ChasePulse
Back to Home

How ChasePulse Works

Complete transparency about our probability models and methodology.

Our Mission

ChasePulse helps collectors track the probability of finding serial-numbered chase cards from Topps products. By aggregating publicly visible sightings from break videos, eBay listings, and other sources, we provide data-driven estimates of how many copies of a card remain "in the wild."

Our goal is complete transparency. Below, we explain exactly how our probability models work and what assumptions they make.

Model A: Deterministic Count
N = serial_total (e.g., 99 for a /99 card)
k = count of APPROVED sightings
publicly_remaining = max(N - k, 0)

Model A provides a simple, deterministic count of how many copies have been publicly documented versus the total print run. This is a lower bound on remaining copies, as it only counts cards we've actually seen.

Limitation: Model A assumes every sighting is unique and accurate. It doesn't account for cards that may never surface publicly (e.g., in private collections).

Model B: Bayesian Estimate
N = serial_total
k = approved sightings
q ~ Beta(2, 6) — detection probability
f = clamp(0.15 + 0.10 × log₁₀(1 + total_sightings), 0.15, 0.85)
D ~ Binomial(N, f) — cards that become public
k ~ Binomial(D, q) — cards we actually see
Compute: P(D < N | k) and E[N - D | k]

Model B uses Bayesian inference to estimate the probability that unseen copies still exist. It accounts for the fact that not all cards become publicly visible, and not all visible cards are reported.

Key Parameters:

  • f (visibility factor): Estimates what fraction of cards eventually surface publicly. Increases with more total sightings for a product.
  • q (detection rate): The probability we catch a card when it does surface. Uses a Beta(2,6) prior, meaning we expect to catch about 25% of visible cards on average.

Output: Model B provides P(still unpulled) — the probability that at least one copy hasn't been opened yet — and E[remaining] — the expected number of copies still in sealed product.

Confidence Score

Each probability estimate includes a confidence score (0-100) that indicates how reliable the estimate is. The score is based on:

  • Coverage: What percentage of the print run has been sighted
  • Source diversity: How many different source types (YouTube, eBay, PSA, etc.) have contributed sightings
  • Product maturity: How many total sightings exist for the product (affects the f parameter)
Low
< 40 points
Limited data
Medium
40-69 points
Reasonable estimate
High
≥ 70 points
Strong confidence
Data Sources

We accept sightings from the following sources:

YouTube
Break videos
eBay
Listings
PSA
Registry
Manual
Personal photos
Other
Forums, social
Verification Process

All submitted sightings go through a verification process before being included in our probability calculations:

  1. Submission: Users submit sightings with source URLs
  2. Review: Our team verifies the source and card details
  3. Duplicate check: We check for duplicate serial numbers
  4. Approval: Verified sightings are marked APPROVED
  5. Calculation: Only APPROVED sightings affect probabilities
Limitations & Disclaimers
  • Probability estimates are based on publicly visible data only
  • Cards in private collections may never surface publicly
  • Some cards may be lost, destroyed, or permanently held
  • Our models make assumptions that may not hold for all products
  • ChasePulse is not affiliated with Topps or any card manufacturer
  • This is not financial advice — always do your own research

Have questions or feedback? We'd love to hear from you.