tl;dr, I think I can predict the best deck. Open the why it's useful spoiler to see.
----------------
I think it's possible to determine ahead of time what creatures will be best in a given format. The same goes for removal. Please read and comment!
For removal, it's simply a matter of ranking how well each removal spell deals with the most commonly played or most-expected-to-be-played creatures in the format. You can make a table with the good removal spells down the side, and the creatures across the top. If a certain removal spell answers a creature, it gets a 1 in that column. If not, it gets a zero (see the link for an example if you're having trouble picturing it). At the end, add up all the values and divide by the number of creatures. Tada, you now have the percentage of creatures that a certain removal spell will answer.
I tried to do this for gatecrash block constructed (I did not include all the creatures or all the removal, feel free download and add to it). I discuss it's usefulness below and give a sample decklist based on my results.
At the far right of the page you will see the rankings that each removal spell has (they're in order). At the bottom you'll see the creature's rankings (they're not in order). The creatures also need to be evaluated on how well they match up against each other, so the rankings for creatures are incomplete.
I picked all (I think I got them all) of the viable removal spells, but only a small subset of the possible creatures we'll see. This could throw off the results. I tried to make the threats representative of a generic meta (aggro, midrange, and control threats, some meh cards, etc).
I chose to evaluate answers based on whether or not they stopped the creature from achieving a certain purpose, NOT on whether they killed the creature. So, for example, a Supreme Verdict does not answer Rakdos Cackler, because Cackler will have dealt 6 damage before you cast verdict, and I would say that by then it's fulfilled it's purpose. Sometimes the creature only needed to live for a turn to fulfill it's purpose (Pack Rat, Niv-Mizzet, Dracogenius), and other times it just needed to cause a huge swing in card advantage (Angel of Serenity, Prime Speaker Zegana). I believe this made the numbers better overall, but it certainly leads to a lot of discussion about the purpose of each creature (after all, the creature's purpose can vary depending on the matchup!).
I gave removal spells 1/2 of a point if they were an answer on the play, but not on the draw. (example, Cancel gets 1/2 a point against Hellhole Flailer).
Remember, this is a qualitative approach, not a quantitative one, so there are certainly inaccuracies and it's very likely to result in some mis-ranked spells (after all, there's a lot more to take into account such as instant/sorcery, synergies, etc). I did not include all possible creatures and answers either. However, I believe that it still produces useful results, and that is good enough for me.
Right now I'm using it as a speculative tool, but it can be adapted once a metagame is known. Every removal spell will receive a weight according to it's popularity, as will each threat. This tool could then be used to find "holes" or "sleepers" that have a high threat rank, but low popularity with which to attack the meta!
As a speculative tool it helps filter out which strategies appear best on paper and should be tried first. Using the results from the chart, it looks like BWR control should be the starting point for control lists. They have access to a number of highly ranked 2 and 3 mana point removal spells in addition to two highly ranked sweepers (Mizzium Mortars counts!). They aren't doing too bad as far as the resiliency of their threats go either, with Obzedat, Ghost Council, Desecration Demon, and Angel of Serenity all ranking high (in terms of fulfilling their purpose, remember we still have to do a threat vs. threat evaluation). Did you need me to tell you BRW was good? Probably not. Did it help make an excellent deck to start beating the format with? Definitely. Finally, I saved a lot of time by reducing the number of iterations it would take for me to get here! I essentially just need to tune it.
Here's what I would try first after examining the results:
There are a number of results that are ranked inaccurately, and I think that it's due to my methodology.
Aurelia's Fury - I ranked this based on how well it answered creatures, NOT on anything else. So it naturally did about as good as Street Spasm, that is, not so well. Ground Assault had the same problem.
Syncopate is the #1 answer, apparently. When I ranked spells, I assumed that the threats were cast on their ideal turn and that players weren't playing around Syncopate. Obviously you can't just leave mana open all the time to cast syncopate, and people can play around it rather easily.
Some facts of magic:
-Terror is an emotion which, when experienced, results in death.
-The pox was a disease notorious for having killed one-third, rounded up, of Europe’s population. Smallpox, on the other hand, killed only a single person.
-A person riding a horse cannot be stopped by foot soldiers, large animals, walls, archers, or even catapults.
So, guys, it looks like there may be some merit to this after all. The deck I predicted isn't too far off from one of the established decks now, AND I didn't do a full analysis. What do you think?
Private Mod Note
():
Rollback Post to RevisionRollBack
Some facts of magic:
-Terror is an emotion which, when experienced, results in death.
-The pox was a disease notorious for having killed one-third, rounded up, of Europe’s population. Smallpox, on the other hand, killed only a single person.
-A person riding a horse cannot be stopped by foot soldiers, large animals, walls, archers, or even catapults.
I glanced at this when you first posted it, but now i'm really looking at it.
I think you got the answers for angel of serenity wrong.
Devour flesh, orzhov charm, grisly spectacle, Ultimate Price, Rapid Hybridization, even street spasm, Killing glare, and Aurelia's Fury are all answers for angel of serenity. An answer being something that can kill it as soon as it hits the field and before anything gets exiled by it. You said that vraska answers it but dreadbore doesn't, aren't both of those basically the same thing? And in the case of AoS neither of them are as good as the ones I brought up (except for the X spells, but they are still instants that can kill her).
Also why did you give syncopate a point against loxodon smiter, and half a point to cancel and psychic strike? Smiter can't be countered.
And Abrupt Decay can't answer Crypt Ghast, spark trooper, ghor-clan rampager, or biomancer. Why does it have a point against them?
If MTG is a part of your life, the formats are like relationships:
Standard/Block = The on-again, off-again holiday fling
Modern/Vintage/Legacy = Stable, homely. A ***** after absence/misreading
Limited/Sealed = Heart breaking free spirit
Commander/Cube = Agreeable, needy and expensive
Pauper/Peasant = Sweet, kind, practical, but shy and boring
I like your approach. It is far from complete because a specific game of magic is all about context and it's hard to take into account all the subtlety before playing the cards a certain number of times, in different decks. Play-testing a deck is still the best way to tweak it in order to achieve best consistency... but that doesn't mean that taking apart the format like you're doing is a bad thing.
As we all know, thinking about a card when it's announced should be enough to figure out if it's trash or playable, but it's viability in constructed ends up to how the meta game plays out. When someone tweaks his/her deck based on which cards are played in his/her metagame. S/he is basically doing what you did with the whole format, but only for the very limited pool that makes the current metagame. Applying this concept to the whole format gives some kind of head start... Overall, I would say it's better than analyzing the card out of context but at the same time, the analysis can give too much importance to cards that won't even being played for another reason. (Position on the curve vs other spells, colors and archetype pairing)
I'm wondering what results you would get if you gave each cards different "weight" based on their popularity in the meta.
I glanced at this when you first posted it, but now i'm really looking at it.
I think you got the answers for angel of serenity wrong.
Devour flesh, orzhov charm, grisly spectacle, Ultimate Price, Rapid Hybridization, even street spasm, Killing glare, and Aurelia's Fury are all answers for angel of serenity. An answer being something that can kill it as soon as it hits the field and before anything gets exiled by it. You said that vraska answers it but dreadbore doesn't, aren't both of those basically the same thing? And in the case of AoS neither of them are as good as the ones I brought up (except for the X spells, but they are still instants that can kill her).
Also why did you give syncopate a point against loxodon smiter, and half a point to cancel and psychic strike? Smiter can't be countered.
And Abrupt Decay can't answer Crypt Ghast, spark trooper, ghor-clan rampager, or biomancer. Why does it have a point against them?
Whew! Looks like I made a lot of mistakes.
I determined the purpose of Angel of Serenity was not permanent removal, but temporary removal and tempo. That's how I ranked it. I know it's weird to have a 7 mana tempo spell, but it's pretty powerful as a "falter." If It netted tempo and card advantage, I gave it a "1." I could be completely wrong, this is one of those cards that's easy to argue either way. Seeing as I predicted it as being in the RBW deck, and it isn't, I suspect I may have overrated it.
Here's my plan for the future:
- I want to make full threat v. threat, threat v. removal, and card to card synergy tables. (Synergy was not considered before, but is actually quite important!)
- Since this is a huge job (150 card = 22500 combinations to consider) I think I want to crowdsource. I need to learn how to make websites (I can program, but I've never done any web stuff), and set one up so that anyone can "rank" one card against another. I'll average the scores and use those to determine the 'expected' meta. (Note, expected! Since it's based on what everyone thinks, it could potentially predict the initial meta!).
-The website will display two cards and simply ask "do these have synergy?" It would allow a yes/no/maybe answer for this question. "Which of threat x and threat y would fulfill their purposes, if each were played on their optimal turn?" It would allow X, Y, both X and Y or neither for this question. Finally, "Would removal X stop threat Y from fulfilling it's purpose, if both are played on their optimal turns?" Again, this will have Yes/No/Maybe.
- Once we have an "initial" meta, I'll start iterating. Assuming that the worst half of the cards aren't played, what changes? Do underperformers become all stars? I'll cut the worst few cards again and see how things change further. This has GOT to produce something useful, even if it's the craziest, most round about way to get back to what people already think are the best cards.
It's a lot of work, and I'll probably have to wait until the next block comes out to really get into it. I'm excited though!
I can't help but thinking that if I had taken this "all out" I could have predicted something like boros reckoner's popularity. If you put him in a threat v. threat chart, he wins basically every time (because at the worst, he's a Lightning bolt!)
Private Mod Note
():
Rollback Post to RevisionRollBack
Some facts of magic:
-Terror is an emotion which, when experienced, results in death.
-The pox was a disease notorious for having killed one-third, rounded up, of Europe’s population. Smallpox, on the other hand, killed only a single person.
-A person riding a horse cannot be stopped by foot soldiers, large animals, walls, archers, or even catapults.
----------------
I think it's possible to determine ahead of time what creatures will be best in a given format. The same goes for removal. Please read and comment!
For removal, it's simply a matter of ranking how well each removal spell deals with the most commonly played or most-expected-to-be-played creatures in the format. You can make a table with the good removal spells down the side, and the creatures across the top. If a certain removal spell answers a creature, it gets a 1 in that column. If not, it gets a zero (see the link for an example if you're having trouble picturing it). At the end, add up all the values and divide by the number of creatures. Tada, you now have the percentage of creatures that a certain removal spell will answer.
I tried to do this for gatecrash block constructed (I did not include all the creatures or all the removal, feel free download and add to it). I discuss it's usefulness below and give a sample decklist based on my results.
Threatdown (Links to a google spreadsheet)
At the far right of the page you will see the rankings that each removal spell has (they're in order). At the bottom you'll see the creature's rankings (they're not in order). The creatures also need to be evaluated on how well they match up against each other, so the rankings for creatures are incomplete.
I picked all (I think I got them all) of the viable removal spells, but only a small subset of the possible creatures we'll see. This could throw off the results. I tried to make the threats representative of a generic meta (aggro, midrange, and control threats, some meh cards, etc).
I chose to evaluate answers based on whether or not they stopped the creature from achieving a certain purpose, NOT on whether they killed the creature. So, for example, a Supreme Verdict does not answer Rakdos Cackler, because Cackler will have dealt 6 damage before you cast verdict, and I would say that by then it's fulfilled it's purpose. Sometimes the creature only needed to live for a turn to fulfill it's purpose (Pack Rat, Niv-Mizzet, Dracogenius), and other times it just needed to cause a huge swing in card advantage (Angel of Serenity, Prime Speaker Zegana). I believe this made the numbers better overall, but it certainly leads to a lot of discussion about the purpose of each creature (after all, the creature's purpose can vary depending on the matchup!).
I gave removal spells 1/2 of a point if they were an answer on the play, but not on the draw. (example, Cancel gets 1/2 a point against Hellhole Flailer).
Remember, this is a qualitative approach, not a quantitative one, so there are certainly inaccuracies and it's very likely to result in some mis-ranked spells (after all, there's a lot more to take into account such as instant/sorcery, synergies, etc). I did not include all possible creatures and answers either. However, I believe that it still produces useful results, and that is good enough for me.
Right now I'm using it as a speculative tool, but it can be adapted once a metagame is known. Every removal spell will receive a weight according to it's popularity, as will each threat. This tool could then be used to find "holes" or "sleepers" that have a high threat rank, but low popularity with which to attack the meta!
As a speculative tool it helps filter out which strategies appear best on paper and should be tried first. Using the results from the chart, it looks like BWR control should be the starting point for control lists. They have access to a number of highly ranked 2 and 3 mana point removal spells in addition to two highly ranked sweepers (Mizzium Mortars counts!). They aren't doing too bad as far as the resiliency of their threats go either, with Obzedat, Ghost Council, Desecration Demon, and Angel of Serenity all ranking high (in terms of fulfilling their purpose, remember we still have to do a threat vs. threat evaluation). Did you need me to tell you BRW was good? Probably not. Did it help make an excellent deck to start beating the format with? Definitely. Finally, I saved a lot of time by reducing the number of iterations it would take for me to get here! I essentially just need to tune it.
Here's what I would try first after examining the results:
1 Mizzium Mortars
4 Dreadbore
3 Orzhov Charm
3 Devour Flesh
2 Auger Spree
4 Desecration Demon
3 Obzedat, Ghost Council
3 Rakdos's Return
4 Angel of Serenity
There are a number of results that are ranked inaccurately, and I think that it's due to my methodology.
Aurelia's Fury - I ranked this based on how well it answered creatures, NOT on anything else. So it naturally did about as good as Street Spasm, that is, not so well. Ground Assault had the same problem.
Syncopate is the #1 answer, apparently. When I ranked spells, I assumed that the threats were cast on their ideal turn and that players weren't playing around Syncopate. Obviously you can't just leave mana open all the time to cast syncopate, and people can play around it rather easily.
-Terror is an emotion which, when experienced, results in death.
-The pox was a disease notorious for having killed one-third, rounded up, of Europe’s population. Smallpox, on the other hand, killed only a single person.
-A person riding a horse cannot be stopped by foot soldiers, large animals, walls, archers, or even catapults.
More facts of magic
-Terror is an emotion which, when experienced, results in death.
-The pox was a disease notorious for having killed one-third, rounded up, of Europe’s population. Smallpox, on the other hand, killed only a single person.
-A person riding a horse cannot be stopped by foot soldiers, large animals, walls, archers, or even catapults.
More facts of magic
I think you got the answers for angel of serenity wrong.
Devour flesh, orzhov charm, grisly spectacle, Ultimate Price, Rapid Hybridization, even street spasm, Killing glare, and Aurelia's Fury are all answers for angel of serenity. An answer being something that can kill it as soon as it hits the field and before anything gets exiled by it. You said that vraska answers it but dreadbore doesn't, aren't both of those basically the same thing? And in the case of AoS neither of them are as good as the ones I brought up (except for the X spells, but they are still instants that can kill her).
Also why did you give syncopate a point against loxodon smiter, and half a point to cancel and psychic strike? Smiter can't be countered.
And Abrupt Decay can't answer Crypt Ghast, spark trooper, ghor-clan rampager, or biomancer. Why does it have a point against them?
Standard/Block = The on-again, off-again holiday fling
Modern/Vintage/Legacy = Stable, homely. A ***** after absence/misreading
Limited/Sealed = Heart breaking free spirit
Commander/Cube = Agreeable, needy and expensive
Pauper/Peasant = Sweet, kind, practical, but shy and boring
As we all know, thinking about a card when it's announced should be enough to figure out if it's trash or playable, but it's viability in constructed ends up to how the meta game plays out. When someone tweaks his/her deck based on which cards are played in his/her metagame. S/he is basically doing what you did with the whole format, but only for the very limited pool that makes the current metagame. Applying this concept to the whole format gives some kind of head start... Overall, I would say it's better than analyzing the card out of context but at the same time, the analysis can give too much importance to cards that won't even being played for another reason. (Position on the curve vs other spells, colors and archetype pairing)
I'm wondering what results you would get if you gave each cards different "weight" based on their popularity in the meta.
Whew! Looks like I made a lot of mistakes.
I determined the purpose of Angel of Serenity was not permanent removal, but temporary removal and tempo. That's how I ranked it. I know it's weird to have a 7 mana tempo spell, but it's pretty powerful as a "falter." If It netted tempo and card advantage, I gave it a "1." I could be completely wrong, this is one of those cards that's easy to argue either way. Seeing as I predicted it as being in the RBW deck, and it isn't, I suspect I may have overrated it.
Here's my plan for the future:
- I want to make full threat v. threat, threat v. removal, and card to card synergy tables. (Synergy was not considered before, but is actually quite important!)
- Since this is a huge job (150 card = 22500 combinations to consider) I think I want to crowdsource. I need to learn how to make websites (I can program, but I've never done any web stuff), and set one up so that anyone can "rank" one card against another. I'll average the scores and use those to determine the 'expected' meta. (Note, expected! Since it's based on what everyone thinks, it could potentially predict the initial meta!).
-The website will display two cards and simply ask "do these have synergy?" It would allow a yes/no/maybe answer for this question. "Which of threat x and threat y would fulfill their purposes, if each were played on their optimal turn?" It would allow X, Y, both X and Y or neither for this question. Finally, "Would removal X stop threat Y from fulfilling it's purpose, if both are played on their optimal turns?" Again, this will have Yes/No/Maybe.
- Once we have an "initial" meta, I'll start iterating. Assuming that the worst half of the cards aren't played, what changes? Do underperformers become all stars? I'll cut the worst few cards again and see how things change further. This has GOT to produce something useful, even if it's the craziest, most round about way to get back to what people already think are the best cards.
It's a lot of work, and I'll probably have to wait until the next block comes out to really get into it. I'm excited though!
I can't help but thinking that if I had taken this "all out" I could have predicted something like boros reckoner's popularity. If you put him in a threat v. threat chart, he wins basically every time (because at the worst, he's a Lightning bolt!)
-Terror is an emotion which, when experienced, results in death.
-The pox was a disease notorious for having killed one-third, rounded up, of Europe’s population. Smallpox, on the other hand, killed only a single person.
-A person riding a horse cannot be stopped by foot soldiers, large animals, walls, archers, or even catapults.
More facts of magic