How OPMT Scores Work: LT, ST and MT Explained

Every card on OPMT carries three numbers from 0 to 100: an LT, an ST and an MT score. They are the backbone of the whole site — the Hot Deals page, the Long-Term Deals page and the Sell Signals page are all ranked by them. This post explains exactly how each one is calculated, because we believe a score you can't inspect is a score you shouldn't trust.

LT Score — the long-term fundamentals

The Long-Term score answers one question: how desirable is this card likely to be in a year or more? It is a weighted average of eight independent categories, each normalised to 0–100. The current default weights:

• Pull rate (21%) — how many booster packs it takes, on average, to open one copy. A Manga rare that appears once in ~2,500 packs has a fundamentally capped supply; a common alt-art that drops every 250 packs does not. We research these rates manually for every special rarity and maintain them per set.
• Character popularity (18%) — a franchise-wide popularity ranking. Luffy, Zoro, Shanks and Nami cards hold value differently than minor side characters, independent of rarity.
• Sales velocity (14%) — how fast the card actually trades over the last 30 and 90 days, measured against the typical velocity of its own rarity class. A Manga rare that sells twice a month is liquid for a Manga rare; comparing it against bulk-rare volume would be meaningless, so each class has its own baseline.
• Scarcity (11%) — how many copies are listed for sale right now, with fewer listings scoring higher on a smooth curve from 25 copies down to 1,000+.
• Rarity type (11%) — Manga and SSP variants score highest, followed by Wanted Posters, SPs, Treasure Rares and alternate arts.
• US/EU arbitrage (10%) — the card's TCGPlayer price versus its CardMarket price. When the US market pays 50% more for the same card, the European price historically tends to catch up rather than the other way around.
• Rarity relative value (8%) — is this card cheap or expensive compared to the median of its own set and rarity class, and does its sales velocity back that price up? This is the score's valuation conscience: it rewards strong cards that are currently priced below their peers.
• Set age (7%) — older sets stopped being printed long ago; new sets are still flooding the market with fresh supply.

Two design decisions matter here. First, every threshold is a smooth interpolation, not a hard step — a card with 26 listings scores almost the same as one with 25, instead of falling off a cliff. Second, class baselines use medians and trimmed means rather than plain averages, so one outlier card can't distort the baseline for every other card in its class.

ST Score — reading the offer book

The Short-Term score hunts for supply squeezes. Instead of looking at price history, it walks the actual list of current offers, cheapest first, and asks: how many copies would buyers need to absorb before the cheapest listing is 25% higher than today? And at the current, confidence-weighted sales velocity, how many days would that take?

A card with only four cheap copies between today's price and a +25% price level, which is selling every day, scores very high. A card with a wall of 300 evenly-priced copies scores near zero no matter how popular it is. The full 100 points break down into sales velocity (22), copies-to-+25% (18), the 7-day price trend (18), copies-to-+50% (14), estimated days to target (13), US/EU arbitrage (10) and scarcity (5).

One detail we're strict about: a card with zero recorded sales in 30 days gets zero velocity points and an infinite time-to-target. Illiquid cards with thin offer books can look like squeezes — they aren't, because nobody is buying.

MT Score — fundamentals meeting momentum

The Medium-Term score is a blend: 40% ST, 40% LT, and 20% from the card's 30-day price trend. It exists because the best medium-horizon opportunities are cards that are both fundamentally strong and already moving — a high LT score tells you what to hold for years, a high ST score what might pop this week, and MT sits deliberately in between.

What the scores are not

No score predicts the future. What they do is compress thousands of data points — offer books, detected sales, pull rates, cross-market prices — into a consistent, comparable ranking that would take hours to research per card by hand. We publish the full methodology in the guide, and Premium users can change every weight and build their own formula. Treat the scores as a research shortlist, not a buy button — and never invest money you can't afford to lose in cardboard, however beautiful.