How to credit us

Free to share and adapt, including commercially, under CC BY 4.0. Please cite:

Poké Life, “Pokémon TCG pull-rate evidence registry,” https://poke-life.com/data/ (accessed Oct 12, 2026). CC BY 4.0.

License: CC BY 4.0. A link back to the set page you used is appreciated.

The files above always hold the latest numbers. To cite a fixed version, link /data/v/2026-10-12/pull-rates.csv; every version is listed in the data changelog.

A look inside pull-rates.csv

First 8 of 157 rows.

SetRarityHitsPacksRate95% rangeConfidenceOpenings
151Double Rare2381,795*13.28%11.75–14.89%Very sureMEW-001
151Illustration Rare1531,795*8.50%7.30–9.88%Very sureMEW-001
151Ultra Rare1171,813*6.44%5.39–7.65%Very sureMEW-001
151Special Illustration Rare561,809*3.11%2.37–3.97%Fairly sureMEW-001
151Hyper Rare351,784*1.94%1.39–2.69%Fairly sureMEW-001
151Holo Energy4451,792*24.83%22.87–26.87%Very sureMEW-001
30th CelebrationDouble Rare9164,356*21.02%19.84–22.26%Very sure30C-001
30th CelebrationIllustration Rare6393,240*19.72%18.38–21.12%Very sure30C-001

* Pack count reconstructed from the published rate and margin of error.

What each column means

set_id, set_name
Which set
rarity_id, rarity
Which rarity, using the set's official names
cards_in_rarity
How many different cards share that rarity
hits, packs
Hits found and packs opened, combined across verified openings
rate
Our headline rate per pack
ci95_lower, ci95_upper
95% range (Jeffreys interval)
confidence
high, moderate, limited, or estimated (shown on the site as Very sure, Fairly sure, Rough estimate, Estimate)
counts_reconstructed
True when a source only published a rate and margin, so we worked out the counts
datasets
IDs of the openings behind the rate; look them up in datasets.csv

What's not included

Card and sealed prices. Our price providers' terms don't allow redistributing their data, so prices stay on the site and out of the downloads. Values derived from prices, like pack value, are shown on set pages but not offered in bulk.

Unverified openings are listed in datasets.csv with qualifying = false, so you can see what we left out, but they never feed the rates.

How every number is made: our methodology.