How cards are scored

Two signals, one score

Vulcan Highlander's approach is inspired by a combination of Canadian Highlander's points list, and EDHPowerLevel's impact scores. Canadian Highlander uses points to limit the number of high-powered "game changer" cards in a deck. Vulcan Highlander takes this concept a step further, to tackle widespread power creep across hundreds of cards.

Scoring tens of thousands of cards based on their merits via human (or AI) judgement would be a dauntingly huge and error-prone task, rife for arguments and disagreements. Instead, we use a formula very similar to EDHPowerLevel's to determine an impact score for every card. This score is a weighted blend of a price rating and a popularity rating, based on data from EDHREC. For nonland cards the price rating is weighted at 1.25 and the popularity rating at 0.75. Nonbasic lands are scored on the same two ratings, but popularity counts about half as much. The most popular nonbasic lands are not constrained by color and can provide useful effects in any deck. Basic lands always cost 0 points.

The price input is the card's market price in US dollars, capped at $100. Past that point, price measures how few copies were printed rather than how strong the card is. We take the second-cheapest vendor quote rather than the cheapest. This mitigates the possibility of a stale or erratic low price listing resulting in an incorrectly scored card.

The popularity input is the card's rank among all Commander-legal cards based on how many decks run it, using EDHREC's data. A legendary creature's rank counts the decks it commands as well as the decks playing it in the 99. EDHREC's data covers only the most recent 2 years on a rolling basis, which is perfect for our purposes. It means older cards can't accumulate more popularity over time: all cards are evaluated on the level playing field of current popularity.

For further details see the "Discover the algorithm behind this tool" section at EDHPowerLevel, as our approach to impact scoring is inspired by theirs.

A note on cards banned in Commander

At least for now, Vulcan Highlander uses the same banned list as Commander. Banned cards inherently lack relevant price and popularity data for Commander. In the future, after we have more real-world feedback on our pointing system, we may unban and manually point some of the currently banned cards.

Translating impact scores to points

A card's number of points is determined by its impact score:

The lower bands are deliberately wider. Power creep lives in breadth: hundreds of good-not-broken cards, not a handful of outliers. About 2,000 cost a single point, about 215 cost 2, and only 48 cards in the entire format cost 3 or more.

Some landmark examples:

• 1 point: Arcane Signet, An Offer You Can't Refuse, and every shock land
• 2 points: Necropotence, Badgermole Cub, and the original dual lands
• 3 points: Cyclonic Rift, Sensei's Divining Top, Roaming Throne
• 4 points: Demonic Tutor, Rhystic Study, The Soul Stone
• 5 points: Chrome Mox, Mana Vault, The One Ring

Why 50 points?

The point budget was determined by analyzing Commander decks from MTGGoldfish, which is a gold mine of historical deck lists. We focused on Commanders that have maintained their popularity over time and support numerous different deckbuilding strategies (such as Atraxa, Praetors' Voice). So the decks we analyzed included a wide variety of cards, and we could see how the composition of decks changed over time for the same Commanders. We focused specifically on decks built in 2016 and 2020. We also made sure to exclude deck lists that had been later modified to include newer cards, and we excluded deck lists that were clearly scratchpads and not real decks.

We determined that a budget of 50 points sits between 2016 and 2020 decks. The median 2016 deck fits with about six points of headroom, experienced as room to upgrade, while a typical 2020 deck does not fit without trimming. This targets the goal of the Vulcan Highlander format: to play Commander at a typical power level from the 2016-2019 era, while still allowing the inclusion of newer cards. But you can't just fill your deck with power-crept cards: with a limited point budget, you must choose carefully.

When a new set gets points

We score the cards in a new set the day it's officially released, or very soon thereafter ("we" is really me - I'm one guy, with a day job). When we display a card's points, we show them with an asterisk, to indicate the points are preliminary. Our assignment of points is derived from EDHREC data, which contains information about players' decks from the most recent two years. To be compared fairly to older cards, we need a full two years of data for a new card. Until then, a card's points may move up or down. We re-evaluate cards on a quarterly basis each year. Analyzing the data, we found that starting from a week after release, the number of decks playing a new set's most-played cards grows by about 56% every six days, and by four months that is down to about 4%. After that the points still drift, but they no longer jump.

Surprisingly, we found that the prices for new cards have mostly settled by the release date (not pre-release). Pre-order prices are where the volatility is. We went back through what Marvel Super Heroes cards actually sold for on each day around its release. At pre-order prices, 23 of its cards would have cost points. By the release date that was down to 8, and it has since settled at 9.

Popularity is the more difficult aspect to assess, as it drifts over time. EDHREC has a rolling two-year window for counting the decks that cards are in. A card printed last week has existed for only a tiny fraction of that window, so only a few days' worth of new decks on EDHREC can possibly contain it. So a one-week-old card is judged against only about 6% as many decks as a card that has been out two years or more. So even if a new card is genuinely popular, it takes time for this to show up in the data.

Interestingly, we found that a card's popularity only ever drifts in one direction: up. We know this because we checked: over a six-day period we tracked the forty most-played cards in each of six sets, ranging from a week old to two years old (The Hobbit to Bloomburrow). In every set, at least nine cards in ten appeared in more decks at the end of those six days. This makes sense: as more time passes since the release of a set, and more new decks appear in EDHREC data, the more opportunity there is for new cards to appear in them. But what we score is a card's rank, not its raw deck count. Every card's count is climbing, so a card can turn up in more decks than it used to and still slip down the ranking. That's how a card's points can move down even though its popularity only ever moves up. Since EDHREC publishes only current deck counts, with no history, and Vulcan Highlander itself is new, the chart below is mostly estimated. It shows the expected number of cards with points from Secrets of Strixhaven (released in April 2026). We currently have 1 data point for comparing our estimates, and it comes out close: at four months we projected 7 of those cards would have points, and the list shows 8.

Cards from Secrets of Strixhaven that would cost points, by the set's age

1. 2 day 7
2. 2 day 14
3. 4 day 30
4. 5 day 45
5. 6 day 60
6. 6 day 90
7. 8 day 120
8. 9 day 180
9. 14 day 365
10. 16 day 730

What the points cannot see

Points are determined card-by-card and do not attempt to take into account infinite deck building possibilities. A deck built as a tight synergy engine plays stronger than its point total suggests, and a random pile of expensive staple cards plays weaker. This has always been true in every Magic format. The elegance of the point system is that we can follow market signals for scoring cards instead of taking on the impossible task of individually evaluating the relative merit of every card in every possible deckbuilding strategy.

Guarding against future power creep

The impact scores are based on a current evaluation of EDHREC data (July 2026), compared to our 2016-2020 baseline. Assuming the current power creep trend continues, that means many older cards will drop further in price and popularity over time, which would lead to their points also dropping. That would defeat the purpose of the Vulcan Highlander format, so we made a rule: once a card is two years old, we lock in a "floor" for its points. Its points can still go up, because a new card can make an older one better, and an old card can still rise on its own. If that rise turns out to be temporary, the points can come back down to the floor, but never go below it.

For example, let's take a look at Soul Warden, first printed in 1998. It's a 1 point card and continues to be a Commander staple. Without a rule to prevent its points from dropping, if a new "strictly better" similar card is printed, then its price and popularity would drop (I'm looking at you, Liliana the Faultless).