As a Pokemon obsessed collector, me and my kids love to rip and open packs.
But we quickly learned that sometimes bad luck can hit OFTEN and HARD.
What are your actual odds when opening packs? Well, it turns out the odds, otherwise
known as "pull rates" were widely available and generally considered very reliable.
So, as a data analyst by trade, I couldn't help but crunch some numbers, including
market prices of cards, combined with pull rates to see which sets had better odds,
and which sets were AWFUL.
Thus began a journey that started with moving some numbers around in a spreadsheet, to
a much broader, more aspirational idea. What if I built a tool that I can use to run
these numbers and it update every day automatically?
And... what if I shared it with the world??
was born.
Collectrics is still pre-1.0, but it is moving fast. This page is a living trail of the bigger product improvements so you can see how the model, market tools, community features, and collector workflows are evolving over time.
Added Collectrics IQ subscription access, account modules, notification support, and admin grant tools. Market movers now account for demand volume and card price, search matching is sharper, and JustTCG history backfills are more controlled with rate-limit backoff and cleanup utilities.
Launched the first V1 partner API for latest card data, including authenticated card endpoints, API key tooling, partner docs, and a full dump example. The leaderboard and market pages also got a major redesign, with refined hero signals, set leaderboards, graded chart controls, and an IR-specific IQ model pipeline.
Added community polls, account watchlists, expansion entitlements, refreshed account/store UX, and a broader workspace feel for signed-in users. This release also introduced market divergence and demand-pressure signals so cards can surface when price, supply, and demand start moving out of sync.
Introduced Collectrics IQ as an experimental price lab for expected raw price and grading intensity. The beta added new model inputs, Google Trends support, grading intensity logic, IQ branding, leaderboard hooks, and model health views so the analytical side of Collectrics could grow beyond simple market averages.
Added eBay listing analytics, ended-status enrichment, raw price signals, homepage market charts, shipping metadata, set logos, and more stable trend views. This was a big step toward showing not just what a card is worth, but what the surrounding market is doing.
Prepared Collectrics for real users with account flows, login polish, checkout pages, store support, legal pages, shared navigation, dark mode, mobile navigation, and email magic link delivery through Resend.
Added the modeled Collectrics pricing layer, JustTCG ingestion, sealed product leaderboards, model parameters, precomputed leaderboard sparklines, and better page performance. This is when the app started turning daily raw inputs into a more opinionated pricing system.
Launched card leaderboards and reorganized the data pipeline into clearer daily and weekly stages. PSA population handling improved here too, along with the rebuild scripts that keep card histories, set values, card indexes, and leaderboards current.
Added card search, card indexes, rarity filters, sorting, default filtering, set metadata improvements, and mobile-friendly UX passes. This turned the site from a set/card viewer into something collectors could actually explore.
Built the first usable Collectrics web app: Node/Express serving, set pages, card detail pages, early data ingestion, PSA-oriented extraction scripts, and the first pass at the public frontend. It started as a practical way to answer one question: what are the numbers behind ripping packs?
