Sets picked by AI
Our algorithm scores every set from 0 to 100 based on price per element vs. the series average, the number of unique minifigures, discount depth and the price history trend.
How does AI Picks work? Methodologyexpand
Every set gets a score from 0 to 100 built from five signals. The higher the score, the stronger the case for buying now, not a month from now.
- Price per element: we compare the current price per brick to the average for the whole series (Star Wars, Technic and so on). The further below the series average, the higher the score for this component.
- Unique minifigures: a set with minifigures that appear only in this set has greater collectible value.
- Exclusive parts: elements you cannot get in other sets (printed pieces, unique molds) push the score up.
- Discount depth: the difference between the current price and the LEGO catalog price. The deeper the discount, the higher the score.
- Position in price history: where the current price sits against the last 30/90 days. The lowest price on record = the maximum score for this component.
Important: we do not take subjective opinions into account (“nice design”, “interesting theme”): the algorithm looks only at the data. A high score does not mean “you must buy”, only “objectively this set is well priced and makes collecting sense”. The purchase decision is always yours.
We update the ranking every few minutes based on fresh prices from the scraper. The weight of each component is shown next to every set. Check the “Why” line under each recommendation to see which signal worked the strongest.
AI scores take into account: price per element vs. the series average, the number of unique minifigures, discount depth and price history. This is not investment advice.





























