🔥How we compute a game's popularity
How do you compute a game's popularity?
The popularity score shows which games are genuinely in demand right now, not just the most visited. Each interaction is weighted differently: a page view is worth 1 point, adding to a wishlist 8 points, a click through to a shop 12 points, and setting a price alert 15 points — the signals closest to real purchase intent count the most. Recent activity weighs more (its importance halves roughly every 14 days, and anything older than 60 days barely counts), each user can contribute at most 20 points per game so nobody can inflate a title artificially, and games with very few interactions are pulled toward the average until they gather enough signals from more people. The score is recomputed weekly and orders games in the rankings, in the “For You” recommendations and across the discovery rails.
↑ Top⚡How “Hot in Romania” works
What is “Hot in Romania” and how is it computed?
“Hot in Romania” shows the games getting the most community attention over the last 7 days. Each game gets a score by summing user actions: adding to a wishlist is worth 5 points, a click through to a partner shop 3 points, and a game-page view 1 point. Unlike the general popularity ranking, here we use a fixed 7-day window with no time weighting — it captures strictly “what's hot this week”, not historical popularity. The list refreshes continuously, so it always reflects the trends of the moment.
↑ Top🏷️How we compute the price verdict
How do you decide whether a price is good (“Is it a good time?”)?
The verdict badge tells you at a glance whether it's a good time to buy, comparing the lowest price available now (in stock, converted to RON) against the average of the daily lowest prices over the last 90 days and the lowest price recorded in that window. We mark an offer “Lowest price in the last 3 months!” when it's within 2% of that low, “good deal” when it's at least 10% below the 90-day average, and “expensive” when it's 15% or more above it; otherwise it's a price in line with the recent average. The verdict only appears when we have price history for at least 7 distinct days — otherwise the data is too thin to conclude. We use the exact same thresholds everywhere: on the game page, in the Trișache assistant's answer, and in the public price data.
↑ Top📊How we compute the price barometer
How do you compute the price barometer?
The price barometer is a market index computed from all active offers in the catalog, split into two segments: all products, and identified board games only. To avoid skewing the average, we only count “live” offers — in stock or updated in the last 30 days — and show both the mean and the median price, all converted to RON at the BNR exchange rate. We also track how many products have an active discount, the average discount, and how many have hit their 90-day low again (within a 2% tolerance), plus a top 20 of the biggest discounts. The figures are recomputed once an hour and can be freely cited as a benchmark for the Romanian board-game market.
↑ Top🎯How we rank deals
How do you rank deals (the deal score)?
The deal score sets the order of discounts by combining how big a discount is with how fresh it is. The discount is measured against the same product's recent average price at the same shop (over the last 60 days), and an offer only enters the ranking if it drops by at least 10% and the product is in stock. Fresher offers rise higher: “freshness” halves every 14 days, so an old discount gradually falls below an equally large newer one. That way the top of the list holds discounts that are both meaningful and current.
↑ Top📉How we compute the biggest price drops
How do you compute the biggest price drops?
The drops report tracks real discounts over the last 30 and 90 days, using our price history. The discount percentage compares the lowest current in-stock price with the same shop's median price over that window — we use the median, not the maximum, so a single old offer can't inflate the discount artificially. We include only games with at least 5 recorded historical prices at that shop and only drops of at least 8%, to cut noise, and the current price must be freshly updated, so an offer left over from a dead feed doesn't show up as a phantom discount. We display the top 25 drops per window, with all prices converted to RON.
↑ Top📈How we compute “Most wanted”
How do you build the “Most wanted” ranking?
The “Most wanted” ranking lists the top 20 games by real community interest over a 30-day window and updates every hour. Each game gets a weighted score in which a click through to a shop counts the most (12 points), adding to a wishlist counts 8 points, and a plain view 1 point; we count each visitor only once, and accessories are excluded. Next to each game we show how far it rose or fell versus the previous 30-day period, with a “new” tag for fresh entries. The 30-day window is fixed, with no time weighting, so it reflects recent interest rather than an average that fades over time.
↑ Top🏪How we compare Romanian shops
How do you compare Romanian shops?
The shop scoreboard compares only active shops with at least 10 unique game titles in stock — below that, the averages become unreliable. For each shop we show how many unique titles it has in stock, what percentage of its range is actually available, the average price in RON (from each title's lowest in-stock price), and on how many games the shop offers the market's lowest price, also as a percentage of its catalog. We add the buyers' average rating (from 1 to 5, on delivery, packaging and service), the number of reviews, easybox delivery availability and the free-shipping threshold. The board is ordered by number of unique titles and updates once an hour.
↑ Top🔍How we compute unmet demand
How do you compute unmet demand (games wanted but out of stock)?
Unmet demand shows the games Romanians search for or add to their wishlist, but that no Romanian shop currently has in stock. We count how many distinct people — both logged-out visitors and accounts — viewed or saved a title in the last 90 days, and a game only appears in the report if it has at least 5 interested people, to protect individual data. We show the top 30 titles with the highest unsatisfied demand, plus a ranking of the top 15 publishers by number of wanted-but-unavailable games. The data refreshes every hour.
↑ Top🏭How we compute publisher demand
How do you compute the publisher demand signal?
The demand signal shows how many people were interested in a game over the last 12 months — that is, visited it or added it to a wishlist — and what percentage of them have not yet bought it (unmet demand). The market estimate starts from the number of people who want the game, multiplied by a conservative 12% conversion rate and by the average price over the last 90 days, to give a realistic rather than optimistic figure. To protect privacy, we only show figures when there are at least 10 distinct interested users; below that threshold we show no number. All data is aggregated and anonymous — we never identify individual people.
↑ Top🧩How we recommend similar games
How do you recommend similar games? What does “similar” mean?
For each game we build a “similarity profile” from title, publisher, designers, mechanics, categories, mood and description. “Similar” weighs several things at once — mechanics and how it plays, theme, designers and complexity level — not just theme; that's why a game with the same mood but much heavier or for a different player count can appear lower than you'd expect. When you like several games, we compute their average profile and show the titles closest to it, breaking ties by popularity; we can exclude expansions or keep only in-stock offers. It all runs on profiles prepared in advance, with no call to artificial intelligence at search time, so the answer is instant. If none of your reference games has enough data, we'd rather show nothing than force weak recommendations.
↑ Top🇷🇴When a game has a Romanian edition
When do you consider a game to have a Romanian edition?
A game is considered to have a Romanian (or Hungarian) edition if at least one of four independent signals confirms it: an official edition recorded in BGG data, a translated rulebook, a community-voted localized title, or an explicit language mention in the product name. We do not guess the language from how the title's letters look, precisely to avoid confusion (for example, “Hive” does not automatically become a Romanian title). Titles proposed by the community through the Play and Help sections only count after at least 3 weighted votes and an 80% consensus (a logged-in user weighs 1.0, an anonymous one 0.25). Based on these signals we show what percentage of the BGG Top 100 and Top 500 has an RO or HU edition, and which top-ranked games are still untranslated.
↑ Top🤖How Trișache verifies rule sources
How does Trișache verify rule sources?
Trișache, our rules assistant, finds answers by combining two methods in parallel — exact-keyword matching and meaning-based matching — then merges the results into a single ranking. Each source gets a trust weight based on who wrote the rules: official publisher material weighs the most, official translated or verified material about 70%, community sources 40%, and “fan” content (unofficial variants, house rules, player-made solo modes) is excluded entirely and never appears in an answer or a citation. We first find the exact passage, then rebuild a broader context around it, and always display the source of each answer at the end — including a translation notice when the text was translated.
↑ Top🗂️How we build the catalog and gather offers
How do you gather every shop for one game onto a single page?
We automatically read Romanian board-game shop catalogs every day and link each product to the correct game by its BGG identity, not just by name. All offers that match the same game land on the same page, sorted by price, and prices are collected in the shop's currency and converted to RON at the official BNR rate on the day they were collected. We re-check offers several times a day; if a price hasn't been reconfirmed for more than 7 days, we drop it from the “lowest price” calculation so we don't send you after an expired offer. We add new shops constantly — if one is missing, let us know.
↑ Top