Price Signal
Know the fair value before you trade.
5 mins
TCG markets move on vibes, Twitter screenshots, and whoever pulled the chase card last night. Price Signal cuts through that chaos with a data-backed estimate of what a card is probably worth, so you can see when a listing looks high, low, or right on the money.
What you get on a card page
Open a covered Pokémon card and Price Signal shows:
- A fair value estimate: what the card looks worth from market structure
- A likely range so you know when a deal looks good (or when someone is asking for the moon)
- What-if sliders for demand, supply, and grading intensity
- Extra clues like cost to pull and PSA population when we have them
Fair-value estimates are for select English Pokémon cards in our latest training coverage. Not every card has one yet, and that's intentional. We would rather show nothing than invent a number.
How it thinks (without the black box vibes)
Price Signal is powered by an AI / machine-learning model trained on real market structure and sales-related signals. In plain English, it looks at three kinds of pressure at once:
- Demand: character heat and marketplace buying pressure
- Supply: how hard and expensive a card is to pull, listing flood, and set age
- Grading intensity: how hard collectors chase graded copies of that card
Then it estimates a structural fair value: what the card should be worth if those signals are doing the talking, not the latest Discord panic.
- Demand: how hot the card looks right now. Character heat and how buyers are absorbing supply in the marketplace both feed this side of the estimate.
- Supply: how hard and expensive the card is to pull, plus how flooded the market looks with listings and how old the set is.
- Grading intensity: how much that card is being graded. Heavy PSA traffic is a strong clue that collectors care enough to slab it, and that intensity feeds the fair value.
The insane part is the gap
Fair value alone is cool. Comparing it to the live market is where Price Signal gets spicy.
- Market way above the estimate: collectors are paying a desire premium (classic chase-card energy)
- Market way below the estimate: the card looks structurally cheap versus peers with similar supply and demand
That gap is the product. It is not a buy button. It is a spotlight that says: “Something interesting is happening here. Go look.”
How good is it? Meet R²
When people ask “how accurate is the model?” we use a score called R² (R-squared). Think of it as: how much of the price differences across cards can the model explain?
Cross-validated R²
≈ 0.86
About 86% of the price variation across covered cards is explained by the model in honest holdout testing.
That leftover ~14% is not a failure. It is where hype, thin markets, and chase premiums live. Those gaps are useful. An Umbreon trading way above structural fair value is not the model “being wrong.” It is the market screaming collector desire.
R² is not a crystal ball for tomorrow's price. It is proof that Price Signal is reading the market's bones, extraordinarily well, so you can trust the comparison when a listing looks off.
What you can do with it
- Drag the what-if sliders to preview hotter demand, flooded supply, or heavier grading activity
- Check Market Pulse for cards that look most undervalued or overvalued versus the estimate
- Use fair value as one input alongside price history, liquidity, and your own thesis
What we don't claim
- Price Signal is informational market data, not financial or trading advice
- The estimate is not a guaranteed listing price. Markets can (and do) diverge
- A gap does not mean the card will crash down toward fair value or skyrocket up to it. Price Signal is explaining what the card looks worth given current circumstances, not forecasting that the market must converge
- Coverage is Pokémon English for fair value today, and expands over time
Stop guessing. Start with a signal.
Try Price Signals free with a 10-day trial. One plan. No upsell ladders. Cancel anytime.