About PCR

Who built this, what it does, and what it doesn't try to do.

Quick facts
  • What it is: live eBay pricing for the top 300 MLB prospects, cross-referenced with minor league stats.
  • Cost: free. No signup, no newsletter, no paywall.
  • Monetization: eBay Partner Network affiliate links. That's the whole business model.
  • Coverage: Bowman Chrome 1st autos, Pro Debut, Bowman's Best, plus the major parallels.
  • Refresh cadence: prices snapshot multiple times per day, stats once daily.

Why this exists

I'm a designer. I started collecting prospect cards and realized the actual hobby tools out there are kind of bad. They're either ugly, paywalled, locked into one brand of card, or full of buy-this-now hype that ages badly. The thing nobody was doing was the one thing I actually wanted as a collector: live prices in one place, next to how the player is actually performing in the minors.

That's the whole site. Stats next to price. If you're trying to figure out whether a prospect is having a real breakout or whether the market has already priced it in, you can compare both halves in one view. That's the only thing PCR tries to do better than anyone else.

Who runs it

Just me. Will. I'm not a card shop, not a podcast network, not affiliated with any breaker, grading service, ranking site, or any of the people you've seen on hobby Twitter. The site started as a personal project and grew. It still runs as one.

If you want to reach me, the easiest way is @PC_Radar on X.

What this isn't

It's worth being clear about what PCR is not, because the hobby has a lot of overconfident tools.

If you want the full mechanics of how prices are sourced, what gets filtered, and the known parser quirks, that's all on the methodology page.

What's open and what's still rough

The data engine is the part I'm proudest of. Prices come from eBay's real API, they refresh through the day, and the cross-reference with stats is the genuine differentiator. The interaction and design language is intentional and consistent. The verdict model (underpriced / fair / overpriced) is honest about its inputs.

What's still rough: the parser that classifies cards into parallels gets fooled on edge cases (some IP-signed bases get tagged as 1st autos, for example), and the sample size on low-ranked prospects is sometimes too small to say anything confident. When the site is uncertain, it tries to say so instead of inventing precision. If you spot something wrong, please tell me.

Last updated: 2026-05-16