I see where you're going with the "challenge people to think critically about their presumptions about a format" - that's definitely a worthy goal and something a lot of us could learn to do better/more often!
I'm also glad you mention the transitional character of the data (unlike the principles derived from it), as I've run into a couple of people on the WOTC forums who will use that exact approach to the stats you present ("Stats say Orzhov is best so we should pick the Orzhov card here" in community drafts... Eugh).
I don't disagree with you. I realize that there is a lot more complexity to an individual draft format than just looking at population ratio compared with win rate. There are many small factors that influence how different decks perform. The problem is that it is incredibly difficult to quantify those small factors. For example, you brought up the point of depth compared with popularity. The problem is that there is no unbiased way for me to determine that; the only thing I can think of is assigning playability numbers to the cards that can be drafted by certain decks, and then try to determine how a deck's popularity matches up against its depth. This introduces a tremendous amount of bias. There is already enough inherent bias in any kind of study I do, and so one of my biggest goals is always to eliminate as much bias as possible. Doing what you suggest automatically takes it out of the realm of what I try to do in my articles and puts it more into the realm of what I want people to do in the comments of my articles.
In any case, I agree with a lot of your analysis; Gruul suffers because its cards can be more easily poached by neighboring guilds. I would argue that there is also the problem that Gruul is just not all that good anyway, but the fact that its cards often go into Boros and Simic decks certainly hurts. Of course, neither of those two decks are performing all that well, so maybe there is a different problem going on.
The point I was trying to make with this article is that people always try to make reductive arguments about why decks are performing a certain way. They do this, not to understand a format better, but in order to confirm what they were already thinking about a format. This is confirmation bias. People look at my data, and they only look at the things that confirm what they were already thinking. I would challenge people to be a little bit more open-minded and try to actually understand the things I am demonstrating.
Finally, I think that the biggest problem with these studies is that they are outdated very quickly. The metagame of a format changes pretty quickly, and it is incredibly difficult to try to stay right on top of its pulse. I think that the principles in the articles are always applicable, so I recommend that people understand the principles, and then apply them to what they actually see. Again, this is the blueprint vs. map argument. Treat this like a map to help you understand where things are generally happening, but be aware of the immediate events that are happening in a specific draft. Don't treat this like a blueprint that you need to follow precisely.
In any case, thanks for the comments. I do appreciate what you're saying, and even agree with you, but it kind of falls outside the realm of what I'm trying to do with my articles.
Fun deck, and congrats on taking down a 6 person table. I don't have the patience anymore to take on a table that big so it's nice to see others do it.
I know you said you didn't like Deadapult, but Goblin Bombardment is a better replacement. It's a free sac outlet, and you have several ways of making a mess of tokens. It's especially nasty with Tombstone Stairwell. It definitely warrants consideration.
And I think you might be underestimating Gravecrawler and Skullclamp. Of course, they work better with a free sac outlet like Bombardment.
Nice work, Matthew! I really enjoy each article you put out, keep up the good work! And looking forward to that new "All in the telling"-episode, too!
While I remain impressed with the usefulness of raw data like this, I think you're a little too fast with your conclusions/impressions this time around; I agree with the basic premise that people probably put too much weight on over- and underdrafting and that it can sometimes serve as a convenient scapegoat for your deck not putting in the results despite drafting the "best deck" (or, what you think is the best deck, at least).
I really liked the population-ratio factor, but I feel like it would have been a much more powerful tool along with what I consider an absolutely crucial factor if we are to discuss over- and underdrafting; color (or guild) depth. One of the reasons I think this is important is that I think that the popularity ratio might actually differ significantly from the "depth ratio", at least in new formats or formats that are opaque or hard to figure out. I feel like Gatecrash is a set that confuses a lot of drafters; both blue guilds, per example, are not completely intuitive to draft properly (I spent some time figuring out the format myself, and I can only imagine there are less experienced drafters out there who spent more time - or haven't "solved" the format yet!).
For simplification purposes, let's say that the "depth ratio", just like the "popularity ratio", must end up totalling exactly 5 (that is, "there are no non-guild decks and there are exactly 8 decks' worth of card depth in an 8-man draft"). This might be a gross simplification but I don't think it makes it a lot less applicable. Your stats tell us that, at an 8-man table, the numbers would look like this:
Boros: Pop-ratio 1.26 - ~2 drafters.
Simic: Pop-ratio 1.11 - ~1,75 drafters.
Orzhov: Pop-ratio 1.03 - ~1,67 drafters.
Gruul: Pop-ratio 0.91 - ~1,4 drafters.
Dimir: Pop-ratio 0.69 - ~1 drafter.
If we go by the (quite faulty) assumption that every guild/color is equally deep, this would clearly be an easy explanation for Dimir's success (and Boros' somewhat lackluster results): This is the simple argument of over- and under-drafting that I think we agree is a gross oversimplification.
However, let us introduce some (purely thought up, by me, right now) color depth ratios. These are guesses off my sleeve of how the color depths of Gatecrash actually look, but aren't meant to be accurate, merely to illustrate my point:
Boros: CD-Ratio 1.1 - ~1,75 drafters.
Simic: CD-Ratio 0.7 - ~1 drafter.
Orzhov: CD-Ratio 1.4 - ~2,25 drafters.
Gruul: CD-Ratio 0.8 - ~1,4 drafters.
Dimir: CD-Ratio 1 - ~1,6 drafters.
If the numbers even roughly match up like this, it becomes apparent that the over-/underdrafted explanation might have at least some merit. Even if more people start drafting Dimir, there's room for quite an increase before the color becomes "overdrafted" though it will certainly get a little diluted.
I don't mean to (and certainly isn't!) disregarding your analysis, I just feel like there are lots of other factors to consider that the framing of these stats could perhaps help illuminate if done a little differently. Another consideration to color depth is the single colors; as an example, it is my impression that both Red and Green have cards that are valuable in both of their guilds; I think Gruul suffers in your stats by the high popularity of Simic and Boros both. On the other hand, the white cards are more firmly seperated by guild (Smite, Dutiful Thrull and Basilica Guards are a lot better in Orzhov, Denizen, Maneuver and Elite in Boros), which means that Orzhov is a little less affected by Boros' popularity (in my opinion, at least).
One of the concern with "color depth" is, of course, how good a deck needs to be to be "good enough": How much depth is needed to support one drafter? I think this problem might be one of the reasons that people keep liking Boros: A "decent" Boros deck (ie one which got eight ninths of a "deck", depth-wise) can still "get there" on speed and good draws some of the time, especially in a draft which is only three rounds. People are slow to learn if their feedback is not consistent. Decks with a more fragile game-plan, like Simic (not getting enough evolve/curve) or Orzhov (not being able to stay alive early on) might be scaring people off a bit faster since their flaws are more obvious even in a short three-round tournament.
I hope you found this comment helpful/insightful and would like to discuss it further, and once more thanks for a really great article!
Rock solid performance Mikey. You have such great instincts for the ebb and flow of a Magic game, and they really serve you well and help you eek out the tough matches. It is always fun to watch your videos. I was actually very surprised you beat Aphnird and his RG Land Destruction deck since your deck relies so much on Cloudpost, of which you only have 4 in your deck while Aphnird backs 15 Land Destruction cards. At first look, it doesn't seem like you have a prayer of keeping a Cloudpost in play. But again, your instincts in the match were great and the way you played it really gave me a lot to think about. Thanks for doing these videos and providing such great commentary.
In the first game when the Zo-zu player cast wildfire, I had Patron of the Orochi untapped. I could have saved my entire team by tapping it, and funneling my extra mana through Kamahl, pumping everyone out of Wildfire range.
In the second game, when the Vorosh player used Diluvian Primordial to grab Krosan Grip in my graveyard to blow up my Survival of the Fittest, I could have saved my Survival. With the Primoridal targeting the Grip, I could have tossed Genesis to Survival to search up Kozilek, then tossed Kozilek to Survival. My graveyard would have been shuffled, and I could have tutored up something else. Diluvian would no longer have a target.
Thanks for the shout out. I am happy to help AJ as you know. I think the promo codes have so many possibilities that could be good for advertising and just players rewards as you mention. Force of wills for all would be nice of course but so would promo codes for new players who take part in special new account tutorials. (Get a new account, take part in a tutorial, get promo code at the completion.) I expect to see magazine ads with inserted codes, and or code cards.
Another interesting price jump is Angel of Serenity which basically doubled in price over the last few days. The card has been featured in a number of top tier decks but as far as I know it is one of a few cards that have been consistently top tier for months. The reason for the jump is about as explicable as that of Speaker.
I am a little disappointed we aren't getting the Farseek promo but Forbidden Alchemy is pretty cool looking. Not sure I like it better than the original art, but if it is cheap enough Ill probably pick up 4.
Yea, Chalice of the Void can be a total nightmare to face. Nearly all the deck is in the 1 CMC range. I think this deck also need Echoing Truth and some artifact hate cards to SB.
I didn't consider to adding Echoing Truth, but it seems very reasonable SB option against the cards that can cripple this deck. I will test it. Thanks again.
This deck looks incredible- once again proving that Wizards should just outright ban Grapeshot once and for all.
As a sidenote, I think this deck needs Echoing Truth as a way to deal problem permanents such as Rule of Law, Ethersworn Canonist, Leyline of Sanctity, etc..
Language is the big barrier irl but local volunteerism and store sponsorship takes the load off of WOTC. Online, I assume the problem is dealing with various international trafficking laws and regulations and satisfying them all at once.
But they do that in the paper world, where the logistics is actually a factor. Every country has plenty of support, Italy has a big WotC presence, I've a couple friends that run game stores and they get absolutely everything an American store gets. Plus I remember when I was a judge, the organization was thorough with tests and communications, and after I stopped judging one of the Italian WotC organized play big honchos even called me by phone to ask me back (which required for me to take an exam and stuff.) I believe Italy, back in the first phase of development, had one of the highest numbers of players for non-US countries, but I think these days dozen of countries are at this level.
Except the MTGO country, where you could actually reach the world with just a link. :)
It's also mightily contradictory: they always tried to woo the international markets, since Rinascimento and Portal Three Kingdoms; they have product localizations into a lot of different languages; and they organize GP all around the world. Then MTGO still feels like a product directed to the American audience. It's weird.
Man, coupon codes for US players only is yet another slap in the face for us foreigners. They really should devise a workaround international rules - many other online games do.
It's just awful to feel yet again like a second-class costumer.
I had thought of that (the Misty Rainforest part, not the Lifetap :P), but then I noticed than in the previous version there weren't fetch lands caring for Forests, and the Savannah was already there.
Great article as always. Just curious, how big of a sample size do you use for something like GTC data?
Your Hollow could used to "regen" the creature you want to target with some other effect, when Glass Spinner is out.
Used to be one of the reasons for running the Kamigawa legendary lands. The other of course being Deaths Wail.
I see where you're going with the "challenge people to think critically about their presumptions about a format" - that's definitely a worthy goal and something a lot of us could learn to do better/more often!
I'm also glad you mention the transitional character of the data (unlike the principles derived from it), as I've run into a couple of people on the WOTC forums who will use that exact approach to the stats you present ("Stats say Orzhov is best so we should pick the Orzhov card here" in community drafts... Eugh).
Keep up the good work!
I don't disagree with you. I realize that there is a lot more complexity to an individual draft format than just looking at population ratio compared with win rate. There are many small factors that influence how different decks perform. The problem is that it is incredibly difficult to quantify those small factors. For example, you brought up the point of depth compared with popularity. The problem is that there is no unbiased way for me to determine that; the only thing I can think of is assigning playability numbers to the cards that can be drafted by certain decks, and then try to determine how a deck's popularity matches up against its depth. This introduces a tremendous amount of bias. There is already enough inherent bias in any kind of study I do, and so one of my biggest goals is always to eliminate as much bias as possible. Doing what you suggest automatically takes it out of the realm of what I try to do in my articles and puts it more into the realm of what I want people to do in the comments of my articles.
In any case, I agree with a lot of your analysis; Gruul suffers because its cards can be more easily poached by neighboring guilds. I would argue that there is also the problem that Gruul is just not all that good anyway, but the fact that its cards often go into Boros and Simic decks certainly hurts. Of course, neither of those two decks are performing all that well, so maybe there is a different problem going on.
The point I was trying to make with this article is that people always try to make reductive arguments about why decks are performing a certain way. They do this, not to understand a format better, but in order to confirm what they were already thinking about a format. This is confirmation bias. People look at my data, and they only look at the things that confirm what they were already thinking. I would challenge people to be a little bit more open-minded and try to actually understand the things I am demonstrating.
Finally, I think that the biggest problem with these studies is that they are outdated very quickly. The metagame of a format changes pretty quickly, and it is incredibly difficult to try to stay right on top of its pulse. I think that the principles in the articles are always applicable, so I recommend that people understand the principles, and then apply them to what they actually see. Again, this is the blueprint vs. map argument. Treat this like a map to help you understand where things are generally happening, but be aware of the immediate events that are happening in a specific draft. Don't treat this like a blueprint that you need to follow precisely.
In any case, thanks for the comments. I do appreciate what you're saying, and even agree with you, but it kind of falls outside the realm of what I'm trying to do with my articles.
Fun deck, and congrats on taking down a 6 person table. I don't have the patience anymore to take on a table that big so it's nice to see others do it.
I know you said you didn't like Deadapult, but Goblin Bombardment is a better replacement. It's a free sac outlet, and you have several ways of making a mess of tokens. It's especially nasty with Tombstone Stairwell. It definitely warrants consideration.
And I think you might be underestimating Gravecrawler and Skullclamp. Of course, they work better with a free sac outlet like Bombardment.
Nice work, Matthew! I really enjoy each article you put out, keep up the good work! And looking forward to that new "All in the telling"-episode, too!
While I remain impressed with the usefulness of raw data like this, I think you're a little too fast with your conclusions/impressions this time around; I agree with the basic premise that people probably put too much weight on over- and underdrafting and that it can sometimes serve as a convenient scapegoat for your deck not putting in the results despite drafting the "best deck" (or, what you think is the best deck, at least).
I really liked the population-ratio factor, but I feel like it would have been a much more powerful tool along with what I consider an absolutely crucial factor if we are to discuss over- and underdrafting; color (or guild) depth. One of the reasons I think this is important is that I think that the popularity ratio might actually differ significantly from the "depth ratio", at least in new formats or formats that are opaque or hard to figure out. I feel like Gatecrash is a set that confuses a lot of drafters; both blue guilds, per example, are not completely intuitive to draft properly (I spent some time figuring out the format myself, and I can only imagine there are less experienced drafters out there who spent more time - or haven't "solved" the format yet!).
For simplification purposes, let's say that the "depth ratio", just like the "popularity ratio", must end up totalling exactly 5 (that is, "there are no non-guild decks and there are exactly 8 decks' worth of card depth in an 8-man draft"). This might be a gross simplification but I don't think it makes it a lot less applicable. Your stats tell us that, at an 8-man table, the numbers would look like this:
Boros: Pop-ratio 1.26 - ~2 drafters.
Simic: Pop-ratio 1.11 - ~1,75 drafters.
Orzhov: Pop-ratio 1.03 - ~1,67 drafters.
Gruul: Pop-ratio 0.91 - ~1,4 drafters.
Dimir: Pop-ratio 0.69 - ~1 drafter.
If we go by the (quite faulty) assumption that every guild/color is equally deep, this would clearly be an easy explanation for Dimir's success (and Boros' somewhat lackluster results): This is the simple argument of over- and under-drafting that I think we agree is a gross oversimplification.
However, let us introduce some (purely thought up, by me, right now) color depth ratios. These are guesses off my sleeve of how the color depths of Gatecrash actually look, but aren't meant to be accurate, merely to illustrate my point:
Boros: CD-Ratio 1.1 - ~1,75 drafters.
Simic: CD-Ratio 0.7 - ~1 drafter.
Orzhov: CD-Ratio 1.4 - ~2,25 drafters.
Gruul: CD-Ratio 0.8 - ~1,4 drafters.
Dimir: CD-Ratio 1 - ~1,6 drafters.
If the numbers even roughly match up like this, it becomes apparent that the over-/underdrafted explanation might have at least some merit. Even if more people start drafting Dimir, there's room for quite an increase before the color becomes "overdrafted" though it will certainly get a little diluted.
I don't mean to (and certainly isn't!) disregarding your analysis, I just feel like there are lots of other factors to consider that the framing of these stats could perhaps help illuminate if done a little differently. Another consideration to color depth is the single colors; as an example, it is my impression that both Red and Green have cards that are valuable in both of their guilds; I think Gruul suffers in your stats by the high popularity of Simic and Boros both. On the other hand, the white cards are more firmly seperated by guild (Smite, Dutiful Thrull and Basilica Guards are a lot better in Orzhov, Denizen, Maneuver and Elite in Boros), which means that Orzhov is a little less affected by Boros' popularity (in my opinion, at least).
One of the concern with "color depth" is, of course, how good a deck needs to be to be "good enough": How much depth is needed to support one drafter? I think this problem might be one of the reasons that people keep liking Boros: A "decent" Boros deck (ie one which got eight ninths of a "deck", depth-wise) can still "get there" on speed and good draws some of the time, especially in a draft which is only three rounds. People are slow to learn if their feedback is not consistent. Decks with a more fragile game-plan, like Simic (not getting enough evolve/curve) or Orzhov (not being able to stay alive early on) might be scaring people off a bit faster since their flaws are more obvious even in a short three-round tournament.
I hope you found this comment helpful/insightful and would like to discuss it further, and once more thanks for a really great article!
Rock solid performance Mikey. You have such great instincts for the ebb and flow of a Magic game, and they really serve you well and help you eek out the tough matches. It is always fun to watch your videos. I was actually very surprised you beat Aphnird and his RG Land Destruction deck since your deck relies so much on Cloudpost, of which you only have 4 in your deck while Aphnird backs 15 Land Destruction cards. At first look, it doesn't seem like you have a prayer of keeping a Cloudpost in play. But again, your instincts in the match were great and the way you played it really gave me a lot to think about. Thanks for doing these videos and providing such great commentary.
This lets us have the ability for gift cards.
A couple play mistakes were pointed out to me.
In the first game when the Zo-zu player cast wildfire, I had Patron of the Orochi untapped. I could have saved my entire team by tapping it, and funneling my extra mana through Kamahl, pumping everyone out of Wildfire range.
In the second game, when the Vorosh player used Diluvian Primordial to grab Krosan Grip in my graveyard to blow up my Survival of the Fittest, I could have saved my Survival. With the Primoridal targeting the Grip, I could have tossed Genesis to Survival to search up Kozilek, then tossed Kozilek to Survival. My graveyard would have been shuffled, and I could have tutored up something else. Diluvian would no longer have a target.
Lazav has gone up (a little bit) as well. It's probably redemption season.
Thanks for the shout out. I am happy to help AJ as you know. I think the promo codes have so many possibilities that could be good for advertising and just players rewards as you mention. Force of wills for all would be nice of course but so would promo codes for new players who take part in special new account tutorials. (Get a new account, take part in a tutorial, get promo code at the completion.) I expect to see magazine ads with inserted codes, and or code cards.
Another interesting price jump is Angel of Serenity which basically doubled in price over the last few days. The card has been featured in a number of top tier decks but as far as I know it is one of a few cards that have been consistently top tier for months. The reason for the jump is about as explicable as that of Speaker.
I am a little disappointed we aren't getting the Farseek promo but Forbidden Alchemy is pretty cool looking. Not sure I like it better than the original art, but if it is cheap enough Ill probably pick up 4.
Yea, Chalice of the Void can be a total nightmare to face. Nearly all the deck is in the 1 CMC range. I think this deck also need Echoing Truth and some artifact hate cards to SB.
Kara, thank you for your comments.
I didn't consider to adding Echoing Truth, but it seems very reasonable SB option against the cards that can cripple this deck. I will test it. Thanks again.
I did an interview with Deluxeicoff and here are his current view on Stompy:
http://www.youtube.com/watch?v=kwOntadLP6g
I need to start putting Chalice of the Void back in my sideboards. :)
This deck looks incredible- once again proving that Wizards should just outright ban Grapeshot once and for all.
As a sidenote, I think this deck needs Echoing Truth as a way to deal problem permanents such as Rule of Law, Ethersworn Canonist, Leyline of Sanctity, etc..
Language is the big barrier irl but local volunteerism and store sponsorship takes the load off of WOTC. Online, I assume the problem is dealing with various international trafficking laws and regulations and satisfying them all at once.
But they do that in the paper world, where the logistics is actually a factor. Every country has plenty of support, Italy has a big WotC presence, I've a couple friends that run game stores and they get absolutely everything an American store gets. Plus I remember when I was a judge, the organization was thorough with tests and communications, and after I stopped judging one of the Italian WotC organized play big honchos even called me by phone to ask me back (which required for me to take an exam and stuff.) I believe Italy, back in the first phase of development, had one of the highest numbers of players for non-US countries, but I think these days dozen of countries are at this level.
Except the MTGO country, where you could actually reach the world with just a link. :)
Maybe because their core audience initially was mostly American. That and the logistics nightmare of having to serve a venue as large as THE world. :D
It's also mightily contradictory: they always tried to woo the international markets, since Rinascimento and Portal Three Kingdoms; they have product localizations into a lot of different languages; and they organize GP all around the world. Then MTGO still feels like a product directed to the American audience. It's weird.
A true Johnny knows how to do billing shenanigans too. :)
Codes for US BILLING addresses, where you have your bills sent are your business.. :)
Man, coupon codes for US players only is yet another slap in the face for us foreigners. They really should devise a workaround international rules - many other online games do.
It's just awful to feel yet again like a second-class costumer.
Ooh ooh also Glissa.
I had thought of that (the Misty Rainforest part, not the Lifetap :P), but then I noticed than in the previous version there weren't fetch lands caring for Forests, and the Savannah was already there.