After a brief break and a lot of procrastination, I’ve successfully entered the CubeCon deck and match data for my cube and done some basic stats and analysis! It’s been interesting to see how my theories and goals hold up to actual play. Before I share any data, three caveats:

1. Experience and feel are much more important to me than picture perfect balance.
2. This data is from exactly 3 drafts from a very specific subset of players (CubeCon attendees interested in nostalgia cubes), and any analysis should keep these factors in mind.
3. My spreadsheet expertise and the dataset both have some serious limitations. Where decklists weren’t provided (pleaseeeee post decklists to Hedron Network), I constructed the deck for analysis myself based on the pool photos.

Source data

Card level analysis

Given that I have just Maindeck% and MW%, I wanted to mainly focus on a groupings of cards that might give me some good indications of pain points or uninspiring cards. I settled on the following:

• Unplayed cards: these cards never saw a maindeck
• Trap cards: these cards were played at every opportunity, but did poorly
• Outliers: these cards with multiple drafts of data did very well or very poorly

Unplayed Cards

First off, an honorable mention to Saffi, who saw play just once despite having 2 copies in each of 3 drafts. I’ll get more into Project X later, but nobody seemed especially interested. Some of these cards are more sideboardy (like Muse Vessel) or appear that way if you underestimate the density of targets (like Trygon Predator), and I’m not going to sweat those. Others are part of archetypes that didn’t come together that frequently (Saffi for Project X, Flame-Kin Zealot for tokens/dredge, the weaker blink cards, and a slew of the UR tempo cards), which I’d like to keep an eye on when tuning. Finally, we have generalist cards that might just be a little too weak (Pongify, Utopia Sprawl). It’s worth thinking about whether these are unappealing, bad, or both and considering cutting them accordingly.

Trap Cards

These cards were played in every draft but averaged a 1-2 record or worse. I’m not going to read too much into the lands (I’m not worried about the performance of Azorius Chancery), but the other cards are all very much in the red oriented aggro/tokens camps, which will show up a lot in the big losers and I’ll discuss at length in the archetype analysis.

Win Rate Outliers

For these categories, I looked for a win rate >= 5/6 or <= 1/6 while being played at least 2/3 of the time. There’s pretty obvious trends on both winners and losers - UG midrange/omnichord type cards did consistently well, and the BR aggro/tokens/madness and The Rack themes did not. Individual cards that surprised me are Stonecloaker as a winner (has a lot of tricks, but is not especially powerful) and Beacon of Unrest as a loser (at its best getting something huge early, but should give decks a lot of grind game if they play any number of top end creatures). I do want to try and replace the gold Rakdos cards with more powerful stuff in those individual colors, while leaving the on theme cards that underperformed.

Archetype level analysis

Here’s the draft results breakdown:

On a high level, I’m pretty happy with this. No one archetype (aggro/control/midrange/combo) failed to make it to the top half. The big hole is in the token strategies - we see some wider use of tokens, but the more focused token/sac decks had a very bad time. We also had strong UGx performances, which is period accurate at a minimum, though I’ll keep my eyes open for ways to make other blue decks more appealing.

As far as color balance goes, there were a lot of U and B decks which underperformed, and a relative lack of W and G decks which overperformed in turn. I’ll get more into the UBx situation later, but I think a pileup of these decks in the Prequel draft is a reasonable illustration of a potential cause. Conversely, in spite of a large number of aggro decks in Drafts 1 and 3, these decks had relatively good results. This is a great relief after hearing from some players that aggro was undertuned - there’s obviously still enough room for these decks to perform, but I will pay attention to see if they continue to excel.

UBx

My somewhat heavy handed nerfs to UBx appear to have succeeded! As I mentioned before Cubecon, my goal was to make UBx draftable but not sufficiently powerful to be dominant when contested. My theory on why this deck is drafted so much despite performing fairly poorly is that the control cards have a very easy appeal, despite the landscape of control differing quite a bit due to the lack of planeswalkers and single card win conditions. This means if a player sees something like a Terror late in a pack, they might think a controlling Black deck is more open than it really is. I might walk back some of the black changes in particular a little, since removal is overly sparse, but the poor rate on hard counterspells in blue appears to have resulted in blue decks with more definition performing better across the board than do-nothing control which does make me happy.

BR sac/token aggro

No bones about it, this is certainly my biggest disappointment with how the Cubecon drafts went. The sac and token aggro decks are some of my favorites, and put a lot of effort in the last updates into giving them a good dose of juice. The difficult part about just working off results and not seeing gameplay is that I’m not certain where these decks are falling short. The additional cards I’ve added that were less played at the time like Nantuko Husk and Fallen Ideal have been pretty disappointing, and a lot of the nice to have elements like burn and standalone creatures are contested by other aggro decks. It’s also possible the extent of board wipes in the cube goes a bit too far, but without seeing the losses it’s hard to make any strong conclusions.

I’ve considered a few different options, such as improving density of the appropriate cards (though I’m running out of good options at this point), looking for gameplay and power level appropriate signposts from outside this time period, and looking at ways to add the damage on the stack-like effects via custom cards. It’s also possible this was just a weird coincidence, so I do want to get some first hand data before doing anything drastic. The other potential thought is that the amount of go wide hate available to sideboards is a little excessive, and I’ll probably clip at least one of those.

Damage on the stack

I’ve had it suggested more than a few times that this cube should be played with damage on the stack, as this meta was played contemporarily. I’m sympathetic to this idea, especially given how helpful it is for a deck built around sacrificing their permanents. That said, I see some significant issues:

• Damage on the stack has not been in the rules for over 15 years now. Most Magic players have never played with it, and the players who grew up in this era are more confident in their memory of the rules than maybe they should be (myself included).
• Other rules about combat damage have changed as well during this time! Deathtouch, lifelink, and assigning damage to blockers have all changed in this time period, and some of the interactions of these with damage using the stack are edge cases most players likely won’t be able to interpret on the fly.
• The effects aren’t just limited to RB sac - blink gets a whole lot stronger to the point where I’d be a little concerned. Card evalution gets an unintuitive twist across the board, and

I’ll probably try and play at least one draft under damage on the stack rules I’ll formalize, but I think it’s wisest to not make this cube any harder to approach than it already can be.

Tarmorack

This is a very cool deck to me, but didn’t come together in any real sense during the event. The goal is to put your opponent in a bind early, playing discard spells to feed your goyfs, with Bob and GY benefits helping you grind. Ideally, this is a deck that should be pretty playable without The Rack, but a lot of players of this strategy gravitated towards a Korlash midrange shell rather than the discard centric one. I’ll keep my eyes on this deck, as 2 slots is a bit intense for something not seeing play, but I do think it has potential.

Blink/Touch

Blink was generally successful at this event, having very good synergies with the UGx core. Something that hasn’t had the same level of success has been the second-line blink cards: Voyager Staff, Ghostway, and Peel from Reality went unplayed in maindecks, so Aethermage’s Touch has proved difficult, since the density of both Touches and big creatures are harder to get right in limited. The US Nats blink-touch list had 6 “great” targets for touch (the dragons), 13 “good” ones (adding Venser and Cloudskate), and 20 “anything” ones, giving it odds of 35% “great”, 63% “good”, and 80% “anything”. In limited, that would equate to ~4, ~9, and ~10, with that middle number seeming the hardest to reach. Including 1 Touch in the cube is pretty low cost, so I’m inclined to let it ride, but I do want to see if it can be successfully used in the Chordless blink strategies. I’ll do some more digging for cards with sick ETBs that won’t primarily bolster UGx.

UR Aggro + the Dragonstorm problem

As I mentioned in my last post, I was punting on storm to give me a less tense situation to try and “walk the tightrope between trap archetype and overpowered combo deck”. The replacement I opted for was a UR tempo or Magnivore slant, with some different options like Gossamer Phantasm and Wee Dragonauts. This completely failed to connect, and given the lack of historical presence of this style of deck post-Time Spiral, I’m going to scrap that and take another look at storm. My goal is to minimize the number of cards that are only for storm while making it something that can be achieved when the stars align. A knock on requirement is making sure “failed” storm decks can still function - so far the outcome has been slightly awkward ramp decks that hardcast Bogardan Hellkite, but I hope to do better. I’ll be looking to Time Spiral Remastered for examples on how this might work.

Dredge + Reanimator

Because of the similar problems faced by Bridge Dredge to Dragonstorm, I tried to move a lot of the dredge support into madness/Solar Flare during cube creation. Neither really succeeded, despite a good amount of slots devoted to reanimation, looting, and fatties. That said, someone attempted this baller deck, 1-2d, and left me a recommendation - let’s give it a shot! In general, I’d like to have more cards with use in the graveyard, so these decks have some additional benefits.

Project X

Project X is also on my list of favorite archetypes, but I do somewhat understand why it wouldn’t see a lot of play. I think Saffi is reasonable enough to have at two even if that’s excessive ignoring this deck, and if I do add Bridge that’s another weird out for the combo. It makes sense that the fiddliness and color requirements would make this deck rare, so I’m mostly planning on letting it be for the moment.

Wrap-up and next steps

This is certainly more than enough to have written, so I’ll spend a few days tinkering on the update and then publish an article explaining my thought process. Here’s my starting thoughts:

• Unplayed/trap/poor performers - Cut stuff that is both bad AND unappealing that’s not archetype based.
• Even out lands - Utility lands into the general land slots and put in the missing pains.
• UBx - Find room for another removal spell, see if some of the nerdier threats can come out.
• Blink/touch - Cut some poor blink cards for more blink targets.
• Tarmorack - Cut some boring discard spells for more beef for the remainder of the gameplan.
• RB sac - Look for ways to make this deck stand out, including replacing some of the underperforming gold cards with monocolor options. Take a quick look at well-fitting but out of format cards here.
• Dredge/reanimator - Add Bridge from Below, think about other potentially useful tools and flashback cards.
• URx aggro/tempo - Out the dang cube, leaving some slots for other UR stuff.
• Storm - Explore ways to build this out with good failure cases.