No TCG publisher tells you the odds. Every pull rate you've ever read — 1 in 33 for a special illustration rare, 1 in 12 for an alt art — is a community estimate built by people opening boxes and counting. Some of those estimates are excellent. None of them are official, and anyone quoting them to two decimal places is selling confidence they don't have.
Here's how the numbers actually get made, and how to use them without fooling yourself.
Where pull-rate numbers come from
Sampling. Someone — a content creator, a forum project, a store cracking cases — opens a few hundred or a few thousand packs, logs every hit, and divides. That's it. There's no leaked spec sheet from the printer.
Sample size is everything. If a card appears at a true rate of 1 in 80 packs, a 500-pack sample expects to see it about six times. Random clumping means that sample could easily show four hits or nine — which reads as "1 in 125" or "1 in 55." Both are wrong, and both get repeated forever once published. Rates for common slots (reverse holos, double rares) converge fast; rates for the rarest tiers carry wide error bars even after thousands of packs.
Print waves add another wrinkle. Collation can shift between production runs, so a rate measured at launch may drift by the third restock. Nobody re-samples old sets, so quoted rates fossilize.
The slot model: how a pack is actually built
Packs aren't a lottery over the whole set. They're built from slots, and each slot draws from its own rarity table. A standard Scarlet & Violet-era pack looks like this in the model our Pack Value Calculator uses — community estimates, flagged as such:
| Slot | What can appear | Estimated odds |
|---|---|---|
| Reverse holo | Reverse holo | 100% |
| Holo rare slot | Rare 55% / Double rare 30% / Ultra rare 10% / ACE SPEC 5% | per pack |
| Hit slot | Baseline holo 84.8% / Illustration rare 11% / SIR 3% / Gold 1.2% | per pack |
Read that hit slot honestly: about 85% of packs give you nothing above baseline in the slot that matters. An illustration rare shows up roughly 1 pack in 9, a special illustration rare about 1 in 33, a gold about 1 in 83. Across a 36-pack booster box you should expect around 1 SIR and 4 illustration rares — expect, not receive. Zero-SIR boxes are entirely normal at those odds.
Rates change set to set — sometimes drastically
The slot model is not one-size-fits-all, which is why per-set numbers matter:
- Prismatic Evolutions guarantees a shiny-slot card in every pack, with a special illustration rare landing there an estimated 16% of the time. That's five times the standard SV rate, and it's the whole reason the set's pack EV looks so different.
- Crown Zenith adds a dedicated Galarian Gallery slot on top of the normal pack, with an estimated 13% chance of an alt art in that slot alone. Boosted rates were the set's selling point.
- Standard sets like Surging Sparks or Stellar Crown run the vanilla table above.
Special sets with guaranteed-hit structures also price differently, which is a big part of the ETB versus booster box math changing from set to set.
What pull rates don't tell you
A pull rate tells you how often a rarity tier appears. It doesn't tell you which card. A Destined Rivals SIR slot hit is Team Rocket's Mewtwo ex at about $559 as of our July 2026 price snapshot — or Team Rocket's Crobat ex at about $70. Same slot, same odds, eight-fold difference in outcome. Set value concentrates viciously in one or two cards, which is why rates alone can't tell you whether a box is worth ripping — you need rates multiplied by prices, which is expected value, and even EV hides the variance that decides your actual night.
How to sanity-check a claimed pull rate
- Ask for the sample size. Under ~1,000 packs, treat rare-tier rates as rough sketches.
- Check it's per-slot, not per-pack mush. Good data separates the hit slot from the holo slot; bad data blends them.
- Check the date. Launch-week data may not match post-restock product.
- Distrust suspicious precision. "3.14% SIR rate" from 800 packs is numerology.
FAQ
Are pull rates the same in every box?
The long-run rates appear consistent within a print run, but individual boxes swing hard. At a 3% SIR rate, roughly a third of 36-pack boxes contain no SIR at all. That's math, not a bad box.
Do ETB or bundle packs have better odds than loose packs?
There's no credible evidence for it. The packs are the same product; only the packaging and the price per pack differ.
How many packs until I'm "due" a big hit?
You're never due. Each pack is independent — after 32 dry packs, pack 33 is still roughly a 3% shot at an SIR. That's precisely why estimated rates only describe long-run averages, not your box.