General Introducing the Cube Map – find out how your cube fits within the Cube Universe!

See the Cube Map here!

I’m super excited to reveal what I’ve been working on for the past few months – the Cube Map! We at Lucky Paper partnered with Cube Cobra to analyze all the cubes on the site and place them on one map. The results are incredible – we see the emergence of Vintage and Legacy Cube “landmasses” as well as the emergence of “islands” of set cubes, combo cubes, and more!

I hope you have as much fun exploring it as I do. Let me know what you find!
 

Jason Waddell

Administrator
Staff member
This was kind of creepy, literally the first cube my mouse hovered over was one of my own:
uVRaKIj.png
 
so Far RTL seems to mostly occupy the southern coast of the main continent. Classic powered cubes like wtwlf123's occupy the northwestern reaches of the same continent.
Screenshot 2021-03-24 112745.png
 
I have claimed the main continent. As the main continent’s first and rightful ruler it is my right to name it and I have named it

Cubanica

All those who would oppose me shall be crushed
 

Onderzeeboot

Ecstatic Orb
I have a good spread between my four cubes :D Apart from Cubanica, I also have a settlement on Peasantine, one of the larger islands in the archipelago, and quite a mellow place, though this particular town seems to be keen on clandestine home improvements! I wonder what's going on there?! A bit to the east we find a quaint little village on Bête-le-Baux, a quiet, little stretch of land whose inhabitants appreciate finely curated 2 player formats and guaranteed land drops. And finally, all the way in the southeast, beyond the Sea of Nostalgia, lies a village under construction on the western tip of Old Cubanica. It's true, many people moved to Cubanica over the past decade, as the power curve of Magic was gently nudged to an elevated level and old cards fell by the wayside one by one, but there are still people left who remember when Cubanica was still New Cubanica, and Old Cubanica... well, you get the idea. It's slow building here with most of the workforce moving northwest, but just you wait, one day this town will be finished, and people will look at it and reminisce about the age where a card that created two 3/3 elephants for 5GG total could become the most expensive card in a set for a bit.

Cubes.JPG
 

landofMordor

Administrator
yeah, this is sweet! If there was some legend overlay of what some of these "masses" are, that would be awesome.

We thought about it, but ultimately decided that it'd be too prescriptive/reductive. Also, we don't even know what some of the clusters are! Partly the intent was to just create a super user friendly map so that people could explore it and discover clusters for themselves :)
 
We thought about it, but ultimately decided that it'd be too prescriptive/reductive. Also, we don't even know what some of the clusters are! Partly the intent was to just create a super user friendly map so that people could explore it and discover clusters for themselves :)

Totally understandable! We seem to be having plenty of fun assigning our own labels (and backstories!) to various areas of the map anyways
 
Capture.jpg


I claim this land Hondenuras


Hello, fellow citizen of Hondenuras! I live a bit further south, in the quiet region of Alternítas. It's the dense blue bit — we have nice quiet winters, but the spring really has a kick to it.

Out of curiosity, do the colors mean anything?

EDIT: Will this be maintained for any future cubes? Or was this done manually?
 
image0.png

I found this on the Cube Brainstorming Discord.

I don't entirely agree with the characterization of the "starter/budget" section, as a lot of our Riptide cubes with non-budget Fetch/Shock manabases fall into that area. Some examples include Onderzeeboot's Wheel of Change, Erik's "Cool Side" cube, and Sigh, a Cube (although that last one kind of reaches into the MTGO modern section).

J4dDJGf.png


For some reason, my unfinished 2007 cube fits broadly in the "unpowered fair stuff" category along side the Cultic Cube, which seems a little bit weird to me given the card pool of each. The only reason I can see this is because I was trying to get to a similar aggro density as the 2020 versions of these types of cubes (which are actually a distinct cluster- whoever made the original reference chart probably missed the nuance there).

In either case, I don't know where the version of my Highball cube that I'm working on right now would actually fall in this categorization. I'm thinking it will fall somewhere between it's current position and the 2020 brainstorming cubes. I could see it falling closer to Sigh's cube as well.
 
This was an interesting find while idly perusing cube names (the average one, not the guild one)
Screenshot 2021-03-24 152344.png
It seems that the middle of Cubanica corresponds with "average" selections of cards, which seems to play out when checking out a few other cubes in that bright green cluster. Aspects of RTL cubes, but also very high power creatures and cards like Griselbrand etc.
 

landofMordor

Administrator
I found this on the Cube Brainstorming Discord.

I don't entirely agree with the characterization of the "starter/budget" section, as a lot of our Riptide cubes with non-budget Fetch/Shock manabases fall into that area. Some examples include Onderzeeboot's Wheel of Change, Erik's "Cool Side" cube, and Sigh, a Cube (although that last one kind of reaches into the MTGO modern section).
Worth noting that the figure you show is not a LP thing, it was a Discord member exploring the map and labeling as they went. So I heartily agree the naming convention could be improved :) and those labels also shouldn't be taken as definitive or objective.

And, to explain about why the clustering tended to spread out the Riptide cubes: if you check out "Cluster Details" you can see the cards which most differentiate cubes here. The Global cards are those which set the Riptide cubes apart from all 20K cubes -- stuff like Wharf Infiltrator and Archfiend of Ifnir. Budget cubes and "modal cubes" like to run these cards, too, and Vintage Cubes usually dont, which is why the Riptide cubes are near to explicitly budget cubes. However, the Locally Defining cards are what set Riptide clusters apart from the 15 nearest clusters (in this case, budget cubes and stuff), and those Local cards include fetches and shocks.

In other words, the algorithm sees some overlap between Riptide cubes and budget cubes, but then the Fetch/Shock manabase you mention is significant enough to warrant a separate cluster :) Tjornan also wrote a nice article on the math which I'll link below:

https://luckypaper.co/articles/mapping-the-cube-landscape/
 

Jason Waddell

Administrator
Staff member
Worth noting that the figure you show is not a LP thing, it was a Discord member exploring the map and labeling as they went. So I heartily agree the naming convention could be improved :) and those labels also shouldn't be taken as definitive or objective.

And, to explain about why the clustering tended to spread out the Riptide cubes: if you check out "Cluster Details" you can see the cards which most differentiate cubes here. The Global cards are those which set the Riptide cubes apart from all 20K cubes -- stuff like Wharf Infiltrator and Archfiend of Ifnir. Budget cubes and "modal cubes" like to run these cards, too, and Vintage Cubes usually dont, which is why the Riptide cubes are near to explicitly budget cubes. However, the Locally Defining cards are what set Riptide clusters apart from the 15 nearest clusters (in this case, budget cubes and stuff), and those Local cards include fetches and shocks.

In other words, the algorithm sees some overlap between Riptide cubes and budget cubes, but then the Fetch/Shock manabase you mention is significant enough to warrant a separate cluster :) Tjornan also wrote a nice article on the math which I'll link below:

https://luckypaper.co/articles/mapping-the-cube-landscape/

This is coming from the context of someone who has a Masters in Statistics (and finished PhD coursework), I am extremely impressed by this. Most statistical analyses you run into in these sorts of contexts are super shaky, but this does such a good job at educating the reader on stuff like PCA and clustering in a way that is so clear and visually accessible. I was not expecting to find something so thorough and concise when I clicked on that link. Well done guys, you knocked it out of the park. Looking forward to seeing what you guys come up with next.
 
Worth noting that the figure you show is not a LP thing, it was a Discord member exploring the map and labeling as they went. So I heartily agree the naming convention could be improved :) and those labels also shouldn't be taken as definitive or objective.
Oh I should have been clearer about that.
And, to explain about why the clustering tended to spread out the Riptide cubes: if you check out "Cluster Details" you can see the cards which most differentiate cubes here. The Global cards are those which set the Riptide cubes apart from all 20K cubes -- stuff like Wharf Infiltrator and Archfiend of Ifnir. Budget cubes and "modal cubes" like to run these cards, too, and Vintage Cubes usually dont, which is why the Riptide cubes are near to explicitly budget cubes. However, the Locally Defining cards are what set Riptide clusters apart from the 15 nearest clusters (in this case, budget cubes and stuff), and those Local cards include fetches and shocks.

In other words, the algorithm sees some overlap between Riptide cubes and budget cubes, but then the Fetch/Shock manabase you mention is significant enough to warrant a separate cluster :) Tjornan also wrote a nice article on the math which I'll link below:

https://luckypaper.co/articles/mapping-the-cube-landscape/
Thank you for such a detailed response! I hadn't see the article yet when I wrote my previous post, it is quite enlightening! Now that I have a better understanding of how the clusters work, I've noticed a few interesting ones which I might go in to for a post at a later date.

Thank you so much!
 
Top