PackRips reads the public live feeds of 19 sites and saves every pull with these details.
Pulls are stored without usernames, wallet addresses or avatars.
All card value ÷ all money spent
At 100%, the cards were worth exactly what people paid for the packs. Every ranking on PackRips uses this number.
The middle pull's value ÷ its price
Half of pulls land below it. If it sits far under the average return, a few big hits carry the pack.
Pack pages also show the average return for today, this week and all time, so you can see whether a pack is running hot or cold. New pulls show up within seconds, and PackRips works the returns out again every five minutes.
A pack with fewer than 500 pulls counted stays visible. It keeps its own page and shows up in the unranked thin samples view, marked provisional, because a few hundred pulls can still swing a return a long way. The API marks the ranked packs ranked.
A site needs 1,000 pulls counted before it gets a rank number. Below that, PackRips lists it as provisional.
Category averages on a site's live page are marked solid at 100 or more pulls, building at 30 to 99 and provisional below 30.
Hits-only feeds. Some sites post only their big wins, which would make returns look far better than they are. PackRips shows those pulls on the Live page but leaves them out of every return. When a site streams every pull, as Tilt Rips does, PackRips reads that stream.
Duplicate pulls. Collector Crypt reposts ONE Arena pulls. Each pull counts once, on the site where it was opened.
Pulls without a price. If PackRips can't match a pull to the price paid, the pull doesn't count toward a return. It still counts toward category averages on a site's live page.
Store credit. Credit you can only spend on more packs is worth less than cash, so PackRips ranks store-credit sites and their packs on a separate board. Cash and crypto sites (USDC, for example) share one board. Every row says what the site pays in.
Payout terms. Grades use average return only. You'll find sell-back rates and fees listed next to each grade.
Card value is the price the site puts on the card. You only get all of it if you ship the card and sell it yourself. Most sites also buy cards back on the spot for a share of that value, and that sell-back value is what lands in your account. Returns on PackRips use card value unless a label says otherwise. When a site states its own expected value after sell back, PackRips labels it that way and shows the sell-back rate beside it so you can convert.
Some sites publish odds for each value tier. Those odds are the site's own claim, and they never decide a ranking. Once a pack has 100 pulls counted, its page shows the published odds next to what the pulls showed, tier by tier.
A site gets a letter from its average return and nothing else, on the same scale for every site, once PackRips has counted 1,000 of its pulls. An A means you get back more than you spend, on average.
Each range is a site's average return. Payout terms sit beside the grade, in full, and never move it.
Every figure is PackRips' own count from the sites' live feeds. A figure under 100 pulls is an early sample and can still move a lot, so check its pulls counted. PackRips launched with figures imported from RipRanks' published data and no longer uses any of them. Some payout terms still come from there, and PackRips is re-checking them against each site's own pages. Every figure and its pulls counted are also free as JSON from the PackRips API.
Card values come from the site or its pricing partner. PackRips reports them and doesn't price every card again.
A live feed is a sample. If a site filters what its feed shows, the sample carries that bias.
A site could fake its feed. That's true of any outside measurement. PackRips checks published odds against the pulls and shows any gap tier by tier.
Send the page and what looks wrong. Fixes go into the data, so the pages, the API and llms.txt change together. Operators can write too, and they can't pay to change a number. About PackRips says more about how the site stays independent.