cards is 31,5%.

Here we have to consider another factor. By turn 6 you have a 12% chance of drawing multiple copies of your 3-of card. This means that a big finisher or something for the late game is probably not a card you want 3-of (Rolling Thunder in Pauper RUG Tron is a good example here).


So when do you run 3 copies of a card in your deck?



You want to be sure that you can draw one copy of the card in most games
This card is good and you want to see a copy of it in almost all games you play.

This also implies that you don't want as many copies of the card as possible because then you would just run 4-of.

You really want to run 7 copies of a card
It is hard to think of a situation where you don't want 8 copies of a card but exactly 7 but those situations may occur. In this case you have found a second card that is similar enough to the first card (that you have 4-of) so you are including 3 more.

Your 4th copy is in the sideboard as a wish target
This is a corner case but if your deck can fetch cards from the sideboard the fourth copy needs to be there to be fetched.

You need to run 3 copies for mana curve considerations
If your mana curve needs exactly three cards of this cost you have a good case for a 3-of.

I think that is what happened with Memnite in this Modern Affinity deck: http://www.mtggoldfish.com/archetype/modern-affinity-8972#online

This includes situations where you have to cut down to 60 cards from 61 and something just has to go. Remember, though, that the likelihood of seeing the card in your opening hand has been reduced by almost a third.

If having more than one copy of the card in your starting hand is a problem, consider running three copies of the card
We mentioned this last week. If a card is critical to your strategy but would be problematic in multiples in your opening hand, you should not run four copies of the card but go for three.

A good example of this is Armadillo Cloak in Pauper Hexproof decks. This card is an excellent card for the deck but having two cloaks in your starting hand will make you very slow. Thus, do not run four of this card.

here is an example deck list with three Armadillo Cloaks for this very reason: http://www.mtggoldfish.com/archetype/pauper-hexproof-22844#online

Wednesday, December 9, 2015

4-ofs

Last week we discussed Hypergeometric Distrubtion and introduced this series. We also discovered that if you run four copies of a card in a 60-card deck you have a 40% chance of having at least one copy in your opening hand.

Here is the introduction: http://mtgolibrary.blogspot.se/2015/12/1-ofs-2-ofs-3-ofs-and-4-ofs-and.html

So when do you run 4 copies of a card in your deck?



You want as many copies of the card as possible
My first rule is that you should consider if you would like 5, 6 or 7 copies if you could. A good example of this is Lightning Bolt in a Modern Burn deck. You would probably play 40 Lightning Bolts if you could. This is the easy 4-of.

In fact this rule is a good reason to run 3-ofs, 2-ofs and 1-ofs to include extra copies of the card. We will return to that example in future chapters.

You want to have a copy of the card in your starting hand
Well, the chance of that is only 40% (see above) BUT it is the highest chance you can get. If the card is critical to your early strategy you probably want four copies of the card.

A good example of this is the RG Landfall deck in Standard. You want to start the game with a Scythe Leopard on turn 1 so you run four copies of the card.

Here is a sample deck list of RG Landfall in BFZ Standard: http://www.mtggoldfish.com/archetype/standard-r-g-landfall-22839#online

If having more than one copy of the card in your starting hand is a problem, DO NOT run four copies of the card
If a card is critical to your strategy but would be problematic in multiples in your opening hand, you should not run four copies of the card.

A good example of this is Spire Golem in Pauper Delver. This card is an excellent card for the deck but having two Spire Golems in your starting hand will make you very slow. Thus, do not run four of this card. This is a basic probability mistake made by players who do not care enough to do the math.