Explore the DeckBuilding and Analysis Tool SorceryDeckDiff.com!

Written by bitshifter

One of the great joys of Sorcery: Contested Realm (SCR) is brewing a new deck. Understanding exactly what changes you make, and their magnitudes during revisions, can be challenging. Once you’ve built a deck, other questions arise:

Has your deck converged toward something similar to what others have already created, or is it relatively novel? Maybe it only appears similar to known decks, but in reality, it’s quite different.

These questions can be answered by defining how we measure and visualize similarities between decks. This tool can compare two Curiosa.io decks:

✨ SorceryDeckDiff ✨

SorceryDeckDiff serves two purposes:

Showing card-level differences between two decks
Computing deck-level similarity scores between two decks

These two types of information serve different purposes:

Card-level information enables fine-grained inspection of differences/similarities between decks.
Deck-level statistics can compare the deck as a whole to provide a broader picture of similarities and differences.

Cards in the collection and maybeboard are not considered in the diff or the deck-level similarity scores.

Join the Eternal Durdles on Patreon

Comparing Decks

At its core, SorceryDeckDiff compares two decks (A & B) by showing the cards present only in deck A in the left column, those only in deck B in the right column, and any cards in both decks in the middle column.

The middle column shows the quantity of the shared cards in each deck, so “2/4” means there are two copies in deck A and four copies in deck B. Cards are grouped by each category of card (e.g., Minion, Magic, Aura, Artifact, and Site), and the difference is shown within each category group.

Using SorceryDeckDiff to compare two decks with the same Avatar highlights key differences in their approaches. To use SorceryDeckDiff, enter the Curiosa.io URLs of the two decks you want to compare and click “Compare Decks.”

The comparison page breaks down each card category using the three-column difference format. At the bottom are summary statistics and a mana curve comparison.

Let’s use two recent extensions of the Harbinger precon (queried on 3/16/2026) as an example:

Budget Harbinger Precon by Lord_of_Itza

Artifacts (3)

3 Toolbox

Minions (38)

3 Hauntless Head
2 Aaj-kegon Ghost Crabs
4 Forsaken
4 Spectral Stalker
3 Winter Nymph
3 Coy Nixie
3 Gibbous Nightgaunts
3 Headless Haunt
2 Nommo Monitor
3 Phase Assassin
2 Hounds of Ondaros
2 Lacuna Entity
2 Shoggoth
2 Dormant Monstrosity

Magics (19)

3 Blink
2 Displace
4 Riptide
3 Grapple Shot
2 Swap
1 Into the Abyss
2 Shapeshift
2 Call of the Sea

Collection (10)

1 Zap!
1 Blink
1 Disenchant
1 Lightning Bolt (scr)
1 Magic Missiles
1 Drown
1 Grapple Shot
2 Horrible Hybrids
1 Shoggoth

Use Our Link to Buy Cards!

Harbinger Upgraded Precon by Retsudo

Auras (2)

2 Falling Star

Minions (35)

2 Aaj-kegon Ghost Crabs
4 Forsaken
4 Willing Tribute
3 Estranged Loner
3 Heretics of Seth
3 Nommo Monitor
3 Failed Mutation
2 Gloam Toads
2 Lacuna Entity
2 Slimy Mutants
3 Gnarled Wendigo
2 Dormant Monstrosity
1 Yog-Sothoth
1 The Doom of Dilmun

Magics (23)

2 Second Wind
3 Desecrate
2 Murder of Crows
3 Scatter
3 Swap
3 Abyssal Assault
3 Four Fat Frogs
2 Into the Abyss
1 Howl from Beyond
1 Call of the Sea

Collection (10)

3 Desecrated Ground
4 Horrible Hybrids
3 Shoggoth

✨ See the Comparison ✨

If the decks change in the future, the contents of this link will change as well. Two different approaches were taken. Let’s focus on minions:

Cursor-hovering over entries displays each card image.

The common minions from the precon base appear in the middle column. Lord_of_Itza’s Budget Harbinger Precon on the left includes more Voidwalk minions than Retsudo’s Harbinger Upgraded Precon on the right.

Lord_of_Itza’s list has many two- and three-mana cost minions. Retsudo’s list has many five-cost or higher minions. This difference is displayed in the mana curve distribution. Retsudo’s is shifted to higher mana costs; visualized as the entire curve shifted to the right:

Deck-level statistics for Spellbook and Atlas similarity are computed as the ratio of overlapping / total cards — Jaccard similarity — which ranges from zero to one. Two variants of this statistic are computed:

One considers only distinct cards (i.e., each card has quantity one).
The other considers total card quantity.

These are slightly different ways of interpreting deck similarity. These two decks are highly divergent, scoring only 0.2 and 0.15 similarity for Spellbook cards using distinct and total card metrics, respectively. Similar decks would score in a range of 0.4-0.6. Highly similar decks would score between 0.7 and 1.

These two precon extensions diverged significantly, as expected, given Gothic’s many options. The ability to perform these pairwise comparisons to different decks clearly highlights differences that may be less obvious than if you only viewed two decks side by side.

For example, the competitive Archimago decks from 2025 were less similar than expected when compared side by side. The “Load Demo” button will randomly select a pair of Archimago decks from that era to show as an example if you would like to delve further.

Comparing the Meta

While this aggregate similarity can be used to compare two decks, the power of the summary statistic lies in its ability to quantify the similarities across many different decks. This large-scale comparison can help better understand the meta.

Since we can determine whether decks are similar or dissimilar, any deck with a similarity above a specific threshold can be associated with a link. Compare enough pairs of decks, and you can build a network of decks connected by similarity. This was the premise of a video at the end of 2025, with decks in the Top 8 finals of competitions, creating a network of the meta and allowing visualization and Exploration of these similarities.

✨ Accompanying Interactive Network* ✨

*Use a computer to scroll in/out of the network with the trackpad/scrollwheel. Additional details on the network’s construction are provided on that web page in the help menu via the “?” button.

Choose Your Comparison

The goal of these tools is to enable deck comparisons of interest to identify which deck-building spaces have not yet been explored. I hope you find these useful on your journey through SCR!