|
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort.
|
On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort.
At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing.
|
It looks like my LGS is going to do a Born of the Gots-only sealed as release event. It's gonna be crazy I think
|
On January 31 2014 23:52 Gorsameth wrote:Show nested quote +On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing.
Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness.
That doesn't necessarily mean you made a bad play - the best play doesn't win every time, it wins most of the time. But if you systematically test your decisions and over time find that you are, in general, too paranoid about your top decks, then you have learned something - at the very least, you have singled out something for reevaluation.
The obvious counter argument is that this is much too general - but then again: alpha strike, leave myself open to lethal and hope he hasn't got fog given that it's turn six and he has two cards in hand and is playing this particular deck OR hope to top deck from 44 cards, two 5-mana win conditions, four 1-mana cantrips, 2 0-mana win conditions - well, that is much too specific.
|
On February 01 2014 00:38 Darkwhite wrote:Show nested quote +On January 31 2014 23:52 Gorsameth wrote:On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing. Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness. That doesn't necessarily mean you made a bad play - the best play doesn't win every time, it wins most of the time. But if you systematically test your decisions and over time find that you are, in general, too paranoid about your top decks, then you have learned something - at the very least, you have singled out something for reevaluation. The obvious counter argument is that this is much too general - but then again: alpha strike, leave myself open to lethal and hope he hasn't got fog given that it's turn six and he has two cards in hand and is playing this particular deck OR hope to top deck from 44 cards, two 5-mana win conditions, four 1-mana cantrips, 2 0-mana win conditions - well, that is much too specific.
Good points.
It's one thing to play to your outs, but its another to play to when you expect your outs to be. Flipping cards is pointless because its like okay my out was 5 cards deep, does that mean I should have been waiting for 5 or less turns to get to it the next time I am in this situation? Obviously not.
Flipping over cards after the fact is and should be a fun thing, not an educational one. If you're losing track of the cards left in your deck, then that's the problem that needs to be fixed, not looking through your deck after the fact.
Anyways I am talking about the card flipping players who go like 20 deep and tell you exactly how many cards down it was, and act like there is no way the board state wouldn't have changed 20 cards later between the two of us.
I might flip cards if I am playing some kind of combo deck and I was missing a piece, but even then its just mostly for kicks.
|
United States24814 Posts
On February 01 2014 00:38 Darkwhite wrote:Show nested quote +On January 31 2014 23:52 Gorsameth wrote:On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing. Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness.
The problem is that seeing if you would have won with the next couple of topdeck cards is that, short of actually recording data from game to game, it will be completely misleading. Humans are notoriously bad at drawing statistical conclusions based on these types of informal observations over a long period of time. You will probably be completely wrong when attempting to determine whether you are being overly paranoid or not.
As Judicator said, some people do it for kicks more than anything else and that's certainly fine. However, one data point by itself is not useful... and if you want to argue that you are accruing useful data, you need to do a better job than just glancing at your topdeck cards after some games, and trying to keep track informally in your mind.
|
If you know what cards are left in your library and exactly how they will impact the game state, while accounting for earlier scrying and second order effects, such as what you can do with what you are expected to draw if you top deck Read the Bones - if you know all of this exactly, then there is nothing more to learn from flipping cards. Checking exactly where your two Lava Axes are in a randomized pile of cards doesn't have any value.
Not everybody plays at the level where the fraction of winning to total cards in the library is obvious, though - in some complicated board states, I'm not sure if anybody does. And again, some people are intuitive learners who play extremely well without ever jotting down a fraction or a probability on paper. And that's when sampling the random distribution is an entirely sensible thing to do, which allows you to get a bit more experience at a very negligible cost.
On February 01 2014 02:01 micronesia wrote:Show nested quote +On February 01 2014 00:38 Darkwhite wrote:On January 31 2014 23:52 Gorsameth wrote:On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing. Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness. The problem is that seeing if you would have won with the next couple of topdeck cards is that, short of actually recording data from game to game, it will be completely misleading. Humans are notoriously bad at drawing statistical conclusions based on these types of informal observations over a long period of time. You will probably be completely wrong when attempting to determine whether you are being overly paranoid or not.As Judicator said, some people do it for kicks more than anything else and that's certainly fine. However, one data point by itself is not useful... and if you want to argue that you are accruing useful data, you need to do a better job than just glancing at your topdeck cards after some games, and trying to keep track informally in your mind.
Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation.
|
United States24814 Posts
On February 01 2014 02:05 Darkwhite wrote:
Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation.
I think you just misunderstood me. People tend to remember extreme outcomes more than others. I always miss all the traffic lights when I'm running late to an appointment! How do the traffic lights know!? Most likely your luck over time is exactly the same as you would predict mathematically... it just seems like you do more waiting at lights on days when you are at risk of running late, and do less waiting on more leisurely days.
If you look at your top card after every game, and someone asks you a few weeks later how in tune you have been with your deck, are you really able to give a better answer than if you recorded the card after every game and did a basic statistical calculation? Of course I do recognize that MTG situations are actually more complicated than this simplified scenario implies.
|
On February 01 2014 02:05 Darkwhite wrote:If you know what cards are left in your library and exactly how they will impact the game state, while accounting for earlier scrying and second order effects, such as what you can do with what you are expected to draw if you top deck Read the Bones - if you know all of this exactly, then there is nothing more to learn from flipping cards. Checking exactly where your two Lava Axes are in a randomized pile of cards doesn't have any value. Not everybody plays at the level where the fraction of winning to total cards in the library is obvious, though - in some complicated board states, I'm not sure if anybody does. And again, some people are intuitive learners who play extremely well without ever jotting down a fraction or a probability on paper. And that's when sampling the random distribution is an entirely sensible thing to do, which allows you to get a bit more experience at a very negligible cost. Show nested quote +On February 01 2014 02:01 micronesia wrote:On February 01 2014 00:38 Darkwhite wrote:On January 31 2014 23:52 Gorsameth wrote:On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing. Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness. The problem is that seeing if you would have won with the next couple of topdeck cards is that, short of actually recording data from game to game, it will be completely misleading. Humans are notoriously bad at drawing statistical conclusions based on these types of informal observations over a long period of time. You will probably be completely wrong when attempting to determine whether you are being overly paranoid or not.As Judicator said, some people do it for kicks more than anything else and that's certainly fine. However, one data point by itself is not useful... and if you want to argue that you are accruing useful data, you need to do a better job than just glancing at your topdeck cards after some games, and trying to keep track informally in your mind. Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation.
Umm you are suppose to know what cards are left in your deck, this is pretty basic stuff if you want to improve at the game. Flipping through the deck shouldn't be the fix to that mistake. As for people being brilliant learners in these situations...when compared to an AI maybe (even then they aren't), nope nope nope. Humans more than anything buy way too much into statistics rather than the methodology behind it, just read published peer-review literature (like Lancet, Nature, etc.), you'd be surprised at how many professionals suck at it.
The problem I have with your line of thought here (and I think micronesia as well) is that you think you can make some kind of conclusion from flipping the cards over then somehow extrapolate some kind of conclusion from there, when in fact you really can't which makes the entire basis of this particular exercise kind of pointless. For example, if your goal is to determine if you were making the right plays in a particular game, then your solution should be to improve your situational awareness of the game state which includes what cards left in your deck, and not to flip cards after the fact to see if you were right because frankly that doesn't do shit for you while the game is still undecided.
Edit:
This discussion is why I tell new people wanting to improve to think of plays in MtG not as individual cards, but lines of play no different from chess (obviously with variance, but not as much as people would like to think).
|
On February 01 2014 02:30 micronesia wrote:Show nested quote +On February 01 2014 02:05 Darkwhite wrote:
Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation. I think you just misunderstood me. People tend to remember extreme outcomes more than others. I always miss all the traffic lights when I'm running late to an appointment! How do the traffic lights know!? Most likely your luck over time is exactly the same as you would predict mathematically... it just seems like you do more waiting at lights on days when you are at risk of running late, and do less waiting on more leisurely days. If you look at your top card after every game, and someone asks you a few weeks later how in tune you have been with your deck, are you really able to give a better answer than if you recorded the card after every game and did a basic statistical calculation? Of course I do recognize that MTG situations are actually more complicated than this simplified scenario implies.
I don't think you can even do the statistical analysis, because many of the plays are entirely unique - different board states, different matchups, different cards in hand, etcetera. Before you can do anything, statistically, you will have to find a way to generalize across these differences, and by then you really just have a some mathematics slapped on top of human judgement anyway. That said - meticulously collecting data and spending time analyzing it will probably be better, mostly because that way you spend a lot more time and think more carefully about things, but not because the statistics themselves are all that helpful. Spending that extra time playing more games would probably be a better investment.
On February 01 2014 02:47 Judicator wrote:Show nested quote +On February 01 2014 02:05 Darkwhite wrote:If you know what cards are left in your library and exactly how they will impact the game state, while accounting for earlier scrying and second order effects, such as what you can do with what you are expected to draw if you top deck Read the Bones - if you know all of this exactly, then there is nothing more to learn from flipping cards. Checking exactly where your two Lava Axes are in a randomized pile of cards doesn't have any value. Not everybody plays at the level where the fraction of winning to total cards in the library is obvious, though - in some complicated board states, I'm not sure if anybody does. And again, some people are intuitive learners who play extremely well without ever jotting down a fraction or a probability on paper. And that's when sampling the random distribution is an entirely sensible thing to do, which allows you to get a bit more experience at a very negligible cost. On February 01 2014 02:01 micronesia wrote:On February 01 2014 00:38 Darkwhite wrote:On January 31 2014 23:52 Gorsameth wrote:On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing. Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness. The problem is that seeing if you would have won with the next couple of topdeck cards is that, short of actually recording data from game to game, it will be completely misleading. Humans are notoriously bad at drawing statistical conclusions based on these types of informal observations over a long period of time. You will probably be completely wrong when attempting to determine whether you are being overly paranoid or not.As Judicator said, some people do it for kicks more than anything else and that's certainly fine. However, one data point by itself is not useful... and if you want to argue that you are accruing useful data, you need to do a better job than just glancing at your topdeck cards after some games, and trying to keep track informally in your mind. Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation. Umm you are suppose to know what cards are left in your deck, this is pretty basic stuff if you want to improve at the game. Flipping through the deck shouldn't be the fix to that mistake. As for people being brilliant learners in these situations...when compared to an AI maybe (even then they aren't), nope nope nope. Humans more than anything buy way too much into statistics rather than the methodology behind it, just read published peer-review literature (like Lancet, Nature, etc.), you'd be surprised at how many professionals suck at it. The problem I have with your line of thought here (and I think micronesia as well) is that you think you can make some kind of conclusion from flipping the cards over then somehow extrapolate some kind of conclusion from there, when in fact you really can't which makes the entire basis of this particular exercise kind of pointless. For example, if your goal is to determine if you were making the right plays in a particular game, then your solution should be to improve your situational awareness of the game state which includes what cards left in your deck, and not to flip cards after the fact to see if you were right because frankly that doesn't do shit for you while the game is still undecided. Edit: This discussion is why I tell new people wanting to improve to think of plays in MtG not as individual cards, but lines of play no different from chess (obviously with variance, but not as much as people would like to think).
The problem with your line of thought is that people do learn from experience. We don't really know how, but people sit down, play some hundreds of games, and gradually they become stronger players, just from seeing lots of different situations and having their board wrathed a few times. How do you think people learn to evaluate new cards, if not by seeing them in play or having seen how very similar cards perform?
Explicitly checking what would have happened if I waited a turn instead after having made these kinds of all or nothing decisions takes on the order of five seconds and is sometimes going to a learning experience - particularly for poorer players. I'm not saying it's going to vastly improve your play overnight, and if you're carefully thinking about good and bad cards before making the call there is much less to gain from it, but then again - that's not how everybody play the game. The point was just that it's a perfectly reasonable thing to do, and really no different from any other kind of simulation.
That said, when you dig ten cards deep and explain that, if only all your draws had been in the perfect order you would have won - at that point, you deserve some mockery for stating the obvious.
|
No no no no, you are assuming people first of all identify correctly what exactly they are suppose to learn, then you are assuming that people learn the right things. That is very clearly not the case, you see this in education, you see this in the workplace, and you especially see this in gaming.
Its one thing to run the simulation, but its entirely different to be able to make a solid conclusion.
Case in point, me and another friend (rated 2k+ before DCI ratings went away) were spectating our friend Jack Fogle at GP Louisville in round 10. Some background, I probably have more experience with Esper Control than my friend does, but he definitely have more experience with MtG in general. Our friend's playing against BW post-board after taking game 1. Keeps a very a sketch hand on the draw (it wasn't strong or even good by any means). We briefly discuss the keep and go back to the game, I wouldn't have kept based on experience and he said he probably wouldn't have either but wasn't sure. Now what's the point of this story? Jack told us after the fact that the post-board match up revolved around Duress/Thoughtseize/Sin Collector and not the Demon/Obzedat/Blood Baron for the Esper player, the hand was a weak one for sure, but that's not relevant since it was very strong against the sideboard cards.
I can run that simulation of an opening hand a million times, but whether I decide to keep is entirely based on context. That's why flipping cards is a pointless educational exercise. Circumstances change game to game to game, there's no value in the answer to the question of "what would have happened if I waited a turn?" because that question has very little value, especially to new players. More experienced players can try to examine their lines which reached that context, and see if they could have done anything differently. New players have no clue on all of the factors that contributed to a loss or win.
Edit:
TL;DR version learning and learning the correct lesson are not the same things.
|
|
|
what do you think those cards are hinting at? a custom cube?
|
Top one looks like a Commander's Arsenal 2 card, bottom one is something out of an Un-set.
|
The uncapitalized 'C' in "council" is driving me nuts.
|
On February 01 2014 12:53 Cel.erity wrote:
The uncapitalized 'C' in "council" is driving me nuts.
I initially thought it was fake because of this. Infuriating.
|
cube confuses me. I lose a lot of time to people who just draft big shit first and build around it (i.e. titans, etc.) but whenever I try that it feels like a trap.
here's a deck i went 2-1 with http://i.imgur.com/2Ywozpq.jpg
any other perspectives on how I should have built it/drafted it? I feel like i might have leaned towards too many counters, as once I got ahead I had a hard time closing the game. However, In the last match I sided into magus and turned my opponents 8 lands into mountains, which was pretty hilarious.
|
On February 01 2014 03:54 Darkwhite wrote:Show nested quote +On February 01 2014 02:30 micronesia wrote:On February 01 2014 02:05 Darkwhite wrote:
Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation. I think you just misunderstood me. People tend to remember extreme outcomes more than others. I always miss all the traffic lights when I'm running late to an appointment! How do the traffic lights know!? Most likely your luck over time is exactly the same as you would predict mathematically... it just seems like you do more waiting at lights on days when you are at risk of running late, and do less waiting on more leisurely days. If you look at your top card after every game, and someone asks you a few weeks later how in tune you have been with your deck, are you really able to give a better answer than if you recorded the card after every game and did a basic statistical calculation? Of course I do recognize that MTG situations are actually more complicated than this simplified scenario implies. I don't think you can even do the statistical analysis, because many of the plays are entirely unique - different board states, different matchups, different cards in hand, etcetera. Before you can do anything, statistically, you will have to find a way to generalize across these differences, and by then you really just have a some mathematics slapped on top of human judgement anyway. That said - meticulously collecting data and spending time analyzing it will probably be better, mostly because that way you spend a lot more time and think more carefully about things, but not because the statistics themselves are all that helpful. Spending that extra time playing more games would probably be a better investment. Show nested quote +On February 01 2014 02:47 Judicator wrote:On February 01 2014 02:05 Darkwhite wrote:If you know what cards are left in your library and exactly how they will impact the game state, while accounting for earlier scrying and second order effects, such as what you can do with what you are expected to draw if you top deck Read the Bones - if you know all of this exactly, then there is nothing more to learn from flipping cards. Checking exactly where your two Lava Axes are in a randomized pile of cards doesn't have any value. Not everybody plays at the level where the fraction of winning to total cards in the library is obvious, though - in some complicated board states, I'm not sure if anybody does. And again, some people are intuitive learners who play extremely well without ever jotting down a fraction or a probability on paper. And that's when sampling the random distribution is an entirely sensible thing to do, which allows you to get a bit more experience at a very negligible cost. On February 01 2014 02:01 micronesia wrote:On February 01 2014 00:38 Darkwhite wrote:On January 31 2014 23:52 Gorsameth wrote:On January 31 2014 23:40 Darkwhite wrote:
In the trivial case where Lava Axe is the only card which matters, and you know that they are two out of 44 cards in your library - well, then you've already done all the counting and further checking is pointless. But a lot of the time, things are much more complicated, and for a lot of people checking is more effective than math'ing.
What if the guy has scried away five cards, but can't even remember if one was a Lava Axe? What if there are four cantrips in there, which don't do anything by themselves, but he has enough mana to cantrip directly into Lava Axe, or cantrip into cantrip into Lava Axe, but can't afford to triple cantrip? What if he forgot that the double landfall triggers from a Terramorphic Expanse would be enough damage to close it out, and there are two of those left in the deck?
Suddenly, you have to choose between a spreadsheet and checking. And checking is very nearly zero effort. At that point you should check your entire deck and not the top 4/5 cards since that teaches you nothing. Nonsense. It tells you (for instance) that you would have won, had you waited and trusted your top deck instead - in this particular situation, with this particular randomness. The problem is that seeing if you would have won with the next couple of topdeck cards is that, short of actually recording data from game to game, it will be completely misleading. Humans are notoriously bad at drawing statistical conclusions based on these types of informal observations over a long period of time. You will probably be completely wrong when attempting to determine whether you are being overly paranoid or not.As Judicator said, some people do it for kicks more than anything else and that's certainly fine. However, one data point by itself is not useful... and if you want to argue that you are accruing useful data, you need to do a better job than just glancing at your topdeck cards after some games, and trying to keep track informally in your mind. Bolded part is literally as false as it gets. Humans are state of the art technology for drawing sensible conclusions from complicated and messy data, particularly incomplete data sets. That's why people are brilliant learners and AI is still terrible. That's why deck matchup evaluations are still based on a combination of playtesting and human evaluation. Umm you are suppose to know what cards are left in your deck, this is pretty basic stuff if you want to improve at the game. Flipping through the deck shouldn't be the fix to that mistake. As for people being brilliant learners in these situations...when compared to an AI maybe (even then they aren't), nope nope nope. Humans more than anything buy way too much into statistics rather than the methodology behind it, just read published peer-review literature (like Lancet, Nature, etc.), you'd be surprised at how many professionals suck at it. The problem I have with your line of thought here (and I think micronesia as well) is that you think you can make some kind of conclusion from flipping the cards over then somehow extrapolate some kind of conclusion from there, when in fact you really can't which makes the entire basis of this particular exercise kind of pointless. For example, if your goal is to determine if you were making the right plays in a particular game, then your solution should be to improve your situational awareness of the game state which includes what cards left in your deck, and not to flip cards after the fact to see if you were right because frankly that doesn't do shit for you while the game is still undecided. Edit: This discussion is why I tell new people wanting to improve to think of plays in MtG not as individual cards, but lines of play no different from chess (obviously with variance, but not as much as people would like to think). The problem with your line of thought is that people do learn from experience. We don't really know how, but people sit down, play some hundreds of games, and gradually they become stronger players, just from seeing lots of different situations and having their board wrathed a few times. How do you think people learn to evaluate new cards, if not by seeing them in play or having seen how very similar cards perform? Explicitly checking what would have happened if I waited a turn instead after having made these kinds of all or nothing decisions takes on the order of five seconds and is sometimes going to a learning experience - particularly for poorer players. I'm not saying it's going to vastly improve your play overnight, and if you're carefully thinking about good and bad cards before making the call there is much less to gain from it, but then again - that's not how everybody play the game. The point was just that it's a perfectly reasonable thing to do, and really no different from any other kind of simulation. That said, when you dig ten cards deep and explain that, if only all your draws had been in the perfect order you would have won - at that point, you deserve some mockery for stating the obvious.
Magic is a game of statistics and limited info - decisions are (in most cases) good or bad before seeing the actual outcome. So your mulligan should be based on a clear thought process, rather than the anecdotical evidence of looking at the top card.
In similar vein, put yourself into a situation where you can either play around a card you know your opponent has 4-of in his deck, or you can play around a 1-of. The decision should always be the same, regardless if the outcome is him holding the 1-of.
|
On February 01 2014 16:09 EMIYA wrote:cube confuses me. I lose a lot of time to people who just draft big shit first and build around it (i.e. titans, etc.) but whenever I try that it feels like a trap. here's a deck i went 2-1 with http://i.imgur.com/2Ywozpq.jpgany other perspectives on how I should have built it/drafted it? I feel like i might have leaned towards too many counters, as once I got ahead I had a hard time closing the game. However, In the last match I sided into magus and turned my opponents 8 lands into mountains, which was pretty hilarious.
Can't really give advice on how you should have drafted it without the draft file (which autosaves in your My Documents/Games/Magic Online directory if you have that option enabled). It seems like you have a lot of underpowered/situational cards. In general, expensive burn/creature removal is sideboard material. Mizzium Mortars, Pyroclasm, Bogardan Hellkite are all really bad, and then cards like Electrolyze, Legacy's Allure, Ral Zerek, Izzet Charm, Daze are not necessarily bad, but having too many of them just makes your deck too weak and reactive. Your deck is all support cards, and it seems like the only way you can win is by hoping your opponent is playing WW, or casting Treachery on something big that they cast.
As for the sideboard: Zealous Conscripts is absolutely insane, a top red pick in cube. You also had two clones, so leaving him in the sideboard was a big mistake. He has the ability to kill out of nowhere, or steal Planeswalkers, or just steal a blocker and attack a Planeswalker with haste. Negate is better than Pyroclasm/Mortars for sure, and probably better than Daze. Otherwise the build looks reasonable, your other creatures are too small and don't have enough support to deal 20.
|
I see. I appreciate the feedback, I was hoping to pick up some fatties that I could stall to, but the dragon was the only thing I picked up (which ended up being insane surprisingly enough). At first I was trying to do some kind of izzet aggro with the conscripts and stuff but I ended up deciding against it. The first match I simply couldn't win the game as you suspected, but the second two were hilarious; an opponent eureka'd a titan out that I treachery'd and copied twice, and the last game magus won me it.
|
|
|
|
|
|
EPT![[image loading]](http://i.imgur.com/Byl2JQz.jpg)
![[image loading]](http://i.imgur.com/C8CFXMX.jpg)
