Guide
Grade, compare, learn.
For a practical starting point, follow our guide to making your own Limited set review.
1. Choose a set
Use Quick Grade for a focused card-by-card flow, or browse the complete Arena Limited pool. Depending on the release, that pool is based on official Draft or Play Booster information plus reviewed Arena differences.
2. Save an evaluation
Give a card a normal F-to-A+ grade. Add a separate build-around grade when its ceiling depends on a specific deck or synergy. Nothing changes until you select Save.
3. Review your patterns
Filter My Grades by set, color, rarity, type, timing, or grade. Compare the active selection with community or available 17Lands-based benchmarks, and import or export grades when needed.
4. Track pre-release calibration
Only the final normal grade saved before a set’s Arena release cutoff can contribute to its prediction score. Build-around and post-release grades remain useful history but do not alter that score.
5. Choose what to share
Profiles, public analytics, and anonymous community participation are controlled in Settings. Guest grades remain on this device for up to 30 days unless you sign in and merge them. Read the privacy notice for details.
What do the grades mean?
- A —
- A standout card that can take over a game.
- B —
- Strong, efficient, and usually an early pick.
- C —
- A solid playable that fills out your deck.
- D —
- A weak or narrow card you prefer to replace.
- F —
- A card you generally leave out.
Use + or − for finer distinctions. Your normal grade assumes a typical deck in this Limited set. A build-around grade describes its potential with the right support.
How benchmark grades are calculated
myLimitedGrades calculates its own benchmarks from 17Lands Public Data, licensed under CC BY 4.0. These are not official 17Lands grades, and 17Lands does not endorse this analysis.
For each card, we count copies in the opening hand or drawn during each game, then divide the observations from won games by all observations. Multiple copies in one game count separately. The current method requires at least 500 copy observations, which does not necessarily mean 500 separate games.
Within each set and event-format dataset, we compare each eligible card’s win rate with the average and standard deviation of eligible card win rates. Each card contributes equally to that average. We assign F-to-A+ grades in bands one-third of a standard deviation wide, centred on C. Basic lands contribute to this calculation when they meet the sample threshold, but ordinary basics do not receive published grades.
For example, with an average card win rate of 56% and a standard deviation of 3 percentage points, 56% maps to C, 59% to B, and 62% to A. These are illustrative values, not fixed win-rate thresholds.
We use the games included in each published snapshot, without adding player-rank or deck-color filters. Snapshot dates, sample thresholds, and the cards included in the calculation can produce different results from the live 17Lands site. These benchmarks describe observed outcomes, not a card’s objective strength or how much it caused a win. Card details show the dataset date and methodology version where available.
Calibration measures agreement with the available 17Lands-based benchmark, not your overall drafting skill. Numeric grades are steps on the F-to-A+ scale, from 0 to 12.
Start grading