As part of Ryan and I's commitment to bringing cold, hard data to the Vintage Community (however that may be defined), we would like to start monthly installments of MTGO metagame reports. Our data sets will include both the results of Daily events and (eventually) the aggregate results of individual Vintage challenges. Before we dive into the data, we should take a moment to explain certain elements of our methodology (which will likely be cut and pasted in subsequent reports).
Methodology and FAQs
What are Vintage Dailies?
Vintage Dailies are four round swiss events that fire, as the name would imply, at least once a day. Wizards relays the full decklists of the 3-1 and better finishers on their website. On Friday, Saturday, and Sunday, there are three daily events - someone at Wizards somehow selects one of them to be published though the details are unknown (Brian Kelly believes they purposefully choose events in which he doesn't 4-0. We are doubtful that this is the case, but wish to humor him by bringing his theory to the collective consideration of the Vintage community). From these lists, Ryan and I categorize decks into archetypes and assign them several descriptive tags. They get entered into a spreadsheet that allows us to do more detailed analysis.
What are the current Archetypes and what decks do they include?
Our list of archetypes will change repeated over the evolution of the format, in part based on what Ryan and I see being played in events, and in part based on feedback from you (hint, hint).They are best used to answer the question "What does the Vintage metagame look like?" The current archetypes are:
- Mentor - This is not any deck with one or more Mentors. This category refers specifically to what was previously Gush Mentor. The current iteration of the deck features Cantrips and restricted cards as the draw engine with Mentor and planeswalkers as threats that win the game. The vast majority of these decks have been UWR, though it would also include Esper and Sylvan (UWRG) Mentor lists.
- Delver - During the first month post-Gush restriction, several versions of more aggressive lists have emerged. The defining characteristic of these decks is the card Delver of Secrets combined with an assortment of two drop threats that complement a slim manabase with Cantrips and restricted cards. Such two drops include Young Pyromancer, Eidolon of the Great Revel, Harsh Mentor.
- Shops - The card Mishra's Workshop is all sorts of broken and so we label those artifact heavy decks after the namesake card. Most decks are aggressive versions feature Arcbound Ravager and Sphere effects. A small minority are Smokestack variants or aggro decks with Null Rod instead of Arcbound Ravager. Of note, we do not consider the decks like 2-Card Monte to be Shops decks. Most Vintage players have a very specific style of play in mind when the term Shops is used, and therefore we categorize Shops-based combo decks like 2-Card Monte into the "Other Combo" category.
- Eldrazi - Aliens invaded every format with the printing of Thought-Knot Seer and Reality Smasher, and Vintage has proven no exception. The two subtypes are White Eldrazi (featuring both Thalias and other hatebears) or Colorless Eldrazi (normally unpowered, with Null Rods and Eldrazi tribal synergies).
- Paradoxical - Paradoxical Outcome decks have quickly emerged as a fixture of the metagame. The actual win condition of these decks tends to be pretty variable. Vault/Key, Mentor, Tendrils of Agony, random creatures, and even Emrakul, the Promised End have all been used (I know I would also use Auriok Salvagers if it was feasible on MTGO).
- Oath - Decks running the card Oath of Druids in the main deck. We do not consider decks that run an Oath transform as part of the Oath archetype (though those decks would have been tagged as Oath). We also give Oath of Druids priority over other archetypes. Paradoxical Oath, Gush Oath, and Oathstill are all considered to be part of this category.
- Dredge - Both Transformational and Anti-Hate sideboard dredge decks. Graveyard based, loads of zombies, Cabal Therapies...you know the deal.
- Null Rod - The remnants of the Null Rod pillar. We consider BUG Fish, Hatebears, 3-5c Humans, and Merfolk as part of this archetype.
- Big Blue - Your Tinker-Vault/Key deck from 3-4 years ago. The most common variant right now is Grixis Thieves.
- Blue Control - Landstill and other control decks. Lots of counterspells, Jace TMS, and random ways of winning the game.
- Other Combo - Non-Paradoxical Outcome based combo decks. For instance, Dark Petition Storm, Ad Nauseam, Gifts Combo, etc.
- Other - If it doesn't fit elsewhere, it goes here. We've used this for Infect, The Mountains Win Again, and that random Modern deck that someone accidentally or intentionally submits.
What are the current tags in use?
Tags are descriptive modifiers with no upper bound. Unlike archetypes, they are not exclusive and most decks are given multiple tags. They are best used to answer the question "How many decks in the field are running _____?" Current list of tags (please chime in with suggestions):
- Blue Control
- Big Blue
- Null Rod
Don't players like Rich Shay, Brian Kelly, and Montolio significantly skew the results of Dailies?
Yes. Yes, they do. Rich had 12 3-1 or better finishes in the month of May. Brian Kelly had 13. What does this mean? Good players can win with mediocre decks, but they will trend towards playing good decks. If Rich plays a deck, does well with it, but decides the deck if flawed in some way, he typically abandons it and moves on to a different deck. It holds that these repeat performances do have some value in determining the best decks in a field. However, it's difficult to assign a quantitative value to that. The answer we arrived at was to create two different columns: one for total finishes and one with redundant results filtered out. Google Sheets allows us to count unique entities within a given archetype, so no matter how many times Brian places with his Paradoxical Dragonlord deck, it only counts as one Paradoxical deck in the end. Use this data as you will.
What is this "Delta % Total" thing?"
Short answer: something we were experimenting with. Long answer: theoretically, it should be possible to infer win rates from the percentage of decks that 4-0 compared to 3-1. If you take a deck with a 50% win rate, the odds of it winning four matches in a row is 1/16. The odds of it winning three out of four matches is 4/16. The ratio of 4-0 decks to 3-1 should be 1:4 if a deck has a 50% win rate. When Ryan tried to work backwards to calculate the win rate, we quickly discovered that the sample size was much too small to derive meaningful data. However, we still wanted a way to measure a compare decks based on these results. We created a weighted category called "total wins" (pretty self-explanatory) and compared that distribution to the unweighted distribution. That is the "Delta % Total" column - the decks that had more 4-0s would have a positive value and decks with relatively more 3-1s would have a negative value.
Alright, that was boring...let's get to the data.
We are going to pass on this with only 2 Vintage Challenges in this month, but in future, we'd take the data from individual challenges and combine them to look at trends and matchup vs matchup data.
All information needs to taken with a grain of salt. I would caution people from drawing rigid conclusions from any individual data set. Is this metagame any different from previous months?
Is any of this significant? No, we need more data and less jumping to conclusions.
On a more positive note, it looks like we have more players playing Vintage on MTGO than in previous months. Reason for that? Who knows...