How the hypemeter works

Every tracked game gets a hypemeter score from 0 to 100 -- our best guess, right now, at how anticipated it is. It's not a review score and it isn't trying to predict sales; it's a single number built from 10 signals we can actually measure, blended together with fixed weights.

The signals

• IGDB anticipation (weight 5) -- how many people have marked the game "hyped" on IGDB. This signal is saturating: it hits half its maximum contribution once a game reaches 25 IGDB hypes, and keeps climbing slowly after that.
• IGDB want-to-play (weight 1) -- IGDB's want-to-play counter, expressed as a share of IGDB-wide want-to-play activity rather than a headcount: a livelier cousin of the anticipation count above. Saturates at 0.0068% of that activity. On a game's own page, this row reports where it stands among tracked games rather than that raw share.
• Want-to-play momentum (weight 2) -- the week-over-week change in the game's IGDB want-to-play share: the first-derivative counterpart to the level above. A negative week -- the share falling -- scores zero rather than a negative number; only growth earns points here. Saturates at 1.7% week-over-week growth.
• Wikipedia attention (weight 3) -- trailing 7-day pageviews on the game's Wikipedia article, pulled from the Wikimedia pageviews API. This corrects IGDB's enthusiast-only skew with a mainstream-attention signal. Saturates at 4,000 views.
• Wikipedia edit activity (weight 1) -- the number of revisions to the game's English Wikipedia article in the last 30 days. Edits complement views the way supply complements demand: editors swarm an article when something is actually happening, and readers follow. Saturates at 6 edits.
• YouTube trailer views (weight 3) -- the highest view count across the game's known trailers. Saturates at 600,000 views. Optional: this source needs a YouTube API key; without one it's cleanly absent (0 points, "no data yet") rather than breaking anything.
• Outside arrivals (weight 1) -- visitors who land on a game's page from outside Hype Climb (a search engine, a link, a bookmark), counted once per visitor per day. Clicks from elsewhere on Hype Climb -- our own homepage, rankings, or charts -- are deliberately excluded, so this signal can never feed on our own ranking. Saturates at 50 outside arrivals in a week.
• Storefront breadth (weight 1) -- how many of five places we check list the game: Steam, GOG, its official site, Reddit and Discord. A slow-moving commercial-footprint prior rather than a measure of anticipation itself. Saturates at 3 of 5 links.
• Recent updates (weight 1) -- how many events we recorded for the game in the last 30 days. Chart debuts and trailer-milestone events are excluded: an event derived from this score, or from a signal that already feeds it, must not feed back in. Saturates at 2 recent events.
• Release proximity (weight 2) -- a dated upcoming game earns more of this credit the closer release gets; an undated game gets a fixed, middling share; a game keeps most of its proximity credit for 60 days after release, then it decays further the longer ago it shipped.

Each signal is saturating: past its threshold, more of the same input keeps helping but with steadily diminishing returns, so one outlier number (a game with thousands of IGDB hypes, say) can't blow the scale out for every other game. A game with no data yet for an external source (no Wikipedia article, no known trailers, no YouTube key configured) simply contributes 0 for that row -- the row still shows, so the score never hides which signals it does and doesn't have.

The 10 signals are combined as a weighted average and scaled to 0-100: each signal's saturated value (0 to 1) times its weight, divided by the total of all the weights (20). You can see this same math broken down per-game on any game page, under "Why this score?".

Hype freezes at release

A game's hype score is a pre-release quantity. It stops moving the day a game releases or is marked cancelled: the score freezes at whatever it reached, we stop collecting the external signals above for that game, and the frozen number stands as a permanent record of how anticipated it was -- not a live figure that keeps decaying afterwards.

Why IGDB-led, for now

Right now the two IGDB-family signals together carry the most weight (6 of 20) simply because IGDB is the source with the most real volume behind it before Hype Climb has a meaningful readership -- our own update counts are still small for most games. Wikipedia pageviews and YouTube trailer views are our first step toward diversifying away from a single enthusiast-catalogue source: Wikipedia adds a mainstream-attention check, and YouTube adds a trailer-reception check -- though YouTube stays optional (it needs an API key we haven't wired up in every environment) so its absence never breaks the score. That's an honest imbalance, not a permanent design choice: as the site gains readers and generates activity here, and as we tune the new signals against real collected data, we expect to shift weight further away from IGDB. The weights above are tunable constants, not settled science.