ClusterFuck

[Ric_Flair]

First, what is board position?  Give us a working definition.  No metaphors not allusions.  Define it.

Second, I thought that it was clear that theory in Magic, after a certain level of competency, is useless and testing should take over.

[PucktheCat]

Second, I thought that it was clear that theory in Magic, after a certain level of competency, is useless and testing should take over.

This statement seems wrong to me.  How can testing accomplish anything without theory to focus it?  What is theory but the results of testing generalized so they are most useful?

Theory is a way to econimize when testing is impossible, but that doesn't make it inferior.  Every time you make a decision in a game situation you haven't seen before you are applying some of the results of your testing in a genearlized or abstracted fashion: theory.  Every time you choose to try one card because it fits your decks strategy over another that doesn't you are again using theory.

Leo

[Matt]

EDIT: Aw, dammit, Puck kind of beat me to most of the points I'm making, and probably did a better job explaining them, too. Oh well.
I would say that you alternate periods of theory and testing to achieve maximum results in minimum time. Not that one gives way to the other ultimately, but that they give way to each other.

The truth, the thing that Kant found, was that theory supports reality.

I don't know much about Kant, but this is exactly what I was saying earlier when I said that CA and T are not esoteric and abstract concepts, but instead are highly connected to actual gameplay (testing).

They also know that cards are so well made today that shooting from the hip and using theory alone is pointless.

You're right to say that testing is necessary, but I couldn't disagree more that theory is pointless. Theory is still extremely valuable. If you disregard all theory as "pointless" you get the Magic equivalent of a cargo cult, and when you try to ape the known successes mindlessly (as you are suggesting) and something goes wrong, you're up shit creek without a paddle.

My quip that WINNING is important is not just an empty phrase.  It is the truth.

It is both true, AND an empty phrase.

Since we can no longer theorize and build great decks we have to playtest,

Maybe you can't. I see no reason why this would hold true for everyone. No, we can't do this using old, outdated theories, but that's what new theories are for.

Swap cards around, test them out, but remember that those cards that WIN are the cards that should be in your deck. And also remember that what wins differs from deck to deck.

This is exactly the uselessness of your statement. Here's sixty slots: what do you put in? If we were to go by your method, the answer would be "whatever wins! teehee!" and nothing would ever develop, because evolutions don't come in stages, they come in huge intuitive leaps, which come directly from theory. Pure empiricism -- testing everything -- would work if you were able to test every little interaction. But there isn't TIME for that, and thus we use theory to more immediately discard poor options and sift out the better ones. Is it foolproof? No, but it gets better and better, and after ten years the theories are pretty damn good.

If you just say that Bazaar nets you -2 cards you are doing a great disservice to the game. If those two cards are Ambassador Laquatas and Worldgorger, the minus two cards is inconsequential if you have the right stuff in play.

Did you not read anything I said? When you're discarding Dragons and Ambassadors, Bazaar is not card disadvantage.
Here's a thought: if Bazaar said "discard three cards, if these are named Worldgorger Dragon or Ambassador Laquatus, remove them from the game", would it be played in Dragon decks?
Moreover, would you have to test to know the answer?

[Ric_Flair]

Since we can no longer theorize and build great decks we have to playtest,

Maybe you can't. I see no reason why this would hold true for everyone. No, we can't do this using old, outdated theories, but that's what new theories are for.

Well, I will start here because this is the funniest quote Matt.  Maybe I can't.  Right, and neither can Rob Dougherty or Ben Rubin or Kibler.  The point I was trying to make and the point that the Pros have proved over and over and over again is that playtesting is the primary method of deckbuilding.  Remember the experiment at the beginning of the year on Star City with Rob and his B/r deck.  The point was that no one, or hardly anyone tested it.  The result was that people sat around and theorized and posted utter SHIT decks.  Rob scolded people for not actually building the deck.  The thing is that Magic cards are so complex now that mere theory is not enough.  Everyone has to playtest to be good.  The reason everyone has to playtest to be good is because at a certain level everyone is roughly equally intelligent and everyone else playtests.  It is the arms race phenomenon.  Look at Finkel v. Budde.  I have met both.  I have seen both play many many many times.  Finkel is smarter than Kai.  Kai ain't no dummy, but Finkel is just smarter.  But Kai kills him every time because at the end of the day Kai has put in the work, made up the difference between their intelligences (which is not that great) and far surpassed Finkel's theorizing about decks.  

New theories are funny.  Theorists postulate what cards are good and what are bad.  They pontificate and sometimes they are on the mark, sometimes they are not.  Then they make up new theories to explain why they were wrong.  What use is this?  Why make up a post hoc theory?  It helps in the future, but it doesn't make the theorist any better at the game.  The "new" theory is quickly found out and suddenly that microsecond of an edge is gone and the theorist is left in the dust by people that read and then do actual playtesting.  

The game is far beyond Mind Twist and Ancestral.  It is much more important to understand card interaction than to think up new theories.

This is exactly the uselessness of your statement. Here's sixty slots: what do you put in? If we were to go by your method, the answer would be "whatever wins! teehee!" and nothing would ever develop, because evolutions don't come in stages, they come in huge intuitive leaps, which come directly from theory. Pure empiricism -- testing everything -- would work if you were able to test every little interaction. But there isn't TIME for that, and thus we use theory to more immediately discard poor options and sift out the better ones. Is it foolproof? No, but it gets better and better, and after ten years the theories are pretty damn good.

The reason I chose Kant and the reason the metaphor works is because Kant realized that theory is necessary to understand reality.  One without the other is impossible.  But in the end it is reality that we are inquirying about, theory just helps us get there.  So to with Magic theory.  Of course blind playtesting would be retarded.  Theory whittles down the choices for decks, too much I would say, then playtesting takes over and fills in the rest.  Playtesting finds cards like Food Chain that defy theories yet break decks.  I believe that combo decks in general were unnaturally suppressed for so long because people clung to theory too much.  CA and Weissman's ideas dominated Magic.  Combo decks were dismissed as inconsistent and full of bad cards.  Combo Winter happened, in part, because Pros tested and found out that cards like High Tide, which was traditionally seen as a waste, were good and with enough tweaking combo becomes dangerously consistent.  

Knowing a basic amount of theory and reading the new stuff that comes out is sufficient to get a person to the point where they can work hard and figure out the rest through playtesting.  I would rather have intimate knowledge of my deck gained through hours of playtesting as opposed to knowing lots of new theory.  

Theory is the basic foundation of the game.  If you don't know basic theory then you can't play at the highest (higher) level(s).  It is just the admission ticket though.  It is not the main event.

[Matt]

The point was that no one, or hardly anyone tested it. The result was that people sat around and theorized and posted utter SHIT decks.

This doesn't necessarily show that theory is no good, because they very well could have just been theorizing wrongly. GOOD theory does not produce SHIT decks, and the starcitygames crowd, from what I've seen, are not exactly the most adept at wielding theoretical concepts.

Playtesting finds cards like Food Chain that defy theories yet break decks.

I'm sorry, but this is just wrong, and makes me question your understanding of what magic theory is. What theory or theories, exactly, do you think Food Chain defies? Old Weissman-style card advantage theory?

This is why I think it is a bad idea to try and look at this issue with a historical perspective, as you have done. You end up comparing old CA theories to new, more complete theories, and concluding that because the old card advantage theory isn't complete, that all theory is malfunctional. This is wrong.

[Ric_Flair]

Look if in all of Star City's readership, not one person found a good theoretical addition to the deck, you have to face facts that theory only gets you so far.  No one playing Magic is enough of a genius to win a PT with all theory and no playtesting.  Not Rob, Finkel, Kai, Kibler, Zvi, no one.  The game is beyond mere theory.

As far as keeping up to date on theory, theory seems to lose its value when every time a new deck comes out new theories need to be invented.  Theory works best and is most useful when it is a prognosticator, not an explicator.  

Also when theories are invented and reinvented after every new deck, there seems to be little difference between theory, which is supposed to be generalized and forward looking, and explanation, which is narrowly tailored and operating after the fact.  If after U/G the theory became tempo, then more narrowly "card quality" and then more narrowly...whatever... and so on, it cease to be genuine theory and starts to look like post hoc explanation.  That is helpful too, but it is neither theory, nor what makes good players successful.  Experience is the ultimate criteria.

[Fëanor]

just wanted to reply to the general thought brewing. first card advantage:

#2 Card Advantage:

The above statements about the main topic have been both mind-boggling and quite fun to read. I must say I have had an interesting journey with card advantage "theory". Ive only been playing since Tempest, but I used to read old Duelists. I remember an article in one of them, this was probably back in '94 or '95. Rosewater went on forever about the importance of keeping yourself a few cards over your opponent. Of course the theory was that if you have more cards you will get a chance to play more cards, cards that are probably good since youve put them in your deck, and ultimately win the game. Of course it was more true if anything back in the days of vinatage magic, Ice Age really brought some massive changes in game style and we're glad it did, because if it didn't, the game might have phased out. Mirage block comepletely changed the meaning of card advantage (CA) to closely what it is today. It is simply a way to identify the power of a card or set of cards based on how it will affect your speed as well as efficiency. It is still something to be considered.
Take a power card for example. Time walk was CA because you could use it to give you one more chance to draw a kill, or use a kill in play. Time Walk, technically doubled your CA.
Enchant creatures on the other hand, gives your creature a boost in some respect (well most enchantments do). Briar Shield will beef your beast, giving you more CA, via board control, but if the creature is killed, you lose card value, and your +2 CA is also lost. This may not affect the game in it's entirety, because maybe the player was looking for more graveyard cards, or maybe he already had enough board control for this -2 CA to not make any difference.
Which leads to card disadvantage (CD). Wild Mongrel is a creature that trades cards in hand for damage and life/board control. there is a CD when the cards leave the hand, but there is equal CA when the +1/+1 is added. Also, Bazzar is a very good card example as has been on this post. losing cards for a win is because of the style of the deck. In old vinatage, what player would have said,
"hey bazzar is good because i can put my yotian soldier in my graveyard, and then i can animate dead it for fun."
the game has changed so much, and more and more cards become synergetic with the bazzar, so  now its second nature to put cards youll actually do something with in your graveyard rather than keep the kill in your hand to be played. this CD is in effect a CA, the cards in the graveyard have been lost to the hand, but they still have power. We could talk on this tread about every single card and every single permutation and discuss what kind of effect it would have on gaining control etc. however, the point is, it always makes a difference. if you bazzar and discard dragon, or anger, then you have advantage. if you bazzar and discard stifle or swords, then unless you have a way to get them back (Yawgmoth's will) you have come to a disadvantage. but this in effect has many more variables you havent considered. what about the cards youve drawn. it seems natural to discard a swords if you end up choosing between that and an instant kill card you found with the bazzar. so CA and CD, while they can ultimately mean control and advantage in general, there are just too many variables to make a general statement like,
"CA is good, and should be considered when making a deck"
it should be more like,
"CA is most of the time a good rule to follow in keeper and should be considered when making that kind of deck under a certain type of meta, on a certain budget."
Or if the stement proves very little, then it shouldnt be said at all.

This is precisely why magic players post their decks and help each other out on places like TMD (including TMD ). some players see the same decks in different ways based on their experience with (since we're talking about CA/CD) CA and CD. people might say to take out Lions Eye Diamond because it nets you horrible CD, and some others would say to use this CD to your advantage (hence madness), teehee.

hope all of that makes sense, sometimes i make very little sense

peace

[Matt]

Look if in all of Star City's readership, not one person found a good theoretical addition to the deck, you have to face facts that theory only gets you so far.

I did not follow the actual incident, so I'll take your word on this. Still, I don't see how it shows that theorization itself is flawed, and not just that no good theorist felt like contributing.

Moreover, why should the results of that contest surprise you in any way? You said it yourself: in the arms race of magic tech, anyone who DID find a good B/r deck would have kept it to themselves. It seems - to me, at least - far more likely that the 'arms race' nature of the game is responsible for the failure of Rob's experiment than any inherent failure of theorization.

In fact, if I was a pro player and a portion of my paycheck was on the line, I might very well disseminate false theory and flawed logic. But perhaps that's just me.

No one playing Magic is enough of a genius to win a PT with all theory and no playtesting. Not Rob, Finkel, Kai, Kibler, Zvi, no one.  The game is beyond mere theory.

No one playing Magic has enough time and resources to win a PT with all playtesting and no theory. Not Rob, Finkel, Kai, Kibler, Zvi, no one.  The game is beyond mere playtesting.

Both of these statements are true, but also neither of them helps in any way. You seem to be trying to set me up as someone who thinks that theory is everything, and if you'd paid any attention to what I've written, you'd know that couldn't be further from the truth.

As far as keeping up to date on theory, theory seems to lose its value when every time a new deck comes out new theories need to be invented.

Aha, we see the crux of the misunderstanding. You seem to think that theories are constantly being torn down and new ones put up in their place.

This is not so.

Magic theory has worked by evolution, not revolution (a tired phrase, but applicable nonetheless). And if you'll notice, the new theories don't come out after every new deck. There are only very rarely new additions to the house of Magic theory, and they come less and less frequently. This is because the theories (which may be thought of as fragments of one central theory) closer and closer approximate reality, which you (rightly) put so much stress on.

If after U/G the theory became tempo, then more narrowly "card quality" and then more narrowly...whatever... and so on, it cease to be genuine theory and starts to look like post hoc explanation.  That is helpful too, but it is neither theory, nor what makes good players successful.

I don't rightly know what you mean this, specifically the "after U/G [madness, I assume] the theory became tempo."

Would it be helpful to discriminate between being a good player and being good at Magic? Routinely, the best players are NOT the best deck designers. You seem to be stressing the "playskill" half, while I'm concentrating more on the "deckbuilding" half. And neither half is useful without the other.

[Carlos El Salvador]

First off, board position is, "Of the things on the board and in your hand, how do things currently look in the game."  It is THERORIZING your current amount of resources (used up and gained) and your oppoents.  Control decks look at board position, as it is how control wins.  Aggro decks focus on building board position fast to push people into corner.  That is about how simple board position is.

Engouth about therorizing VS Decktesting debate.  Both are needed to attain the perfect deck (as testing ALL options is truely, impossable.  Lets look at a card that is excellent and terrible:  Force of Will.  What makes Force good?  Of course the ability to counter 0, first, and second turn plays.  What makes the force bad?  A tradeoff 2-1 and a life.  Now, without force of will, T1 would never be anything more than combo.  So, in order to stop some major decks int he format, your 2-1 tradeoff is well worth it, as it gives you board position to build on.

[Smmenen]

Look if in all of Star City's readership, not one person found a good theoretical addition to the deck, you have to face facts that theory only gets you so far.  No one playing Magic is enough of a genius to win a PT with all theory and no playtesting.  Not Rob, Finkel, Kai, Kibler, Zvi, no one.  The game is beyond mere theory.

As far as keeping up to date on theory, theory seems to lose its value when every time a new deck comes out new theories need to be invented.  Theory works best and is most useful when it is a prognosticator, not an explicator.  

Also when theories are invented and reinvented after every new deck, there seems to be little difference between theory, which is supposed to be generalized and forward looking, and explanation, which is narrowly tailored and operating after the fact.  If after U/G the theory became tempo, then more narrowly "card quality" and then more narrowly...whatever... and so on, it cease to be genuine theory and starts to look like post hoc explanation.  That is helpful too, but it is neither theory, nor what makes good players successful.  Experience is the ultimate criteria.

I completely agree with everything said here.  But I would go even further, theory helps explain what happened - it helps us understand WHY - but it beyond that, we have to understand the dynamics of every match and every deck itself.  That doesn't mean we can't apply principles - but understanding a match and testing must come FIRST.  Then we can see how principles might apply instead of applying them blindly or relentlessly.

Steve

[BreathWeapon]

I took the time to read all of the installments in your Long.dec and the Winding Rode article. I have to say its probably the single best article i've ever seen written on Magic, easily Top 5.

One comment I do disagree with you on tho', is your assumption that the most powerful deck in the environment is the deck every one should play. This simply isn't an absolute truth, and you seem to "sermonize" it in your article. The most powerful deck in any given meta is always relative, IMO. Dragon was atrocious 4 months ago, many considered it the most powerful deck in the format, but where is it now? To illustrate, once people come to a unanimous decision, spoken or unspoken, on what "The Most Powerful Deck" is, people over load hate in their SB and even their MD to snuff the life out of it, I.E. Dragon. I would rather say, "Play the most powerful deck in the environment that isn't directly meta'd against." If that isn't the definition of Hulk I don't know what is.

At any rate, I think you should certainly expand your reasoning on the matter and take into consideration, if you haven't already, that the comment is much more relative than you alluded to in your article. Perhaps the subject is worth an individual editorial all by its lonesome?

Thank god people have something to read other than Oscar Tan's rantings.

[Ric_Flair]

@ Matt:

Magic theory has worked by evolution, not revolution (a tired phrase, but applicable nonetheless). And if you'll notice, the new theories don't come out after every new deck. There are only very rarely new additions to the house of Magic theory, and they come less and less frequently. This is because the theories (which may be thought of as fragments of one central theory) closer and closer approximate reality, which you (rightly) put so much stress on.

I am relatively familar with the way theory, in the abstract sense, evolves.  Thomas Kuhn's famous books, Structures of Scientific Revolutions, outlines the basics.  In the book he says that there first is an experience.  In order to understand the experience, people make theories.  The theories last for a while until things that happen in reality do not comport to the theory.  Either the theory slowly morphs over time into something entirely new, or there is a period of eruption in which the scientific paradigm shifts to an entirely new theory.  This is where that new age business term comes from.  The book also explains that theories, in order to be valuable must be useful.  And for a theory to be useful it must have some ability to predict future outcomes.  This is my first beef with so called Magic theory.  Theories such as CA or even virtual CA or other off shoots like card quality, are too explanatory.  They have no predictive powers, or their predictive powers are limited to common sense observations.  For example, common sense, that is a basic intuition present in the minds of all of the relevant, somewhat knowledgeable participants in the game of Magic, will tell you that the more cards you draw the more resources and options you have and the better you will do in a given game.  This is not something so sophisticated as a theory.  If this is a theory then "go fast" is a theory for car racing.  CA in is most solid, well conceived form has gone from theory to simply basic knowledge.  This happens too in Kuhn's model.  Eventually over time things that were theories, like the Earth being round, just move into the realm of fact.  Other theories like Carlos' "board position" theory lack the predictive capacity to rightly be called a theory.  

The second beef I have with theory is that often times it is really useless when it comes down to the game itself.  The idea of "who is the beatdown" is a good concept, but in the game, resources are randomly distributed and only through playtesting will you understand what role your deck plays.  In this sense theory is flawed because it provides imperfect data.  Again back to the racing analogy: theory about who is the beatdown is like saying don't wreck is a theory in car racing.  It is not a predictive engine so much as a statement of direction.

@Carlos

board position is, "Of the things on the board and in your hand, how do things currently look in the game." It is THERORIZING your current amount of resources (used up and gained) and your oppoents. Control decks look at board position, as it is how control wins. Aggro decks focus on building board position fast to push people into corner. That is about how simple board position is.

So here is my problem with board position as you define it:  it is meaningless.  At the point of deckbuilding, trying to maximize "board position" is no different than having a game plan and using cards to make sure it happens.  And this, in my mind, is simply called deckbuilding.  If you slap together cards without understanding how the deck is supposed to win and how to make sure it does you are not deck building.  At the point of playing "board position" is also an empty concept.  If a Stasis player has the lock in place he has, according to your "theory" good board position.  But if he has protection in hand of the lock he has "better" "board position."  But if this is true what is the difference between "board position" and winning?  Until you can show me how "board position" contributes something relevant and yet to be analyzed to the game, it strikes me as a meaningless and useless concept.

@ Breath Weapon

I think that what Steve means by best deck is that if you assume a fully powered meta, equal skill, card access, and familiarity with the top tier decks, which deck would a good player choose?  I think that all of these assumptions are fair to make.  A fully powered, or nearly fully powered meta, is a good default because a) it is the most common metagame and b) it has the highest power level, meaning that if you can win here you can win anywhere.  Similarly, card access and play skill should be assumed to be the highest possible when analyzing the format because there is no other way to make generalizations about the game.  So given these parameters I think it is fair to say that sometimes there is a best deck.  GAT, pre Gush restriction, was pretty clearly the best deck.  Then Tog was the best deck, though not as decisively.  After Tog I think Steve made a good case that Long was the best deck.  They best deck may not win every event because of anomalies in play skill or card access (especially if the decks include tier 1 and tier 2 power like MUD, which is nearly impossible to build without the p9 even in 5 proxy events).  Also random bad luck and matchups play a role.  But the statement is that all things being equal there is sometimes a deck that is the best deck.  It is not necessary that there be a best deck, but sometimes it happens.    So given that this does occur, the question is why NOT play the best deck?  I can't think of a good rational reason.  Playing Keeper all the time can make you better with that deck, but if the deck itself is weak, lots of skill can only make up for so much.  When you face an opponent with equal skill and a better deck you will lose.  I guess the issue is if you want to win, do the most you can to offset luck and skill difference by playing with that deck that has the most raw power.

[BreathWeapon]

Well, my major point is: even if there is such a thing as the best deck in T1, doesn't the Bull's Eye on its head and the influx of hate to deal with said best deck make it the inferior deck to play?

This postulate was my reasoning behind Long doing soo poor in the meta when it was supposed to be "The Best Deck." Once you categorize a deck as being the best, the amount of attention it gets after that assertion may very well cause it to completely bomb out in the meta.

[Ric_Flair]

@ Breath Weapon:

There are two issues here:  1) deck v. anti-deck and 2) Long's "weakness" in the metagame.  

Issue 1

This is an ancient debate in Magic, but I think that there is enough data out there to support the conclusion that mere hate decks, as opposed to decks with hate in them, cannot win.  The foremost example of this is the PT: Rome or PT: Academy decks.  Tommi Hovi, perhaps the greatest designer of foil decks in the history of the game, devised a tool to beat a format full of Academy combo decks.  His creation bears a striking resemblence to what we call Fish.  Hovi placed second in the event, ultimately losing to the very deck his deck was designed to beat.  Similar fates befell those that tried to foil Necro during Necro Summer.  Dennis Bentley ultimately powered through 3 other decks, each designed solely to beat Necro, on his way to the National Championships.  In the end raw power can just win sometimes.  Hate, on the other hand, has no "lay down hand."  Hate is always reactive.  It plays its best game when the best deck is working as it should.  This is a dangerous proposition.  As has been said many times before:  There are no bad questions, only bad answers.  This paraphrase of the famous Dave Price quote is the truth.  In Magic there are no bad threats, only weaker ones.  Any threat, unchecked can win.  

Another piece of information bears this point out from another angle.  Hovi got second at a major event with a hate deck.  This, it would seem, should indicate success.  But it didn't.  No one else came close to repeating his amazing performance that season.  The reason:  Hovi had skill that let him win.  Hate alone is never enough.  Hovi is (was) such a good player that even if he is playing a slightly inferior deck his skill can overwhelm his opponent and let him win.  Unless you are that skilled, I think the hate route is just too dangerous and inherently weak.  

Again, the question remains:  why NOT play the best deck?  If all things are equal and you and your opponent have equal luck (which, by definition you must, if seen from a generalized perspective), equal skill, and equal access to cards, but you have a weaker deck than he does, he wins.  Why necessarily weaken yourself, when, instead you can just play the best deck?

Issue 2

I first saw Lon being played by a local Vintage player who, while not Kai, was a good player.  He had a good grasp on the deck, but in the end he did not win as much as others.  I then saw another player, one that was a great player play the deck and this made all the difference.  For the great player Long was like a puzzle and the pieces fell in order in a certain way and then he won.  Long's biggest enemy is itself, not the opponent.  The deck is HARD TO PLAY.  This is one reason why it did not win all the time.  Secondly, the deck is expensive to make.  It requires ALL of the Power and a number of other hard to find cards.  And unlike other decks, where missing a card here or there is not a big deal, Long is CRIPPLED without each and every card.  So that is why Long did not win all the time.  If the skill was there and the cards were there though, it was a fucking MONSTER.

[cssamerican]

I have to agree with Ric Flair (Cause I want to be Styling and Profiling Woooooooo!) Long was such a monster if the right person was playing it, it could combo through hate like no other deck. And if it wasn't for the skill and the cards required to play it, it would have been just as rampant as 4 gush GAT.
My opinion is if there is a clear best deck like Long, 4 gush GAT or BBS, playing hate might win a few rounds, but in the end the best deck wins the tournament. The reason why no deck (Including Dragon) is not dominating tournanments is because there is not a clear best deck. Instead right now there are several very good decks, and the best is dependent on the metagame that shows up at the tournament. Whenever a deck becomes the clear best deck in the format, DCI will hit it over the head with a Hammer. That is part of their job, and they are very good at it.

[nietzreznor]

"Theory without playtesting is empty, playtesting without theory is blind."

       I think this spin on Kant's famous quote (since we seem to be using many philosophical references) sums up the relationship between CA/tempo theory and playtesting.   We can't seriously think that we could playtest without any theory whatsoever.  How would we ever get started?  Without basic concepts like CA and tempo (BOTH of which are important), we wouldn't be able to play test.  Hell, even if we had decks given to us to test, how could we play the individual cards without a bascis theory on how to win?
    But, as the quote suggests, all of this is only useful in the context of actual decks and actual cards.  Playtesting ccan help us determine whether or not a tempo boost is worth the CA disadvantage that comes with it.  Maybe this differs from match to match.

   The point is, I'm not sure there is really anything such as "defying theory," or "useless theory," or anything like that.  We might have some cards in our deck that are there with seemingly no theoretical justification, just because "they win."  But that does not mean that theory cannot explain them.  An average player may do just this: he may net-deck, then use cards only because they seem to do well, without any idea WHY they work so well.  A great player, I would think, would always be able to analyze such a situation, and understand the theoreticla reasosn for the success of a given card/deck.  He can do this, because the actual playing/testing, and the theory, aren't foreign concepts.  They are related--we need them both to even begin playing.

[BreathWeapon]

That was a well spoken reply to the question at hand Ric.

One thing I would like to clear up tho', is i'm not talking about a "Hate Deck," but a general rise in the amount of hate in SB's for specific archetypes. For instance, the 3 Coffin Purge that allowed Hulk to dominate Rector at Gencon last year or the massive influx of Stifle (Pseudo Hate really) in Keeper and Fish etc that seems to be restricting Dragon/Rector from performing in the format. The same can be said for Null Rod and Damping Matrix and my personal favorite, Blood Moon. These are just bullets being used as compliments to already strong decks, and they ultimately perform better with them vs the field, IMO.

[Saucemaster]

I am relatively familar with the way theory, in the abstract sense, evolves.  Thomas Kuhn's famous books, Structures of Scientific Revolutions, outlines the basics.  In the book he says that there first is an experience.  In order to understand the experience, people make theories.  The theories last for a while until things that happen in reality do not comport to the theory.  Either the theory slowly morphs over time into something entirely new, or there is a period of eruption in which the scientific paradigm shifts to an entirely new theory.

Here's the thing about this: does no one else see a fundamental difference between theory meant to describe a game and theory meant to describe "reality"?  Because there is a fundamental difference: it is debatable whether or not "reality" is composed of a set of rules, logical statements, etc., but it is NOT debatable that a GAME of the sort of Chess, Magic, etc is actually composed of such a set of rules.  A game is NOTHING BUT a set of rules and logical statements.  It is CONSISTS of such.  To that end it should be entirely, wholly describable by theory.  It should be solvable.  Because this game we play is incredibly complex, such a complete theory of M:TG is nothing like being at hand, and certainly both CA and Tempo are simply sputtering first efforts at solving it, but it is solvable.

None of that precludes a Kuhnian analysis of the development of theory in Magic or any other such game (Chess theory can actually be fairly fruitfully analyzed in Kuhnian terms), but the point is that the object under scrutiny is of a completely different character than the ground-level "reality" that Kant was attempting to deal with, or that scientific theories purport to explain, etc.

[zero]

I like to think of it in these terms : if it's just a game composed of rules and logical statements, then it should be a trivial task to program it. Turn it into an "A.I". Does anyone think it would approach the level of chess based programs, the weakest of which would kick my ass anyday of the week?
     One of my other hobbies is to play a strategic level WWII wargame (World in Flames, if your curious). It is in essence, a set of rules and the equipment (maps and counters) needed to use them, similar to Magic. There will not be, in my lifetime, any way for a computer to play that game at the level of a human opponent. There is way too much randomness for that, too many permutations. It is not solvable. Maybe you could give it goals ( '41 barbarossa? build armor and mech; '41 gibraltar? build airplanes and artillery), and maybe parts of the game rely on logic (the attritional naval system, how much to contend air to air combat when in an inferior position), but games like this and magic rely on intuition and experience as well, to combat the randomness that comes from dice rolling and card shuffling.
     I may be wrong, but wasn't the original design theory behind magic combinatorial mathematics? Certainly sounds solvable. Chess can be brute forced through running a set of finite permutations. In magic, you can statistically estimate the chances of a person having a mana drain in hand, you can't come up with a statistic on how likely he is to use it on a spell you just cast. What if you don't mana drain that savannah lion, and lose the game to it because you don't end up being able to deal with it? The correct call may seem to be let it go, and try to find StP, but if you lost the game it obviously wasn't the correct call. So is it the correct call 90% of the time? 85%? How would you detemine this, given that you posit that it should be entirely describable in theory?

[Ric_Flair]

@ Saucemaster

Because there is a fundamental difference: it is debatable whether or not "reality" is composed of a set of rules, logical statements, etc., but it is NOT debatable that a GAME of the sort of Chess, Magic, etc is actually composed of such a set of rules. A game is NOTHING BUT a set of rules and logical statements. It is CONSISTS of such. To that end it should be entirely, wholly describable by theory. It should be solvable. Because this game we play is incredibly complex, such a complete theory of M:TG is nothing like being at hand, and certainly both CA and Tempo are simply sputtering first efforts at solving it, but it is solvable.

Thanks for the reality (or unreality) check.  I am surprised that I never saw this point of view.  For a long time I have been thinking about/working on the concept of a game in general and the notion of solvable v. unsolvable games.  I know that this will seem tangential to some people, but I really believe that it is not, just bear with me.  

There are some games out there that include no elements of randomness, like Chess or Tic Tac Toe or Connect Four, games that are entirely random, and games that have elements of randomness mitigated by other elements.  After taking a negotiations/game theory class I learned that the two kinds of games are different because one has "perfect" information and other has imperfect information.  I am not an expert on this by any means and most of my thinking comes from a philosophical/logic background, but here is what I have determined.  Games that are are solvable and games that are unsolvable are, for the most part, equally fun/entertaining and so on.  However, on an abstract level or a very advanced specific level, games like Chess are "broken."  Eventually they will make a computer that can complete and totally "solve" chess.  It is just a matter of time.  Chess, with its myriad of moves, is infinitely more complex than a game like Tic Tac Toe, but because the move set is not infinite, Chess can ultimately be figured out.  Tic Tac Toe is a retarded game because if two people know the trick no one can win.  Chess is more sophisticated, but eventually we will have two supercomputers staring each other down, stuck without a move and the game will be like Tic Tac Toe.  As I said before, in general between two human opponents this is not an issue.  But when making theories about such games, such theories are ultimately "completeable."  That is, the theory can isomorphically overlay the game and explain EVERYTHING.  It is just a matter of understanding the game and "solving" it.  Programming these chess computers amounts to making a theory of the game and once an unbeatable computer is made, the theory of Chess will be complete.  

On the other extreme are games like War, where there is no skill whatsoever.  These have no theory, because you can't predict anything or do anything to change your fate.  However in games where the element of luck is mitigable, like Poker, where betting and checking and folding minimize the impact of luck, or Magic, where preparation can minimize luck, I think that theories can get close to being complete, but can never be isomorphically explanatory/predictive like they are/will be in Chess.  The open element of random luck makes it impossible.  So in a way, Magic, while having rules, is not "solvable" and thus cannot be entirely explained by theory.  Magic is an open set, whereas Chess is a closed set.  Chess is a large closed set, but ultimately it is a closed set.  

So back to the topic.  I think that theory in Magic is useful.  It is a necessary precondition to deck construction.  But ultimately it is a secondary concern.  I can understand all of the negative imports of Kaervek's Spite, but if I consistently win in a given metagame over and over again with Kaervek's Spite eventually I have to realize that traditional theories could not be used to predict this success.  This is because of the random element in the game.  Eventually, yes so theory can come BACK and explain why Spite was not awful, but that is ex post facto reasoning and as such is nothing special.  A deck or card that does not comport to an existing theory does not mean that that deck or card is useless.  Only playtesting can deem a deck or card entirely useless.  And I think that this is the point.  Vintage has evolved in the past year because people like Oscar Tan, with his obsession with abstract thinking about the game, have fallen from grace and playtesters have taken their place.  Oscar played on App, playing old decks, and constantly propounding the unimpeachable superiority of Keeper.  This sort of thinking dominated the format for a long time and led to stagnation.  This is why theory is dangerous.  Playtesting can never blind you.  It always gives you useful information.  The information may eventually be marginally useful, but it is still useful.  Theory, on the other hand, can blind you to whole regions of possibility.  

Finally I think that Magic theory is so awful in comparison to traditional theory because first, Magic theorists are no where near the caliber of scientific or scholarly theorists.  Most of us do this as a hobby and no one is a professional theorist.  Second, Magic theory is so awful because the people that do this for a living, the pros, rarely share good predicitive theory with us.

[Saucemaster]

Thanks for the reality (or unreality) check.  I am surprised that I never saw this point of view.  For a long time I have been thinking about/working on the concept of a game in general and the notion of solvable v. unsolvable games.  I know that this will seem tangential to some people, but I really believe that it is not, just bear with me.  

Well I love this kind of discussion.  I hope if Steve thinks we're hijacking his thread, he'll let us know.

There are some games out there that include no elements of randomness, like Chess or Tic Tac Toe or Connect Four, games that are entirely random, and games that have elements of randomness mitigated by other elements.  After taking a negotiations/game theory class I learned that the two kinds of games are different because one has "perfect" information and other has imperfect information.  I am not an expert on this by any means and most of my thinking comes from a philosophical/logic background, but here is what I have determined.  Games that are are solvable and games that are unsolvable are, for the most part, equally fun/entertaining and so on.  However, on an abstract level or a very advanced specific level, games like Chess are "broken."  Eventually they will make a computer that can complete and totally "solve" chess.... But when making theories about such games, such theories are ultimately "completeable."  That is, the theory can isomorphically overlay the game and explain EVERYTHING.  It is just a matter of understanding the game and "solving" it.  Programming these chess computers amounts to making a theory of the game and once an unbeatable computer is made, the theory of Chess will be complete.

We're in total agreement here, then.  What I believe I disagree with (I haven't thought about it a tremendous amount yet either, and I too am simply approaching it from a logical/philosophical perspective) is what follows:

However in games where the element of luck is mitigable, like Poker, where betting and checking and folding minimize the impact of luck, or Magic, where preparation can minimize luck, I think that theories can get close to being complete, but can never be isomorphically explanatory/predictive like they are/will be in Chess.  The open element of random luck makes it impossible.

I suspect we may be at cross purposes here and find out that we're stressing slightly different aspects of the same ideas, and that we really agree after all, but here's how I see it.

There are two ways in which Magic is fundamentally different from something we all agree is solvable, like Chess.  One is that it incorporates a certain degree of randomness.  The other is that it is a game of incomplete information.  A game with randomness but complete information would be Risk, for example; a game with incomplete information and no randomness is harder to think of.  I actually can't think of one, which interests me, but is beside the point right now.

Let's take the "Tic Tac Toe" of pseudo-random, incomplete information games: Blackjack.  The rules of which Blackjack is composed are incredibly simple, though put enough money on the fall of the cards and it will be entertaining all day long.  Especially when you're winning.  Trust me.  Anyway, the point is, assume for a second that you're not counting cards (the illustration works even if you are, but that will needlessly complicate the process, as you come closer to complete information).  Blackjack has nevertheless been solved: there is a set of basic strategy rules (and they're easy, and everyone should know them before they go to Vegas) that optimize your chances to win, and they are completely known.  Some of them are counterintuitive for most people (if you have a 16 and the dealer is showing a 10, for example, you hit every time), but they're proven and they work.  The real difference between a game with an element of randomness and a game without is that the correct strategy in any given situation will not necessarily lead to a win; it just gives you the best chance at a win.  Nevertheless, I don't see how that changes the status of the theory describing the game; the game has still been solved, it's just that solving the game doesn't result in a wholly determinable outcome, just certain statistical probabilities.

Now, given that simple games of incomplete information and mitigated randomness can be solved, there doesn't seem to me to be any immediately apparent reason why a much more complex game with such properties can't also be solved.  Magic is tremendously, absurdly more complex than Blackjack, or Poker, etc., but it should nevertheless be solvable.  Unless--and this is kind of interesting, but I don't have nearly the math background for this kind of speculation--you think there's a complexity threshold for games of this sort, and beyond that threshold it ceases being fully describable in logical/analytical terms.

The "incomplete information" part of the game is actually more interesting, I think, since depending on a player's acumen it may actually end up being correct to deviate from the theoretical "best play" (see Poker here).  Usually, however, this is because some quirk of your opponent's behavior has actually given you some extra piece of information, so the situation has actually changed somewhat.  That, or you want to disguise your hand and mislead your opponent.  Anyway, in the end the theory should be able to encompass the incomplete information aspect of the game as well.

So back to the topic.

Amen!

I think that theory in Magic is useful.  It is a necessary precondition to deck construction.  But ultimately it is a secondary concern.  I can understand all of the negative imports of Kaervek's Spite, but if I consistently win in a given metagame over and over again with Kaervek's Spite eventually I have to realize that traditional theories could not be used to predict this success.  This is because of the random element in the game.  Eventually, yes so theory can come BACK and explain why Spite was not awful, but that is ex post facto reasoning and as such is nothing special.  A deck or card that does not comport to an existing theory does not mean that that deck or card is useless.  Only playtesting can deem a deck or card entirely useless.  And I think that this is the point.  This is why theory is dangerous.  Playtesting can never blind you.  It always gives you useful information.  The information may eventually be marginally useful, but it is still useful.  Theory, on the other hand, can blind you to whole regions of possibility.  

On this point, I think we are in *almost* full agreement.  After all, I'm on a team which is almost solely dedicated to playtesting.  I think that the reason that playtesting is so valuable is actually pretty simple, and as a fellow Heidegger lover you may end up agreeing with me: humans are creatures who are able to form incredibly thorough, detailed, and accurate "knowledge" of something simply through repeated experience.  Not "knowledge" in the form of a set of unconsciously applied rules and logical statements, but a certain basic altering of the way we see possibilities, the import and meaning of certain observations, which observations we privilege over other observations, etc, which manifests itself (in those who are highly skilled and have alot of experience) in an *intuitive* grasp of the game.  It's not the same thing as internalizing theory, and it pretty much always results in a significantly better *player* than someone who is trying to internalize theory.

Playtesting does exactly that.  It gives you a "feel" for a certain matchup, and an intuitive grasp of the possibilities of the game, etc. that only someone with an incredible analytical mind exploring a fully-solved theory of the game in incredible detail would ever come to.  And since we have no such complete theory of Magic, and probably (no offense to anyone) no such incredible minds--in any case, someone with a mind of that caliber probably has some sort of social duty to use it for something that will actually, you know, help someone somewhere--anyway, since we don't have such an entity, and since in any case playtesting is just *quicker*, playtesting is the way to go.

But theory does have its place, and like it or not is still guides card selection and such to a huge degree simply because there is not time to consider and test every card.  Theory still guides us to a large degree, and *to that end*--since we'll always be dependent to some degree on theory--it serves our interests to develop that theory until it is as broad and nuanced as possible, so as not to let it blind us to other possibilities in the future.

My much-more-than $0.02.

Finally I think that Magic theory is so awful in comparison to traditional theory because first, Magic theorists are no where near the caliber of scientific or scholarly theorists.  Most of us do this as a hobby and no one is a professional theorist.  Second, Magic theory is so awful because the people that do this for a living, the pros, rarely share good predicitive theory with us.

Again, amen.

[Jacob Orlove]

a game with incomplete information and no randomness is harder to think of. I actually can't think of one, which interests me, but is beside the point right now.

You can play bridge variants where you set up the hands for all players before the game starts--sometimes for more than one game. There's no randomness, but there is incomplete information.

[Gothmog]

While I follow Steve's chain of logic on playing the best deck, that logic misses an extraordinarily crucial element that frankly is why Type I is so interesting; the cards are really good.

We will never evolve the metagame to the point Steve is looking for because of 2 main factors:

1) The game changes too quickly and there are too many new interactions (too many new cards are printed).

With new sets being printed at the rate they are, Type I is constantly evolving.  Does anyone truly feel like they've explored all of Mirroden as it relates to Type I?  Yet Darksteel is here, and while I imagine it won't bring as many cards of interest to Type I as Mirroden did, like most sets it is bringing cards certainly worth evaluation.  

WoC also seems to have taken a heightened interest in Type I printing more cards relevent to the format.  We can discuss whether or not Trinisphere will be good in Type I but clearly the format it has the most obvious ability to affect is our format.  I can't think of many Invasion or Odyssey era cards that was true for.

Lastly, each new set makes you weigh that sets new cards and their interactions with all the previously printed cards.  Obviously, this becomes a more complex task (and a more interesting one) as more sets are printed.

2) So many cards are available, and they are so good (or so swingy as R&D would say) that decks many would consider "suboptimal" really are very good and can win matches from even the top decks.

Even if the best deck can be identified at a current time in Type I, the fact that so many powerful cards are available means that decks won't be nearly as dominant in Type I as in other formats.  In other formats the interactions of that format's keyword abilities (ie Madness) can be so much better than the other cards printed a best deck or a couple of best decks clearly emerge and dominate.

In Type I, if a great deck like Long emerges and threatens to dominate, the Restricted List will be amended.  In November, when everyone was saying that there was no reason to play anything but combo and that Type I would be a combo-only format because so form of critical mass had been achieved, people ignored one of the most fundamental rules of Type I, there is a restricted list.  A critical mass of cards was achieved years ago for combo to dominate, frankly that's the major reason a restricted list exists, is to keep combo from dominating.  There's no reason to believe that the restricted list won't serve this function in the future.

That's my case for my opinion that Type I will never be as well defined a metagame as Smmemen believes is possible.  As far as I'm concerned this is a wonderful thing.  Type I is a deckbuilders paradise, able to be explored while using the current metagame as a backdrop for developing new deck ideas and having that metagame be ever evolving.

Sure, its a good thing to try to find the best deck.  My main point here is to understand that trying to do so is the fun of the game, and rarely if ever, will that be successful for any length of time because of the complexity and raw power inherent to the format.  If you do find the best deck, enjoy it for a few weeks until the entire metagame is redefined by new restrictions, a new set release, or someone finding some interactions amongst older cards we hadn't considered.

[Ric_Flair]

@GothMoth:

2) So many cards are available, and they are so good (or so swingy as R&D would say) that decks many would consider "suboptimal" really are very good and can win matches from even the top decks.

This comment is just patently untrue.  While there are currently 6133 legal cards in Magic, see www.mtgpics.com, there are only a few hand of cards, relatively speaking that make up the de facto cardpool in Vintage.  De facto cardpool is a loose concept that I have been using to mean those cards that are legitimately useful in Vintage in light of the incredibly powerful metagame.  The reality is, and I have explained this elsewhere before, that the comparative power level of certain cards like, for example, Force of Will and Mana Drain, make it such that few if any modern counterspell type cards will ever be useful.  This arms race of power occurs in all spell types, permanents, and spell functions.  Lightning Bolt is the top of the heap direct damage spell, and so on through the cardpool.  Thus, as Dr. Sylvan's statistically analysis has borne out, there are really very few cards in the Vintage cardpool.  That is, the de facto cardpool is very shallow.  

This ties into your first argument, stated thusly:

1) The game changes too quickly and there are too many new interactions (too many new cards are printed).

With new sets being printed at the rate they are, Type I is constantly evolving. Does anyone truly feel like they've explored all of Mirroden as it relates to Type I? Yet Darksteel is here, and while I imagine it won't bring as many cards of interest to Type I as Mirroden did, like most sets it is bringing cards certainly worth evaluation.

As I pointed out above the number of cards that are printed and the number of cards that are ACTUAL USED IN VINTAGE is very different.  Suppose the most liberal estimates of de facto card pool size are true and there are 400 cards that are regularly used, this still equates with only 6.5% of the games cardpool.  I would venture a guess and say that this is roughly the number of Vintage viable cards we get a set, with larger sets having slightly more, and smaller sets, especially middle block sets, having less.  Even with the conscious focus of attention on Vintage I still think we are getting only a small handful of useful cards per set.  So again I don't think there is that much of a change.  Not like Block or T2 where whole Blocks rotate out resetting the power level of the format each time, thus changing the minimum threshold of tournament viable cards.  

I think we can get close to better information of the caliber that Steve references if we PLAYTEST more.  It is not as daunting a task as it sounds.

[Razor]

Card Quality is key, too.  Cards like Bazaar and Compulsion allow the 'cycling' or exchange of less useful cards for more useful cards.  Decks like 'Tog abuse this concept very well.

[Dr. Sylvan]

Suppose the most liberal estimates of de facto card pool size are true and there are 400 cards that are regularly used, this still equates with only 6.5% of the games cardpool.  I would venture a guess and say that this is roughly the number of Vintage viable cards we get a set, with larger sets having slightly more, and smaller sets, especially middle block sets, having less.

A set with 6.5% playability would be uber-sexy. In January, 263 different cardnames made Top 8s at large tournaments. A good bit less than the 352 from November/December. Over time, I think 400-500 is a reasonable estimate.

It's fascinating to read TMD's resident philosophers go nuts. Hijack Smmenen's threads more often, and you know who to request empirical data from. :)

[Saucemaster]

It's fascinating to read TMD's resident philosophers go nuts. Hijack Smmenen's threads more often, and you know who to request empirical data from.

What is this "empirical data" of which you speak?  Philosophers never touch that stuff--it's like poison.    Actually, I only have an undergrad degree in Philosophy, and IIRC Ric_Flair has a graduate-level degree.  And while I'm sure neither of us would dare use the term "Philosopher", Ric_Flair obviously has more claim to it than I do.

[Gothmog]

Ric_Flair wrote:

This comment is just patently untrue. While there are currently 6133 legal cards in Magic, see www.mtgpics.com, there are only a few hand of cards, relatively speaking that make up the de facto cardpool in Vintage.

A short list of older cards deemed unplayable or suboptimal two years ago that are now in winning decklists:

Bazaar of Baghdad
Lion's Eye Diamond
Ancient Tomb
City of Traitors
Accumulated Knowledge
Intuition
Berserk
Maze of Ith
Sleight of Hand
Brainstorm
Hurkyl's Recall
Nevinyrral's Disk

Of course some of these cards were played by someone two years ago, but they are all cards that were generally ignored that are played to a greater or lesser extent now.

I certainly will not make the case that all 6000+ cards may be playable in Type I, but limiting your analysis of potential cards for decks to ones currently played absolutely will lead to a stagnant metagame where a best deck or two emerges.  It also will limit your ability to find the "best deck" as new cards are printed for potential use.

[Dr. Sylvan]

The contention isn't that new cards don't become good, or old ones interacting in new ways don't expand the card pool. Instead, it's undoubtedly true that when you start using a new card, you have to cut an old one. If TnT starts popping in a new card from Darksteel, then maybe one of the old standards falls by the wayside. Same for Mud and Trinisphere; perhaps Sphere of Resistance will cease to be played.

Skeletal Scrying replaced Braingeyser/Stroke of Genius.
Decree of Justice replaced Goblin Trenches.
Orim's Chant replaced Abeyance.

There's, of course, more. The point is that a metagame will consistently be using the top several hundred cards available. Even in an environment like Type 1, where decklists automatically have more different cards on them due to restrictions, those same restrictions pressure out the use of second-rate cards.

BTW, Saucemaster: As one of the utterly Sauceless effete intellectuals without any piece of paper backing me up, you're miles ahead anyway. Besides, I'm a bad student: I read Sparknotes of Plato instead of the real thing. =-O

[Ric_Flair]

No claims of being a philosopher here.  I think Heidegger permanently kicked that mantle of historical encumbered concepts to the curb.  Thinker maybe, that or John Rawls barn.  I even saw him in person once when I went to H for a talk.  He was old.  Coincidence or divine irony:  Rawls and Nozick died within a month or two of each other in the same year.

Sparknotes on Plato...oh rend my heart in two.  Plato is probably my favorite writer of all time.  Going to a Jesuit school, I read the Republic, oh, I don't know, 6 times (no joke), once in Greek.  The Laws are good too, but have none of the seditious majesty of the Republic.  If people ever read that book and understood even the surface meaning I think they would be disgusted.  No families, no equality, it is a tough pill for Americans to swallow.  Even the Founders, the conservative and gentile landed elite that they were, would be shocked by the Republic.  

As for the Magic issue, and yes I think we are still in a Magic forum, Dr. Sylvan is right about the swapping phenomenon.   But the other mechanism that keeps the number of new cards in check is the fact that most new cards cannot be printed at the same power level as old cards.  The comment from R&D was that Wizards would reprint Mana Drain the day they all got hit by a bus (irony: this comment was made around the time that they were "playtesting" Urza's Block, and I use "playtesting" in the most liberal of senses).  They simply cannot make a non-niche counterspell, ie not Stifle, that can compete with Force of Will and Mana Drain without wrecking every other format.  This forcing of their creative hand limits the number of "good" Vintage cards they can make.