[Strategy] Mulligans

[GreenOne]

I'm quite good as a combo player, being able to focus on different strategies, knowing when to go off, and seeing easily the paths to victory through the cantrips.

Unfortunately, I never had the mulligan skills.

Till now, I did well piloting decks that are forgiving with mulligans, like DDFT and Solidarity.
Thanks to the high manipulation power of those decks, you can almost keep every hand with some number of lands, some number of cantrips and shape your hand as game progresses, taking in consideration what your opponent is doing meanwhile. When you'll finally combo, your hand is completely different from your starting 7, and often resembles the hand you want to have at that time. The decks rewards greatly your playing skills, but not that much your mulligan skills.

Even with belcher the mulligan is quite simple: look at your hand and ask yourself if this hand can combo right now or the next turn. If it can't, throw it away.

But now, I'm approaching TES, and this deck is not as forgiving with mulligans.

How can I improve my mulligan skills in general?
And how, specifically, with storm combo decks?


EDIT: Here's some literature about mulligans:
The Art of the Mulligan, by Ken Krouner - 2003
The Art of the Mulligan: 8 case studies by Ken Krouner - 2003
Fundamentals: the Mulligan by Rogier Maaten - 2005
SCG Daily - The Kid’s Perspective: Some Mulligan Exercises by Julien Nuijten - 2006
SCG Daily - The Mulligan by Zac Hill
Finding Meaning in Mulligans by Jeff Stewart - 2006
Feature Article - Sixes and Sevens: Mulligans in Constructed by Dan Paskins - 2007
Unlocking Legacy - Mulliganing and Deck Design by Christopher Coppola - 2008
The Philosophy of the Mulligan by Terry Soh - 2005

[Bryant Cook]

I'm the exact opposite on an aggressive mulligan-er (?). If it's somewhat keepable I roll with it, although, it's not the best idea.

[frogboy]

You should mulligan hands that won't let you win the game. I shipped a bunch of ostensibly good hands with lands and spells last weekend because they didn't have any relevant cards in the matchup.

With a combo deck I just count up mana and business and keep if I'm good on both counts.

[sauce]

i think knowing when to mulligan and doing so with discipline is the hardest and most relevant/necessary "next step" in going from being avg to being a good player.

basically most of the time i tend to keep a hand that is avg, i ship hands with inadequate land count (depending on the deck im playing), whether its too much land or too little land -- should cause the same mulligan decision.

threshold decks can run on 1 land, but if you know that youre up against LD heavy opponent, you may want to mull to get at least 2 lands.

with any deck, i assume the strategy is the same, as mentioned by a previous poster, you want to figure out if you can win with a given hand or not.

the easiest way to figure it out, is obviously experience w/ the deck.
goldfish the deck youre trying to build and keep all sorts of hands.

you will start to see what seems to work and what doesnt.
that is what has helped me in the past in my attempt to figure out if i should mull or not...


also, i think the hardest part of competitive play is taking emotions out of the mulligan, sideboarding, and general decision making process.
just like in poker, if you go on tilt, your decisions will be severely hampered and you will most likely not win.

sometimes taking a mulligan or two, or three... in a row... may put you on tilt. (ie make you pissed/cloud your judgement)
this is where i think everyone is different in their personality type.
some people can manage to not tilt, others cannot avoid tilting and do so loudly.

but i digress.

goldfishing or talking to other people/watching others play the archtype seems to help alot when trying to figure out what type of hand is keepable and what hand is autoship.

[SuperBean]

You are correct with Belcher, if you can't go off the first or second turn you dump the hand and get a new one.

With T.E.S I do a few things if I have to Mulligan.

The first thing I do is count my mana sources in my opening hand: LED = 3, D. Ritual = 2, etc. etc. And see how much mana I have to work with, and is it enough to go off based on what else is in my hand. If not, can I go off the second turn? These are both important factors especially when you are trying to out race your opponent.

I always think about my match up, how many disruption spells they are playing, how weak am I to those spells when and if they are cast. If they are playing Counterspells, I'm going to need Orim's Chant and Duress to back my combo up, and will I have the available mana to disrupt them enough and combo off? Will it be enough?

Hopefully that helps you a little, I have a feeling it's a rough explanation, but it's always a hard thing to describe when you have to mulligan with this deck.

[Kuma]

also, i think the hardest part of competitive play is taking emotions out of the mulligan

Truth.

I don't know how many otherwise excellent Magic players turn stupid when it comes to mulligans. They'll make bad decisions out of anger because they were "forced to mulligan." They make silly rules for themselves like, "I'll never mull to five." They look at the top card(s) of their deck after deciding to mulligan, even though doing so can only lead to second guessing and bad decisions in the future.

The main question you need to be asking yourself is "Does this hand give me a chance to win this game?" or alternatively, "Does this hand allow me to make a gameplan that has a chance of winning?" If it doesn't, then you need to mull.

The next question you need to ask yourself is "Is this hand better than an average hand of one card less?" If it isn't, you need to mulligan.

When you start getting in the 4-5 card hand area, the questions usually turn into, "Is this hand playable?" or "Does this hand have land?", but that's a function of diminishing returns.

The only thing I do differently with storm combo, is that I'm more likely to keep iffy hands vs control because you're probably not going to be able to win in the first few turns, and you'll need all the cards you can get.

However, I'll still mull if the hand is atrocious.

[KrzyMoose]

I agree with Kuma's post above. To add to that, ponder this:

Four steps to mulligan'ing correctly:

1. Think about your opponent's deck.
2. Think about whether or not your current hand actually does anything relevant given what your opponent might do.
3. Think about whether or not your current hand may be better than a hand with one less card.
4. Think about the probability that you can draw a card (or series of cards) that will make your current hand good. (This is actually very, very, very important; people with SCG Premium should read Richard Feldman's article about playing AIR, which discusses this point. Even though I think Feldman is kinda mediocre, this article does a fairly decent job on this topic. Obviously, I understand that most of the people on here don't have Premium. But whatever).

There is a very delicate balance between 2, 3, and 4, and there really isn't a "most important" step.

[Pulp_Fiction]

This is the trick to mulligans, you have to know a fully understand the deck you are playing. When you look at a hand and think "ehh, I don't know" throw it away, everytime. Never, ever rely on the ability to topdeck anything. Naturally assume the deck is going to fuck you over. Here is a prime example, when I play Zoo I run 21 lands, if the hand doesn't have at least 2 I throw it away. I have absolutely no reservations about going down to 5 or 4 cards if the situation really merits that type of action.

Most people are afraid to mulligan down to 4 and 5 card hands, which is fine, to each their own. I just ask myself the question, is this hand better than a random hand -1 card? I have thrown away hands and watched people around me staring at me with wild expressions wondering what the hell I was doing. You have to have the right hand for a certain matchup, and I have no problems about going to 5 cards in order to try and find the one I need for a matchup. Like with Aggro Loam, when I know I am up against Leyline of the Void, Im going to mull into Burning Wish or Krosan Grip every time, or go down to 4-5 cards trying. Most people will mull once trying to find their answers and if their next hand is really good but has no answers they will most likely keep it anyway thinking they can play around the hate and their current hand is way better than a random 5. Fair enough, but when that hate card comes down followed by additional things the opponent plays the story changes and most of the time you wish you mulled into a different hand.

Some people like Bryant can keep semi-decent hands and have them work out, I won't take that chance, because I truly am one of the most unlucky people you will ever meet at a card table, and that bad luck has taught me to never rely on the deck giving you what you need when you need it. You have to mull into it. This is JMOP. Hope this helps :)

[GreenOne]

You should mulligan hands that won't let you win the game.

Nice, but how do I get there?

With a combo deck I just count up mana and business and keep if I'm good on both counts.

Yeah, I keep those hands too, and usually results in turn 1/2 wins. This are, actually, highly keepable hands that I won't discuss keeping.

i think the hardest part of competitive play is taking emotions out of the mulligan, sideboarding, and general decision making process.

I see magic as a chance game, so I continually calculate odds of drawing cards/opponent have card X if playing deck Y, etc. This, I believe, is one of the strongest parts in my game. The bad thing about this is that with mulligan decisions you usually have little or none informations about the opponent's hand and what he's playing. This screws my math bad.

To give a simple example, I'm on the play with this hand:
Petal, Petal, AN, Chrome mox, Burning Wish, Rite of flame, Tendrils.

This is a good hand against anything non disruptive. I can actually combo if I draw Land or Petal or Chrome Mox or LED or Rite of Flame or Dark Ritual, giving me 1/2 of the probabilities to draw a card and win each turn.

Against anything blue this is a mediocre hand. Force is going to fuck me, and even daze is a problem (unless I draw Dark Ritual, Rite of Flame or LED), he has a 8 card hand, he gets to play his first turn cantrip and he'll also know that I'm playing combo when I pass the turn playing nothing. If I don't draw the needed card on the first turn the opponent is going have a chance to play more cantrips, CB, Standstill and put me in a bad position.

If I'm against a Prison deck he can go Chalice@0 or trini and fuck me.

A discard spell screws me too.

Should I keep this hand or not? What if one of the petal was a land (possibility of playing cantrips I draw, but opens up to waste)? what if I had a 6 card hand (no tendrils)?

Evaluating things when you don't know opponent's deck is hard.

goldfishing or talking to other people/watching others play the archtype seems to help alot when trying to figure out what type of hand is keepable and what hand is autoship.

Here comes one of my problems. Unfortunately, I'm the only legacy combo player where I live. Other combo players generally doesn't know anything about the format and never played it. Other non-combo players don't really know how the deck works.

This is the trick to mulligans, you have to know a fully understand the deck you are playing. When you look at a hand and think "ehh, I don't know" throw it away, everytime. Never, ever rely on the ability to topdeck anything. Naturally assume the deck is going to fuck you over.

So I guess you always mull the hand I presented above. This is actually an interesting advice. I'll try this and see how it works. Maybe this should correct my "keep-er" attitude.

Four steps to mulligan'ing correctly:
1. Think about your opponent's deck.
2. Think about whether or not your current hand actually does anything relevant given what your opponent might do.
3. Think about whether or not your current hand may be better than a hand with one less card.
4. Think about the probability that you can draw a card (or series of cards) that will make your current hand good. (This is actually very, very, very important; people with SCG Premium should read Richard Feldman's article about playing AIR, which discusses this point. Even though I think Feldman is kinda mediocre, this article does a fairly decent job on this topic. Obviously, I understand that most of the people on here don't have Premium. But whatever).

There is a very delicate balance between 2, 3, and 4, and there really isn't a "most important" step.

Yeah, as I said things screw up when you have no knowledge of #1 and #2. Does this works anyway without 1 and 2 for you?
Got a link for this article? Gonna read it in 90 days.

Is there something already written that halps with those decisions? I remember there was an old article on starcitygames about mulligans, but nothing about mulligans and combo decks? things are working quite differently in the world of combo.

[Joe_C]

Evaluating things when you don't know opponent's deck is hard.

This is the same situation I always run into. Not knowing what your opponent is playing, what would you do with this hand?

forbidden orchard, dark ritual, dark ritual, AN, duress, brainstorm, LED

would you:

1. attempt to win turn 1
2. cast duress turn 1. down side allows them to wasteland your land, maybe draw into force on their turn if they didnt have it
3. play land and pass. giving us knowledge of their first land drop, we can brainstorm their EOT

It seems like the BEST option is to attempt to win turn 1. They either have Force or dont... Ive gotten alot better at making seemingly dull hands work out... being able to resculpt your hand after a failed attempt is where the playing of the deck can be a little rough

[GreenOne]

This is the same situation I always run into. Not knowing what your opponent is playing, what would you do with this hand?

forbidden orchard, dark ritual, dark ritual, AN, duress, brainstorm, LED

would you:

1. attempt to win turn 1
2. cast duress turn 1. down side allows them to wasteland your land, maybe draw into force on their turn if they didnt have it
3. play land and pass. giving us knowledge of their first land drop, we can brainstorm their EOT

If the opponent is unaware of what you're playing, the probability of opponent having FoW+blue card are less than 20%, so it's obviously the right call to take the chance and go for the throat. You'll NEVER have more than 80% win against a random deck.

However, I'd never mull this hand, so it's roughly relevant in this thread.

[Elfrago]

If the opponent is unaware of what you're playing, the probability of opponent having FoW+blue card are less than 20%, so it's obviously the right call to take the chance and go for the throat.

However, I'd never mull this hand, so it's roughly relevant in this thread.

Are you're sure it's only 20%?
I remember somthing like 40% or so.

[MSC]

This is the same situation I always run into. Not knowing what your opponent is playing, what would you do with this hand?

forbidden orchard, dark ritual, dark ritual, AN, duress, brainstorm, LED

Forbidden Orchard means you have to try the First Turn Win or the will try to Optimize their Play against Combo.

It would be different, when you had a Fetchland or USea in your Hand. Then you would play Duress. This way you get a maximum amount of Information about their Deck and only give them a little Hint about yours...

[GreenOne]

Are you're sure it's only 20%?
I remember somthing like 40% or so.

40% is the probability of an opponent to have at least a card of a 4x in the first 7 cards.

The only data at my disposal is that of the 40 decks that top8ed this month, 20 had FoW in it. Decks that did not top8 are likely to have less FoWs: casual decks, extended ports, etc. are present in non-top8 decks.

So I'd say that in the first rounds/lower tables the probability is less than 20%, where it becomes more and more approximate to 20% as you reach the top tables.

This doesn't take account of the chances the opponent also has a blue card to pitch, or if the opponent mulligans or plays less than 4 FoWs for some reason. It's also based on an average metagame built with data I have. It's obviously worse for you if you're in a blue-heavy meta than in a blue-light one.

[Bryant Cook]

I think the type of hand Greenone is talking about are the hard decision hands... such as...


Mox, Ponder, Mystical, Infernal, LED, Dark Ritual, Infernal. Do you keep this? I would. It's got tons of potential, although, at the same time tons of fail.

On the play, Mox (Imprint Mystical), Ponder (Pray to hit 2nd mana source). 2nd Turn 2nd mana source, LED, Ritual, Infernal, Ad Nauseum.

On the draw, if you draw a land, Mox, or Ritual. You cast your hand and Ad Nauseum or do the upkeep trick.

or...the Mulligan.

[GreenOne]

I think the type of hand Greenone is talking about are the hard decision hands... such as...

Mox, Ponder, Mystical, Infernal, LED, Dark Ritual, Infernal. Do you keep this? I would. It's got tons of potential, although, at the same time tons of fail.

On the play, Mox (Imprint Mystical), Ponder (Pray to hit 2nd mana source). 2nd Turn 2nd mana source, LED, Ritual, Infernal, Ad Nauseum.

On the draw, if you draw a land, Mox, or Ritual. You cast your hand and Ad Nauseum or do the upkeep trick.

or...the Mulligan.

Exactly.
Often it comes up to a limited number of cases:
1- Land light hands with no immediate combo, like the hand you described or the one I took as example before.

I generally keep those if about 1/2 or 1/3 cards of my deck gives me the win.

2- Acceleration light hands with cantrips, lands and tutors, like: [Petal, City of brass, City of brass, Brainstorm, Brainstom, Burning Wish, IT]

I generally keep those.

3- Acceleration light hands with lands, tutors, a good protection and no cantrips, like: [B.Wish, Duress, Duress, Gemstone Mine, City of Brass, Chrome Mox, Petal]

I mull them. It's a hand that rewards keeping only against control/aggrocontrol decks. And I usually want to be the aggressor against those anyway.

4- Hands with no tutors nor cantrips, but loads of acceleration and protection, like: [City of Brass, Gemstone Mine, Duress, Chant, LED, D.Rit, D. Rit.]

Those are one of the most difficult hands. A single tutor or AN (12 cards) win you the game, a cantrip (7 other cards) put you in a good position. You also have some turn 1 plays anyway (duress) and some way to stall the game against aggro (chant).
I usually keep them on the draw, mull on the play.

5- Hands with loads of acceleration, but no tutors and just 1 cantrip: [City, Mine, LED, LED, Brainstorm, D.Rit, Petal]

A revealed tutor/AN and you win the game in a short time. I usually keep them if I got a Ponder, and ship them if I have a brainstorm: if brainstorm reveals chaff, I almost lost the game. Maybe I should keep it anyway but slowroll the brainstorm, so I can see more cards?


Another thing I really don't understand is:
If my unknown opponent mulligans, should I keep worse hands? Should I instead try to be more aggressive and aggressively mulligan?
What if my opponent is playing randomFoW.dec instead?

EDIT: I searched for some articles about mulligans and put the links in the opening post.

[KrzyMoose]

The link to the aforementioned Feldman article: http://www.starcitygames.com/magic/e...AllIn_Red.html

AIR is basically a combo deck, so this article also applies to other formats/decks.

[Bryant Cook]

I'd keep all of those hands regardless.

[jrp]

I'm with Bryant. I'd keep all the hands listed above. TES specifically does not mulligan well, if it's at all keepable, I go with it.

[GreenOne]

I'd keep all of those hands regardless.

I'm with Bryant. I'd keep all the hands listed above. TES specifically does not mulligan well, if it's at all keepable, I go with it.

So I guess I'm an aggressive mulligan-er?

EDIT: I just remembered I hand I kept, but I questioned it much:
[City of Brass, City of Brass, Undiscovered Paradise, Rite of Flame, Lotus Petal, Chrome mox]

It ended up in a great failure, without being able to combo off before the opponent killed me on turn 4, playing an aggro deck.