Is going 4-0-drop in a league considered "hiding data"?
— Aaron Forsythe (@mtgaaron) October 13, 2017
Aaron Forsythe is the Senior Design Director for Magic: the Gathering, and on Friday the 13th he threw this gem of a tweet about MTGO Data into the ether. The idea here is pretty clear. Wizards publishes the deck lists that go 5-0, so players who don’t want their super secret sauce published to the public will drop out of the league after four wins, sacrificing prizes for secrecy. Aaron, a former pro player, wanted to know more about this practice, and how the community felt about it. I think he was genuinely surprised to learn how rampant this problem is.
The community responses were pretty consistent along two basic lines of thought. The first is the idea that pro players want the luxury of being able to test on MTGO without the threat of revealing information to the public. The second is the idea that Wizards of the Coast are the ones “hiding data” from the community.
To avoid being part of the potential public set. Your question conflates two sets of data. 4-0 drop may hide data from you, but not players.
— Joe Spanier (@FoundOmega) October 13, 2017
The only people hiding data are the @MagicOnline team by not publishing everything.
— Jeff Hoogland (@JeffHoogland) October 13, 2017
How did this all begin anyways? Well in case you missed it, Player of the Year and Pro Tour Hall of Fame member Paulo Vitor Damo da Rosa started writing a new column for Wizards of the Coast titled “Almost There” in which he featured two Standard decks that went 4-1 in league play. Paulo was interested in the decks, and they wouldn’t have been part of the usual dump of undefeated deck lists.
All this discussion started because people were annoyed that their 4-1 lists might end up in PV's article.
— Aaron Forsythe (@mtgaaron) October 14, 2017
My guess is that Wizards of the Coast 100% did not expect this kind of backlash to the new column by Paulo. I didn’t expect it either. I read Paulo’s article in the morning and thought to myself that it was interesting to see them start exploring quirkier decks and that this was a natural progression of content like the Pirate Stompy incident and more general stuff like all of Seth (better known as Saffron Olive)’s off-beat stuff (if you haven’t seen his Against the Odds series you’re doing yourself a disservice).
There’s another angle here that didn’t get too much coverage: how much access to “hidden data” has Paulo, a member of team Channel Fireball, been given? Forsythe claimed on Twitter that he was only given the two deck lists that were included in the article and while I’m inclined to believe him there’s certainly reason to be concerned. Why does a member of one pro team, or any pro player for that matter, have privileged information about the metagame that other players don’t have access to?
It’s important to note at this point that the very reason Wizards doesn’t release more metagame data, and why they’ve continued to pare down the data that they do release, is over concerns that it results in the meta being solved more quickly. It’s highly unlikely anything malicious is going on, and it’s highly doubtful that Paulo gained anything of value here, but having the door open for this possibility seems like a potential conflict of interest.
I’m not going to start a crusade to remove active pro players from Wizards bullpen of content creators, I’m just saying that the door they’ve opened here goes down that road.
What I do want to talk about is several data points I think Wizards should start publishing. I think that 99.99% of the community is really narrowed in on the idea of publishing deck lists. I’m bored of deck lists. I have 25 years of boring deck list data I can look at on the internet. Magic doesn’t really change a whole lot. Deck lists exist. Let’s move on with our lives.
What is "everything?" If we published all the 5-0's, that would not capture any of the best players' decks.
— Aaron Forsythe (@mtgaaron) October 13, 2017
1. Mulligan Data
Forget the contents of decks. Lets talk mulligan decision-making. Consider a format like Ixalan limited which by now has been played thousands of times on Magic Online. How many players mulligan to six-card hands? How many of those mulligan decisions are for hands that contain one or fewer lands? What’s the win-rate of a player who mulligans to five, on the draw, in game one? What about post-sideboard?
This is the kind of data that Magic Online has, that we’re not getting access to. You better believe that pro player teams work to build up this kind of data. Wizards could publish it so that the public could gain a better understanding of mulligan decisions. This seems to me like something that wouldn’t break any meta game. If anything it would probably give Frank Karsten something to write about.
2. Missing Land Drop Data
Along the same lines I’ve always been interested in how “forgiving” some formats or even decks could be. What’s the win-rate of a Ramunap Red deck that never hits a fourth land? What about a fifth land? In individual games, or even in a sample set with a hundred games, luck could play a very large factor in this sort of analysis. But what about the thousands or even millions of games worth of data available from Magic Online?
Even if, at a minimum, we could conclusively say that if you miss your third land-drop you need to start praying, it would be an improvement over what we know today, which is nothing.
3. Retired Format Metagame Data
This one starts to probably encroach on what Wizards is trying to avoid, but I’m genuinely interested in the metagame data for retired limited formats. Sure, they come back occasionally as flashback formats, but I want to know what the best deck ever in Rise of the Eldrazi draft was. I want to know what the most successful first-pick common in Ravnica-Guildpact-Dissension drafting was.
I think the community would enjoy pouring over this data. Magic fans across the globe host casual drafts of old formats all the time at their kitchen tables. This kind of data could provide a lot of insight into limited Magic in general. The list of reasons to do this is pretty long. The obvious concern is how it could impact the currently live formats, but I still think there’s some amount of information we could get about old formats.
The same is true for completely retired Standard formats. At this point is there any reason not to release every single deck list from a Standard meta where none of the sets are Standard legal anymore?
A lot of people say they want to see “Everything” but I think they’re just talking about deck lists. Magic is so much more than the 75 cards you sleeve up for a tournament. I want to know more about the actual game of Magic and I think the community would like to know more about it as well, and Magic Online should be able to give us that information without compromising the current Standard metagame.
So come on Aaron Forsythe, give us everything (just without deck lists for currently live Standard sets).