I really just chose this list because it was from someone much more respected than myself and I wanted to avoid criticism about personal choices that I would have made that differed.

Do you happen to have a preferred Flash decklist?  I had a slow day at work, so I was working through all the possibliities of Turn 1/Turn 2 kills with/without protection.  I would love to see what it was for a tourney proven deck. 


I actually couldn't agree with you more.  The reason I gave this example is because people have been claiming it is analogous recently and some just throw out percentages.


I agree that sometimes these things are a little like pornography....hard to define but we know it when we see it.  That approach is fine for me too.

The precise numbers comes from more of just how I think about building decks and Magic in general.  For the most part, I am perfectly happy with well-reasoned arguments; however, when some people start posting win percentages and making claims based on small sample sizes, I generally like to know exactly what is going on (honestly, this is not a reflection of the quality of those posts, but a mere reflection of how I solve most problems I deal with on a daily basis).  I guess some of this is also just a tendency to "go all the way" on the nerd factor as well -- I actually find the exact win percentage possibilities to be quite interesting (hell, I might even assign problems like this to some of my graduate students to work on to get them ready for some tricky comprehensive exam questions).

Finally, I think that a set of precise statements about why the deck is bad for Magic as a whole is stronger when supported by some quantification.  Of course, if the ridiculousness is salient enough, then the quantification is just "gravy", so to speak.  On the other hand, it is a fallacious argument to say that since a B/R decision hasn't been based on such evidence in the past that it is unnecessary -- that doesn't invalidate the process and possible importance of the quantification for the question at hand.  Dismissing it prior to the completion of the exercise, makes the argument appear weaker as it seems supporters of the restriction are worried that the numbers won't support their claims -- although I have a very strong sense that they will in the end.   

For further clarification, upon rereading my original post, I can see how the tone indicates that I think that 7.2% is no big deal so let's get on with our lives.  This is really meant to be a first pass at things and get some feedback (which I already feel as been successful) to get a more detailed analysis.  Once Turn 2 is taken into account, the numbers will look a lot different.  I am one that agrees that this does take a little fun out of magic, but I don't know what the right restriction is -- I lean more and more towards Merchant Scroll because I am inherently opposed to it being unrestricted regardless of the deck.     

A couple of things...agreed on the statistics.  Thanks for updating this, I was commuting this morning and couldn't put in the desired table.  I definitely didn't mean to get people into an uproar.  Although, I don't know how far some one would realistically mulligan or if they would keep a hand of 5 with three leylines, so there is always a little fudge.  People could use the first table I posted and this table to get bounds on the probabilities to get some idea of what they may/may not be willing to do.

Agreed on Merchant Scroll.  This card is way overpowered.  It effectively sets up turn 2 kill for Flash almost everygame.  With a 40% chance you see it in your opening hand, you can happily just go along and grab a counter or Flash itself.  This card is the engine that recent decks of note have been built around.  Whether it is too powerful or not, I think it should be restricted based on the basis that you put a card directly into your hand for 2 mana.

Personally, I think that Flash (the deck, perhaps not the card) is too powerful.  I understand that it is not the list that optimizes the best configuration; however, it was an objective list to consider and I would be glad to consider another list as well.  I would be glad to consider different lines of play as well that are not as obvious.  The calculations are not that difficult.  As a point of clarification, I assure you that nothing was "convenient" to skew the results in any way shape or form, just trying to use an objective list from a notable person.  If I would have made up my own that optimized some mana considerations, I am sure that would have drawn some complaints as well.

I am fairly certain that a turn 2 probability of a kill is quite high and very disconcerting.  I just think that it would help make more clear arguments if some of them were grounded in fact and backed up by clear numbers instead of anectdotal evidence. 


This is not meant to be deceptive.  I am just pointing out the probability of the "combo" (which I use loosely since, as you stated, you can just play lands to get out of it) that people often are comparing it to.  Note, I didn't say first turn kill.  I try to be exact in my wording so some sense of reference from prior experience can be gleaned and an overall approach to the argument can be formulated. 

I really think that something needs to be done about the ridiculousness that is Flash; however, I think that the arguments against it need to be supported with more than anectdotal evidence.  Otherwise, the arguments will be dismissed out of hand as (1) some people just don't like to play against Flash because it ruins their pet deck -- I find this argument ridiculous in the extreme, (2) Vintage is broken, so get on board (another ridiculous argument), (3) only the vocal people are letting their opinions be known--most of vintage is fine with Flash (this may or may not be true), etc.  If this can be backed up with objective evidence, then the argument becomes all the stronger.

As I said, please outline some common lines of play that you would consider part of the Flash powerhouse -- I am very curious to see how likely they are. 

This interpretation is correct.  Sorry if that was a little unclear.  Also, it is the probability of seeing exactly one in your hand as I thought it would be unlikely to desire to keep a hand with more than one.



If you consider instances of seeing multiple copies, then the probability of seeing 1 in your opening hand is 33.63%, of seeing 2 is 5.93% of seeing 3 is .38% and of seeing 4 is .007% -- which is 39.95% overall.  I considered the possibilities of seeing only one for Leyline of the Void; however, for the 1st turn Flash kill I considered multiple copies of everything.

You have to set up the math to ignore playable hands because that would be impossible to determine and so subjective depending on the skill (or lack thereof) of the player.

I know there has been a lot of talk about the nastiness of the first turn kill and the frequency at which it arises, but I don't think it is quite as common as people are perceiving or testing.  In fact, we can just work it out mathematically (I refer you to the hypergeometric distribution if you are curious).  In any event, I am basing this off Steve's decklist for consistency. 

Now, the minimum for a turn one kill before your opponent does anything is having 2 mana to cast Flash.  That sets us up with several alternatives:

(1) Lotus, (2) Flash, (3) Hulk, (4 - 7) whatever.  Naturally, we will assume there are 4 Flashes, and 6 Hulks (we will count the 2 Summoner's pact since it can just grab it for free).  Some variation of this will happen 1.84% of the time.

Now, we can also get a kill off a draw with one land and one mox, with the same chances at Flash and Hulk.  This variation will happen 4.98% of the time.

Finally, we could get the kill from a random mox + mox sapphire/lotus petal .... that will happen .28% of the time.

This puts us at 7.20% of the time 1st turn kill when playing -- which may or may not be too much for Vintage.  Of course, this could be boosted a little through aggressive mulligans or adding ESG/SSG, but nothing too much.

Compare this to an unrestricted Trinisphere with Mishra's Workshop.  The same analysis shows that had a 16.18% chance of a first turn Trinisphere.

Additionally, is is a little information for Leyline of the Void if you are determined beyond all reason to open the game with one against Flash.

Probability you get one in opening 7:  33.63%
Probability if you mulligan to 6:           53.89%
Probability if you mulligan to 5:           66.29%
Probability if you mulligan to 4:           73.96%
Probability if you mulligan to 3:           81.25%
Probability if you mulligan to 2:           83.62%
Probability if you mulligan to 1:           84.71%

*All of these probabilities are computed directly from formulae, so it isn't depending on specific samples, computers, or playtesting.



(I don't play Flash often, so if I am missing a potential first turn kill from this decklist, just let me know).

Considering the best decks ever is a really difficult; much like trying to compare athletes from different eras.  Since the cardpool is constantly evolving and growing, I think it would be none too surprising if the current Top Tier decks are always going to be better (at least on average) than older decks.  As more cards are added to the game, the number of combinations grows exponentially.  Keeping the older decklists static puts them at an inherent disadvantage. 

Perhaps a more fair way to test the "Banned/Restricted" decks would be to allow the core of the decks to be built around with the existing cardpool.  For instance, just build a deck with 4 Balances -- I imagine that could be transformed into quite a force; however, the downside is a lot of effort required to build decks that cannot be realistically played with.   

These numbers are not quite correct -- I only bring this up since this is what you seem to be forming your decisions.

If there are 12 disruption/protection spells there is about 13.5% chance you will see 3 or more in your opening seven.  Cutting this to 11, reduces the percent to 10.8% (almost 3%).   In any event, the probability of seeing 1 or 2 in the opening 7 with 12 cards is about 67.4%; whereas, with 11 cards the probability of seeing 1 or 2 is reduced very slightly to about 67%. 

However, I think the take home message is the same -- free up the slot for some business.

I am actually not too troubled by Flash....I tried to portray that in my first post.  It really is a challenge to play against and I think it does make players consider more protection for their combo (if they are flash) and more disruption against flash.  At some level, this can be considered interacting with each other.  I really think that Merchant Scroll is a lot of the fuel right now.  For instance, it can tutor up Gush and draw two cards off just two mana--add fastbond to the mix and that is when everything reaches new levels of ridiculousness.  While it is a separate discussion, Merchant Scroll seems to go against some of the principles of restricting other tutor cards.

I do like diversity and it is nice to see something new once in awhile.  I agree that it hasn't reached dominating levels yet, and if you aren't prepared for it, then woe be to you.  The same principle with Ichorid.  It requires more thought to how to build your deck to account for what advantage it is trying to exploit.  However, I don't think everything is quite as diverse as you portray.  We still have only a few big time players.  From Steve's article:

Workshop (25%)
Gush (25%)
Flash/Ichorid (16%)

That is 3 archetypes (if I count Flash and Ichorid being the same type of Archetype as the same type of Hate will get rid of them and they both use the graveyard for the their business) dominating 2/3 of the field (65.6%) to be exact.  Furthermore, the number of flash decks increased from 5 to 11 -- which is 120%.  Perhaps this is only a coincidence and will be evened out during the next sample, but it might be a trend and we will have to see.

@Shock Wave:  Are you in favor of fully restricting Flash, or do you think that restricting Merchant Scroll would be a reasonable solution (perhaps first) to slow the deck down without cutting it off at the knees, with the benefit of also balancing out the power of the colors?  Also, I think that the point you made about interaction is interesting.  I definitely am a fan of more interactive games, but I don't know how to define interaction.  Almost any combo deck tries to race the opponent while minimizing interaction with the opponent that is beyond counters.  Granted, the majority of cards in the classic combo decks are more difficult to play and most of the cards are restricted -- are those two conditions enough to leave them be even though they don't really interact.  I see the same lack of interaction problem with Ichorid and to some degree Oath.  It just seems like interaction is a nebulous term to figure out how to start banning types of cards. 

@diopter:  I think the underlying argument of "pet deck getting blown out of the water" is really about the division of whether you should try to create/play a deck that you are fond of and tweak it to your metagame or if you should just lay out the five best Tier 1 decks and choose the one that is best positioned for the tournament/event you plan on attending.  Personally, the second choice is less appealing to me because it feels like I am no longer in control, just piloting an already well-tuned deck.  The difference is going from a true master of a certain archetype (which is maybe undesirable) or a jack of all trades (perhaps a master of all, but that is a much higher bar to hit).  I think that it some of the unfun comes from some people feeling they have to choose either to build a deck designed to beat Flash/Ichorid or play Flash/Ichorid. 

Personally, I think that Merchant Scroll is the much lethal card that needs to restricted (if any card does).  This is the card that makes flash (or almost any deck that runs blue) absurd.  It is easily as good as Enlightened, Vampiric, and Mystic Tutor.  Just a little worse than Demonic Tutor in that you are limited to blue and you have to announce it.  One can easily snag a pact of negation or FoW to protect the combo, just get flash to make it happen.

On the plus side, I have been working on a new deck and my play testing partner runs Sliver Flash.  I do not run Leylines in the main because they don't really go with the deck and I am on a toss up right now between Leylines and Planar Void for the SB.  Irrespective of that, I am going about 50/50 against him (9 - 11 to be exact), but I do run 4 FoW, 4 Durress, and 4 Mana Drain.  As in the argument above, the turn 1 kill is not near as frequent as perceived, it is just that when it happens it feels like a kick in the face.

@Harlequin -- I definitely understand what you are saying about there not being wrong threats only wrong answers.  However, I don't think your argument really addresses the point I was trying to get at by a type of global hate.  Drawing cards is never bad....in fact, most decks try to get card advantage some how (even if it is just a one of with Ancestrall Recall).  The point is that if you can shut down drawing cards, then you have crippled an opponents strategy because you have crippled a core component of the game itself.

For instance, if you had Maralen of the Moonsong and a Shadow of Doubt imprinted on an Isochron Scepter (way too clunky, but for illustrative purposes), you would be completely able to shut your opponent out of the game.  No more draws, period.  This is a type of hate that doesn't care about which deck it plays against...it shuts down the whole mechanic of drawing cards.  Of course, this is a sliding scale.  A slight less clunky, but less of a lock would be Maralen with Mindscensor....severly cripples the opponent while letting you dig for answers, but it doesn't shut down the mechanic to the extent that one would want.  So hating out a mechanic vs. hating out a deck is sort of a sliding scale of hate.  (Note: This is meant as a theoretical point,  ) 

Chains isn't necessarily "bad" against workshop aggro.  It might be, depending on the deck it is in; however, it also can be perfectly viable as well -- it really depends on the overall design of the deck.  For instance, if one gets out Chains and a Puzzlebox against workshop aggro, stax, or MUD, then they are locked out of doing anything else in the game....

Hauntedechos -- just wanted to compliment your thread.  I am enjoying the theoretical level of it.

First, I would be glad to see some more discussion about an idea I mentioned in my first post.  Mainly, that we can have the "evolved" hate decks that decide to hate a key component of the game of magic.  I view this as "breaking" that aspect of the game in the opposite direction.  For instance, take the Gush draw engine -- this aims to break the basic rule of magic in such a manner as to exploit the normal restriction of drawing a card per turn (not to mention using mana to play for spells).  This is very good at breaking this restriction completly and going all out.  Now, on the other hand, one could imagine a deck that prevented players from drawing cards at all (like some old life gain/Zur's Weirding decks back before people knew better)-- this would be considered a hate deck, but it would be a hate deck built on removing a key component of the game, drawing cards.  One of my contentions is that it is easier to exploit a component of the game mechanics in a positive manner than to shut it down completely.  

This difficulty in shutting down a core component leads people to using decks that are designed to shut down a specific deck (or engine).  As the deck becomes more aligned with stopping a specific deck than it misses out on a larger amount of the metagame considerations.  Unfortunately, if the larger metagame is focused on, a hate deck can become more vulnerable to the "Tier 1" decks.  This is what makes the hate decks discussed above more of a Tier 2 type of deck for the reasons discussed in this thread.

Now to answer the questions:

1.  I select the decks I play because of player interaction.  I want the game to be one of interaction -- a hard fought battle where the outcome is teetering back and forth.  For myself, I don't forgo a deck just because it is too expensive to build.  Granted, I don't have a Black Lotus (or Workshops), but I have enough of the powerful cards that I can play around with most archetypes.  Some of the decks I just find more boring to play than anything.  For me, Flash is ridiculously boring to play--but I think it is okay to play against if the notorious first turn can be avoided).  The high-powered combo decks have some learning curve, but I think most people at TMD probably could master them given some practice.  

2.  Currently, I am working on a deck that abuses Chains of Mephistopheles.  I am drawn to it because I have liked the card since it came out in Legends.  I think it is absurdly powerful, but it takes a high-level of skill to play properly and to play around, especially in a Type I environment.  I also like it because it is off the beaten path a little.  I really like to try to create workable decks that are innovative and viable in the environment (unfortunately, I live in an area with a pretty piss poor Type I community).  I do have a friend that pilots Flash, and my deck runs about 50/50 with it as of now.  We sleeved up GAT last week and my deck went 2 - 2.  We are working through the metagame considerations, and when it is done I will post to TMD so I can have reliable expectations.

3.  Like hauntedechos, I really like Fish.  I favor U/W/B as well for various reasons.  I like player interaction (btw, I don't think that FoW and Duress count as player interaction) and think it is best if everyone is playing on more of an even playing field.  I like to see skill in reacting to an opponent -- and I think it shines in Type I if a lid can be put on some of the ridiculousness.

1st -- @hauntedechos -- congratulations on your promotion to full member status and what a nice article to provide for discussion in celebration.

To preface this post, I would like to say that it is what I consider my first real post, so if I stray to far from the spirit of TMD, I apologize in advance and just want to note that some of the minor disagreements that I mention below are intended to further the discussion.  I have only been a member for a short time, but I have been an active "reader" of TMD for quite some time, and I value the community as a whole; consequently, I tried to make the below post as well thought as possible.



To some extent, I disagree that control elements should be globally labelled as crutches (not a claim that you alone have made, by the way).  Whether it is a crutch or not really depends on its specific synergy with the rest of the deck's goals.  Sometimes decks are overly loaded with control elements as a form of denial; however, I would contend that control pieces can be important pieces of a forward advancing strategy.  I think that one could go too far in either direction -- over control or under control, with the former direction being more of a "crutch" and the latter just being reckless.  For me, the key is to find the balance between the two such that risk is minimized while leaving open viable avenues for winning.  Then, one has to weigh the comparative advantages for trading in/out control for speed.  Within this line of reasoning, I don't really think the "crawl" ---> "walk" ---> "run"  &  "reading articles" ----> "hate decks" ----> "tier one"  is really that simple either.  To be fair, I think it might hinge on the definition of "hate" deck.  If we limit the definition to building a deck specifically for a given meta, then I completely agree with almost everything you have stated.  On the other hand, I feel that we could expand on your discussion if we widened the definition of hate -- mainly because I think the restriction of the definition can limit development of new decks, which is difficult enough in and of itself.   

Insted of considering hate to be directed at a specific meta, I have always conceptualized a hate deck to be one that builds some win conditions within a hate framework, where the framework is built upon hating different core elements of the game itself (whether it be land, card drawing, creatures, etc.).  I suppose the simplest example would have to be the Fish ideal (I do endorse that label over the label of archtype).  As a concept, Fish provides a framework that is capable of hating almost any element of the game that another specific deck may try to exploit.  Now we can think of "hate" as a general type of deck and then the options for building an overall framework become more variable.  It could be argued that tier one decks that lock the opponent out of the ability to do anything (like Stax) are just more evolved hate decks -- hate decks that have evolved from focusing hate on one type of deck to a hate deck that "hates on" a core component of the game itself.  This is almost like global hate vs. local hate.  Within the global hate, the deck can build-in other win conditions.  Thus, I think hate is like anything else, different levels of complexity -- but still hate nonetheless.

I almost see this discussion as one of proactive vs. reactive decks.  Top tier decks are about as proactive as you can get, establishing the win conditions as soon as possible.  Reactive decks are built to a specific environment and aim at reacting to the key elements of the opponent's deck, whether that be true reaction (like a Swords to Plowshares) or preemptive reaction (like a Meddling Mage).  For me, some of the appeal of playing a reactive deck is forcing your opponent to play Top Tier decks that require multiple win conditions.  If your opponent is relying a single threat, and you hate it out then they deserve to be beaten by a 2/2 over 10 turns.  The difficult part is evolving that reactive deck into a deck with viable, multiple win conditions beyond the standard swarm of little guys for the slow beating.  I think that mastering this evolution and combination of winning with certain core hate elements can result in truly innovative decks that are quite powerful.  However, speculating about and designing such a deck are really two different things, so perhaps I am a little delusional.       

The editor's blurb definitely caught my attention.  I am a professor of statistics with a research emphasis on combinatorics (which is really the bread and butter of the math underlying magic).  Unfortunately, I am not a subscriber to SCG, but I would be immensely interested in hearing more about mathematical proof of luck enhancing skill.

Beyond recognizing card combinations and interactions, the central theme, in my opinion, of deck construction is probability theory.  A great deal of information can be gleaned from deck matchups by just understanding the probabilistic relationship between sets of cards that are contained in each deck and how they will interact and at what frequency.  Additionally, understanding these probabilities enables more fine tuned deck construction.  However, I would argue that skill in this vein is about minimizing luck and exploiting the fact that your opponent may know less about these probabilities.

At any rate, I am very interested in the "fleshed out" contention mentioned in the editorial blurb.

Hi TMD,

I am actually pleasantly pleased to recently have found this community; a mature, online magic community -- who would have guessed.  I have been playing magic since around 1996; however, due to a fairly fast track through college and graduate school, I have little time to attend tournaments.  I am hoping to rectify this at Gencon 2008.  I am currently a Professor in a tenure track position in Statistics, and I absolutely love the deck building aspect of MtG, with most of my love given to the underlying combinatorics problem that is the game.  My personal preference is to try to incorporate previously thought unworkable cards into a viable deck.

Best,
Doug