MTG ELO Explained: A Plain-English Guide for Magic Players

MTG ELO is a single number that tries to summarize how good you are at Magic: The Gathering. The system came from chess, it's a few decades older than most card games, and on Nerd Leagues it's the engine behind every player rating and every deck rating you'll see on the platform.

The short version: when you win, your number goes up. When you lose, it goes down. How much depends on who you beat — knocking off a higher-rated player earns you more than running over a lower-rated one. That's the whole concept behind MTG ELO. The interesting part is the math underneath, and the way Nerd Leagues stretches it from a 1v1 chess system into something that handles four-player Commander pods, multi-deck pools, and league-specific scoreboards.

Where ELO came from

ELO was invented by Arpad Elo, a Hungarian-American physics professor and chess player. The U.S. Chess Federation adopted it in 1960, FIDE picked it up in 1970, and it's now the same system you've seen in chess.com, ranked StarCraft, and pretty much every competitive online ladder built since.

The core idea is straightforward. Every player has a rating. When two players meet, the system calculates an "expected score" for each based on the rating gap between them. If you outperform what the math expected, your rating climbs. If you underperform, it drops. Beating a 1500 when you're a 1200 is worth more points than beating a 1500 when you're a 1700 — because the first result was less expected.

What an ELO number actually means

ELO is relative, not absolute. There is no objective "1500 means good" — a number only means something compared to the rest of the population in the same rating pool. In chess, 1200 is a casual club player, 2000 is an expert, 2700+ is world-class. MTG ELO benchmarks don't transfer directly from those numbers because the player pool is different and the variance is wider.

On Nerd Leagues, every player and every deck starts at 1000. From there, the rating drifts up or down with every game you record, and the gap between the top and bottom of any given leaderboard is what tells you how stratified that pool is.

How MTG ELO works

MTG ELO runs into a few problems chess didn't have to solve. Magic isn't always 1v1 — Commander pods are usually four players, Two-Headed Giant is two-on-two, and a single game can have three losers and one winner. Chess ELO assumes one winner and one loser; MTG forces the system to spread credit and blame across more bodies.

Nerd Leagues handles this by shrinking the rating change as the pod grows. The system uses a base K-factor of 20 (the "K" is just the maximum number of points that can swing in a single one-on-one match-up). A three-player free-for-all drops K to 19, a four-player to 18, a five-player to 17. Two-Headed Giant works differently — each teammate's rating still updates independently, but the K-factor is split across the team so a 2HG win moves your individual rating about as much as a 1v1 win would.

The full math, including how the rating change is calculated for every winner-loser pairing in a multi-player game, lives in The Math Behind Your Rating.

Players and decks both have ELO

ELO on Nerd Leagues isn't only attached to you. Every deck you build has its own rating, calculated independently from the same game. Win a Commander pod with your Atraxa deck and both your ELO Rating and your Atraxa deck's ELO Rating climb. Switch to a different deck the next week and your Atraxa rating sits where you left it while your new deck builds its own track record.

This matters because deck strength and pilot skill are two different things, and most players want to see them separately. Player ELO vs Deck ELO goes deeper.

One game, several scoreboards

ELO ratings on Nerd Leagues live on more than one scoreboard at a time. A single recorded game updates several of them at once:

• Your all-time ELO — the running tally for your entire history on the platform.
• Your yearly ELO — every calendar year starts fresh at 1000, so each year gets its own leaderboard.
• Your per-league ELO — every league you join tracks its own rating for you, isolated from other leagues.
• Your per-season ELO — within a league, commissioners can run time-bounded seasons that have their own rating reset.

Same game, multiple ratings updated. Win a Commander pod inside a league with an active season and that single record bumps your all-time, your yearly, your league, and your season ELO together.

When does an ELO change become real?

ELO changes apply immediately. The moment a game is recorded, every affected scoreboard updates and the change is final unless someone flags the game and a commissioner archives it. There is no provisional period or 7-day waiting window. Older versions of Nerd Leagues used an attestation system that gated ratings until participants confirmed, but that was retired — recording a game today is the commitment.

If a commissioner does archive a game later (because it was wrong, fake, or a duplicate), Nerd Leagues recomputes the affected ratings from scratch by replaying every remaining game in chronological order. What Happens to ELO When a Game Is Archived covers the cleanup flow.

Where to go from here

If you want the math, read The Math Behind Your Rating. If you care about the difference between your ELO and your deck's, read Player ELO vs Deck ELO. If you're curious about the yearly reset, read Yearly ELO. Otherwise, the rest of this section walks through how ELO updates after each game, what plus/minus means, and how to read your rating chart.