Why the Data Matters
Every tier list, meta share, matchup read, and archetype page on PremodernHQ stands on one thing: the decks underneath it. Get the data wrong and everything built on top of it lies.
The Foundation Everything Rests On
The numbers you read here aren't hand-typed opinions. They're computed from thousands of real tournament decklists. When a tier reads “Tier 1,” it's because the decks say so. When a matchup leans one way, it's the placements talking. That only works if the underlying corpus is clean — and clean data is harder than it looks.
A single mislabeled event, a mis-parsed card count, or two spellings of one player's name quietly distorts a tier list, inflates a meta share, or splits one person across two leaderboard rows. Small corruptions compound. So before any of it reaches a page, every deck runs the gauntlet below.
Built on the Community's Work
Premodern exists because of the players, organizers, and communities who record its tournaments and preserve its history. We don't replace any of them — we bring together publicly-posted community results, standardize them, and keep a link back to where each came from. Our thanks to:
You - the community
Players, organizers & contributors who report results, submit decklists, and send corrections
Magic Online (MTGO)
League & Challenge results
Moxfield
Representative sample decklists
Community tournament archives
Publicly-posted paper event results & decklists
Community metagame coverage
Published metagame-share trends
Scryfall
Card data, images & price estimates
MTGJSON
Per-retailer price history (incl. MTGO tix)
premodernmagic.com
The official ban list & format rules
What Clean Data Demands
Aggregated decklists arrive messy — formats blended together, placements garbled, counts inflated, duplicates everywhere, provenance lost along the way. Trusting them as-is is how a metagame picture goes subtly, confidently wrong. Each deck on this site passes these gates first.
Format Legality
Premodern is a precise window — 4th Edition through Scourge, minus the ban list. Decks carrying cards outside that pool are caught and held out, so a stray Modern or Legacy list never sneaks into the metagame as if it belonged.
Parse Verification
Card counts, mainboard size, and sideboard limits are checked against what a real 60-card deck can be. Inflated quantities and malformed lists — the kind that quietly skew aggregate counts — are flagged rather than averaged in.
Player & Deck Deduplication
The same event reported by two trackers, or one player spelled three ways, collapses into a single canonical record. One person, one row; one deck, one entry — so standings and meta shares count reality, not echoes.
Archetype Classification
Every deck is matched to a real archetype by its actual card profile, not its self-reported label. Generic and mislabeled buckets are resolved to the deck they truly are, so tier lists reflect the format as it's played.
Built to Stay Right
Placements are renumbered consistently, dates reconciled against event records, and banned cards flagged rather than silently dropped. The pipeline's job isn't to gather the most decks — it's to gather the ones that are actually true.
New results flow in, run the same gauntlet, and join a corpus that only gets cleaner over time. Nothing reaches a page until it has earned its place in it.
What Makes It Different
Plenty of sites track Magic data well. These are the things we choose to put work into — not a knock on anyone else, just where we try to add something worth having.
Every result traces to its sources
When duplicate reports of one tournament are merged, we keep the link back to every site that reported it - provenance is preserved, not flattened away.
The format, on a map
Events are resolved to a city, country, or online and aggregated into a world heatmap, so you can see where Premodern is played - not just who won.
Price history, not just today's number
Cards carry a price trend over time across multiple retailers, including MTGO tix - a moving picture rather than a single snapshot.
Open and free to read
The underlying data lives in a public mirror with no paywall and no login. Corrections and additions are welcome via the contribute path.
We report the format, we don't decide it
Legality follows premodernmagic.com exactly. Tier lists and evaluations come from the data alone - never from a commercial relationship.
Guarded against silent breakage
An anomaly guard flags unexpected data drops and a review queue surfaces anything needing a human, so quality problems get caught rather than published.
Spot something wrong or missing? The fastest way to improve the dataset is to send a correction or a missing decklist.
Honest Limitations
No gauntlet makes data perfect. Bad data in is still bad data out — if a tournament report is wrong at the source, our cleanup can normalize it but can't invent the truth. We surface what we can verify, hold back what we can't, and would rather show a smaller, trustworthy picture than a complete but corrupted one.
- •Coverage is partial. Not every event is reported, and online results are far easier to capture than paper ones. Meta shares reflect the decks we can see, not every deck played.
- •Source errors propagate. A wrong placement or misspelled card in the original report can survive cleanup. We catch the obvious, but we don't claim to catch everything.
- •Classification is a judgment. Archetype labels and tiers are derived from card profiles and placements — useful, but a model of the format, not gospel. Edge cases and brews won't always land where a purist would put them.
- •Small samples are noisy. A handful of results can swing a number. Treat thinly-represented archetypes and recent shifts as signal worth watching, not settled fact.